UNIVERSITY OF FLORIDA LIBRARIES
PHILOLOGICAL MONOGRAPHS
PUBLISHED BY THE
AMERICAN PHILOLOGICAL ASSOCIATION Number XVIII
EDITED BY
JOHN L. HELLER University of Illinois
COMMITTEE ON THE PUBLICATION OF MONOGRAPHS
BERTHE MARTI Chairman
PHILLIP H. DE LACY JOHN B. McDIARMID
JOHN L. HELLER JOHN B. TITCHENER
THE BRONZE TABLES OF IGUVIUM
BY
JAMES WILSON PpULTNEY
The Johns Hopkins University
PUBLISHED BY THE AMERICAN PHILOLOGICAL ASSOCIATION
1959
To be ordered through the Secretary of the Association
JAMES W. POULTNEY, THE JOHNS HOPKINS UNIVERSITY BALTIMORE 18, MARYLAND or through B. H. BLACKWELL, LTD, 50 BROAD ST., OXFORD, ENGLAND
ROLAND I GRUBB KENT MANIBUS
Digitized by the Internet Archive
in 2011 with funding from
LYRASIS IVIembers and Sloan Foundation
http://www.archive.org/details/bronzetablesofigOOpoul
PREFACE
There is no scarcity of editions of the Iguvine Tables, as a glance at the Bib- liography will show. The study of these tables and of Italic dialect monuments in general, which was intensified during the thirties and which lapsed during the Second World War, has been revived again, especially in Italy. But the older grammars of von Planta and Buck have been out of print for many years and can now be obtained only with considerable difficulty, and moreover until quite recently all translations of the tables were in a Latin which adhered with extreme fidehty to the Umbrian text, reproducing many of its obscurities and translating some of its words merely by reconstructions invented through ap- plication of phonetic laws. Professor Roland Kent both in print (Language, XIV [1938], p. 213) and in conversation with me emphasized the need for a translation into a modern language and it was at his suggestion that I undertook to prepare an edition of the type which he contemplated, including an English translation. Devoto's Le Tavole di Giibbio, which appeared in 1948, actually includes a translation into Itahan beside the Umbrian text. In 1954 there ap- peared Bottiglioni's Manuale dei dialetti italici, which contains not only the Iguvine Tables and other dialect texts with Latin translation and brief com- mentary but a fairly comprehensive grammar as well. Yet the importance of the Iguvine Tables is so great both for the hnguistic and for the religious history of pre-Roman Italy that another edition seems not superfluous, especially in view of the fact that reading knowledge of Itahan is not as widespread in the English- speaking countries as it should be.
In the present edition, as in previous works, the grammar has necessarily been presented from a historical standpoint throughout, for only by careful attention to the history of the sounds can we set up valid etymological equations which, in combination with other evidence within and outside the LTmbrian text, can give us the correct sense. The interpretation of the cuneiform Hittite texts and the recognition that Hittite is related to the Indo-European languages have radically altered our notions of proto-Indo-European phonology, especially as a result of the conviction shared by the great majority of Indo-Europeanists of
VIII PREFACE
the present day that the phonemic system at some stage must have included several " laryngeal " consonants. Yet there is still considerable disagreement with regard to the number of laryngeals (three or four?), their precise character and effect on the vowel-system, and whether Hittite is to be regarded as an Indo- European language roughly comparable to Sanskrit, Greek, Latin, etc., or as a member of an " Anatohan " group which, with Indo-European, was derived from an earlier " Indo-Hittite." For these reasons it has seemed safest and most convenient to follow the traditional reconstruction adopted in such standard works as Brugmann's Gnmdriss, Buck's Comparative Grammar of Greek and Latin, or Kent's Sounds of Latin, while giving a brief account of the earher laryngeals and their part in the subsequent development of the vowel- and consonant-system (see II, §§ 35, 36). The principal deviation from the traditional system is in the recognition of two, not three, series of /c-sounds. It is improb- able that the " plain velars " originally existed as a series distinct from the palatals, and in any case the distinction is unimportant for Oscan-Umbrian, as for Latin, Greek, Celtic, and Germanic.
Following the practice of other recent editors I have presented the Umbrian text without capitalization and without punctuation other than the word- dividers, but where words are written together on the bronze without the divider I have inserted a space between them if they are known to be properly separate. The customary practice of using itahc type for texts and individual words re- corded in the Latin alphabet and bold-face type for those in the native alphabet has been followed throughout the present work for Oscan as well as for Um- brian. Oscan words recorded in the Greek alphabet are regularly preceded by the abbreviation "0." Greek words are not designated, the alphabet being sufficiently distinctive to identify them as Greek.
In order to provide an easily readable translation I have avoided placing question marks after those words whose meaning is doubtful, although such marks are shown in the Lexicon when the meanings seem sufficiently uncertain to justify it. The practice of leaving words in the translation in their Umbrian form has been held down to a minimum, but in a few cases it has seemed unavoid- able; so, for example, in the case of erus, persondro, and the names of several kinds of offering-cakes. Certain personal and divine names and other words not capable of being fully translated into English have been presented in Latin dress; so, for example, adfertor (for Umbrian arsfertur), prinuati (for prinuatur prinuvatur with native nom. plur. ending). But it is impossible to achieve and maintain an absolutely logical and consistent standard of usage in this regard.
For those texts which exist in two versions, an earher and a later, the fullest commentary is on the later (Via, b, Vila, b), both because their greater length renders them more important and because, if my impression is correct, instructors often make it a habit to present them to students before the earlier versions.
PREFACE IX
Yet la and b are provided with commentaries designed to summarize their content and to explain matters wherein they differ from the later tables.
The Lexicon is at the same time an Index Verborum, and is intended to account for every instance of every word in the Tables.
The selection of tables to be illustrated by plates was made with a view toward the variety of alphabets and letter-forms represented. Tables la and Va show somewhat divergent forms of the native alphabet. Table Vb shows the native and Latin alphabets both on the same surface of one table, while Vila shows the Latin alphabet in an especially clear and legible form. All twelve plates are to be seen not only in Devoto's Tabulae Iguvinae, but also in the ediiio minor of the same work which contains, in addition to the plates, simply the Umbrian text and Latin translation.
My debt to earlier editors, especially Devoto, is far greater than one might infer from the references to them which appear here and there, but I have en- deavored to maintain independence of judgment and have occasionally arrived at a new interpretation of my own.
Several publications of importance for the study of the Tables have ap- peared too late for use to have been made of them in the preparation of the present work. Among them may be mentioned : Giovannangelo Camporeale, La terminologia magistratuale nelle lingue osco-umbre (Atti delV Accademia Tos- cana di Scienze e Lettere ''La Colombaria"), pp. 78, Florence, 1957 ; K. Olzschka, Das umbrische Perfekt auf nki (Glotta, XXXVI [1958], pp. 300-304) ; Ugo Coh, // diniio pubblico degli umbri e le tavole eugubine (Circolo Toscano di Diritto Romano e Storia del Diritto, 1), pp. 98, Milan, 1958.
My obligations for help in bringing my work to completion are many. In the early stages of my study of the Iguvine Tables Professor Kent gave me much valuable assistance in conversation and by letter, and it is a matter of great regret to me that he did not live to see the finished work. I am deeply indebted to the Committee on the Pubhcation of Monographs for accepting the manuscript for publication, to the referees for their many valuable suggestions toward the improvement of the work, and to the American Philological Association itself for bearing a large share of the cost. I am deeply grateful to Professor Giacomo Devoto for graciously allowing the plates in his own edition to be used in the preparation of plates for mine, and to Professor Herbert Bloch and Mr. Ernest Nash for generously exploring various means of reproducing the plates. To the Imprimerie Universa I am greatly indebted for patience, care, and skill in the printing of an extraordinarily difficult work. For the reading of the proofs I take the present opportunity to express my thanks to Suzanne Young and Robert E. A. Palmer, both of whom relieved me of the tedious and risky task of reading alone. Finally, to John L. Heller as the editor of this monograph my gratitude is due in a very exceptional degree, not merely for his having assisted me with
X PREFACE
his philological learning and with the full range of his familiarity with the intric- acies of editing and printing, but far more for the unnumbered hours which he has patiently devoted to the minute examination and correction of the manu- script and proofs. As an ancien eleve of Professor Kent's, he desires to join me in dedicating this work to his memory.
September 29, 1958. James W, Poultney.
TABLE OF CONTENTS
Abbreviations and Bibliography xiii
1. Editions of the Iguvine Tables, and works containing them in their entirety or in part 2. Lexicons, grammatical works, monographs, selected journal articles, etc. 3. Periodicals 4. Dialects 5. Grammatical terms 6. Con- cordance of dialect inscriptions cited
I. Introduction 1
1-2. Description of the Tables 3-4. History and topography of Iguvium 5-6. Umbria and the Umbrian dialect 7-9. Other Italic dialects; concept of the group 10-11. Language and style of the Iguvine Tables 12. Discovery and history of the Tables 13. Interpretation of the Tables; outline of con- tents 14-15. Religion and ritual in the Tables 16. Organization of the Atiedian priesthood and of the state of Iguvium 17-19. Chronology of the Tables
II. Grammar 25
A. Phonology (1-63) 1 25
1-3. The alphabets, variants, and pronunciation 4. Accent 5-25, The vowel system; diphthongs 26-35. Lengthening, shortening, syncope and other phenomena of the vowels 36-59. The consonant system; consonant clusters 60-62. Other phenomena of the consonants; juncture 63. Sources of the Umbrian sounds
B. Formation of nouns and adjectives (64-89) 84
64-84. Stem classes 85-89. Nominal compounds
C. Declension of nouns and adjectives (90-103) 97
D. Numerals (104-105) 105
E. Pronouns (106-111) 107
F. Indechnables (112-113) 113
G. The Verb (114-134) . . . 116
114. General survey 115. Personal endings 116-121. Present tense forma- tion and conjugational classes 122. Irregular verbs 123-125. Other tenses 126-127. The moods 128-129. Infinitive and supine 130-131. Participles 132-133. Gerundive; periphrastic forms 134. Verbal composition
^ Throughout the book, boldface numerals refer to sections of the Grammar.
XII CONTENTS
H. Syntax (135-155) 142
135-139. General remarks; ellipsis, agreement, etc. 140-146. The cases 147. Prepositions and postpositions 143-153. Verbal syntax 154. Subordinate clauses 155. Word order
III. Text, translation, and commentary 157
IV. Lexicon and index verborum 295
Plates I-IV 334
Tables la, Va, Vb, Vila
ABBREVIATIONS AND BIBLIOGRAPHY
1. Editions of the Iguvine Tables, and works containing them in their entirety or in part
A.-K. = Th. Aufrecht und A. Kirchhoff, Die umbrischen Sprachdenkmaler.
Berlin, 1849-51. Hu. = E. Huschke, Die iguvinischen Tafeln, Leipzig, 1859. Br. = M. Br^al, Les Tables Eugubines. Paris, 1875. Bli. = F. Biicheler, Umbrica. Bonn, 1883. PI. = R. von Planta, Grammatik der oskisch-umbrischen Dialekle. 2 vols. Strass-
burg, 1892-97. Co. = R. S. Conway, The Italic Dialects. Cambridge, 1897.
* R. S. Conway, Dialectorum Italicarum Exempla Selecta. Cambridge, 1899. H. Jacobsohn, Altilalische Inschriften, ausgewdhlf. Bonn, 1910.
Bk. = C. D. Buck, A Grammar of Oscan and Umbrian. New printing with addi- tions and corrections. Boston, 1928.
Blum. = A. von Blumenthal, Die iguvinischen Tafeln. Stuttgart, 1931.
Rosenzweig = Irene Rosenzweig, Ritual and Cults of pre- Roman Iguvium. Lon- don, 1937 (with Buck's text and translation).
Dev. = G. Devoto, Tabulae Iguvinae. 2nd edition. Rome, 1940.
Dev., T.G. = G. Devoto, Le Tavole di Gubbio. Florence, 1948.
• H. H. Janssen, Oscan and Umbrian Inscriptions. Leyden, 1949.
♦ Angel Montenegro Duque, Osco y Umbro. Madrid, 1949.
Pis. = V. Pisani, Manuale storico delta lingua latina. Vol. IV, Le lingue dell' Ita- lia antica oltre il latino. Turin, 1953. Vet. = Emil Vetter, Handbuch der italischen Dialekte. Vol. I. Heidelberg, 1953. Bott. = G. Bottiglioni, Manuale dei dialetti italici. Bologna, 1954.
* contains only selected passages of the Iguvine Tables.
2. Lexicons, grammatical works, monographs, selected journal articles, etc.
Brugmann, Gdr. = K. Brugmann, Grundriss der vergleichenden Grammatik der
indogermanischen Sprachen. 2nd edition. Strassburg, 1897-1916. Buck, C. D., The Oscan- Umbrian Verb System. Chicago, 1895. CLE. — Corpus Inscriptionum Etruscarum. C.I.L. = Corpus Inscriptionum Latinarum. Devoto, G., Gli antichi italici. 2nd edition. Florence, 1951. Dumezil, G., Les dieux des indo-europeens. Paris, 1952. E.-M. = A. Ernout et A. Meillet, Dictionnaire etymologigue de la langue latine.
3rd edition. 2 vols. Paris, 1951. Fest. = Sexti Pompei Festi de verborum significatu quae supersunt cum Pauli
Epitome. Ed. W. M. Lindsay. Leipzig, 1933. Hofmann, J. B., Bursian's Jahresbericht, CCLXX (1940), pp. 80-100 (Literatur-
bericht).
XIV ABBREVIATIONS AND BIBLIOGRAPHY
Kent, R. G., The Sounds of Latin. 3rd edition. Baltimore, 1945.
Kent, R. G., The Forms of Latin. Baltimore, 1946.
Kent, T.C.I. = Kent, R. G., The Textual Criticism of Inscriptions. Philadelphia,
1926. Leumann-Holmann = M. Leumann und J. B. Hofmann, Lateinische Grammatik.
Munich, 1928. Mu. = F. Muller Jzn, Altitalisches Worterbuch. Gottingen, 1926. Pighi, Umbrica = Memorie dell' Accademia delle scienze, lettere ed arti di Bologna,
Classi morali, ser. V, vol. IV (1953). Pokorny = J. Pokorny, Indogermanisches etymologisches Worterbuch. Bern, 1948-. R.-E. = Pauly-Wissowa, Realencyclopddie der klassischen Altertumswissenschaft. Ribezzo, I = Fr. Ribezzo, Problemi iguvini, Rivista indo-greco-italica, XVIII
(1934), fasc. 3-4, pp. 55-88. Ribezzo, II = Fr. Ribezzo, Problemi iguvini, Rivista indo-greco-italica, XX (1936),
fasc. 1-2, pp. 75-106. Rosenberg, Arthur, Der Staat der alien Italiker. Berlin, 1913. Schulze, Lat. Eigenn. — W. Schulze, Zur Geschichte lateinischer Eigennamen.
Berlin, 1904. Sommer, F., Handbuch der lateinischen Laut- und Formenlehre. Heidelberg, 1948. Walde-Hofmann — A. Walde und J. B. Hofmann, Lateinisches etymologisches
Worterbuch. Heidelberg, 1938-54. Walde-Pokorny = A. Walde und J. Pokorny, Vergleichendes Worterbuch der indo-
germanischen Sprachen. 3 vols. Berlin & Leipzig, 1930-32. Whatmough, J., The Foundations of Roman Italy. London, 1937.
3. Periodicals
A.G.L = Archivio glottologico italiano.
A. J. P. — American Journal of Philology.
A.L.L. — Archiv fiir lateinische Lexikographie und Grammatik.
B.B. = (Bezzenberger's) Beitrdge zur Kunde der indogermanischen Sprachen.
B.S.L. — Bulletin de la Societe de linguistique de Paris.
Ber. kon. sdchs. Ges. Wiss. = Berichte der koniglichen sdchsischen Gesellschaft der
Wissenschaften zu Leipzig. Philologisch-historische Klasse. C.P. = Classical Philology. Gl. = Glotta. His tor ia. Milan, 1927-35. I.F. = Indogermanische Forschungen.
K.Z. = (Kuhn's) Zeitschrift fiir vergleichende Sprachforschung. Lang. = Language. Latinitas.
Mnem. — Mnemosyne.
N.Jb. = Neue Jahrbiicher fiir Philologie und Pddagogik. R.F.I.C. = Rivista di Filologia e d' Istruzione Classica. R.I. G.I. = Rivista indo-greco-italica. Rom. Mitt. = Mittheilungen des deutschen archdologischen Instituts, Romische
Abtheilung. T.A.P.A. =^ Transactions and Proceedings of the American Philological Association.
ABBREVIATIONS AND BIBLIOGRAPHY
XV
4. Languages and Dialects
Aeol, = Aeolic Av. = Avestan Dor. = Doric Eng. ^ English Etr. = Etruscan Fal. = Faliscan Fr. = French Germ. = German Gk. = Greek Goth. = Gothic Hitt. = Hittite Hom. = Homeric IE = Indo-European Ital. = Italian L. = Latin Lett. — Lettish Lith. = Lithuanian Marruc. = Marrucinian
|
Mars. |
= Marsian |
|
0. = |
Oscan |
|
ocs = |
= Old Church Slavic |
|
OE = |
: Old English |
|
OHG |
= Old High German |
|
Olcel. |
= Old Icelandic |
|
Olr. = |
= Old Irish |
|
OL = |
Old Latin |
|
ON = |
-- Old Norse |
|
Pael. |
= Paelignian |
|
pIE = |
= proto- Indo-European |
|
pit. = |
= proto-Italic |
|
Russ. |
= Russian |
|
Sab. = |
= Sabine |
|
Skt. = |
= Sanskrit |
|
Ved. |
= Vedic |
|
Vest. |
= Vestinian |
5. Grammatical terms
abl. = ablative ace. = accusative act. = active adj. = adjective conj. — conjugation dat. = dative decl. = declension denom. = denominative f. = feminine fut. = future fut. pf. = future perfect gdve. = gerundive gen. = genitive impf. = imperfect imv. = imperative ind. = indicative
inf. = infinitive
loc. = locative
m. = masculine
neut. = neuter
nom. = nominative
pass. = passive
pcpl. — participle
pf. = perfect
pi. = plural
postpos. — postposition,
postpositive pres. = present sg. = singular subj. = subjunctive voc. = vocative
The abbreviation "cf." is used in etymological equations in the grammar, commentary, and lexicon to introduce forms which are cognate with the forms being explained, but not full etymological equivalents with them. The abbreviation is omitted before forms which are full etymological equivalents of those being explained, or which show only minor differences of stem-formation while still forming parts of the same paradigm. E.g., (10a) "$imo gimu, give: cf. L. citra," (Lexicon) "arkani ... cf. L. ac-cino, ac-centus," but (42 b) "calersu kalerxif : L. calidus," (Lexicon) "fertu... L. few."
XVI
ABBREVIATIONS AND BIBLIOGRAPHY
6. Concordance of dialect inscriptions cited
|
Conway |
von Planta |
*Buck |
Pisani |
Vetter |
Bottiglioni |
|
28 |
17 |
2 |
9 |
2 |
79 |
|
63 |
49 |
17 |
13G |
25 |
75 |
|
95 |
127 |
1 |
18 |
1 |
51 |
|
105 |
141 |
23 |
77 |
33 |
|
|
107 |
137 |
33 |
22B |
83 |
44 |
|
108 |
138 |
25 |
23 |
94 |
36 |
|
109 |
134a |
26 |
24 |
81 |
37 |
|
113 |
133 |
29 |
25A |
86 |
39 |
|
114 |
134 |
30 |
25B |
87 |
40 |
|
115 |
131 |
27 |
26A |
84 |
41 |
|
116 |
132 |
28 |
26B |
85 |
42 |
|
117 |
135 |
31 |
27 |
88 |
45 |
|
130 |
128 |
19 |
28 |
6 |
27 |
|
137 |
119 |
40 |
30 |
5 |
54 |
|
164 |
182 |
55 |
37 |
161 |
19 |
|
169 |
188 |
50 |
149 |
13 |
|
|
175 |
200 |
45 |
34 |
147 |
7 |
|
208 |
146c |
45D |
206 |
131 |
|
|
216 |
254 |
47 |
213 |
122 |
|
|
239 |
253 |
50A |
216 |
125 |
|
|
248 |
280 |
54 |
227 |
120 |
|
|
253 |
242 |
56 |
223 |
132 |
|
|
264 |
308 |
228b |
134 |
||
|
352 |
292 |
82 |
61 |
230 |
116 |
|
353 |
293 |
232 |
|||
|
354 |
295 |
83 |
62A |
233 |
112 |
|
355 |
296 |
34 |
63 |
236 |
113 |
|
182 |
13E |
28 |
78 |
||
|
35D |
141 |
5b |
* The same numbering is followed in Buck's Elementarbuch der oskisch-umbrischen Dialekte, deutsch von E. Prokosch (Heidelberg, 1905).
I. INTRODUCTION
1. The seven bronze tablets which were discovered at Gubbio, the ancient Iguvium, in 1444 and are now preserved there in the Palazzo dei Consoli, are the only extant records of any considerable extent in the Umbrian dialect;^ that is, in that language which, with Oscan, Latin, and several other dialects, makes up the Itahc branch of the Indo-European family.
2. The largest of the tablets, VI and VII, ^ are each about 33 in. in height and 22 in. in width, the smallest, III and IV, each about 16 by 12. All except III and IV are inscribed on both sides; I, II, III, IV, and the obverse and part of the reverse of V are in a native alphabet very similar to the Etruscan alphabet, the remainder of the reverse of V, as well as VI and VII, in the Latin alphabet. The total number of words is over 4000. The tablets contain instructions for the religious ceremonies of a college of priests known as the Atiedian Brothers who flourished in Iguvium during the period of the Roman repubhc. No other body of liturgical texts from pre-Christian Europe can compare with the Iguvine Tables in extent. They have therefore an extraordinary importance both for the linguistic and the rehgious history of early Italy.
3. Gubbio or Iguvium is located in the province of Perugia about 20 miles north-northeast of the city of Perugia and on the western slope of the Apennines. It is known from Cicero, Balh., 47, to have been alhed to Rome as a civilas foe- derata, and it was there, according to Livy, XLV, 43, that the Illyrian king Gen- tius was sent by the senate after the battle of Pydna to live in exile with his wife and children. After the Social War the inhabitants of Iguvium were granted Roman citizenship and the town became a municipium in the tribus Clustumina. It suffered a disadvantage as a result of the fact that the Via Flaminia, instead of passing through it, lay some distance to the east. Yet it had some importance during the Middle Ages, under the name Eugubium. Charlemagne is said to have passed through it on his return north after his coronation in Rome, and it had been an episcopal see at least as early as the fifth century. Ubaldo, who was bishop of Gubbio from 1129 to 1160, was canonized in 1192 and became the patron saint of the town. The Festival of the Ceri, which is celebrated in his honor on May 15, has certain features which are sometimes regarded as survivals from the pagan cult described in the Tables.^ In 1384 Gubbio passed under the
* Umbrian inscriptions apart from the Iguvine Tables liave been found at Amelia (Ameria), Assisi, Fossato di Vico (Helvillum), Foligno (Fulginiae), Gubbio (coins only), Spoleto, and Todi (Tuder). There are scarcely more than ten in all, and the longest, Co. no. 355, contains only twenty words counting abbreviations. For the minor inscriptions see Bii., pp. 172-7; Co., pp. 397-9; PI. II, pp. 554-7; Bk., p. 310; Vet., I, pp. 164-70; Pis., pp. 214-6; Bott., pp. 323-5; Whatmough, Harvard Studies in Classical Philology, 1939, pp. 89-93.
* For the standard numbering of the tablets see § 13 below.
* Rosenzweig, pp. 44-5.
2 I. INTRODUCTION
control of the counts of Urbino, and with Urbino it became a part of the papal territory in 1631, and so continued until the plebiscite of 1860.
4. Of the places mentioned in the Tables few can with any reasonable certainty be fixed in relation to the topography of the modern town. The spot from which the auspices were taken, and hence the Fisian Mount itself, is probably to be placed not on Monte Ingino, which lies on the northeast side of Gubbio, but rather on Monte Foce directly to the north, for only by this theory can the ob- servation of the auspices have been made in the southeastward direction which the text of Via 1-18 seems to demand. A direct eastward facing is hardly possible. The three gates at which sacrifices were performed can be fixed to a certain degree from their names. If uerir Uehier means ' the Veian Gate, the gate which leads toward Veil, ' this gate would naturally be on the south of Iguvium. The Tese- nacan Gate, if it faced toward a town on the site of the modern village of Tesse- nara about 18 km. southwest of Gubbio, must have been on the southwestern side of the town. The name of the third gate (the first in order in the Tables themselves) contains a root treb- which was used extensively in Italic place-names, but two of the possible sources of the name, Trebula Suffenas and Trebula Mu- tuesca, were in the Sabine country and hence imply that the gate may have been on the southeastern side. This distribution of the three gates is strikingly uneven, but it must be remembered that the mountains on the northern and eastern sides of the town precluded the building of important roads. The location of the gates is connected with the question in which direction the procession moved during the purification of the Fisian Mount.* Miss Rosenzweig favors a coun- terclockwise direction, as in the lustral procession depicted on the Column of Trajan (Strong, La scultura romana, fig. 99), Devoto a clockwise direction, the turning toward the right being supported by Solinus, p. 195 Mo. nee ante substitit [quadriga] ... quam Tarpeium lovem trina dextratione lustrasset. Neither of the opposing arguments is sufficient to settle the question, but if the location proposed for the gates with the help of place-names is correct, the counterclockwise direction seems better, since Devoto {T.G., p. 17) is obliged to place the third or Veian Gate on the northwest side of the town.
The remaining localities mentioned in the text as places for the performance of ceremonies or as landmarks for marking out the augural templum are discussed in the notes on the passages where they occur.
5. There was an ancient tradition according to which the Umbrians were the ear- nest inhabitants of Italy. This belief is attested by Dionysius of Halicarnassus,^
* For the two problems see Rosenzweig, pp. 13-6; Devoto, pp. 177-8; Devoto, T.G., pp. 15-7. It is difficult to see why Miss Rosenzweig places the Porta Trebulana at the northeastern corner of the town wall rather than on the eastern or southeastern side.
^ Ant. Rom., I, 19 oi de did rfjQ /neooyeiov TQano/jievoi [sc. IleXaayol] rrjv dgeivrjv rfjg '/ra- At'ag vjisQ^aXovreQ, eig ttjv 'OfA^Qixoiv dcpixvovvrai ;tc6gaj' rd)v o/iogovvTcov 'AfioQtylai.
I. INTRODUCTION 3
Florus,6 and Pliny.' This mention of an extensive conquest of Umbrian territory by Etruscans is in partial agreement with the statement of Hero- dotus, I, 94, that the Etruscans after crossing the sea from Smyrna arrived in the land of the Umbrians, Moreover the use of the name of the 'Ofi^Qixoc to designate the inhabitants of a very large portion of northern Italy receives addi- tional confirmation from Herodotus, IV, 49 ex de rrjg xarvjieQds xoiQ-qc, 'Oju^qi- K&v KaQTiiQ norafjbdq >cal dXXoq "AXnig TtQog Boqeiqv dvefiov xal ovroi geovreg endidovai eg avrov [sc. xov "laxQov]. Umbria, however, is generally un- derstood much more narrowly, as the name of that portion of north central Italy separated on the west from Etruria by the Tiber and bordering Aemilia on the north, Picenum on the east, and the country of the Sabines on the southeast, and mcluding in the northeast a small portion of the Adriatic coast between Ancona and Ariminum. It was numbered sixth among the regiones of Italy by Augustus. The greater part of Umbria lies to the east of the line which ex- tends northward and southward across Italy in the longitude of Rome, Gubbio, and Rimini, separating the cremating people of the west from the inhuming people of the east,^ and in general the iron-age material shows a close resemblance to the Picene culture. But the few sites which have yielded inscriptions in the Italic dialect known as "Umbrian" mostly lie along the eastern fringe of the cremation area.^ The term " Umbrian dialect " then must not be taken as coextensive in area with the Umbrian territory as understood by Herodotus, or even with the sixth regio of Augustus.
6, The question of the position of Umbrian among the ancient languages of Italy is essential in any general treatment of the language of the Iguvine Tables. The linguistic diversity in Italy in the pre-Christian era was far greater than in Greece from the beginning of the period from which we have inscriptions. From Sicily and the coastal regions of southern Italy there are abundant records, both epigraphical and literary, of the speech of Greek colonists. From Apulia we have nearly two hundred inscriptions in the language of the Messapii, who are generally regarded as immigrants from the eastern side of the Adriatic, speak- ing a dialect of the Illyrian branch of the Indo-European family. Much farther to the northwest the Etruscans held control not only of the region between the Tyrrhenian Sea and the Tiber but, during the seventh and sixth centuries,
Tio^^d de xai aXka xoiQia xrJQ 'IraMag coxovv 'O/h^qixoi, xal ^v rovro to sdvoq iv rolg Tidvv /j,eya re xai aQxalov.
® p. 30, 10-13 (Rossbach) Etruscorum duodecim populi, Umbri in id tempus intacti, antiquis- simus Italiae populus, Samnitium reliqui in excidium Romani nominis repente coniurant.
' N.H., III, 112-3 Umbrorum gens antiquissima Italiae existimatur, ut quos Ombrios a Graecis putent dictos, quod inundatione terrarum imbribus superfuissent. Trecenta eorum oppida Tusci debellare reperiuntur.
* Randall-Maclver, Italy before the Romans (Oxford. 1928), pp. 44-5; VV^hamough, The Foundations of Roman Italy (London, 1937), p. iS4.
" Whatmough, op. cit., p. 194.
4 I. INTRODUCTION
of portions of the Po valley and Campania as well. Around ten thousand Etrus- can inscriptions have been discovered, but all except a very few are brief epi- taphs. The meager extent of our knowledge of the structure and affinities of the language even todays" is the more regrettable because of the undoubted influence of the Etruscans on the culture and even the language of their ItaUc neighbors, especially the closely adjacent Umbrians. None of the other non- Itahc dialects of northern Italy is of comparable importance with Etruscan, yet we have some records of Gaulish in the Po valley, of Ligurian and Lepontic in the northwest, and of Raetic in the Raetian Alps. Venetic was formerly grouped as Illyrian along with Messapic, but has more recently been shown to belong in the Italic group.^^ In this group must be placed all the languages of ancient Italy not hitherto mentioned.^- These languages were spoken especially but not exclusively in the interior throughout the peninsula before its romanization, from the Venetic area in the north to Sicily in the south, and include not only Latin, the most important member, and its closely related neighbor Fahscan, but also Umbrian to the north, Oscan to the southeast, and a series of lesser dialects in the central region around Latium : Paelignian, Marrucinian, Vestinian, Sabine, Aequian, Marsian, and Volscian. Although these last seven are known only from the scantiest records, there is not the slightest doubt as to their clas- sification as Italic. Volscian in particular shows a strikingly close relationship to Umbrian if we may judge from the language of the four-hne inscription of Velitrae which constitutes the only certain example of the dialect.^^ Oscan is far better known than the members of this intermediate or " Sabellian " group. Over two hundred Oscan inscriptions are preserved, from Samnium, Campania, Lucania, Bruttium, Messana, and Sicily, including several of fairly considerable length. The close relationship between Umbrian and Oscan is universally re- cognized; the two have in common a series of phonological and morphological features which must have developed during a period of unity but which are
^° The evidence in favor of the view that the Etruscans came to Italy by sea from Asia Minor is now so strong that it is practically impossible to entertain any other theory as to their origin. Their language was almost certainly not Indo-European, at least not in the traditional sense. If we follow the opinion of some recent scholars, as, for example, Sturtevant (Indo-Hitlite Lanjngeals, p, 23), in recognizing an "Anatolian" group, which would include Hittite, Hiero- glyphic Hittite, Luwian, and Lycian and would constitute a sort of sister-group to Indo-Eu- ropean, Etruscan may some day find a place here.
" M. S. Beeler, The Venetic Language, Berkeley and Los Angeles, 1949. His conclusions were accepted by Whatmough, Lang., XXVI (1950), pp. 302-4.
" Except the language or languages of the "Old Sabellian" or "East Italic" inscriptions discovered at several places on the eastern side of the peninsula. Certain scholars have regarded some of them as Umbrian in language, but their meaning and the classification of their language are not known to any appreciable extent.
" Co. 252. Marsian, Aequian, and Sabine may represent intermediate stages reflecting a migration from Umbria to the Volscian territory southeast of Rome, perhaps in the sixth century. See A.J.P., LXXII (1951), pp. 113-27.
I. INTRODUCTION 5
absent from Latin. Of these the most striking is the conversion of the labiovelar stops inherited from proto-Indo-European into plain labial stops: e.g. 0. pid, U. pif-e, but L. quid. Another important point of agreement against Latin is in the treatment of the voiced aspirates: in Latin these stops become voiceless spirants only in the initial position, while in the medial position they merely lose their aspiration and become plain voiced stops, but in Oscan and Umbrian they become voiceless spirants medially as well as initially: e.g. 0. tfei, U. tefe, but L. tibi, with the stop shown in its original form in Skt. tubhyam. When a labial or guttural stop stands immediately before a dental, Latin maintains the stop, but in Oscan and Umbrian it is converted into a spirant: e.g. L. scriptus but 0. serif tas, U. screhto (-ht-<-ft-<-pt-). The group -nd- is assimilated to -nn-: e.g. 0. upsannam, U. pihaner, but L. operandus. The groups ns and rs give rise to / and rf under certain conditions; for the details, which are comphcated, see 58 b, c, d, f, 59 c, e; Latin shows no parallel to these sound-changes. Among the vowels a is raised in certain situations in the direction of o: e.g. 0. molto, U. mutu, but L. multa (with shortening of final a instead of qualitative change). Vowel syncope is much more frequent even than in Latin: e.g. 0. hurz, U. Iku- vins (minor inscription no. 369 Co.), but OL. hortiis, Sabinus. Among declen- sional forms the genitive singular of o-stems has been remodeled after the i-stems: e.g. 0. sakarakleis, U. katles, but L. catiili; the accusative singular of conso- nant-stems has been remodeled after the o-stems: e.g. 0. tanginom, U. curnaco {<~om), but L. cornicem. Features of the Oscan-Umbrian verb system not found in Latin include the third plural ending in -ns, the /-perfect (e.h. aikdafed, U. pihafi), and the infinitive in -om. The foregoing features are all innovations shared in common by Oscan and Umbrian in contrast to Latin, where the original conditions have been preserved or the innovation has proceeded in a different direction. The reverse instances, in which Oscan and Umbrian have escaped the changes which appear in Latin, have less weight as posi- tive evidence for close Oscan-Umbrian unity, yet they are important in any consideration of the Italic group as a whole. The vowel-weakening which is such a characteristic feature of Latin is largely absent from the other two dialects, s before nasals is maintained in such forms as 0. fiisnam, U. fes- naf-e in contrast to its loss in L. fdniim, etc. In noun-inflection the a-stems have kept the s of the genitive singular (e.g. 0. vereias, U. tutas, Gk. ydiQa;;, Goth, gibos), which in Latin has been almost entirely displaced, except in expressions like pater familias, by the ending i (al> ae) of the o-stems. In the nominative plural of both d- and o-stems both Oscan and Umbrian have retained the -s originally characteristic of nouns, as in Sanskrit and Gothic, in contrast to the originally pronominal ending of L. feminae, viri, etc. In the ge- nitive singular of the third declension, just as Oscan-Umbrian has extended the ending of i-stems into consonant-stems (e.g. 0. medikeis like aeteis), so Latin has extended the ending of consonant-stems into i-stems (e.g. hostis like regis).
6 I. INTRODUCTION
Moreover the original short -vowel ending of the nominative plural of consonant- stems, which is preserved in such forms as O.humuns {<-nes),\J.f rater (-r<-rs <-res) was replaced in Latin by -es <-eyes from the i-stems. In the verb Oscan and Umbrian preserve the distinction between primary and secondary endings in the third singular and plural, which has been effaced in Latin, and there are other differences, but some of these may be due to the accidental lack of examples in the dialects. In vocabulary also Oscan and Umbrian frequently agree with one another against Latin; for example, 0. touto, U. tola take the place of civitas, U. mefs (contained also as first member in 0. meddiss) takes the place of L. ius; in place of ignis U. has pir (: Gk. nvQ), a derivative of which appears in 0. purasiai; 0. veru and U. uerir (abl. pi.) take the place of porta as words for 'gate.' Latin words for 'building, home, dwell' are largely replaced by deriva- tives of the root treh- seen in 0. triibum, U. tremnu, trebeit, etc. (but found also in L. trabs). The verb represented by L. volo, velle, apart from specialized uses, has disappeared in favor of 0. herest 'volet/ U. heri 'viilt,' etc.; and the list could be extended. Naturally Oscan and Umbrian show certain divergences between themselves. The anaptyxis which appears in Oscan forms such as ara- getud = L. argento is largely absent from Umbrian, as from Latin. But where Oscan differs from Umbrian it is almost always the former which shows the earlier stage of development. The change of medial s to r, which Umbrian shares with Latin, is not found in Oscan, which advanced in this direction only to the extent of voicing the s: e.g. U. pracatarum, L. dearum, but 0. egmazum; and the Umbrian of the later Iguvine Tables even has rhotacism in the final position: e.g. nom. pi. 0. aidilis, L. aediles, but U. foner. Conversion of A: to a sibilant before front vowels is clearly attested in Umbrian by the use of a special character: e.g. 0. kersnu, L. cena, cenati, but U. sesna, gersnatur (paralleled in the Ro- mance languages in such forms as Ital. cento, Fr. cent). The change of intervo- cahc d into f rs is an especially peculiar Umbrian feature; so also the change of medial / to the same f rs under certain conditions, and the change of initial / to V « in uapef-e (L. lapides), etc. The loss of / before /, which may be observed in the contrast of U. muta with 0. molto, L. multa, is also peculiar. Where Oscan has preserved the diphthongs intact, Umbrian has gone farther even than Latin in converting them to monophthongs: e.g. 0. avt aut, L. aut, but U. ute ote.
7. The Sabellian dialects in general show a decided resemblance to Oscan and Umbrian; the conversion of labio velars to labials is only one among many evi- dences of such agreement. Paelignian, which is the best known of these dialects, shows an especially close similarity to Oscan, and the same appears to be true of Marrucinian and Vestinian, as far as it is possible to determine from their extremely meager remains, except that they occasionally show the agreement with Umbrian against Oscan which their geographical position would lead us to expect. The close resemblance between Volscian and Umbrian has already been mentioned above in § 6.
I. INTRODUCTION 7
8. When we turn from this closely interrelated group of Umbrian, Sabellian, and Oscan dialects to Latin and the question of its relation to them, we are confronted with a problem that has become increasingly controversial in recent decades. Earlier works, such as Brugmann's Grundriss and the Oscan-Umbrian grammars of von Planta and Buck, unhesitatingly affirmed the notion of Italic as one of the principal branches of the Indo-European family, with Latin, Oscan, and Umbrian as closely related members within it, and in fact it is doubtful whether the interpretation of the texts could have been successfully accomplished by any other means than the assumption of a close linguistic relationship to Latin, accompanied by a careful comparison of Latin literature and antiquities. The special features common to Latin and the Oscan-Umbrian group of dialects include: representation of the Indo-European initial voiced aspirates by voiceless spirants; representation of the vocalic liquids f, I by or, ol and of the vocalic nasals n, m by en, em; merging of the diphthongs eu and ou as ou; extensive syn- cope of unstressed short vowels; frequent use of the extended suffix -tion- in place of -ti- in abstract nouns (with gradation -/ion//fn- in Oscan and Umbrian); extension of the ablative singular ending -d from o-stems into most other stem- classes; an imperfect indicative based on the suffix -hha- or more probably -bhwd-, which, however, in the dialects is attested solely by 0. fufans. Moreover in the verbal system of the two branches, despite many important differences and despite the fact that some common features are found even outside of Italic and hence are less significant than innovations peculiar to Latin and Oscan- Umbrian, nevertheless there is a fundamental similarity, which is especially apparent when we attempt to' arrange the verbs in conjugational classes. The system of four conjugations which is traditional for Latin and which has been successfully extended to Oscan and Umbrian, could not possibly be adapted to the verbal system of Greek or of the Indo-Iranian or Germanic languages. Several recent scholars, however, have emphasized the differences rather than the re- semblances between Latin and Oscan-Umbrian, and in fact the differences are unquestionably greater than between, for example, any two Greek dialects." Moreover, according to those scholars who favor a wide separation, some of the features which the two branches share in common are the result of cultural borrowing after the Italic-speaking nations were settled in the peninsula. Latin words of dialectal origin, especially those possessing unexpected labials in place of the normal product of labiovelars (e.g. popina : genuine Latin coquina, bos : Skt. gaus), or medial / from bh, dh (e.g. Alfius : genuine L. Albius, U. alfu, Gk. aXcpog; rufus : genuine L. ruber, U. rofu, Skt. rudhiras) are well-known.
" Kretschmer, Einleitung in die Altertumswissenschaft,^ I (Leipzig and Berlin, 1912), p. 554, found a difference of 10 to 15 words per hundred between the Greek of the Gortynian Law Code and Attic, but a difference of 60 to 70 words per hundred between the Iguvine Tables and Latin. See also Beeler, Lang., XXVIII (1952), pp. 435-43, emphasizing the divergence between Oscan-Umbrian and Latin.
8 I. INTBODUCTION
Among borrowings in the opposite direction, from Latin into the dialects, it is possible to mention 0. aidil (<L. aedilis; if this were native we should expect -/-), 0. kenzsur (beside native keenzstur), 0. kvaisstur, U. kvestur, unless the initial consonant of these and of L. quaestor was kiv (which is unlikely, since these words are probably somehow related to the interrogative pronoun stem with labiovelar k^) and unless kw became kv instead of p in the dialects; certain oth- er words, such as L. vinum, U. vinu, give evidence of borrowing from dialect to dialect, and of ultimate Greek, Etruscan, or other origin. But in general the evi- dence for extensive lexical borrowing among the Italic dialects is not very strong, although a careful phonological study of the vocabulary of Latin, Oscan, and Umbrian in close connection with the chronology of the sound- changes might lead to interesting results.
9. Naturally the extent of kinship between Latin and Oscan-Umbrian is a matter of degree and not of absolute truth against absolute falsity, unless it is asked whether or not we have the right to assume "proto-Italic" as a common ancestor of the Italic languages, spoken after the breaking up of proto-Indo- European, but uniform within itself. In the light of present-day knowledge of dialect geography it is unlikely that we have the right to assume any intermediate stages of this sort except in the sense of groups of closely related dialects, mu- tually intelligible but not absolutely uniform unless their area was an exceedingly small one. In this sense it may be quite proper, and for practical convenience is sometimes desirable, to recognize "proto-Italic" as we recognize "proto-Ger- manic" and similar ancestor-stages for other groups. But it is quite improbable that "proto-Italic" was ever spoken within the Italian peninsula ; the two main divisions had probably become at least partially distinct somewhere beyond the Alps. It is a familiar assumption that among the prehistoric Indo-European dialects from which the separate languages of later times were descended some occupied a position on the periphery of the original home, and that the speakers of these dialects migrated especially early to their new homes, and further that these dialects, even after wide geographical separation, preserved certain archaisms of morphology and vocabulary which disappeared, sometimes being replaced by new forms, in the more centrally located dialects and the languages descended from them.^^ Now at those points where the Latin and Oscan-Umbrian vocabulary diverge from one another it not infrequently happens that the dialects show a striking agreement with Greek or Germanic, two branches generally regarded as being descended from centrally located Indo-European dialects. As examples it may be sufficient to cite U. pir, 0. *pur- in purasiai, Gk. nvo, OHG fuir , Eng. fire (but L. ignis with Skt. agnis); U. trebeii, iremnii, Goth, paurp, Germ. Dorf. In the treatment of the labiovelars also the dialects agree with Greek
*^ For this theory, or variations of it, see H. Hirt, Die Indogermanen (2 vols., Strassburg, 1905-7), I, pp. 158, 162; Meillet& Vendryes, Traile de grammaire comparee des langues classiques (Paris, 1924), p. 17; Devoto, Gli antichi italici^ (Florence, 1951), p. 49.
I. INTRODUCTION 9
(especially with Aeolic, where the labialization does not depend on the quality of the following vowel) and with the Cymric division of Celtic, which, to judge from its geographical position, appears to be a descendant of a more centrally situated dialect than that which developed into Irish. The notion that the pre- Latin speakers were early migrants into Italy, followed by a much later wave of other migrants, accords well with the distribution of the Italic peoples in the latter half of the first millenium B.C., when " Latinians " occupied portions of the west coast, including not only Latium but regions farther to the southeast before they were submerged by the Oscan-speaking Samnites, and left possible remnants in Sicily. It is now possible to arrange the Itahc languages and dia- lects according to the following stemma :
proto- Italic
Latin-Faliscan Oscan-Umbrian
Sicel
Venetic j j
Oscan-SabeUian Umbrlc
Latin Faliscan
Volscian Marsian ? Umbrian
Oscan Paelignian Marruclnian Vestinian
10. Neither the passages of the Tables in which the ritual is described nor the liturgical formulas interspersed among these directions can be considered as pohshed literary compositions comparable with passages of similar content imbed- ded in Greek epic or tragic poetry or in Latin literature of the Golden Age. But if the prayers of the Atiedian Brothers are deficient in elegance and smoothness, they have nonetheless a certain rugged and archaic dignity. Their style can best be compared with that of the prose of Cato the Elder, the fragments of the Twelve Tables, and the ancient formulas of a ritualistic character scattered through the works of Roman prose writers of a later age.^^ The sentence-structure in the Tables is generally simple; imperatives, sometimes in series of some length, are used in the basic directions for the ritual, interspersed with temporal clauses of resumptive character, containing future perfect verb forms and serving to connect the successive steps of the ritual in their proper sequence. There is rarely a second level of subordination and scarcely even an approach to what is known as a 'periodic sentence. ' Asyndeton is the rule when substantives follow one another
" For such texts in general see G. Appel, De precationibus Romanorum (Religionsgeschicht- liche Versuche und Vorarbeiten, VII), Giessen, 1909.
10 I. INTRODUCTION
in series, and imperative forms are often similarly connected: e.g. totar louinar nome nerf arsmo ueiio pequo castruo fri pihatii, or the imperatives persnihmu vestikatu ahtrepufatu afpeltu statitatu. But if there is an appearance of striving for brevity in the sparing use of conjunctions, the opposite is true of the accumulation of near-synonyms which is such a marked feature of legal or litur- gical formulas, where no omission may be allowed which might nullify the effect of the prayer or injunction. So, for example, in Via 27-8 in which Jupiter Grabo- vius is entreated to accept the ox as a propitiatory offering in case of any defi- ciency or flaw in the ritual: Dei Craboiiie persci tuer perscler uaseto est pesetom est peretomest / frosetomest daetomest tuer perscler uirseto auirseto uas est...; or the frequent formula futii fons pacer pase tua ocre Fisi tote liouine erer nomne erar nomne.
11. In making a comparison between the Iguvine Tables and early Latin legal and liturgical prose it is necessary to take account of one particular artistic device which is especially affected by both; that is, the use of alliteration. The most striking example of its use in the Tables occurs in the curse against the enemies of Iguvium in VIb 60 = Vila 49 (with only orthographic differences) tursitu tremitu hondu holtu ninctu nepitii sonitn sauitu preplotatu preuilatu. Other examples include stahmei stahmeitei, praco pracatarum, fato fito perne postne sepse sarsite. It is not necessary to cite Latin parallels, which range from the tour de force 0 Tite tiite Tati tibi tanta tyranne tulisti of Ennius (fr. 109 Vahlen) to the restrained use of such phrases as magno cum murmure montis by Vergil. The theory that alliteration was especially favored in early Germanic, Celtic, and Italic verse because of the initial position of the accent in the early stages of these languages is plausible enough," but the reduced use of this device in Latin literature of the best period is certainly not the result of the new system of accent, but rather of a more refined and sophisticated standard of taste, combined possibly with the influence of Greek verse-structure, where alliteration never had a prominent role. A rather closely related m_atter is the possible use of rhythm in the liturgical formulas of the Tables, but here we are on very un- certain ground because of our ignorance not only of quantitative and accentual conditions in the Italic dialects but even of the prosodic principles of Latin Satur- nian verse, which is the verse-form that most naturally suggests itself as a basis for comparison in studying the question of rhythm in the dialect texts.^^
12. Our knowledge of the facts concerning the discovery of the Iguvine Tables^'
" See E. H. Sturtevant, The Pronunciation of Greek and Latin (Philadelphia, 1940), p. 180.
^* For a rhythmical analysis of the prayer in Cato, R.R., 141 together with certain passages of the Iguvine tables, see Norden, Antike Kunstprosa, I, pp. 156-63, For similar treatment of several Umbrian and other dialect texts, see L. H. Gray, A.J.P., LXIII (1942), pp. 433-43.
1^ For a much more comprehensive account of the study of the tables from their discovery in 1444 down to the present time, see Breal, Les tables eugubines, pp. 1-XVIII, and Devoto, Tabulae iguvinae, pp. 5-27. The present is a condensation of these accounts.
I. INTRODUCTION 11
is derived from certain addenda to Statuta Civitatis Eiiguhii, the work of a seven- teenth-century jurist Antonio Concioli, a native of Gubbio. The passage in question reads: ... confirmat (sc. Eugubium civitatem fiiisse antiquissimam, ac reges ibi resedisse) veierrimum theatrum: confirmant novem aheneae tabulae fama percelebres, ac litteris ac sententiis nemini cognitis exaratae quae totius orbis, nedum Italiae, antiquissima creduntur monumenta. Anno 1444 has tabellas ex aere purissimo fortuna detexit in subterranea concameratione miris emblematis tesssllata, quae quidem tarn egregio picturatur artificio, ut regalis aulae specimen praebeat, apud theatrum, hoc est in planitie ubi antiquitus sedebat Eugubium. Atque illae notissimum sibi nomen compararunt apud eos, qui vetustate cognita delectantur, multique crediderunt leges ibi caelatas esse primorum regum, qui in hac provincia dominarentur. Centesimus trigesimus tertius agitur annus ex quo illarum duae Venetiis in armamentaria ducalis palatii inter rarissima custodiuntur, tanquam pretiosa caligantis, sed pulchrae antiquitatis monumenta, ac Tabulae Eugubinae vocantur. Delatae fuerunt in earn urbem a clarissimo viro ut nobilibus eorum temporum antiquariis traderentur inter pretandae: et licet brevi remittendas fuisse promiserit, nee brevi, nee ullo unquam tempore redierunt in societatem aliarum septem, quae in secreto palatii communis archivio asservantur. The remains of the Roman theatre in which the tables were discovered are located just outside the town of Gubbio, on the western side. The notion that there were originally nine tables has been questioned because of a notice of the sale as early as 1456 of seven tables to the town of Gubbio i^"' ... vendidit... magnificis dominis Gonfalonerio et consulibus dicte civitatis et michi Guererio Canccllario infrascripto recipientibus pro dicto Comuni tabulas septem eburneas (sic) variis Uteris scriptas latinis vide- licet et segretis. But there is further evidence to support the belief that the number of tables orignally discovered was nine. Leandro Alberti (1479-1552) in his Des- criziom d'ltalia^^ writes of the tables: Praeterea loci vetustatem tabulae quaedam ex aere, partim Hetruscis, partim Latinis characteribus inscriptae probant, quae non pridem inter antiquas urbis ruinas inventae, nunc in Curia magna cum honore servantur. Mihi cum aliquando Eugubium venissem, Priores urbis summa cum reli- gione septenas ostenderunt, binas adhuc deesse addentes. Breal sought to solve the problem of the conflicting traditions by supposing that the two tables which eventually disappeared in Venice after 1540 had been in other hands from the time of their discovery or else had been sold to other persons at the time when the seven were sold to the city. In any case the view that there were originally nine tables is generally accepted today. Their possible content is discussed brief- ly below (§ 16).
'" See Conestabile, Giornale di Erudizione Artistica (Perugia, 1872), vol. I, pp. 177-81; Devoto, p. 6.
2^ First edition publislied 1550. Breal, p. 311, cites from the Latin translation F. Leandri Alberti Bononiensis Descriptio totius Italiae. Interprete Guilielmo Kyriandro Hoeningeno (Co- logne, 1567).
12 I. INTRODUCTION
13. The religious character of the tables was recognized as early as the sixteenth century. Justus Lipsius, who in 1588 pubhshed the Inscripiiomim aniiqiiarum quae passim per Europam liber of Martinus Smetius after the latter's death, included Tables IV and VI, with the remark tahiilas hasce ambas Joannes Metellus Burgundus vidii et exscripsit, quas eisi nemo plane inielliqit, quia lamen de rebus sacris agere quidam crediderunt, ideo hoc loco ponendas existimavi. The clue to this religious character could conceivably have been found in such expressions as buf trif fetu marte or in other passages where Latin-appearing divine names and animal names are used in close association, but it was impossible to make any satisfactory sense from the texts until the development of comparative linguistic method in the nineteenth century. It was not even clearly understood what the language of the inscriptions was, and attempts were made to decipher the texts with the help of Dutch (so van Scrieck in 1614, taking the language to be that of the ancient 'Belgians') or of Old Enghsh, Old High German, and Celtic languages (soBardetti, 1688-1767), or of Greek and Hebrew (so Passeri, 1691- 1780). That the language was often taken to be Etruscan is not surprising in view of the type of writing used on five of the seven tables, the proximity of Umbria to Etruria, and the exaggerated enthusiasm for Etruscan antiquities which affected not only some Italians of Tuscan origin but certain non-Italians as well. This view was held by Bernardino Baldo (1553-1617), who in 1613 published a work entitled Divinaiio in tabulam aeneam Eugubinam lingua Hetrusca veteri perscrip- tam, and by the Scottish scholar Thomas Dempster (1559-1625), whose work De Etruria Regali was written in 1619 but first pubhshed in Florence in 1723 with the help of Thomas Coke, Earl of Leicester. Dempster regarded the Umbrian language as one of four dialects of Etruscan, of which the others were Oscan, Raetic, and Faliscan. Filippo Buonarroti, who was entrusted with the task of editing Dempster's work, performed a valuable service by adding a set of re- markably accurate facsimiles of the tables. He also denied the theory that their language was Etruscan, basing his argument on the total absence of the termina- tion -al which is so common in Etruscan inscriptions. Although the Etruscan theory did not fully die out until the early nineteenth century, some advances were made during the seventeenth and eighteenth. Scipione Maffei (1675-1755) correctly interpreted some of the words in the tables, called attention to the resemblance between their content, which he recognized as being liturgical, and that of the Roman Acta Fratrum Arvalium, and defended their genuineness against those of his day who regarded them as forgeries. Hannibal degli Abbati Olivieri (1708-89) recognized the words atiersiur, etc., as the name of a priestly college, which he designated Ateiatiorum, and saw that the word ikuvina in the tables, like ikuvins on a coin from Gubbio (no. 369 Co.) was derived from the name of the town Iguvium, even though he made the mistake of translating tola ioiiina 'tota inventus.' Passeri's work was important because of his under- standing and exposition of the combinational method of decipherment, his re-
I. INTRODUCTION 13
cognition of the language as the ancient language of Gubbio, and his arrangement of the tables, even though this arrangement was later to be superseded by that of Lepsius, which is universally adopted as the standard today.^^ In spite of his dependence on Greek and Hebrew for purposes of etymological comparison, to which reference was made above, he first made correct interpretations of several words the credit for which has generally been given to much later scholars. Aloisius Lanzi (17v32-1818) affirmed the ritualistic character of the texts, which, even though it had been recognized much earlier, had by no means been admitted by all scholars during the two centuries which preceded the publication of Lanzi's Saggio di lingua etrusca e di alire aniiche d'lialia (Rome, 1789), He further succeeded in reaching a sound interpretation of a number of individual words: feiiu 'facito', desenduf 'duodecim,' poi 'qui,' ape ' postquam,' kumiaf 'foetas,' nertru 'sinistro/ fons 'volens,' pir 'nvg.' A generation later Carl Otfried Miiller (1797-1840) made it fully clear that the language of the Iguvine Tables was not Etruscan and that the two alphabets used in the tables do not represent two distinct dialects. ^^ Miiller's pupil Richard Lepsius (1810-84) partly advanced the interpretation of the tables, especially by perceiving the true value of several of the characters in the Etruscan alphabet, but partly went astray by affirming that the use of the Etruscan and Latin alphabets in the inscriptions of the same people must always indicate a difference of language, and by assigning to the tables a greater antiquity than any scholar of the present day would admit. ^ His most important contribution to the study of the Iguvine Tables, however, lies in the chronological numbering which he assigned to them, for even though this numbering may not represent the true relative age of the tables themselves, still less of the texts which they contain, nevertheless it has become the accepted standard for all subsequent editors.^^ The earliest critical edition of the Iguvine Tables still cited by contemporary scholars in their treatment of problems in the Iguvine Tables is the Umbrische Sprachdenkmdler of Th. Aufrecht and A. Kirch- hoff (3 vols., Berlin, 1849-51). Although many of their emendations and inter-
*^ Passeri counted twelve tables by taking separately the two sides of the five which are in- scribed on both sides. His first six tables are nos. I, VI, and VII according to the system of Lepsius; his VII-VIII = our II; IX-X = our V; XI-XII = our III-IV; identity of content was the ground for placing I with VI and VII despite the difference of alphabet.
" Die Etmsker, revised by W. Deecke, 2 vols., Stuttgart, 1877. See especially I, pp. 38-49.
** 6th century b.c. for those tables in the Etruscan alphabet, 4th century b.c. for those in the Latin. Earlier views of their age had varied considerably. Concioli's account of the discovery of the tables, cited above, may be remembered for his bold statement that the tables were the oldest records in the world, not to mention Italy. Bourguet (1678-1742) regarded them as ear- lier than the reign of Romulus and as the earliest extant monuments with the exception of some from Egypt. Passeri with more restraint had ranged himself between those who believed the tables to be earher than the Trojan War and those who believed them to be no earlier than the time of Cicero. For modern views, with the criteria used for dating the tables, see below (§§17-19).
" Huschke assigned the designation Ila to the table normally numbered lib and vice versa.
14 I. INTRODUCTION
pretations can no longer be accepted today, the work, with its full Latin translation, detailed commentary, grammar, lexicon, and facsimiles of all the tables, still has value. Of the various treatments of the tables during the century which has elapsed since Aufrecht and Kirchhoff some idea can be gained from the bibho- graphy. These may be divided roughly into two classes: works on the Itahc dialects, such as those of Conway, von Planta, Buck, Vetter, Pisani, and Botti- glioni, most of them containing a more or less detailed grammatical description of the dialects; and editions of the Iguvine Tables alone, such as those of Huschke, Breal, Biicheler, von Blumenthal, and Devoto (with the minor Umbrian inscrip- tions included in Huschke and Biicheler, as well as the Tables). The value of Huschke's work is seriously impaired by the many sound-changes which are erroneously assumed, and by the invalid etymologies based on these sound- changes. Breal's work has much greater value. Biicheler's edition is excellent, and on it to a large extent are based the interpretations of Conway and of Buck. Von Blumenthal's commentary does not cover the whole of the Tables but con- tains a penetrating analysis of certain selected passages. Along with some inter- pretations w^hich must be discarded there are others which show great ingenuity and are preferable to those of any of his predecessors. Devoto's edition is the most important that has appeared during the twentieth century; it deserves special praise for the acute and independent judgment applied to the innumerable problems which arise in the interpretation of the Tables, for its orderly exposition of the parts of the ritual and their relation to the ceremonies as a whole, for the liberal citation of Latin, Greek, Avestan, and Sanskrit passages relating to parallel cult -practices, and for its abundant bibliographical references. Devoto's smaller work, Le tavole di Gubbio (Florence, 1948), contains the first translation of the Iguvine Tables to appear in any modern language. A number of important shorter studies, especially by Italian scholars, have appeared during the past two decades, of which it is impossible to give an adequate account here. There are still many unsolved problems in the interpretation of the Iguvine Tables, and for many of these it would be too much to hope that with the knowledge we now have we shall ever find an entirely satisfactory solution, although if some fortunate chance should ever lead to the recovery of the two lost tables or of some new Umbrian inscription of considerable length, our knowledge of the language and meaning of the extant tables would be substantially increased. Actually, however, since the middle of the nineteenth century so much of the tables has been clearly un- derstood that it is quite possible to give a detailed and accurate outline of their content. In the following outline the line-numbers of Via, b, Vila are shown in a parallel column after those of la, b, since the content is nearly the same except that the version in VI and VII is much more detailed than that in I. In the re- mainder of the outline Ila, lib, III, IV, Va, and Vb are given in order.
I. INTRODUCTION
15
|
la 1-Ib 9 |
Via 1-VIb 47 |
Purification of the Fisian Mount. |
|
la 1-2 |
Via 1-18 |
Taking of the auspices. |
|
18-21 |
Instructions on the ceremonial wand and on the Ughting of the fire. |
|
|
2-6 |
22-57 |
Sacrifice of three oxen to Jupiter Grabovius before the Trebulan Gate. Text of the in- vocations is given in Via but not in la. |
|
7-10 |
58-59 |
Sacrifice of three pregnant sows to Trebus Jovius behind the Trebulan Gate. |
|
11-13 |
VIb 1-2 |
Sacrifice of three oxen to Mars Grabovius before the Tesenacan Gate. |
|
14-16 |
3-5 |
Sacrifice of three sucking pigs to Fisus Sancius behind the Tesenacan Gate. |
|
16-19 |
5-18 |
Supplementary offerings to Fisovius Sancius. Text of the invocations is given in VIb but not in la. |
|
20-23 |
19-21 |
Sacrifice of three oxen with white foreheads to Vofionus Grabovius before the Veian Gate. |
|
24-27 |
22-23 |
Sacrifice of three ewe-lambs to Tefer Jovius behind the Veian Gate. |
|
27-34 |
23-42 |
Supplementary offerings to Tefer Jovius. Text of the invocations is given in VIb but not in la. |
|
lb 1-4 |
43-44 |
Sacrifice of three bull-calves to Mars Hodius at the Grove of Jupiter. |
|
4-7 |
45-46 |
Sacrifice of three bull-calves to Hondus Serfius at the Grove of Goredius. |
|
7-9 |
47 |
Instruction to repeat the auspicium and the ceremony of purification, if there has been any omission. |
16
I. INTRODUCTION
|
lb 10-45 |
VIb 48-VIIb 4 |
Lustration of the people of Iguvium. |
|
lb 10-11 |
VIb 48-49 |
Taking of the auspices. |
|
11-12 |
49-50 |
Fire-ceremony. |
|
13-14 |
51-52 |
Second auspicium. |
|
14-15 |
52 |
Procession by the Augural Way to Acedonia, the adfertor being accompanied by two primiatur. |
|
16-18 |
53-55 |
Banishment of aliens. |
|
19-22 |
56-65 |
Procession of the citizens of Iguvium with the sacrificial victims, making the circuit three times. Text of the imprecations a- gainst aUens and prayers for favor toward Iguvium are included in VIb but not in lb. |
|
22-23 |
65-VIIa 1 |
The two prinuatur depart. |
|
24-26 |
Vila 3-5 |
Sacrifice of three boars to Serfus Martius at Fontuli. |
|
27-30 |
6-40 |
Sacrifice of three female pigs to Prestota Serfia of Serfus Martius at Rubinia. Vila includes supplementary offerings to Pres- tota Serfia with black vessels, accompanied by imprecations against the enemies of Iguvium and prayers for favor toward Iguvium, and with white vessels, accom- panied by prayers for the protection of Iguvium. The text of the imprecations and prayers is given in full. lb has nothing corresponding to these supplementary of- ferings except for a brief reference to the |
|
' |
black and white vessels and to the moving of the bowls. |
|
|
31-39 |
41-45 |
Sacrifice of three heifer-calves to Tursa Serfia of Serfius Martius at Trans Sanctam, and commands to be passed between the three stations where sacrifices are being performed. |
|
46-51 |
Nearly equivalent to VIb 57-62: imprecations against aliens. |
|
|
40-44 |
51-VIIb 4 |
Ritual pursuit of the heifers through the town. |
|
45 |
Authentication by the quaestor; nearly iden- tical with I la 44. |
I. INTRODUCTION
17
|
Ila 1-14 |
Sacrifices to be performed in case of unfavorable aus- |
|
pices. |
|
|
Ila 1-4 |
Introduction and propitiatory formula. |
|
4-5 |
Sacrifice of a young pig to Vesticius Sancius, an |
|
ox to Jupiter, and a victim to Spector. |
|
|
6-7 |
Sacrifice of a ram to Jovius. |
|
8-10 |
Supplementary offerings to Dicamnus Jovius. |
|
10-11 |
Sacrifice of a sheep to Ahtus Jupiter. |
|
11-13 |
Sacrifice of a young boar to Ahtus Mars. |
|
13-14 |
Sacrifice to the Di Ancites. |
|
Ila 15-44 |
Sacrifice of a dog to Hondus Jovius. |
|
Ila 15-17 |
Proper time for the festival of the Hondia. |
|
17-20 |
Paraphernalia for the festival. |
|
20-32 |
Slaughter and porrectio of the victim at the altar. |
|
33-39 |
Ceremonies at the Obelisk. |
|
39-43 |
Return to the altar and conclusion of the sacrifice. |
|
44 |
Authentication by the quaestor; nearly identical |
|
with la 45. |
|
|
lib 1-21 |
Sacrifices at the decurial festival of Semo. |
|
lib 1-2 |
Selection of the victims, a pig and a goat. |
|
2-7 |
Naming of the decuviae for whom the sacrifices are |
|
to be performed. |
|
|
7-10 |
Consecration and slaughter of the pig. |
|
10-21 |
Consecration and sacrifice of the goat. |
|
lib 21-29 |
Sacrifice of a bull-calf to Jupiter Sancius. |
18
I. INTRODUCTION
|
III-IV |
Sacrifices in honor of Jupiter, Pomonus Poplicus and Vesona. |
|
III 1-3 3 4-11 12-20 20-25 26-30 30-31 31-IV 14 IV 14-16 17-22 23-32 32-33 |
Proper time for the sacrifices. Purification of the jar (or vat?) in the grove. Appointment of the auctor, inspection of the victims and procession to the Field. Construction of the htter for carrying the sacrifi- ficial victims. Slaughter of the young pig in honor of Jupiter. Slaughter of the sheep in honor of Pomonus Poplicus. Further instructions with regard to the young pig. Continuation of the sheep-sacrifice, with special of- ferings to Pomonus Poplicus and Vesona. Offering of a libation and cake to Purdovif. Offerings to Hula and Tursa. Further ceremonies in honor of Pomonus and Vesona. Injunction against use of the jar (or vat?) after its purification. |
|
Va 1- Vb 7 |
Two decrees regarding the duties and privileges of the adfertor. |
|
Va 1-3 3-10 10-13 14-15 15-22 22-Vb 7 |
Authority for the first decree. Responsibility of the adfertor to procure and inspect victims for sacrifice. Prescribed rate for the procurement of vegetable offerings. Authority for the second decree. Rewards granted to the adfertor at the three stages of the sacrifice. Vote on whether the adfertor has performed his duties satisfactorily, and fine to be imposed on him in case he has failed. |
|
Vb 8-18 |
Contracts between the Atiedian Brothers and two of the decuviae. |
|
Vb 8-13 13-18 |
Contract with the decuvia Clavernii. Contract with the decuvia Gasiias. |
I. INTRODUCTION 19
14. Among the deities known from the Tguvine Tables several are thoroughly familiar in Rome, This is notably the case with Jupiter and Mars, who are honored during the purification of the Fisian Mount with sacrifices before the Trebulan and Tesenacan Gates respectively, although for the epithet Grabovius which stands with the names of these two gods most recent editors assume an Illyrian origin and no longer admit any connection with Latin (Mars) Gradivus: see on Via 22, This epithet also stands beside the name of Vofionus, who is honored with three oxen at the Veian Gate. The three gods therefore make up a triad comparable to the Roman triad consisting of Jupiter, Mars, and Quirmus, and this grouping in throes appears in a number of other connections in the Iguvine Tables; there are, for example, three gates at which sacrifice is performed in the purification of the Fisian Mount, three sacrifices in the lustration of the people of Iguvium, three principal stages in the sacrifice (caesio, porrectio, superjectio), three victims in each of the sacrifices of the purification and lustration; the sentence of banishment against aliens is pronounced three times, and the lustral procession makes the circuit three times, Roman parallels are not difficult to find; for a treatment of the subject from an Italic and Indo- Iranian viewpoint see Benveniste, Revue de Vhistoire des religions, CXXIX (1945), pp, 5-16. The deities of the Iguvine cult are occasionally grouped in triads in which the second and third are subordinate to the first; so at least in the case of Serf us Martins, Prestota Serfia of Serf us Martins, and Tursa Serfia of Serfus Martins, to whom sacrifices are offered at three separate spots in the ceremonies of the lustration. The second and third deities, which are female, evidently represent two opposite aspects of Mars, Presto- ta standing for his protective power, Tursa for his hostile power v/hich terrifies enemies. Another apparent instance of deified abstraction is found in I la 10, 11, where sacrifices are offered to Ahtus Jupiter and Ahtus Mars; Ahtus represents the oracular power of the two gods whose names follow in apposition. In the name Vesune Puemunes Pupfices in IV 3-4, etc., Vesuna is probably to be regarded as the consort of Puemunus, both being deities of fertility ; but in general the precise interrelation between deities, when one name accompanies another in genitive or adjective form, is difficult to define, and the tables give little direct help. Cer- tainly it is not possible here to discuss the problem of the degree of anthropomor- phism present in Itahc religion before it had become strongly permeated with Greek influence.
15, The ritual of the Atiedian Brothers was a complex one, partly because the animal sacrifice, which was the most important feature, was generally accom- panied by supplementary offerings of various kinds, and because considerable attention was given to the kinds of vessels and other implements used in the cere- monies. Libations are offered with wine and more frequently with a substance known as puni = poni, which has been variously identified but was probably some other alcohohc beverage, possibly mead; sometimes poni alone is prescribed, sometimes a choice of wine or poni. Several types of cakes are called for, one or
20 I. INTRODUCTION
frequently two types being prescribed for a single sacrifice. A full list would in- clude the mefa, with which the mefa spefa may be identical, the ficla, the strusla, the tesedi, farsio or barley-cakes, the ar^lataf or crescent-shaped cakes, and the petenata or comb-cakes. We can have only a rough notion of the nature of these different cakes, chiefly from etymological comparison and from data regarding Roman offering-cakes in Cato, Festus, etc. The description of the sacrifices also includes instructions on the disposition of the various parts of the victim and on the modulation of the voice and the content of the prayers to be used. Among the passages relating to the purification and lustration the two versions in la, b, and Via, b, Vila in general show remarkably close agreement, the chief differ- ence being that only the later version gives the actual text of the prayers, while the earlier merely gives a hint of their content through the words ukriper Fisiu tutaper Ikuvina ' on behalf of the Fisian Mount, on behalf of the Iguvine State '. But there is no regular sequence in giving the instructions to use wine or mead, to offer cakes of a specified type, to pray silently, to use the same formulas as before the Trebulan Gate, etc. ; the sequence not only varies from one sacrifice to the next, but even between the earlier and later versions of the same sacrifice. The priests must have known in what order to proceed and merely needed to be told what type of cakes to offer, whether they must use mead or whether they had the choice of mead or wine, etc. Certain ritual acts not mentioned in con- nection with the purification and lustration are found in the description of the sacrifice in I la, b, III, IV. Especially striking among these is the anointing of the Obelisk in Ila 38.
16. A topic related to the cults of Iguvium and even more beset with difficul- ties is the organization of the Atiedian priesthood and of the Iguvine state. It is fully clear that the Tables are the records of a priestly organization whose name latinized would be Fratres Atiedii; it is they who passed the resolutions in Va and who contracted with the decuviae Clavernii and Casilas in Vb, and the sacrifices in III are offered on their own behalf —fratruspe(r) Atiierie(s). The number of members was probably twelve, since in VI lb 1-2 twelve unblemished (heifers) are to be procured for the ritual heifer-chase, unless indeed we take desenduf directly with fratrom Atiersio itself instead of with seuacne. Twelve was also the number of the Roman Salii according to Dionysius of Halicarnassus, Ant. Rom., II, 70, 1 ; Livy, 1, 20, 4, and of the Fratres Arvales according to Masu- rius Sabinus (Aul. Gell., VII, 7, 8). There is some difficulty in reconcihng the number twelve with the practice of voting in groups of five, as when in III 4-5 the Brothers rising in puntis 'pentads' appoint an auctor, but Devoto, p. 305, T.G., p. 11, finds a fairly satisfactory solution by recognizing two pentads plus the fratricus plus the adfertor, making twelve in all. The fratricus appears to be a sort of presiding officer within the brotherhood; it is he who in Va 23 calls for a vote as to whether the adfertor has performed his duties satisfactorily, although the quaestor may replace him in this function. The fratricus also has the responsi-
I. INTRODUCTION 21
bility of obtaining victims for the ritual heifer-chase, according to the instructions in Vllb, and like the adfertor in Va, he can be penalized if he fails in his duty. The language of Vila 1-2 — 'whoever at any time shall be fratriciis for the Atiedian Brothers... during his term of office as fratricus' — makes it seem that the office was subject to rather frequent rotation. No single official mentioned in the Iguvine Tables has such an important role as the adfertor. It is he who performs the lustration; so much is plain from Vllb 3 'when the Atiedian adfertor has performed the lustration of the people', compare also lb 40-1. The taking of the auspices in the opening portion of Via, which precedes the series of sacrifices performed during the purification of the Fisian Mount, is the responsibility of the adfertor and augur together; at least the person who is euphemistically designated ' he who goes to observe the divine messengers (the birds) ' may safely be regarded as the equivalent of the Roman augur. During the instructions for the sacrifices of the purification in Via 22-b 47 = la 2-b 9 the adfertor is never directly mentioned, but this avoidance of his title is evidently deliberate; see especially on VIb 50. In general the instructions throughout the Iguvine Tables have the form of future imperatives without expressed subject, and it is to be assumed that the priests knew who was to carry out each particular instruction ; usually it was the adfertor, although occasionally it must have been some other person, as, for example, in IV 29, where the subject to be understood for kanetu is probably a flute-player. Whether the actual slaying of the victim was performed by a victimarius, as ultimately in Rome, it is impossible to say. In Tables la, b, Ila, b, III the instructions are normally given in the second person, in Via, b, Vila in the third, if we may judge by the distribution of s and st in the endings of the future and future perfect forms which frequently appear in the subordinate clauses, for the imperative forms in -tu, like Latin amato, habeto, etc., are ambiguous (115 c). During the lustration and the heifer-chase the adfertor is accompanied by two men called prinuatur. From the meagre data in lb, VIb, Vila it is not clear whether they are similar to acolytes or are civil magistrates of some kind whose office entitles them to a part in the ceremonies, but from the fact that they wear a purple robe (if ponisiater VIb 51, punigate lb 15 means 'one dressed in purple ') and that they join with the adfertor in uttering the imprecations against the enemies of Iguvium in VIb 57 ff. and Vila 46 ff., it appears that their position was an important one. Of two other officials, the auctor and the quaestor, more will be said below (§18) because of the significance of these officials for the chron- ology of the Tables. For the organization of the citizens of Iguvium the most important passages are lib 1-7 and Vb 8-18. The sense of these two passages is discussed in the notes, but may be briefly summarized here. The population of Iguvium had been divided into tekvias, a term which may be latinized as decuviae and which clearly indicates a division into ten groups, as may be seen from the ten names Atiiefiate, Klaverniie, Kureiate, etc., which follow. But as the population increased, the decuviae had to be redivided; hence the use of the ex-
22 I. INTRODUCTION
pressions etre Atiiefiate 'second Atiedias, ' etc., and every decuvia has been divided in this way except for the Casilas, which was divided into three instead of two, and the Peraznanii, which was left undivided. There were thus twenty decuviae after the reorganization. Vb 8-18 contains contracts between the Atie- dian Brothers and two of these decuviae, the Clavernii and the Casilas, the former being understood to include its two and the latter its three subdivisions. Since we should expect to find similar contracts with the remaining eight decuviae, it has been reasonably conjectured that they formed a part of the content of the two lost tables (so Buck, p. 302; Blum., p. 36; Dev., p. 412). Vb 8-18 would then be not the beginning but the conclusion of the series of contracts, since nearly half of the reverse of the tablet is left blank. Some lower grouping of citizens than the division into decuviae must have existed, and of such a grouping we have some evidence in the terms abl. sg. natine and nom. pi. famerias. The sense of natine is not so much that of its etymological equivalent L. natione as of its more remote cognate gente, gens. Names of two of these gcntes appear in the text, the Petrunia Ila 21, 35, and the Vuciia (Lucia), lib 26. To judge from the use of famerias in lib 2 it is not likely that it stood for a division lower than natine, like L. familia in relation to gens. Probably the two terms correspond to two dif- ferent systems of division, but their exact relation is not clear; see on lib 2. One other passage bearing on the organization of the citizens is VIb 56 arsmahamo caterahamo louinur (= lb 19-20), in which the people are directed to arrange themselves in formation for the procession of the lustration. The two verbs have been taken to refer to larger and smaller military units or to ranks of cavalry and of infantry, but the view preferred here is that arsmahamo refers to priestly ranks and caterahamo to military ranks of the rest of the citizens.
17. It is impossible to determine the exact date when any of the tables was written, but it is possible to form some idea of their relative age, partly from script and language, partly from content. Since IV is a continuation of III, both must be of the same age, and similarly Via, VIb, and Vila, which form a connected body of text, must be of the same age. It is obvious that Tables I, II, III, and IV, which are in the native alphabet, were written earlier than VI and VII, which are in the Latin, and also that Vb 8-18, which is in the Latin alphabet and has no connection in content with Va and Vb 1-7, must have been written later in the blank space on the reverse side of V. Hand in hand with this difference of alphabet goes one important difference in language: final s has been changed by rhotacism to r in VI and VII, and even in V, most of which is in the native alphabet, but the change is not yet shown in I, II, III, and IV (see 57 e). Va 1-b 7 is therefore later than I, II, III, and IV, but earher than VI and VII. At the same time it appears from the shape of certain characters that 1 1 1- IV were written later than I and II; so especially from the e, which in III-IV and V has the three cross-bars parallel, as a Roman e would appear if reversed and shghtly aslant, while in I and II the e has no true cross-bar at the bottom, but merely a
I. INTRODUCTION 23
curve in the vertical bar from the point where the middle bar meets it. Within II the content of a 1-14, a 15-44, and b 1-end is different; these three portions are therefore not necessarily of the same age, and the crowding toward the end of I la gives the impression that the reverse had already been filled and hence could not be used.
18. The content of the tables occasionally provides some evidence of their age through references to Roman institutions. Of the two official titles of the uhtur and the kvestur the former is assumed to be the earlier because the word has not so strong an appearance of being a Latin loan.^^ Now Tables lb and Ila were authenticated by the same official during his quaestorship (kvestretie lb 45 = II a 44), while the decrees in Va, which on linguistic grounds must be regarded as later than I and II, are dated by the name of the official holding the office of auctor (uhtretie). If this view of the relative age of the two magistracies in Iguvium is vahd, this discrepancy suggests that the earher table may not neces- sarily contain the earlier text, and that our problem is a twofold one : to determine the order in which the texts were composed, and the order in which the tables were engraved. Breal, pp. 224-7, showed that VI and VII were not merely copied, with additions, from I, but that both depended on an earher text now lost. His arguments from the spelling of the texts are partly of doubtful vahdity, but the archaic accusative plural form abron. in VI lb 43, to which he makes reference, probably points to an original of early date, and the absence in VI-VII of anything corresponding to la 25 puste asiane fetu, b 20 lures et pure, 36, 38 antakre, 41 super kumne strengthens the probability that VI and VII were not directly copiedfroml. Ill and IV, even though the form of certain letters approximates that in Va, may contain the oldest text of all; so at least it appears from the ratio of a and u representing final a : among 3rd pi. imperative forms of the type of fertuta 5 have -a and none have -u, while among neut. ace. pi. forms 8 or 9 have -a, none -u. To assign a relative age to the original texts reflected in each of the tables would be impossible. So far as the tables themselves are concerned, the order proposed by Lepsius and now taken as standard is probably correct, except that lib appears to have been written earlier than Ila.
19. The actual age of the tables can be determined only within wide limits. The alphabet used in the earher tables resembles those used on Etruscan inscrip- tions of the 5th, 4th, and 3rd centuries, especially that of Pohmartium (Bomarzo). Since a certain lapse of time must be allowed for the introduction of the Etruscan alphabet into Iguvium and for its adaptation to the local dialect, and also for the composition of the lost originals on which our texts are based, we may be safe in
^* uhtur could conceivably be from L. auctor, with u (= d) < au and ht < ct, but in this case we should have to assume that the borrowing occurred some time previously, giving the sound-changes time to develop. There is nothing in the phonetic form of the word which could not easily be native Umbrian, while the kv- of kvestur cannot show a native development of IE k^ and it is not even certain that it can represent the rather rare cluster kiv (see 46 h).
24 I. INTRODUCTION
tentatively assigning a date not earlier than the beginning of the 3rd century. For the dating of the later tables some shght evidence may be found through changes in Roman currency. What is explained at greater length in the notes on Vb 9 and Vllb 4 may be briefly summarized here: in the former passage an ap- parently excessive sum has been fixed as the monetary equivalent of the dinner which is to be furnished by the decuviae to the two men who come to fetch the grain for the Atiedian Brothers, and in the latter au excessively high fine is im- posed on the fratricus in case he fails to procure heifers for the heifer-chase. In 89 B.C. the Roman as w^as reduced to 1-24 of its original value. Therefore if we admit for the tables in the Latin alphabet a date shortly after the Social War, we may find a reasonable explanation for the sums of money mentioned in the two passages. The very sparing use of doubled consonants, when examined in connection with Roman usage, also points to a date not far from the Gracchan period, or even later if we assume that the adoption of the habit must have been later in Umbria than in Rome. Even the three instances ponne, appei, issoc in Vllb do not include every case in that table where doubling is etymologically justified. The question of when the local dialect was replaced in daily use by Latin cannot be answered precisely, but for our purpose it is less important than may appear at first glance, for in a liturgical text such as the Iguvine Tables, the property of a priestly corporation, it is by no means surprising that the native language should continue to be used not only in the prayers but even in the ritual instructions. We may then adopt for the latest tables a date somewhere in the first half of the 1st century e.g., a time when the local dialect of Iguvium, even if it had been largely replaced by Latin, must surely not have been completely forgotten.
II. GRAMMAR A. Phonology
1. The Alphabets.^ The following table shows the characters of the native alphabet (used in Tables I, II, III, IV, Va, Vb 1-7; written from right to left) and of the Latin alphabet (used in Tables Vb 8-18, VI, VII; written from left to right). At the right of each column are the characters (excluding the capitals) normally used in modern printed texts.
|
Native alphabet |
Latin alphabet |
|
|
f\ a |
A |
a |
|
8 b |
B |
b |
|
^ k |
G |
c |
|
D |
d |
|
|
3 |
E |
e |
|
8 * |
F |
f |
|
G |
9 |
|
|
e h |
H |
h |
|
1 i |
I |
i |
|
>! 1 |
L |
I |
|
W(A) m |
M |
m |
|
M n |
N |
n |
|
O |
0 |
|
|
1 P |
P |
P |
|
Q |
Q |
|
|
a r |
R |
r |
|
q ^ |
RS |
rs |
|
2(M) s |
S |
s |
|
d 5 |
S |
s |
|
>1(0) t |
T |
t |
|
V u |
V |
u |
|
> |
X |
X |
|
4r z |
a. The native alphabet lacks characters exactly equivalent to D, G, 0, Q, and X. D and G are represented respectively by t and A , the symbols for the corres- ponding voiceless stops; 0 by V ; the rare Q, like C, by >| ; X by 3>| . The Umbrian character Cj is represented in the Latin alphabet by the digraph RS. The Latin V is used both in the vocalic function of the Umbrian V and in the consonantal function of the Umbrian J3 . For the ^ of the native alphabet $ rather than Z is used in the Latin alphabet.
b. A is regularly used in place of fH in Table V; M ("san") appears in place
26 II. Grammar
of the usual 3 in salu Ila 18 and seritu Ila 24; and Q appears in place of >', y in furfa© lb 1 and purtuviSu IV 20.
1 The native Oscan alphabet is essentially similar to the Umbrian, but possesses characters for g and d. On the other hand it has no character equivalent to Umbrian q (f) or d (s), since the Umbrian sounds indicated by these characters had, as far as we know, no counterpart in Oscan. I, the equivalent of Umbrian 1; (z), represents the sound ts, but z in Oscan inscrip- tions in the Latin alphabet represents voiced s (= Engl. z). Special mention should be made of the characters (- (-| h), printed as i, and sy , printed as vl, since both occur in many of the Oscan forms cited in the grammar and lexicon, i represents a mid-high front vowel originating from I or e, or from e before another vowel; it also represents the second component of an i- diphthong. u represents a mid-high back vowel originating from 0 and, under certain conditions, from a.
2. Orthographic Variations. In this section are treated the principal varia- tions in spelling occurring in identical or closely related forms. The listing of examples is not complete, but more details on the origin, distribution, and signi- ficance of the variations may be found in the sections on pronunciation and phonology and in the Lexicon.
a. The following variations exist between forms represented in the two alpha- bets and arise simply from the absence of certain characters from one or the other of these alphabets:
u and 0: ukar ocar, pune pone, persklum persclo, etc. k and c: ukar ocar, kuvertu couertu, etc. k and q: tekuries dequrier. k and g: Krapuvi Grabouei, kumiaf gomia. t and d: testre desire, titu ditu, hutra hondra, etc.
9 and 5 or s: gimu Hmo, Qerfe Serfe, purtingus purdinsust, skalgeta scalseta, etc.
z and s: tagez tases, piliaz pihos, puze puse.
f and rs: kafetu carsitu, tefa dirsa, peri persi, etc.
b. Although b is not, like *dand*g, lacking in the native alphabet, it is some- times replaced by p through analogy with the use of t and k for the voiced dental and guttural stops: hapinaf habina, kapru kabru, kumpifiatu combifiatu, Krapuvi Grabouei, Treplanes Treblaneir.
c. The character ^ which is used in the Latin alphabet as the equivalent of d (9) of the native alphabet is sometimes written S without the diagonal stroke: Serfie Serfie, Sansie, purdinsus purdinsust.
d. Variations between vowels:
a u 0 when the vowel is etymologically long: nom. sg. of the first declension: muta mutu, panta etantu; nom. -ace. pi. neut.: vatuva vatuvu, veskla vesklu; pf. pass. pcpl. nom. sg. masc: piliaz pihos, kunikaz conegos.
e i e i: from original e: tases tasis, vesti9a visti9a (on these forms see 8 e),
A. Phonology 27
krenkatrum krikatru ; from original e: feliuf filiu, kafetu carsitu; from original i: vea via uia, steplatu stiplatu, tefa dirsa, etc.; from original i: pehatu pihatu, pehaner pihaner, purtuvetu purtuvitu.
ei ei in variation with e e or i i : aves aueis, veres uereir, fetu feitu, liouine lioueine, stahmitei siahmeitei, pehaner pihaner peihaner, uiro ueiro, etc. Forms with ei are especially frequent in Table Via 1-39.
e. Vowel-length is generally not indicated, but there are some cases of length, original or secondary, shown by writing vowel -{- h. h or vowel -f ^ + vowel. The latter combination occurs only in the Latin alphabet, hatu hahtu, man- traklu mantrahklu, sate sahaie, sahta sahata, etuf stamu eheturstahamu.
f. The variation meersta mersia is exceptional and the vowel is not an original e, but the doubling in meersta Via 17 together with the frequent omission of r r in the writing of forms containing rs suggest that r in such forms was often weak- ened or lost, with compensatory lengthening of the preceding vowel.
g. The origin and meaning of ooserclom are uncertain, but by any of the etymo- logies proposed the initial is a long o of secondary origin.
h. ii, uv and i, u. When i or u is followed by a vowel the semivocalic glide which naturally intervenes is generally represented by i, v in the native alphabet but is not indicated in the Latin alphabet: triia trio, heriiei heriei, Atiiefiur Ailersur, etc., tuves duir, kastruvuf castruo, prinuvatus prinuatur, etc. The practice in Umbrian agrees closely with that in Oscan. Single writing, however is occasionally found in the native alphabet where ii or uv is expected: tekvias, via, arvia, aviekla; purtuetu, iveka. Conversely uu occurs in the Latin alphabet in saluuom, saluua, tuua. For details and for special explanation of some of the aberrant spellings the proper sections in the phonology must be consulted.
i. Omission of medial consonants :
Variants with and without preconsonantal n n: antentu atentu, ustentu ustetu, krenkatru krikatru, Sansi Sage, tenzitim tesedi, anzeriatu azeriatu, onse uze.
Variants with and without t in medial consonant clusters : pustnaiaf pusnaes.
Variants with and without r r before s s : persnimu persnimu pesnimu pesnimu, persuntru pesuntru, Turse Tuse, farsio fasio.
Variants with s for rs (= f): Akefuniam Acesoniam.
Variants with r r for f rs: arsueitu arveitu arueitu.
j. Omission of final consonants:
Variants with and without final k c: erek erec ere ere.
Variants with and without final f /: ace. pi. treif tre, sif si, etc.; pres. pcpl. zeref serse.
Variants with and without final m m : ace. sg. persklum persclo, sim si, tuta totam, etc.
28 n. Grammar
Variants with and without final s: nom. pi. Ikuvinus Ikuvinu; conjunction heris heri.
k. The following variations are abnormal and not easily explained:
Final p for normal final f : kutef kutep, turuf turup, vitluf vitlup. The aber- rant forms with p all occur in lb 3-4.
Initial m for normal h: menes, cf. benus. Here the m may have been extended into the simplex from a compound "^kommenes, in which the b would have been assimilated to m.
Final m for n: numem but umen.
h and /: erahunt erafont.
1. Variants with and without h h, apart from h /i as indication of vowel length : heritu eretu, hebetafe ebetrafe, anhostatu anostatu.
3. Pronunciation. The phonetic value of the characters cannot be determined with any degree of exactness, because of the scantiness of material, the absence of metrical texts, and the lack of descriptive information from native or Roman sources. It would not be safe to attempt to reconstruct a phonemic system in the form which is customary in the treatment of living languages, since neither the native nor the Latin alphabet can be considered an entirely adequate repre- sentation of such a system. But in general where there is no conflicting evidence it is safe to assume that the values of the characters did not differ strikingly from the values of the equivalent characters in Latin.
Among the orthographic variations some have no significance for the pronun- ciation. The use in the native alphabet of k, t, p, for example, in forms spelled with g, d, b in the Latin alphabet certainly does not mean that the stops in ques- tion were unvoiced in the earlier stage of the language, but merely reflects the lack of distinction between voiced and unvoiced stops in the language of the Etruscans, from whom the alphabet was borrowed. In many cases the variations may arise from conflict between traditional and phonetic spelling; so possibly in forms hke ustentu ustetu, persnimu pesnimu. The distribution of final s in Tables I- IV and r r in V-VII, however, probably shows that the sound-change commonly known as rhotacism had taken place in final position between the time of writing of the earliest and latest tables; at least this conclusion is inevi- table unless we prefer to assume that the earlier tables merely failed to adapt themselves to a change already in effect. But it is altogether improbable that they would have represented r as different from s in the medial and not in the final position.
The following points in regard to pronunciation may be briefly noted:
a. a in those forms which have orthographical variants with u o probably had a raised sound like that of a in Eng. call. For the distribution of the sound see 7 b, c.
A. Phonology 29
b. Among the mid-vowels e was higher than e, and o higher than 6, if we may judge from the more frequent use of i i for e than for e and of u for o than for o. Since this use of i was especially frequent in the tables written in the Latin alpha- bet, we may suspect that the e was progressively raised as the language developed, and it is not unlikely that the o had a parallel history, but the lack of a special character for the o-vowels in the native alphabet renders it impossible to make a comparison of this sort.
c. Among the high vowels I appears to have been more strongly raised than f, since the use of e e is frequent for the short i but decidedly rare for the long i. It is likely that a similar difference existed between u and u : while u under certain conditions is represented by o (u of the native alphabet being of course ambigu- ous), u not only maintained its raised quality but assumed such a front position as to be written i i. Whether it was fully identical with the long i or had a value approximating that of French u, German ii, is uncertain.
d. The stops written p t k in the native alphabet should be voiced in those forms where orthographical variants or etymological evidence shows that the stop in question was actually a voiced one. The orthographical or etymological evidence may be found in the Lexicon. The following is a list of the most impor- tant of such forms:
With p : hapinaf, kapru (pr > br by 60 e), kumpifiatu, Krapuvi, Trepla- nes.
With t: teitu, tenzitim (medial t), tigit, tuplak, tuva (numeral), atru (tr > dr by 60 f), tekuries, tekvias, tesvam, testru (initial t), tertu (initial t), tetu (initial t), tuf, tupler, Kureties, utur, ustentu (second t), ententu, antentu, ampentu, sutentu (second t), pertentu (second t),tikainne, tigel; probably also terkantur.
With k : Krapuvi, kumiaf , krenkatrum krikatru (second k), kunikaz (sec- ond k), Ikuvins, iveka.
e. The character d (9), to which corresponds S {s) in the Latin alphabet indicates a sound which was derived from an unvoiced palatal or velar stop before a front vowel. Its exact nature can of course not be determined, but it was presu- mably a palatal sibilant or a sibilant like Eng. sh. Plain s does not appear as an orthographic variant of 5, but s appears so frequently as a variant of s, es- pecially in medial position (for example, 15 times pase, never *pa5e, in the etymolo- gical equivalent of L. pace), that by the time of the writing of the later tables the sound may have become almost identical with that of the inherited s, with a development parallel to that seen in L. centum > French cent.
i. The character q (r) indicates a sound arising from intervocalic d and, under certain conditions, from intervocahc Z. The corresponding use of rs in the Latin alphabet suggests a sound similar to the Czech f or Polish rz. But the rs in such words as persclu, persnihimu, lursitu,farsio, which has a different origin and corres-
30 II. Grammar
ponds in certain quotable instances to rs in the native alphabet, must have consti- tuted a cluster and not a single phoneme.
4. Accent. On the quality of the accent in Umbrian and the other Italic dia- lects we have no direct information, but from the extensive syncope of unstressed vowels we must conclude that at least during a period in the history of these dialects, the accent was characterized by a considerable element of stress. Of the laws regulating the position of the accent in the word and phrase we know next to nothing, but it is generally assumed that the initial stress which was domin- ant in Germanic, Celtic, and prehistoric Latin must at one stage have characterized Oscan and Umbrian as well. The question then is whether the accentual system remained in this stage or whether it was replaced, as in Latin, by a new system whereby the penult, if long, received the accent, while the antepenult received it if the penult was short. The distribution of double and single consonants in cer- tain Oscan forms led von Planta (I, p. 594) to argue in favor of a system hke that which prevailed in classical Latin (e. g. upsannam^ like L. operdndam but eehiia- nasum like L. amanddrum) but Umbrian provides no evidence of this kind. The vocalism of Puemune may give some indication of penultimate accent: if it is taken as cognate with L. Pomona, we must assume a stem ^Powemonojd-, with weakening of the unstressed initial syllable, in order to account satisfactorily for the Umbrian form along with Sab. Poimunien; and to this evidence von Planta adds several other forms, but they do not constitute a convincing proof. We must therefore leave the problem still unsettled.
Of sentence-accent it is possible to form some notion from the frequency of enclitic forms written without separation from the preceding word. It may safely be assumed that these enclitics were unaccented as in Latin. Examples are :
The indefinite pronoun pis: svepis.
The 3 sg, pres. ind. est and subj. si of the verb to be: peretomest, ortoest, parsest (with rs preserved as in medial position, in contrast to far with the final treat- ment), mersi. Writing as one word is more frequent than writing as two.
The fossilized 3 sg, verb form her (<*herit) in pisher (cf. L. quilibet).
The postpositions af a, kum ku com co, em e em e, per (rarely pe) per, ta tu to: spinamaf, vukukum, izerisco, va.peiem, destrame, ukriper, skalgeta. This type of enclisis is very frequent, and only a few of the examples are given here.
Certain pronominal enclitics for which see 106 b, c, d and 113 f, g, h, i, j,k, 1.
^ The marks over the vowels in these Oscan forms are not indications of stress but of vowel- quality; see 1 n. 1.
5. The Vowel System. In describing the origin, history, and etymological correspondences of the Umbrian vowels and diphthongs the following pIE phon- emes may be assumed as a basis from which to begin :
|
A. |
Phonology |
||||
|
Vowels: |
pure |
semi-consonantal |
|||
|
short |
a |
e |
0 |
i |
u ^ I m n |
|
long |
a |
e |
6 |
i |
u I 1 m n |
|
reduced |
d |
b |
|||
|
Diphthongs: |
|||||
|
short |
ai |
ei |
oi |
||
|
au |
eu |
ou |
|||
|
long |
ai |
ei |
6i |
||
|
an |
eu |
6u |
31
a. The obscure vowel represented by 9 and known by the Hebrew name shwa became in Italic, as in all the European branches of the IE family, fully merged with a and had the same history, but forms which originally contained it rather than a can be distinguished to a large extent by correspondence with Skt. forms containing i or by their ablaut-relationship with forms containing long vowels.
b. The character b is used to represent the sound known as " shwa secundum", which results from the reduction of short vowels in situations where complete loss of the vowel would result in difficult consonant-clusters. In place of b some comparative grammars use a e o since there is evidence that the reductions of a e o did not all have precisely the same history; this, however, may be partly explained by analogical restoration of the original vowel. In any case b is adequate for the few Umbrian forms with reduced short vowels which call for treatment in the present work.
c. For the explanation of the long vocalic liquids r J and nasals m n see 18.
d. In principle there is some justification for admitting diphthongs, both short and long, with liquids and nasals as second element (e. g. ar, el, om, en, etc.), corresponding to the i- and u-diphthongs, since there is a close similarity between the liquids, nasals, and y and w in relation to the pIE phonemic system as a whole.^ Yet in view of the history of the sounds in question in most of the individ- ual languages it is far more convenient to treat ei, oi, en, etc. as diphthongs but to give separate treatment to the separate components, vocalic and semi-con- sonantal, of ar, el, om, etc.
e. The following table is designed to show in the simplest form the history of the vowels and diphthongs in some of the principal IE languages. The "shwa secundum", the long vocahc hquids and nasals, and the long diphthongs, however, have been omitted. No attempt is made to show here the full history of the vowels in any language or to include the results of secondary changes such as, for example, the vowel- weakening of Latin. In general the development of the vowels in stressed syllables and between consonants is taken as typical. They are given in the standard orthography of the respective languages or in the Roman characters conventionally used to represent the standard orthography, but marks of length have been placed over Greek, Latin, Gothic, and Lithuanian vowels when such vowels are known to constitute distinct phonemes, and natural- ly over the Sanskrit long vowels as well.
32
II. Grammar
|
pIE |
Umb. |
Osc. |
Lat. |
Gk. |
Goth. |
Lith. |
OCS |
Skt. |
|
a |
a a |
a a |
a |
a |
a |
a |
0 |
a |
|
a |
a a U 0 |
a aa a U U 0 |
a |
d ^ 7] ^ |
0 |
6 |
a |
a |
|
e |
e e i i |
e e |
e |
£ |
i ai 4 |
e |
e |
a |
|
e |
e, i e, i (ei) (ei) |
i ii i |
e |
V |
e |
e |
e |
a |
|
i |
i i e e |
i I |
i |
I |
i ai 4 |
i |
I ' |
i |
|
I |
i i (ei) ih ihi |
i ii i |
i |
I |
ei ^ |
y |
^ |
I |
|
0 |
U 0 u |
U 0 |
0 |
o |
a |
a |
0 |
a |
|
6 |
VL 0 u |
u uu u |
6 |
CD |
6 |
uo |
a |
a. |
|
u |
u u |
u iu u |
u |
V 3 |
u au ^ |
u |
ii ' |
u |
|
u |
i i |
u uu u |
u |
V ^ |
u. |
u |
y |
a |
|
d |
a a |
a a |
a |
a |
a |
a |
0 |
i |
|
r |
ur or 1 |
ur or ^ |
or |
aQ Qa |
ur ru |
ir |
ri ru |
r |
|
I |
ul ol 1 |
ul ol^ |
ol |
aX Aa |
ul lu |
il |
Ii lu |
r |
|
m |
em em ^ |
em em ^ |
em |
a afj, |
um |
im |
? |
a am |
|
n |
en en ^ |
en en ^ |
en |
a av |
un |
in |
? |
a an |
|
ai |
e e |
ai ai ae |
ae |
ai |
ai |
ai ie |
e |
e « |
|
ei |
e e |
ei ei |
I |
ei |
ei |
ei ie |
i |
e 8 |
|
oi |
U 0 |
ui oi |
oe u |
OL |
ai |
ai ie |
e |
e 8 |
|
au |
U 0 |
av au |
au |
av |
au |
au |
u |
0 8 |
|
eu |
U 0 |
uv ou |
u |
sv |
iu |
au |
u |
0 8 |
|
oil |
U 0 |
uv ou |
u |
ov |
au |
au |
u |
0 8 |
A. Phonology 33
* The history of the vocalic liquids and nasals in the Italic dialects cannot be adequately reconstructed from actual examples. This portion of the table is partly filled out on the basis of comparison with Latin and with forms containing original or, ol, em, en, with which f, I, iji, n appear to have become identical in proto- Italic.
* ?y in Ionic and generally in Attic; a after e, i, q in Attic and in all positions in other dialects. ' t;, t; in Attic and Ionic had the quality of French u, German ii.
* The Gothic reflexes of e, o are in modern transcription written ai, aii, in order to distinguish them from the diphthongs ai, au.
* ei in Gothic, as in Greek in the time of Ulfilas, seems to have had the phonetic value of i.
* y in standard Lithuanian orthography represents a long i, but in Old Church Slavic repre- sents the Cyrillic character bi, the phonetic value of which may have had some resemblance to I and to u without being fully identical with either.
^ I and u stand for the Cyrillic characters b and 5 and designate short raised vowels of pal- atal and non-palatal quality respectively. ' e and 0 represent Skt. i- and u- diphthongs respectively.
6. pIE d generally remains as a a: auij avif : L. ayis ;kabru: L. caprum; argla- taf (initial a): L. arculata; etc.
For lengthening of a. see 26 c, d. For possible weakening of a. see 28 a, b, c. Of syncope of a. in Umbrian there are no knov^n instances. Oscan provides a few examples, and there is no reason to suppose that Umbrian a. was exempt, but a in comparison with e, 6 was a relatively rare sound in those situations where syncope was likely to occur.
7. a. pIE a is represented by a a or by 0 u. The unaltered a a appears in (abl. sg.) asa asa: L. drd; f rater frater: L. fratres; Tarinate Tarsinater (second a) like L. Arpinati, -is; imv. portatu purtatu: L. portato; pcpl. kuratu: L. curatum; etc.
b. The change to 0 u, which indicates a raised pronunciation of the vowel and a merging, partial or complete, with 0 u from pIE 0 appears in first declension nom. sg. mutu (but also muta): L. multa; etantu: L. tanta; neut. pi. alfu: L. alba; Casilos like L. Arplnds; pass. pcpl. nom. sg. m. conegos (but = kunikaz) hke L. amdtus; etc. More problematical is the penultimate vowel in Presto te and Tesenocir. The correspondence of 0 with the a of the older forms Prestate and Tesenakes makes it probable that the forms are derived from stems contain- ing d, and in fact there appears to be no objection to this view in the case of Tesenocir, hut Prestota might be more easily derived from -stdt-< -stdt-, the vowel- grade which L. antistita, etc. lead us to expect.
c. It is not possible on the basis of our extant material to formulate a satisfact- ory rule regarding the change of a to 0 u. It is best attested in final position, and before / in the nom. sg. of -Zi-stems and passive participles, which lost the vowel of the final syllable by syncope. It does not occur in the oblique cases of a-stems or of formations in -dti- or -dto-, and it does not occur in d arising by secondary lengthening of a (e. g. sahatam sahta). Its distribution is closely parallel to that in Oscan.
34 II. Grammar
8. pIE e.
a. e generally remains as e e: emantur: L. emantur; est: L. est; destru testru: L. dextro; prusekatu : L. prosecato; fertu: L. ferto; etc. Sometimes e is maintained even where Latin shows secondary changes : kvestretie like L. duritia (both from -etia-); veltu: cf. L. volo, etc.
b. e before another vowel appears as i i in farsio fasiu: L. farrea; tursiandu: L. terreantur; but not in the pronominal forms earn, eaf, etc.: L. earn, eas, etc.
c. e became i before r in ostensendi (fut. pass. 3 pi. based on earlier ending -Titer); herti beside earlier herter.
d. e became i i before n followed by a velar stop in krikatru (beside krenka- trum) cringatro; the vocalism is like that of L, tinguo beside rsyyco.
e. In the following forms e became i i when followed by a sibilant (including 9 and s from k by 46 b): tigit: L. decet; isegeles with prefix e{n); vagetum-i- se with postpositive e{n); tasis, hut against 21 cases with e e; vistiga, but against 33 cases with e e in this and cognate forms.
f. If uitlwvithx, L. vitulus,SLTe connected with L. vetus, ferog, we must assume a change e > i for which there is no certain explanation. Thurneysen {K.Z., XXX, p. 487) suggested borrowing of these words from an Italic dialect in which the change would be regular, but no such dialect can be identified.
g. e became o u under certain conditions in the proximity of labial sounds: pumpefias, puntes, cf. 0, pumperiais, Puntiis, IJojutztlei; pomtis, all with 0 u u probably after the cardinal "^pompe 'quinqiie', which must owe the change e > 0 to the following -nkw-, since the initial k^e- does not undergo vowel change in petur-, 0. petiro- 'quadri-'; sonitii: cf. L. sonere, sonare, all from ^-swen-^ unless the verbal forms are denominatives to the ablaut-variant swono-, which is improb- able for all these forms.
h. For lengthening of e see 26 c, f, h. For syncope of e see 29 a-f, m.
^ Oscan, against Umbrian and Latin, preserves e after sw. sverriinei, cf. Eng. swear, answer.
9. pIE e is represented by e e or by i i, very rarely by ei ei. The frequency of the i-spelling, which is used to the total exclusion^ of e in the imperative of second- conjugation verbs written in the Latin alphabet, is evidence for a closer pronun- ciation for e than for e.
a. e e from e is exemplified by plener: L. plenis (initial syllable); preve (suffix as in L adverbs in -e(rf); rehte: L. recte-^ imv. karetu, habetu: L. habeto; etc.
b. i i from e is exemplified by frite: cf. L. fretus; prusikurent with -sik- < -sek- in a long-vowel perfect stem; eitipes, if < *eit(om) hepens, also a long-vowel perfect; tigit with second i < e as in L. decet < "^deket; tigel if from *die-kelo-; imv. habitu: L. habeto, etc.
A. Phonology 35
c. ei ei occurs in heriiei v/ith e the sign of the subjunctive, in nesimei with the same adverbial ending seen in rehte < -ed, and in seipodrupei with se- as in L. sedulo, se fraude, etc. This rare spelling may be intended to indicate a high vowel intermediate between e and i.
1 Unless we count maletu Ila 18 as imv., as Devoto does. Most take it as a pass. pcpl.
2 Original long e in the suffixal syllable and possibly also in the radical syllable, with which MuUer, p. 382, compares Av. raSta- < raSta 'in gerader Richtung, gerades Wegs'; otherwise the lengthening in the first syllable is secondary.
10. pIE i.
a. i generally remains as i i: tikamne: cf. L. dicdre; uirseto, pass. pcpl. of verb equivalent to L. video; simo §iinu, give: cf. L. citra; pis : L. quis; etc. In the native alphabet doubling is frequent in the prevocalic position (2 h) : triia (but trio): L. tria; heriiei (but heriei); etc.
b. Yet I regularly appears as e e when final, as in Latin : in neut. nom.-acc. sg. of I'-stems: uerfale: L. verbale; in abl. (originally loc.) sg. of consonant-stems: pure: Gk. dat. tivqI; curnase: L. cornice; etc.; in the conjunction ote ute: 0. auti, L. aut, all from ^auti.
c. i also becomes e e in certain other forms, under conditions which are partly obscure. There is considerable fluctuation in the spelling, but e for non-final i is very rare in the Latin alphabet: vea twice, but via once, uia 5 times, abl. : L. via; neut. plurals sakreu, perakneu with -eu < -id; steplatu once, stiplatu twice: L. stipulate; reduplicating present verb forms sestu: L. sistito, tefa subj. as = L. ^didat (but original e in fut. pf. tefust !); imv. ahtrepufatu (forms with e 5 times, with i 3 times, erroneously with o once) : cf . L. tripodato ; pronomin- al and conjunctional forms pefe perse persi persei pirse pirsi: L. quid + enclitic I; fratrexs fratreks, fratrecate: L. fratricus, ^fratricate (Jratric- with Buck, § 45, rather than fratrec- as suggested by von Planta, I, p. 102). The proximity of f r r may be an influencing factor in the change to e e, but uirseto, auirseto (each 4 times with i, never with e) and kapife capirse, etc. (16 times with i i, never with e e) are unfavorable to such a view.
11. pIE i generally remains as i ihi i ih; e e are much rarer for long i than for short i, and this fact may be taken as evidence for a closer pronunciation of i than of i; ei is also sometimes found.
a. i ihi i ih: Ikuvina: L. Iguvina, cabriner: L. caprlnl, etc; fourth conjugation imv. forms persnimu persnihimu persnimu persnihmu, anouihimu, amparitu, amparihmu.
b. e e: pehatu, pehaner (but 37 instances with i i in forms of this verb and of the noun pihaklu): cf. L. pius, piare, 0. piihiui, the last form lending support to the vowel-length of the Umbrian forms; screhto (h here not a mere mark of
36 n. Grammar
vowel-length) ; fourth conjugation imv. purtuvetu purtuetu, but much more frequently with i i.
c. ei: peica: L. picam; peico: L. plcum; peihaner: cf, L. piandi; screihtor: cf. L. scrlpta; ueiro: cf, Skt. vlras, Lith. vyras (L., Olr., and Goth, forms are based on short i). Most instances of ei for I, including all those here cited, occur within the first 39 hues of Via. This portion includes, for example, 3 occurrences of ueiro against 8 of uiro in the rest of Via, VIb, and Vila.
12. pIE 6.
a. 0 generally remains as o u:^ ocar ukar: L. ocris,oxQig; oui uvef (acc.pl.): L. ovis, ofig; poplo puplu (both syllables) : L. populum; porse pure: L. quod, + enclitic; ostentu ustentu: L. ostendito; etc.
b. 6 before r + consonant becomes u. Naturally the change can be observed only in forms appearing in the Latin alphabet, since u may stand for either o or u. Examples are: curnaco: L. cornicem; purdouitu with prefix equivalent to L. por-^; tursitu <*torsey-et6d, causative-intensive form of verb represented by L. terreo.
c. furo : L. forum is an instance of the change o> u before prevocalic r, unless furo, like Ovga, represents the zero grade *dhuro- in contrast to *dhworo-.
d. portatu, portaia, portust (once each in the Latin alphabet: L. portato, etc.) have only o, and the fut. pf. forms couortus, etc. (cf. early L. vortus, advortit) have twice o and once courtust for *couurtust.
^ The development of o in Oscan agrees for the most part with that in Umbrian; both, for example, maintain 6 unchanged before It in contrast to Latin, which changes it to u: nom. pi. molar; cf. 0. molto, but L. multa, all from *mol(k)ta. Umbrian, however, does not share with Oscan the change o> u before final m: e. g., pres. infinitive O. deicum, acum, etc., but U. erom, afero, all based on original ending -om.
' *por- may be from earlier *pf-: see 17 a.
13. p IE 0 is represented by o u and u, the forms in the Latin alphabet being alone significant, since u serves for both values. The choice between o and u does not depend on phonetic environment alone; rather certain words and certain grammatical terminations have a strong predilection for one or the other vowel. The great frequency of abl. sg. and "future" imv. forms may produce an exag- gerated impression of the extent to which the 6 was raised toward the position of u, but yet it is reasonable to suppose that d was higher than 6, just as e was higher than e.
a. 0 occurs in nome (98 occurrences of its various forms with o, none with u) : L. nomen; homonus: cf. Oh hom6nes;poni (13 times in various forms with o, never with u): cf. L. potus, potio; abl. sg. somo against more than 100 instances of abl.
A. Phonology 37
sg. forms with -u; gen. pi. Atiersio, peracrio (< -6m). esono (10 times in various forms with o, never with u) probably also belongs here as from *ais{e)s-6n.6-, but the vocalism of the suffix is not certain.
b. u occurs in arsfertur, ace. arsferturo, noun like L. scriptor, etc.; abl. sg. pihaclu, poplu, uinu and numerous others; nom. pi. louinur, Clauerniur; imv. etu, habitu, etc., extremely numerous and never with -o.
c. dupursus, peturpursus 'hi-, quadripedibus' are usually taken as containing penultimate u < o, as in Dor. ncbg, Goth, fotus, but it may be possible to derive them from the stem with the ablaut-grade d, as in nodog, nodi etc., if we admit the change 6 > u before rs from intervocalic d as well as before the cluster rs (12 b).
d. ulo probably has u < o, as in L. olim.
14. pIE ii.
a. u in general remains as u u, but only forms occurring in the Latin alphabet are unambiguous. Examples are: puse: cf. L. ut, both < kwu-; subotu and other forms containing sub-: L. sub-; super: L. super, vjieg; pure-to: cf. tivqoq, tivql; Rufrer: L. rubri, cf. egvOgog, Skt. rudhiras; etc.
b. u became o before m, which in some cases was subsequently lost: somo (first o): L. summum, summo; fourth declension ace. sg. trijo: L. tribum; supine aseriato: cf. L. observatum.
c. u became o before p in sopa (of 4 various forms of this word 3 have o, 1 u): L. suppam; but observe super, superne, subra, where the change fails to occur.
d. u became o, apparently, in sorsom, a word whose etymology is obscure, but which according to the prevailing view contains su (as in L. siicula against sus) + -do-.
15. piE a.
a. u is normally represented by i i, indicating a change through ii toward a sound nearly or fully identical with that of I. Examples are: sim, sif sif : cf. L. sus, i)g; pir pir: nvg; frif ace. pi.: cf. L. fruges; abl. sg. of u-stems mani mani, trefiper, afputrati.
b. A possible example of i < u is pass. pcpl. statita if = L. statuta. pur- titu purdito is also sometimes taken as containing i i < u, but the vocalism is not clear, and in any case the vowel here could scarcely have been primary ii. The derivation of the subjunctive form dia from a form with original u is also improbable.
c. There are perhaps no real exceptions to the above change, mugatu and its pcpl. muieto are taken by Buck, § 58, as examples of preservation of u, as in L.
38 II. Grammar
muglre, but it is probably better to assume a short vowel, as in ejuvkov, with von Planta, I, p. 135. In strusla struhgla the length of the u, indicated by h, pro- bably arose too late to participate in the change u > i, which, to judge from traces in Oscan and the minor dialects, may have begun in the period of Oscan-Um- brian unity.
16. pIE 9 shwa and b shwa secundum.
a. d, the reduced grade of e, a, 6 in the long-vowel ablaut series (35 c, d), had precisely the same history as a in all branches of the family except Indo-Iranian. Instances of a < a in the Italic dialects are to be recognized chiefly through their etymological correspondence with Latin and other forms where a alternates with e, a, 6: U. fagia: L. facial, cf. feci, U. fetu (whether fe- or fek-); U. stafli: L. stdbilis, cf. 1 pi. larafisv but 1 sg. larrj/ui, Dor. tarajui. For the occasional secondary replacement of a < 9 by a short vowel of different quality see 35 d with n. 12, 119 b, c.
b. A possible instance of a from b is represented by tapistenu if we admit the explanation in n. on IV 30 whereby the form contains a reduced grade of the root seen in L. tepor, etc.
17. Short vocalic liquids and nasals. The allophones of r and I with vocalic, that is, syllabic function, which actually existed in Sanskrit and are commonly represented by the symbols r and I, must be assumed also for pIE, where they alternated, as ablaut-variants, with er, or, el, ol, just as i alternated with ei, oi, and u. with eu, ou.
a. In pit. f , I before consonants became or, ol, and consequently it is not always certain whether actual forms containing or, ol reflect original or, ol or their ablaut- variants r, I. In some forms, however, the original vocalism can be inferred by comparison with forms in other languages, especially Sanskrit and Greek. The principal Umbrian forms for which we must assume original r, I are: orto, urtas: L. ortus, Skt. rtds; trahuorfi: L. transvorsus, Skt. villas (but fut. pf. kuvurtus couortus may contain original or, since the IE perfect had o-vocalism in sg., zero-vocalism in pi., with possibilities for later leveling); pur- pur- (in purtuvitu, purdinsust, etc.): L. por-rigo, tiuq a; molar mnta. < */77o/(A:)/-: L. multa, Skt. pass. pcpl. m^stds.
b. f , I before vowels — in which position we should write ^r, II — apparently became ar aZ:karu: L. caro, carnis, cf. Skt. kfndli 'injures'; possibly also maletu in contrast to L. molitum with o-grade.
c. The vocalic nasals m, n, unlike the vocahc liquids, did not exist in Sanskrit, but nevertheless are assumed for pIE, where they alternated with em, om, en, on.
d. In pit. m, n became em, en before consonants; consequently em, en from
A. Phonology 39
m, n can be distinguished from original em, en only by comparison with forms in other languages. The principal Umbrian examples are: desenduf '12' with -en- < -em- < -m-: L. decem, dexa, Skt. dasa; iveka iuenga: cf. L. iuvencus, Skt. yuvasas; nom.-acc. sg. of neut. stems in -men: nome numen (with -e -em for -en): L. nomen, Skt. nama, etc.
e. n before vowels — in which position we should write nn — apparently became an; at least this has been assumed in order to explain the discrepancy between the negative prefixes L. in- and O.-U. an-. We may suppose then that Latin generalized the preconsonantal in- < en- < n for all situations: infamis, imbellis, inermis, etc., and O.-U. the prevocalic an- < nn for all situations: U. antakres, anhostatu, etc.: cf. a-, dv-, Skt. a-, an-.
18. The long vocalic liquids and nasals symbolized by f, I, m, n were assumed by Brugmann in order to account for such correspondences as Skt. pHrnds, Lith. pilnas, etc. Their precise character is unknown, but it is clear that their occur- rence is in forms containing disyllabic ablaut-bases (*pele-, etc.) with accent on the suffix, so that the two syllables of the unaccented base appear in the reduced and zero-grades respectively. Hirt symbolized them by brd, bb, hmd, etc. in order to show their original disyllabic character. The history of the long-vowel ablaut series as outlined in 35 is apphcable here, including the statement in regard to the laryngeals as the source of a in zero-grade forms. The origin of i, u is partly similar to that of r, etc., but their history in the separate languages is sufficiently distinct to call for separate treatment.
a. r, Z, m, n are represented in Italic by rd, la, md, nd or, under certain conditions which are not altogether clear, by ar, al, am, an. Probable examples in Umbrian are: kumates comatir < *kom-malt- < '^-mJ-to-: cf. Skt. milrnds;an clar <*an- kld-s < -kl: cf. L. cld-mo (denominative to "^kld-md, like fd-ma), calare, xaKsa), the last two with different gradation ;i natine: cf. L. ndtio, ndtus <*gnd-t-, Skt. jdtds, all from gn-; naratu naratu: cf. L. ndrrdre, gndrus *gnd-rdlro- < gn-: cf. L, i-gno-ro, co-gno-sco with different gradation, mandraclo mantrahklu is sometimes similarly explained, as from *man-trdk-klo- < *man-trdg-tlo <*-tfg- with gradation of the same root found in L. tergeo, although its root is not of the type which customarily furnishes ablaut-grades with f .
1 See Brugmann, Berichte der kon. sacks. Ges. der Wiss., 1890, p. 206.
19. Diphthongs. pIE ai.
a. ai is normally represented by e e: esono esunu: cf. 0. aisusis; pre pre: 0. prai, L. prae; sve: 0. svai; dat. sg. of a-stems (ai < earher di by 25 with a) tote tute, etc.
b. ai is represented by ei in eikvasatis, eikvasese if these forms are connected with L. aequus, but their etymology and meaning are uncertain.
40 II. Grammar
c. pernaiaf, pustnaiaf, persaia per aia are exempt from the change ai > e e, the diphthong being preserved before the consonantal i in the suffix -ayyo-.
20. pIE ei.
a. ei is normally represented by e e: preve: cf. 0. preiuatud, L. prlvus; etu etu: L. ito, cf. Skt. etu; gen. sg. ending in o-, i-, and consonant-stems kapres, ocrer, farer, etc.: cf. 0. ending -eis -eis (no Latin parallel): dat, sg. of i- and con- sonant-stems ocre ukre, luue luve, etc.; loc. sg. of o-stems onse uze, etc.: cf. 0. terei, comenei, L. humi; probably also loc. sg. of i-stems ocre, etc.
h. e e from ei is in all probability to be recognized in kletra: Goth, hleipra.
c. ee from ei appears in eetu VIb 54 beside 16 instances of e e in imv. of this verb and its compound enetu enetu; and also once in fut. eest < *eiset.
d. The extremely rare use of i i to represent the vowel derived from ei occurs in one instance of irer (Via 25) against the frequent spelling of this pronoun with initial e e.
21. pIE oi developed differently according as it stood in initial or final syllables. The development in medial syllables is not attested by any examples, and in fact the evidence for initial syllables is meagre, partly because of the twofold value of u in the native alphabet and partly because of the doubtful etymology and in- terpretation of some of the forms used as evidence.
a. 0 from oi in initial syllables appears to be contained in pora if fully equiva- lent to 0. poizad (110 e); nosue if < *noiswai, an ablaut-variant to 0. nei suae; in ponisiater (punigate) if < earlier poinikiato- : cf. L. punicus, but this etymo- logy is by no means certain.
b. uocu vuku might be an example of o u < oi if equivalent to Folycoc,, but the view here adopted, by a slight margin of preference, is the one which connects it with L. lucus, with wok- < louk- (24 a, 55 a)^.
c. The following forms, known only in the native alphabet, in all probability contain u — o< oi :kuratu: cf. Pael. coisatens; muneklu: cf. 0. muiniku, OL comoinem (communem); unu: OL oino, cf. oIvt] 'one' (on dice), Goth. ains.
d. For oi in final syllables reference is made to 25 b, since in the dat. sg. and dat.-abl. pi. forms of the second declension which furnish the examples the oi results from shortening of the long diphthong oi.
^ uinu vinu, despite its obvious connection with folvog, does not constitute valir evidence against o < oi in initial syllables. The word is of non-IE origin, and all forms o curring in ItaHc languages (e.g. L. vlnum) and Etruscan have i. See Sturtevant, Lang., X (19? 4), pp. 6-9.
22. pIE au is represented by o u: ote ute: 0. auti, L. aut; uhtur: L. auctor; fons: cf. L. Fones, Faunus; frosetom: cf. L. fraudo, fraussus; torn turiif: L. tamos;
A. Phonology 41
also kutef, if it is pres. pcpl. of a denominative verb from pcpl. equivalent to L. cautus, whence our form would = L. *cautens.
23. pIE eu is represented by o u. Since eu had precisely the same history as ou in all Italic languages, it is not possible in every case to distinguish forms with original eu from those with original ou. On grounds of comparative etym- ology it is probably correct to assume original eu for totar tutas: 0. touto, Goth. piuda, and possibly the Illyrian personal name Teuta. On eu < eu see 25 c.
24. pIE ou.
a. ou, showing the same development as eu, appears as o u. It is probably to be recognized in uocu vuku if these forms are connected with 0. luvkei, L. lucus rather than with FoXkoq. For vocalism see Muller, p. 243 undei loukos.
b. comohota apparently has -oho- representing 6 from secondary ou after syn- cope of e in *moweto-: cf. L. commotus.
c. In the following forms it is uncertain whether o u is derived from eu or from ou: rofu: cf. L. rufus (of dialectal origin; on vocalism see E.-M.^, p. 1021); VuvqIs: 0. Liivkis, L. Lucius, with vocalism either as in Aevxog or as in L. lux, which is itself ambiguous. *
25. The ie long diphthongs did not survive as such in any Italic language; in fact, apart from the Greek case-endings commonly written a,r],q),they scarce- ly exist in any European branch of the family, having suffered either shorten- ing of the first element or loss of the second, so that they became identical either with the plain diphthongs or with the long vowels and had the same subsequent history. The change to plain diphthongs was doubtless early, and the change to long monophthongs may have begun in pIE under conditions of juncture which are in part obscure. Forms originally containing long diphthongs are generally distinguished by their correspondence with Sanskrit forms or with Greek case- endings in g, rj, cp.
The following forms contain vowels arising from plain diphthongs arising from earlier long diphthongs:
a. ai > ai > e e: dat.-loc. sg. of a-stems: tote tute, etc.: cf. 0. deivai, viai, L viae, dyoga ; also dat.-abl. pi. of a-stems: aseriater anzeriates: cf. O.kerssnais, L. terris, if based on a case-ending -dis made after the instrumental pi. of o-stems in -ois while the long diphthong was still intact; otherwise it should be included above under ai (19).
b. 6i> oi> eeiiei: dat. sg. of o-stems: luvie, pople, Tefri, fratreci, Grabouei: cf. 0. hurtui, OL Numasioi, Xvxoi, dat.-abl. pi. of o-stems: ueris veres, sehmenier, esoneir: cf. 0. Nuvlanuis, zicolois, OL privicoles, L. hortis, if all the Italic forms
42 II. Grammar
are to be derived from IE instrumental pi. in -ois: Skt. devais. The forms with ei are restricted to the first 42 lines of Table Via.
c. eu > eiz > 0 u in dat.-loc. sg. of u-stems manuv-e, trijo : cf . L. dat. curru, Skt. loc. sunau.
The following forms contain etymologically long vowels arising from long diphthongs with loss of the second element:
d. ef > e > U. e i in fifth declension forms dat. auie, ri: cf. L. fide dat. in Hor., Serm., I, 3, 95; abl. ri, ace. pi. iouie. The fifth declension in general is based on stems in -ei {-eu in the case of L. dies). For details see 79.
e. 6u > 6 > v. u u (13 with b): ace. bum, ace. pi. buf buf : Dor. f^cbv, Skt. gam.
i. There is a slight possibility of a similar origin for the dative ending in the divine names Trebo and Fiso (and perhaps Ahtu), if these are derived from forms in -ai and -6i, like OL Matuta, C.I.L., I, 379, and the standard dat. sg. in L, o-stems eqiio, etc., with further examples in the minor Italic dialects; see von Planta, II, pp. 94-5, 111; not admitted by Buck, § 171, 3 a, 185, 2, who treats the Umbrian forms as fourth declension, in which case the development is through plain diphthongs and not through long monophthongs.
26. Lengthening of Vowels. In Latin it is a well-known fact that vowels originally short were lengthened before certain consonant-clusters, partly in com- pensation for the loss of one of the consonants, partly where no such loss had occurred. The authority of ancient grammarians, which furnishes a part of our information for Latin, is lacking for the Italic dialects, but occasionally the or- thography gives evidence of vowel-lengthening under conditions similar to those in Latin, and it is Ukely that the phenomenon was by no means limited to the few forms in which we have direct evidence of it but extended to categories of which these forms are merely isolated examples. The evidence in Umbrian in- cludes all the means by which vowel-length is indicated: doubling of the vowel, use of the vowel -[- hh, or of the vowel -j- h -\- vowel, and also the use of ei for e and of u for o, indicating a change toward a more raised position as the e, o was lengthened.
a. The consonant-group -nkt- was reduced to -t- with lengthening of the pre- ceding vowel: sahatam sahta: L. sdnctam; sihitu, ansihitu: L. clnctos.
b. A similar lengthening before -nk- is probably to be recognized in gihgefa, if it is from *kinkel{y)a or *kinkedd and related to L. cingo, KiyxXibeg, etc., according to the usual view.
c. The group -kt- became -ht- by 46 i, and in such forms as apehtre, ehueltu it is possible to regard the h as sounded, but in amprehtu, where it has no etym- ological justification, and in forms with the spelling consisting of vowel -f ^ + vowel (e. g. sahatam, eheturstahamu) it is not possible to regard the h as anything
A. Phonology 43
more than a mark of vowel-length. It was used originally, then, to indicate the aspiration arising from k before / (and also from / < p before t by 38 b) and con- tinued to be used sporadically to show the length of the vowel after the aspiration had disappeared. Our principal examples are forms containing ehe- eh-, the prefix equivalent to L. ex, e: apehtre: cf. L. ab extra; ehueltu, ehvelklu, eheturstahamu, and finally the preposition ehe itself. These forms present difficulties when we attempt to explain them satisfactorily for Umbrian, Oscan, and Latin, e cannot be from *ex before I, m, n in Oscan or Umbrian, since s is not lost in this position in these dialects as it is, for example, in L. e-mergo < *ex-mergo. It is necessary therefore to derive O.-U. e-< eh-<*ek (: ix), the e- arising in this way being then extended analogically to other situations. A further difficulty is the fact that in Oscan the h arising from k before / is in general not lost; at least forms like Uh- tavis: L. Octavius; ehtrad do not show variants without h. Von Planta, I, p. 209, suggests that Oscan h may have been lost before spirants (cf. eestint), the prefix then being generalized in the form e~. Actually the Oscan forms with h written before t are too few to allow any generalization as to its value.
d. aanfehtaf may exemplify lengthening of the vowel before nf, as in L. inficio, etc., but its value as evidence is weakened by the fact that the first a fell at the end of a line, perhaps causing the engraver to begin the word anew on the following line.
e. Lengthening before final original ns in the ace. pi. forms toru, rofu (o-stems), and aueif, treif (i-stems) is inferred from the spelling with u, ei unless we follow the less probable view that these forms ended in -ons, -ins originally. The lengthen- ing is common to the Italic languages (cf. L. viros, avis, etc.) and is certainly older than the Umbrian change of -ns > -/.
f. Lengthening before medial ns appears in 0. keenzstur (with nz for nts < ns; for length cf. L. censor) and probably occurred in similar situations in Umbrian, but there is no direct graphic evidence of it.
g. ooserclom has initial o resulting from compensatory lengthening if it is from ^ob-serklom, but there is an almost equally good possibility of taking it from *au-serklom < *awi-serklom, in which case oo would be explained by develop- ment of au as in 22, after syncope had occurred.
h. The reduction of the group rs is accompanied by lengthening of the preced- ing vowel in frateer (= L. fr aires) < "^f raters < ^frateres. Similar lengthening is probable before medial rs, in which the loss, or at least the weakening of the r is shown by frequent omission in spelling (2 i). The doubling in meersta (with rst< rsst with secondary rs = v< d according to one widely accepted etymology) may be an indication of such length, but this doubling occurs only once against 14 instances of merst-.^
1 meersta is admitted as evidence of length by PI., I, pp. 207-8, and Buck, § 76, 1, but taken as dittography by Kent, T.CJ., p. 41.
44 !!• Grammar
27. Shortening of Vowels. Since vowel-length itself is so irregularly shown in Umbrian orthography, it is especially difficult to detect those cases where vowels originally long have become short. On the other hand there is moderate- ly abundant evidence that in both Oscan and Umbrian long vowels are main- tained in a number of situations where these vowels would be shortened in Latin.
a. Final a is altered in quality but keeps its quantity: mutu, 0. molto, but L. multd, etc. (7 b; etymologically short a does not change to o u).
b. The length of the vowel is maintained before final m in the gen. pi. praca- tarum, since -um here can represent -6m, but not -dm.
c. The retention of long vowels before final t is attested in Oscan by the verb forms kasit: L. caret, fusid: L. foret, all from -et, and possibly in Umbrian by trebeit, if ei stands for e (second conjugation) or for i (fourth conjugation). It is much less probable that ei stands for e (third conjugation), both because ei rarely represents a short vowel and because the intransitive meaning of trebeit favors the belief that the verb belongs to the second conjugation.
d. Of the retention of long vowels before final r Umbrian provides no unam- biguous examples.
28. Vowel-Weakening. Umbrian and Oscan present very few parallels to the weakening of vowels so frequent in Latin in syllables which were unstressed under the early system of initial accent. Moreover there are clear cases of ab- sence of such weakening in dialect forms closely analogous to weakened Latin forms (e. g. procanurent against L. oc-cinui, tagez against L. tacitus). Yet there are some instances of vowel-change which cannot be explained by any other supposition than that of weakening in unstressed position. The weakening, when it occurs, is usually in the direction of o u, not of e, i as in Latin, and it seems to be especially favored by the proximity of a labial consonant; possibly some cases of actual weakening are concealed as a result of the lack of a special o-cha- racter in the native alphabet, a being preferred to u, and possibly other cases were eliminated by recomposition, but it is not possible to formulate any compre- hensive rule.
a. The most probable instance of weakening is prehubia (varying, at least in spelling, with unweakened prehabia): L. praehibeat.
b. kumaltu (3 times) kumultu (once) comoltu (5 times) form a similar case if the proper vowel-grade is a/< J in the imperative as in the participle kumates (18 a), but ul ol may represent ol, kumaltu then being remodeled after the par- ticiple. In sumel also we have to choose between the possibility of weakening of *semel (cf. L. simul) in enclitic use or of ablaut-variation in the first syllable.
c. The o-forms Prestota, etc., show weakening if Prestate contains a < 9, but the vowel may be o < a (7 b). atropusatu against 5 instances of -ire- and
A. Phonology 45
3 of -tri- is probably not a form with an obscure labial-colored vowel, but simply an error (cf. Kent, T.C.L, p. 43).
29. Syncope. Loss of unstressed short vowels by syncope is much more fre- quent in Oscan and Umbrian than in Latin. All the short vowels except possibly u^ are subject to syncope, but the analogy of related forms occasionally prevents the loss of the vowel or restores it to its original place. Even for Latin it is not possible to lay down a complete set of rules describing the occurrence of syncope or its failure to occur, and still less is it possible for the dialects.
Since the Italic languages had initial stress during the period when syncope was taking place, we cannot expect to find loss of vowels from initial syllables. The principal instances of syncope in medial syllables are treated in paragraphs a to h and in final syllables in paragraphs i to m, but loss of absolutely final vowels is treated separately in 31 .
a. e is lost in the suffix of comparison and contrast -tro-< -tew-: destre testre: L. dextra, but de^irsQa, nertru, postra, etc.
b. e is lost before t in imperative forms of verbs of the third conjugation: kuvertu < *kom-wertetdd: L. convertito; ostendu ustentu < *ohs-tendet6d, with nd< nt< ndt (60a): L. ostendito; ditu titu< *didet6d (119b): cf. didorco; sestu lib 22: L. sistito (but not pres. ind. sestu lib 24: L. sisto); andersistu < *-siz- detod: L. *intersldito; etc. sumtu < *sub-emet6d shows loss not only of the e before the imv. ending but also of the e of the radical syllable, fertu fertu probably shows not syncope but unthematic inflection, since Latin, which does not syncopate imv. forms in general, has ferto.
c. e is lost in 3 sg. pass. ind. tefte < *dedeter.
d. e is lost before s in onse uze< *omesei: L. umerus; felsva< *feles-wa (etym- ology not quite clear, and / here may show dialect borrowing, but in all probabil- ity connected with L. holerd); mersto if < mefs-to- < *medes-to-, but this etymo- logy is very uncertain; ose if < *opse< ^opesi, but here again the etymology is doubtful.
e. e is lost before I in arglataf : L. arculatas; tiglu < *dik-elom; ereglum; struhgla: L. *strmculam; preuislatu: L. *prae-vinculato. In all these forms g s < k (46 b) shows that a front vowel, in all probability e, must originally have followed. In other cases we must reckon with the possibility that no vowel original- ly stood before the /, hence no syncope. In some cases the former presence of e can be supported on other grounds: katlu < *kat-elom because -tl- would have become -kl- (41 e) ; similarly in vitlu, where the original e also receives some sup- port from 0. Viteliu, L. Vitellius. But the vocalism of several forms is uncertain, and Latin cognates in -cuius, etc., are not decisive, since the u can be from e or from o or can have arisen through anaptyxis.
46 II. Grammar
f. e is lost in the noun-suffixes -men- (oblique cases) and -meno-: nomne: L. nomine; tikamne: L. *dicam{i)no; pelmner if < '^pelpmen-; menzne if < *mens- en-i; termnom-e: L. terminum.
g. f is lost in todcom-e < *toutikom; percam: L. perticam; struhgla < *stru- wikelam with u< uw after syncope: cf. L. struem; scalse-to: L. calice.
h. i is lost in the reduplication after the prefix in restef < "^resistens; at least the transitive sense, generally rendered by L. restaurans, favors the view that the present stem was originally reduplicated.
i. 0 is lost before final s in the masc. nom. sg, of words of the second declension: tigel < ^diyekelos (with s lost after the syncope); similarly katel ; fratreks fra- trexs if < ^fratrikos; pf. pcpls. tases tagez: L. tacitus; pihos pihaz: L. piatus; con eg'os kunikaz ; stakaz ; gerundive pelsans < ^'pelsannos < *pelsandos.
j. The vowel o/e has been lost in the forms nom. sg. mers mefs, abl. pi. mersus, from stem ^medos-medes-.
k. 6 is lost before final s in the dat.-abl. case-ending of the third and fourth declensions, the / which preceded the o being then assimilated to the -s (61 d, 101 c, d, e). Typical examples are: aueis aves < ^awifs < *awibhos: L. avibus; fratrus: L. fratribus; nerus: Skt. nfbhyas; berus: L. verubus; etc.
1. i is lost before final s in nom. sg. of the i-stems Casilos < "^kasildtis: cf. L. Arpinas; fons if < '^faunis; pacer < ^pdkris with samprasarana (32 d) and loss of -s: cf. L. acer < dcris.
m. e is lost in future verb forms between the tense-sign s and the personal ending -s (2 sg.) or -/ (3 sg.): anpenes, heries, purtuvies; prupehast, ferest, fust; etc. Syncope must likewise be recognized in the same situation in the future perfect, unless we regard the second component as an auxiliary form in which the syncope was already accomplished previously,
^ Nom. sg. forms of u-stems, "which could provide evidence, are lacking in both Oscan and Umbrian, but erus, if it represents a stem in -us, shows absence of syncope in comparison with o-stem forms like katel, etc. Von Planta, I, p. 231, compares Goth, u-stem handus unsyncopated and i-stem fisks syncopated.
30. Failure of Syncope. Although it is not always possible to explain those instances in which syncope fails to occur when expected, a few cases deserve mention.
a. The e in the final syllable of 3 sg. pf. ind. forms is maintained: dede ^ < *deded< *dedet: cf. 0. kumbened, L. dedit, convenit. The same is probably true in the 3 sg. pres. ind. of verbs of the third conjugation, on the evidence of Marruc. feret, Vest, didet, although no Oscan or Umbrian forms are quotable.
b. The imv. kanetu is often taken as an unsyncopated third-conjugation form.
A. Phonology 47
on the supposition that a vowel between n and t escapes syncope; but actually there is no support for such a view unless we count 0. Genetai. In view of the almost invariable syncope in imv. forms of the third conjugation, therefore, it seems preferable to assign kanetu to the second conjugation ( < *kanet6d < ^kaneyetod, 118 b).
c. Several pf. pcpl. forms offer difficulty because of their failure to syncopate the vowel before the t. uirseto (with aiiirseto) might be explained as an example of extension of e into the non-present forms of a second-conjugation verb, as if L. video had pcpl. ^videtiis instead of visus, and tagez tases might be similar- ly explained, but this explanation is more difficult for verbs whose present stems belong to other conjugations: prusegetu to prusekatu; muieto to mugatu; also daetom, peretom, frosetom, pesetom, vagetum. Von Planta's explanation^ may be tentatively adopted for want of a more satisfactory answer: syncope occurred in the imv., where the medial syllable stood between the accented initial syllable and the heavy ending -tod, but failed to occur before the endings -ts < -tos and -torn in the pcpl., the unsyncopated form then being generalized.
1 On minor inscr. 352 Co. = 292 PI. = 82 Bk. (from Todi).
* I, p. 215. Devoto's view (pp. 156-7) that long e in pcpl. alternated with a in pres. of the first conjugation is tempting, but this type of alternation is not attested by the evidence of cognate languages, while several Latin verbs of the first conjugation do possess pcpls. in -{i)tus < -{d)tos: domitus to domare, sectus to secure, etc.
31. Loss OF Final Vowels. Many inherited short vowels in final position were lost in the Italic languages, but the instances of loss and of preservation are not distributed in a clear and consistent manner either among the separate dialects or within the same dialect. Apparently different positions in the sen- tence favored one or the other treatment, depending on the initial sound of the following word, and of the doublets which thus arose one or the other was gener- alized for the form in question in all situations. For the most part the type of loss here described differs from syncope both in distribution and in origin and is usually treated separately.
a. -a. is preserved in the voc. sg. of the a-stems Prestota, Serfia, louia, Tursa, the shortness of the a being shown by its contrast with the nom. in -o -u < -a.
b. -e is lost in neip neip: 0. neip etc. but L. neque; ap (but usually ape etc.): L. atque but also ac; erek and other pronominal forms made with enclitic equi- valent to L. -ce.
c. -e is preserved in the voc, sg. of o-stems Sage, Grabouie, etc.
d. -I is regularly lost in the 3 sg. and 3 pi. active primary personal endings: est est: L. est, but sari, Skt. asti; habiest, fefacust, etc.; -f is lost also in et et: L. et but en, Skt. ati; post: 0. pust, L. post, all from ^posti (like ^^anti); etc.
48 II. Grammar
e. f is regularly preserved (as -e -e) in abl. sg. of consonant-stems capirse kapif e, natine, etc.; in nom.-acc. sg. neut. uerfale and most other neut. i-stem nouns and adjectives; in ote ute: 0. auti but L. aut; pufe pufe^: L. ubi but 0. puf.
f. -6 is preserved in supu IV 17, if von Planta (I, p. 567) is right in equating it with V7z6, Skt. upa, L. sub. See note on IV 17 for the two possibilities.
g. For the loss or preservation of -u there is no conclusive evidence.
^ It is not certain what the final vowel of these forms originally was; the Oscan and Umbrian forms may have had suffix -dhe: Skt. ku-ha, or -dhi: no-di; L. ubi may be < ubi < OL ubei.
32. Samprasarana^ is the name used to designate the change of a semivowel to the corresponding vowel when the vowel which originally followed is lost by syncope. The word affected thus suffers no syllabic loss. The phenomenon is treated here for convenience, instead of being made a part of the history of the consonants, because of its close connection with the syncope treated in the fore- going sections and because of the vocalic nature of the sound which results. In principle the semivowels capable of undergoing the change are y, w, r, I, m, and n ; in Umbrian the chief examples involve y and r.
a. In the nom. and ace. sg. of stems in -yo- the o is syncopated, whereupon the y becomes i: nom. Vuvgis: 0. Luvkis, but L. Lucius; ace. Fisim < *Fisyom ; nom. arsir if equivalent to L. alius, alis,'^ the view preferred here; but the alter- native view has considerable support; see on Via 6. tehtefim also probably belongs here, since it is much more easily derived from a stem in -dyo- or -lyo- than in -di- or -li-.
b. If Teteies is equivalent to L. Tetteius, both from ^Tetteiyos, e in the final syllable is a rare spelling for i developing from the y.
c. iveka iuenga: L. iuvencas, with iv- < yuw-, should be regarded as an example of samprasarana in the initial position.
d. If a vowel is syncopated after r, the r assumes the function of a vowel and appears in the standard orthography as er er: ager^ < *ag^s < *agros: L. ager dyQog ; pacer < *pak^s < *pakris; enclitic (and therefore unaccented) postposi- tion -per < *pf < *pro: L. pro.
e. Samprasarana of n has occurred in the development of Padellar < *Paden- Ids < *Patnlds < *Patnolds, if we accept the usual view as to its etymology.
^ A Sanskrit term, literally ' a drawing asunder. '
2 Sail, apud Gharis., Gramm. Lat. (Keil), I, p. 159, 31.
^ Minor inscr. no. 355 Go. (from Assisi).
33. Contraction and Hiatus.
a. UnUke vowels are left uncontracted, the resulting hiatus being frequently indicated by /i h: stahu^ but L. sto; ahesnes: L. ahenis; etc.
A. Phonology 49
b. Like vowels are contracted. The known instances mostly involve vowels left in juxtaposition after loss of ij (53 b): nom. pi. pacrer < *pakreyes: cf. L. acres; second-conjugation verb forms of the type of tursitu tusetu< ^torseyetod: cf. L. terreto; etc.
c. In the dat. sg. and dat.- abl. pi. of r/o-stems the close e < oi contracts with the preceding y, but this contraction is sometimes neglected, or at least is not consistently shown by the speUing: sg. Grabouei Krapuvi, luvi, but also luvie, etc. ; pi. Atiersir but also Atiersier, etc. The contraction is undoubtedly connected with the raised quality of the vowel following the y, and it has a parallel in dur < *duos (cf. L. duo), where contraction took place after 6 had been raised in the direction of u. Significantly, the contraction never occurs when i is followed by the open e in voc. sg., or by e e < ei in gen. and loc.sg., or by e e < ai in dat.- abl. pi. of the first declension.
^ Minor inscr. 355 Co.
34. Elision. On the loss of a final vowel before the initial vowel of the follow- ing word it is impossible to lay down any general rule, but the loss of final short vowels treated above in 31 may be in origin a manifestation of this tendency, secondarily extended to situations before an initial consonant. The loss of final vowel + ni (cf. L. animadverto < animum adverto) may be exemplified by eitipes if < *eitom hepens (124 c), but observe the failure to show elision in vasetomest, pesetomest, etc. neifhabas, if < *nei arhabas, shows loss of the initial vow- el after the preceding final.
35. Ablaut. The variations of vowel in etymologically related forms, known collectively as ablaut or vowel-gradation, are older by several millennia than the weakening and syncope of the Italic languages. Their origin was partly similar — the effect of a stress accent — but they must be assigned to a stage no later than proto-IE, since every language of the family has inherited them to some degree. During the period when the ablaut-variations were developing every syllable of a word was subject, at least in theory, to this type of varia- tion, but the leveling of related forms has greatly reduced the actual scope of the phenomenon. In the Italic languages particularly ablaut has been reduced in importance, as a comparison of the Latin tense-system, for example, with that of Greek or any Germanic language will quickly show.
The study of Hittite and the recognition of its relationship to the Indo-Euro- pean languages have led to an increased understanding of the pre-history of the IE vowel-system. Even before the decipherment of Hittite, Ferdinand de Saussure (Memoirc sur le systeme primitif des voyelles dans les langues indo- europeennes, Leipzig, 1879) had assumed two "coefficients sonantiques" in order to explain the long vowels in such forms as tardfti, ri6r]/j,i. Subsequently Hittite was found to contain two sounds customarily transcribed h (voiced)
50
II. Grammar
and hh (voiceless). Correspondences like Hitt. lahha- 'campaign, war': Aao'g, Hitt. mehur ' time, occasion ' : L. metior suggested that IE e and a arose when an earlier short normal-grade vowel was followed by a laryngeal, so that de Saus- sure's theory in its essential features was vindicated. Sturtevant in The In- do-Hittite Laryngeals (Baltimore, 1942) recognized four "laryngeal" phonemes ' [ X y (cf. p. 22 with correspondences between his own and other systems, some of which recognize only three laryngeals). The most important features of his doctrine, so far as IE vocaUsm is concerned, are: x and ' imparted an a-color to a contiguous e-vowel, whereby a and a. came to exist as distinct phonemes after and only after the disappearance of the a-colored laryngeals; the vowels e, a. in the "long-vowel" series (to be distinguished from e, a as lengthened grade in the e-, a-series) arose when normal-grade e was followed by a laryngeal of e- or a-color, which in turn was followed by a consonant; in zero-grade forms when the e, a disappeared, the laryngeal itself became a, which subsequently alternated as zero-grade with e, a after these long vowels had arisen from the combination of e + laryngeal. Some scholars regard o, o, when not in qualita- tive ablaut-variation with e- and a-vowels, as reflexes of e in contact with the laryngeal y (cf., for example, hehmann, Proto-Indo-European Phonology [Austin, Texas, 1952], pp. 92, 93, 96; Messing, Harvard Studies in Classical Philology LVI-LVII [1947], pp. 217-220). Others admit o, 6 only as ablaut-variants of e- and a-vowels (cf. Sturtevant, Lang., XIV [1938], pp. 104-11).
The implication of all this is that the six ablaut-series presented below were in origin one. The fundamental distinction was between accented syllables con- taining e and unaccented syllables containing no vowel unless a liquid, nasal, semivowel or laryngeal assumed vocalic function. The differences of quality and quantity in the vowels characterizing the six ablaut-series were then large- ly the result of the quality and position of contiguous laryngeals. These laryngeals, however, are not attested in the IE languages which were known before the decipherment of Hittite, while on the other hand there are nume- rous instances of vowel-correspondences of the type of dexa, L. decern, Goth. taihun; ayqag, L. ager, Goth, akrs; edrjxa, L. feci; Dor. /^drf]Q, h. mater, Skt. mdtd, etc. It is safest therefore in describing the origin of the Italic vowels to begin from the fully developed system outlined in § 5, the vowels of which can be arranged in the six ablaut-series shown below.
normal grade reduced and lengthened grade
e-grade o-grade zero grade e-grade o-grade
|
e |
0 |
i. O |
e |
6 |
|
ei |
oi |
i |
ei |
6i |
|
eu |
ou |
u |
eu |
6u |
|
er |
or |
7 |
er |
or |
|
el |
ol |
I |
el |
61 |
|
em |
am |
m |
em |
6m |
|
en |
on |
n |
en |
on |
A. Phonology a-grade o-grade a-grade o-grade
51
|
a |
0 |
b |
O |
a |
6 |
|
0 |
b |
o |
0 |
normal grade reduced and
e-grade o-grade zero grade
"«
|
e |
0 |
d |
O |
|
a-grade |
o-grade |
||
|
a |
0 |
d |
0 |
|
6 |
d |
o |
The table does not show reduced grade as distinct from zero grade, "shwa secundum" (written 6 ; cf.5b,16b) is included, however, to account for unstressed vowels in syllables where complete loss of the vowel would lead to unpro- nounceable clusters of stops. In the long-vowel series it is often customary to distinguish between a reduced grade with d (e. g. L. status, Skt. sthitds) and a zero grade with syllabic loss (e.g. Skt. dadhmds) which vary under conditions which are not fully clear. The two are not separated in the table, but see n. 12 in regard to purdouitu and dersa tefa.
a. A full account of ablaut should include the combinations ye, yo, we, wo, re, ro, etc., since here also the semivowels may appear as vowels in zero-grade forms (e.g. Skt. comparative nav-yas, superlative nav-is-tha-), but actual instan- ces are far less numerous than those showing the variations ei: i, eu: u, en x, etc.
b. In the a-series a takes the place of e as normal grade and a of e as lengthened grade. In the o-series there is no distinction between normal and o-grade nor is there any qualitative distinction in the lengthened grade. In the a- and o-series as well as the e-series it is necessary to take account of the subdivisions in which the vowel is followed (or preceded) hy y w r I m n, but the combinations ai, au,
52
II. Grammar
e-senes
I lo
too
I 2
13
|
(0 1 |
00 1 |
||||||
|
prusikurent 0. triibum |
|||||||
|
tapistenu( ?) L. (rabs |
•43 |
rufru L. ruber L. iuvenis |
P O Qi 1h ft 5^ h-s ft Q, .i4 S |
o |
a •2 •"2 |
||
|
L. modus dupursus ^ sukatu pruzuf e( ?) |
3 to o |
•■—' CO 3 1^ o 3 o |
3 o |
1-4 Q 2 o P 555 to |
a 0 a. J |
||
|
L, tepor mefs L. bipedibus L. in-sece L. prae-side tremnu |
-u a d |
-w' CO 3 -Si 2 .2 |
perse lo couertu L. circulus |
3 2 ?» =0 |
u 0 g a |
tj C!
A. Phonology
53
etc. can be easily inferred from their similarity to the corresponding c-forms, and the o-forms are the same as those found under o, o in the e-series.
c. The origin of the lengthened grade is partly obscure, but the e, a, 5 probab'y came from e, a, 6 by compensatory lengthening under certain conditions involv- ing loss of the syllable immediately following. The lengthened-grade vowels, from the historical standpoint, should be carefully distinguished from the vowels of the long-vowel ablaut-series, which owe their length to the loss of a following laryngeal: for example L. pes, Dor. tzwq represent lengthened grades in relation to L. pedis, Gk. nodog in the e-series, but the long vowels in ridrjjui, didcofii represent normal grades in their respective long-vowel series. The long-vowel series naturally do not possess lengthened grades distinct from their normal grades.
d. The tables herewith show some of the principal examples of ablaut-varia- tion between Umbrian forms or between related forms in Umbrian, Oscan, La- tin, and Greek. The series are kept distinct, but it has not seemed necessary in the table to show the further subdivision into series with ei, eu, er, etc. The variations exemplified are in the root-syllable except when otherwise indicated in the notes.
a-senes a-grade
o-2rade
reduced and zero grade
a-grade
o-srade
|
angif L. hasta axQog |
L. uncus hostatu ocar |
L. acer |
dy.a)>cr] |
o-series : Umbrian shows no good examples of o-grade forms not in ablaut-va- riation with e- or a-grade forms in the same language or in other IE languages. In fact, the right to recognize such a series at all ("non-apophonic o") is questioned by many scholars at the present time. But see below in the o-series, where forms derived from the root *d6-, doiv 'give' appear to contain "a non-apophonic" o- vowel.
e-senes
e-grade
o-grade
reduced grade zero-grade
|
fesnaf-e 0. fiisnu fetu " erietu(?)ii |
L. fanum ^ fa9ia L. aries |
54
II. Grammar
a-senes
a-grade
o-grade reduced grade zero-grade
|
stahmei |
staflarem |
o-series
d-grade
reduced grade zero-grade
|
dumim^^ |
purdouitu |
dersa tera |
^ Or possibly o-grade, as in Goth, fotus.
* ei is not a direct survival of IE ei (see 20 a, 46 j). 3 See 21 a.
* It is not certain whether these forms show the grade eu or ou, since the two became merged as ou in pit. (23).
^ It is not certain whether these forms show the grade or or r, since the two became merged as or in pit. (17 a).
* The gradation is in the suffix trltor, etc.
^ From a disyllabic base, perhaps *mele-, with comoltu < *kom-moldt6d and comatir < *-Tnal- tois < *-mltois (18 a), 8 Suffix -ydnl'in- (81 b).
* From *fdsnom < *dhdsnom.
10 e-grade whether we derive from the unextended root *dhe (cf. ri-drj-fii) or from the extended *dhe-k (cf. L. feci).
^ The only evidence for long initial e here is Lith. erytis 'lamb'.
" dunum (inscr. no. 352 Co. from Todi): L. donum. pur-douitu shows reduced grade of dou, a diphthongal form of the root *dd; o instead of a < a is difficult but may represent the same type of qualitative leveling seen in didofxev beside didcofii. Pf.j)cp\.pur-ditom < *-d'"'ltom has zero-grade of *d6u, as in L. duint, etc. dersa tef a < *deddt has zero -grade of *dd + a, sign of the subjunctive.
36. General Survey of the Consonant System. In describing the origin, history, and etymological correspondences of the Umbrian consonants the fol- lowing pIE phonemes are to be assumed as resting on the surest evidence:
STOPS
Voiceless
Voiced
Voiced Aspirate
Nasals
SEMIVOWELS
LIQUIDS
SIBILANT
|
ibial P |
Dental t |
Velar or Palatal k |
Labiovelar |
|
b |
d |
9 |
gU, |
|
bh |
dh |
gh |
g^h |
|
m |
n |
||
|
y |
w |
||
|
I |
r |
A. Phonology 55
a. The laryngeals are not included in the table. For their part in the develop- ment of the IE vowel system see 35. So far as the consonant system is concerned the rare phoneme b is sometimes regarded as a reflex of p followed by the voiced laryngeal y (cf. Sturtevant, Indo-Hittite Laryngeals, p. 87). For the plain velar stops see 45.
b. The voiceless aspirates ph th kh k^h, which are commonly assumed in addi- tion to the voiceless, voiced, and voiced aspirate stops of the various classes, have been omitted from the table here given. They are normally included in the pIE phonemic system in order to account for sets of related words which show ph, th, etc. in Indo-Iranian but which show reflexes of the plain voiceless stops in Celtic, Germanic, and Balto-Slavic and of either the plain voiceless stops or the voiced aspirates in Italic and Greek (e.g. Skt. tisthati, Gk. tarrj/ui, L. sisto, Goth, standan; Skt. nakhas, Gk. 6Vu|, ovvxog, OE nxgel; etc.). Some recent scholars have derived the voiceless aspirates from combinations of plain voiceless or voiced stops followed by any one of the three voiceless laryngeals ' ! X (in Sturtevant's system). Certain variations between aspirated and unaspi- rated stops, or their reflexes, are then traced to doublets showing the laryngeal immediately after the stop (leading to aspirates) or separated from the stop by an intervening vowel (leading to unaspirated stops). Cf. Sturtevant, Indo-Hit- tite Laryngeals, pp. 83-86; Lehmann, Proto-Indo-European Phonology, pp. 80- 84. The Umbrian forms in which IE voiceless aspirates have been assumed are extremely few: 1) with the labial ph: afepes if we derive from root *ep/i- with Devoto {T.I., p. 201; see Lexicon), but this is very uncertain; 2) with the dental th (after von Planta, I, pp. 470-1): itek if it contains the same suffix as Skt. katham, yathd; urfeta if, with L. rota, it is connected with Skt. rathas; uatuo vatuva if connected with Skt. kvath-, but this is extremely uncertain.
c. For the question of the plain velar stops, their relation to the palatals and labiovelars, and their significance for Latin and the Italic dialects see 45.
d. Strictly speaking, the nasals m, n, the liquids /, r, and the semivowels y, w are allophones of the vowels m, n, I, r, i u respectively, since the conditions under which m or m, r or r, i or y, etc. can occur are mutually exclusive. For the sake of clarity, however, separate symbols are used as if m and m, i and y, etc. were distinct phonemes. The symbols g and n, however, are not used, since in pIE and the early IE languages in general they are merely allophones of n occurring before (labio)velar and palatal stops respectively. (Contrast their phonemic value in Eng. lawn: long, sinner: singer).
e. Some recognize z, a voiced sibilant in contrast to the voiceless sibilant s. Whether or not it is justifiable to admit 2 as a separate phoneme, it is adequate for our purposes to treat z as an allophone of s arising when a voiced stop imme- diately follows, as in zero-grade derivatives of the root *sed- ' sit ' (for Umbrian forms see 29 b, 61 c).
56
II. Grammar
37. History of the stops in Italic and other IE Languages. The fol- lowing is a simplified diagram showing the history of the stops in some of the principal IE languages. The Umbrian and Oscan sounds are shown only as they appear in the Latin alphabet, except that the Umbrian characters f, k, and g are included. The voiced stops arising in Gothic under the conditions of Verner's Law are shown at the right of the voiceless spirants, and the products of pala- talization before front vowels in Sanskrit, Lithuanian, and Old Church Slavic are shown at the right of the k and g symbols, but the unaspirated stops arising in Greek and Sanskrit by Grassmann's Law are omitted.
Umbrian Oscan Latin Greek Gothic pIE Lith. OCS
Skt.
Labials
Dentals
Palatals or Velars
Labio- velars
|
p |
P |
P |
71 |
f b |
P |
P |
P |
P |
|
b |
b |
b |
i5 |
P |
b |
b |
b |
b |
|
f |
f |
fb |
<? |
b |
bh |
b |
b |
bh |
|
t |
t |
t |
T |
P d |
t |
t |
t |
t |
|
dr rs |
d |
d |
d |
t |
d |
d |
d |
d |
|
f |
f |
fd |
6 |
d |
dh |
d |
d |
dh |
|
kegs |
k c |
c |
X |
h 9 |
k |
k s |
k c c s |
k c s |
|
g i |
9 |
9 |
y |
k |
9 |
9 z |
g z z |
9 J |
|
h |
h |
h |
X |
9 |
9h |
9 z |
g z z |
gh h |
|
P |
P |
qu |
71 T X |
h. |
J^W |
k |
k c c |
k c |
|
b |
b |
V |
^ <5 y |
9 |
gW |
9 |
g z z |
9 1 |
|
f |
f |
t |
(p Q X |
9 |
g^h |
9 |
g z z |
gh h |
a. When Umbrian is compared with Latin, two striking differences become apparent: 1) in Umbrian, with Oscan, the voiced aspirates become voiceless spi- rants both initially and medially, while in Latin they become voiceless spirants initially but voiced stops medially; 2) in Umbrian, again with Oscan, the labio- velar stops become labial stops, but in Latin they mostly remain unaltered as velar stops with lip-rounding, at least where the voiceless stops are concerned.
b.When the table is examined as a whole, it is seen that in the group of langua- ges at the left the "palatals or velars" have generally remained as stops, while in those at the right they have largely been changed to sibilants.^ At the same time the labiovelars have generally remained as stops, partly of labial character, in the languages at the left, but have lost their lip-rounding in the languages at the right and appear as plain A^- or ^--sounds, or as sibilants resulting from secondary palatalization of these sounds in the separate languages. This diver- gence in the treatment of the palatals and labiovelars is the basis of the familiar division into " cen /um-languages" and "safam-languages ", which is discussed more fully in 45.
A. Phonology 57
* The Lithuanian, Slavic, and Sanskrit k and g shown in the table as reflexes of pIE k g gh belong to etymological groups like xgeag, L. cruor, Lith. kraujas, Skt. kravis, for which earlier works generally assumed "plain velars" as distinct from both palatals and labiovelars. See 45.
38. Labial Stops. pIE p.
a. p normally remains as p p initially and medially: patre: L. pater, Tiartjg; peri persi: L. pes, novg; pir: nvg, Eng. fire; pequo: L. pecu, Goth, faihu, Skt. pasu; etc.; supei-: L. super, vtieq; dupla: L. duplas; etc.
b. The cluster pt (including pt < bt, bht by very early assimilation), which remained unaltered in Latin, became ft in proto-Oscan-Umbrian, and the result- ing / was further altered in Umbrian to h, the further history of which was par- allel to that of the h in the cluster ht< kt { 46i): screhto: 0. serif tas, L. scriptum; sutentu < *sup-tendet6d; probably also imv. hatu hahtu, from a form with se- condary pt after syncope of the intervening vowel, and perhaps subolu subahtu.
c. For the assimilation of p before s see 61 d. For the assimilation of p before m see 61 e.
39. pIE b, which was a very rare phoneme, generally remains unaltered as b b (for representation by p, which does not there indicate a voiceless stop, see 3 d): trebeit: cf. 0. triibum, L. trabs, Goth, paurp; habe habe: cf. L. habeo: the phonology of the various forms of this verb in the Italic dialects presents a complex problem (122 e), but the Umbrian forms containing b b must be re- ferred to an original b, since bh would not provide a satisfactory solution for Oscan-Umbrian, nor g'" for Latin; kebu: L. cibus, but because of the k unpalata- lized before e the Umbrian form is under strong suspicion of being a loan; if it is a Latin loan, the b could be from bh; afputrati: L. arbitratu, a reasonably sure equation; the ultimate etymology is unknown, but the U. p and L. b must be from pIE 6 unless one or the other of the forms is a dialect borrowing.
For the assimilation of b before n see 61 e.
40. bh normally becomes / f in Umbrian and Oscan not only in the initial position, as in Latin, but also in the medial position, where Latin has b with loss of aspiration: fertu fertu: L. ferto, (psgerco; fratrom fratrum: 0. fratrum, L. fratrum; futii futu: cf. 0. fust, L. fui; etc.; trifo but L. tribum; iefe tefe: O.tfei but L. tibi; alfu: aXcpovQ but L. albus; etc.
For the change mbh > mf > mb see 60 d.
41. Dental Stops. pIE t.
a. / initial and medial normally remains as t t: totam tuta: 0. touto, Goth. piuda; tases tagez: L. tacitus; tenitu: L. teneto; trif tref: L. tris, tqeIq, Eng. three; etc.; fertu fertu: L. ferto, qjEQerco; etantu: 0. etanto, L. tanta; et et: L.
58 II. Grammar
et, en, all from *eti; 3 sg. and 3 pi. verb forms in primary tenses, with -t< -ti: est: L. est, sari; sent: cf. L. sunt, OCS sqti-, etc.
b. t when originally final became d in Oscan and Umbrian and subsequently disappeared in Umbrian: 3 sg. verb forms in secondary tenses: dede^: 0. deded, L. dedit; fagia: 0. fakiiad, L. faciat.
c. Final nt nt in pronominal and adverbial forms containing the suffix -(h)ont, -font is preserved in 29 instances against 2 without the -nt -nt. Since it is alto- gether unlikely that final / would be preserved in -nt when it is otherwise lost, we must assume that a vowel, perhaps -i, had previously followed the t.
d. Initial tl, which in Latin is simplified to 1-, is preserved in the Umbrian form Tlatie, which may be cognate with L. Latium or possibly with latus< *tla- tos, pf. pcpl. to fero, tuli.
e. Medial tl became cl kl: pihaclu and other nouns made with the same suffix (75 b). 0. pestlum with stl< sktl is an exception to which there is no Umbrian parallel unless persclo is made with -tlo- rather than with -lo-.
f. An Oscan-Umbrian change tn > kn is commonly assumed as a convenient means of bringing several forms into connection with L. annus, Goth. dat. pi. apnam, both < *at-no-. The forms are 0. akenei 'in anno\ akun., both with anaptyctic vowel before n ; also U. seuacne sevakne if it contains the same root, but see on Ila 21. The change gains to some extent in plausibility through its resemblance to the change -tl- > -kl- (e above), but cannot be regarded as pro- ven. Padellar might be expected to show the same change, if the reconstruc- tion given in 32 e, 60 b, is correct. It may therefore be necessary to assume that *Patnola resulted secondarily after syncope of *Patenold.
g. Medial t is lost before k, after syncope has occurred, in percam. < '^pertikam: L. pertica.
h. For the loss in t in pusnaies see 2 i. For the change -tr- > -dr- see 60 f. For -nt- > -nd- see 60 a. For -tn- > -dn- (as a stage in the development of Pa- dellar) see 60 b, with f above. For -tt- and groups containing t with other dental stops see 44 with a, b, c, d.
^ From minor inscr. no. 352 Co. (from Todi).
42. pIE d.
a. d in the initial position generally remains as t d, the t in this case of course indicating a voiced stop (3 d). dm, tuf : L. duo; desenduf: cf. L. duodecim; destru testru: L. dextro ; ditutetu. : cf. didco/ui, L. do, dare; Di Dei: cf. OL Diovis.
b. d between vowels normally became a sound represented in the native al- phabet by the character which we write as f and in the Latin alphabet by rs (3 f). Another source of this f rs is intervocalic / (55 b), and in certain forms of
A. Phonology 59
doubtful etymology and interpretation it is not certain whether we have to do with original d or I. In both cases the change is a peculiarity of Umbrian not shared by Oscan, although the character ^ , which occurs 4 times on a single Paelignian inscription from Corfinium (216 Co.), may have indicated a similar sound. The surest Umbrian examples of f rs < d are: erse efek: 0. idic idik, cf. L. id; pirse pife; cf. 0. pidum, L. quid; persi peri: L. pede, nodi; from the same root dupursus, peturpursus: cf. L. bi-, quadrupedibus; capirse kapife: L. capide; calersu kalefuf : L. calidus = XevKoixeTconoc, (cf. Isid., Or., XII, 1, 52); serse zefef: L. sedens; dersa tef a < *didat, pres. subj.: cf. didojjui; similarly an- dirsafust atefafust; ahatripursatu ahtrepuf atu : cf. L. tripodato.
c. When f rs occurs in positions other than intervocalic, we must generally assume that a vowel was lost by syncope after the change d> r rs had been com- pleted, or that the r rs was extended by analogy from forms where it arose nor- mally. The former explanation is possible for mers raefs < *med(o)s: cf. 0. med-diss, L. modus; afpes beside the more frequent arepes. In dirstu teftu the rs r probably arose by analogy with related froms (e. g. dirsa, etc., not sub- ject to syncope), since the syncope in the imperative forms in question is assumed to belong to an earlier period than the change d > rs r, the normal development appearing in ditu titu < ^dittod < *did(e)t6d. ars- ar- in afkani, arsfertur, af peltu, af putrati can be explained as having originated in forms where the prefix was followed by a vowel, but it is difficult to separate this group of forms from L. arbiter, OL arfuise, arveho, etc., whose precise explanation is obscure, but which are sometimes regarded as dialect variants of ad-. It should be observed that in Umbrian arfertur (beside arsfertur), arueitu (beside arsueitu) rs f may be replaced by the more advanced stage r, as in the Latin forms, and also that in both languages the alteration of ad- seems to occur chiefly before labials. The postposition -af could arise normally in close juncture before a word with initial vowel ; of the 5 actual examples 3 are before vowels and 2 before conso- nants, while the variant form -a occurs 8 times before consonants and nowhere before a vowel.
d. In contradistinction to rs f non-intervocalic, we sometimes find d main- tained under conditions that seem to call for the change to rs f . In Coredier, tuder, tuderato, utur the change was apparently prevented by the dissimilatory effect of an r elsewhere in the word (r in tuder, tuderato is from s by 57 c, e; in eturstahmu etuf stamu, where s did not change to r, the change of d to rs f was not prevented). In Padellar the d may be from t (60 b), but unless the d developed so late as to escape the change to rs, it may be better to assume that we have here a case of resistance to sound-change in a divine or personal name. In tesedi tenzitim no explanation of the d is possible, since the etymology is unknown.
e. Final d is lost: abl. sg. tota tuta, etc.: 0. toutad, cf. OL sententiad; imv. fertu, etc.: cf. 0. deiuatud, OL dated. For loss of secondary -d from -t see 41 b*
60 II. Grammar
f. For the treatment of -df- see 44 b. For -df see 61 b. For -dn-, and also for -nd-, see 61 e.
43. pIE dh normally becomes / f in Umbrian and Oscan not only in the initial position, as in Latin, but also in the medial position, where Latin has d (or, under certain conditions, b): fagia: 0. fakiiad, L, facial, cf. Skt. da-dhdmi, ridrj/btr, Jiliu feliuf: L. filius, cf. d'r]h], OrjXvg; furo: L. forum, cf. Ovga, Eng. door, Russ. dvor, all based on stems with initial dhw-; uerfale: L. verbale (b < dh after r), cf. Eng. word; staflarem: cf. L. stabulum, based on suffix containing -dhl- {-QXo-) and L. h resulting from earlier proximity of /; rufru: L. ruber, sqvQqoq, Skt. rudhiras (/ in L. rufus is dialectal in origin) ; combifiatu, with / whether we accept connection with L. fido, neidoo, or the less probable one with nvvddvofiai.
For dht see 44 d. For the possible development of ndh through np to nd see 60 a, d.
44. Dental Clusters. The clusters which arose through the addition of den- tal suffixes to stems ending in dentals underwent very early alteration, probably during the period of Indo-European unity, as a result of the intrusion of a sibi- lant between the two dental stops. Consequently some works include tst, etc., among the proto-IE phonemes. Yet for the sake of etymological transparency it is best to present the combinations in their original form. Since the suffixes with which we have to deal are in almost all cases either the -to- used in form- ing the passive participle or other related f-suffixes, our dental clusters are for practical purposes limited to tt, dt, and dht. In some cases, however, the cluster is preceded by n or r or followed by r, and the normal development of the dental groups may then be altered or followed by secondary developments.
a. tt became L. ss: messus < ^met-tos, pcpl. of meto. Of the development of simple // in Umbrian there are no examples, but trahuorfi must have passed through the stages *-wrt-ted > ^-worssed > ^-uorfi (59 c).
b. dt first became tt by assimilation and had precisely the same subsequent history as original tt: frosetom, if derived like Latin iterative verbs by addition of -toi'td- to the verb-root: cf. L. fraus, fraudis, fraudare, etc.; but actually it is not certain whether the root-final was originally d or dh, since Latin medial d is ambiguous and forms outside of Italic sometimes have root-variants ending in gutturals instead of dentals; spefa < ^spend-to. through the intermediate stages '^spent-td > ^spenssd (58 c); similarly spafu (58 c with n. 3).
c. The cluster ttr, whether original or resulting from the assimilation of dtr, is represented by str in Umbrian, Oscan, and Latin: castruo kastruvu: 0. cas- trous, L. castrum, castra; despite certain semantic difficulties the equation of these forms is generally accepted; for details see Lexicon.
d. The history of dh -\- t is more complicated ; in several of the IE languages
A. Phonology 61
it shows two different reflexes distributed in such a way that there must have been two successive developments in proto-Indo-European itself, nor is there universal agreement among scholars as to which development was normal and which was the result of special factors. According to the view favored here the normal development was that one whereby a transfer of aspiration took place resulting in d^dh > zdh, then Skt, dh (or h) with compensatory lengthening of the preceding vowel, Gk. ad, Italic st. This development may be seen in U. hos- tatu: L. hastatos, cf. Goth, gazds, all from *ghodh-to-, -ta-; ufestne, adj. in abl. pi. 'sealed', if < *op-fest{o)-no- < *ob-bhndh-to-no: cf. L. offendix 'knot, band', Eng. bind, Av. basta-, Skt. baddhas, both = 'bound'. The other development Involved the analogical reintroduction of t (which was the characteristic feature of the perfect passive participle, the largest component among forms contain- ing dental clusters) into the stem, the result in Italic being ss, as from original //, dt: Fise, Fiso, if < ^bheidh-to: cf. L. fido, confisus, nsldofxai', perhaps also nesimei ' proxime', if < ^nedh-tmmed: cf. 0. nessimas, Olr. nessam; the evidence for the aspirate is Skt. naddhas 'tied'. The development of dhtr is illustrated in Oscan by nistrus ' propinquos' (same root as in U. nesimei above, with suffix -t{e)ro-; t would be preserved by either development described above); there are no Umbrian examples.^
* On the development of dh + t in general see Buck, § 138 a; Kent, Lang., VIII (1932), pp. 18-26; Kent, Sounds of Latin^, § 130, 2; § 144.
45. Palatal or Velar Stops. In Brugmann's Grundriss and other standard works on IE comparative grammar or on the historical grammar of individual languages the stops assumed for pIE have included not only a palatal series (commonly designated^, g, gh) and a labiovelar series (qu, gii, guh, or now more usually, /c"', g^, g^h) but a "plain velar" series (k, g, gh) as well. The palatal series provides for such etymological groups as: (with k) L. centum, Gk. exarov, Eng. hundred, Skt. saiam, Av. saidm-} L. decem, U. dequrier, tekvias, Gk. dexa, Skt. dasa, Lith. desimt; U. ocar, Gk. dxgig, Skt. asris; (with g) L. ager, U. ager, agre, Gk. dygog, Skt. ajras; (withg/j) U. hondomu, hondra, L. homo, humus, Lith. feme, OCS zemlja. Etymological groups containing the labiovelars are treated in 49. The plain velars were assumed in order to cover those groups which show unlab- ialized k-, gf-sounds in the western group of languages but also show A'-, g-sounds (sometimes with secondary palatalization) in the eastern languages, and conse- quently fail to fit either the palatal or the labiovelar series: e.g. (with k) L. cruor, Gk. xQsaQ, Skt. kravis; (with g) L. tego, Gk. areyco, Skt. sthagayati; (with gh) Gk. areixcD, Goth, steigan, Skt. stighnoti, OCS stignq. Yet for some time the propriety of recognizing originally separate k- and /c-series has been ques- tioned.2 Such a distinction is completely lacking not only in Greek, Italic, Celtic, and Germanic but in Hittite^ and Tocharian* as well. Probably a single /c-series split into a velar (k-) and a palatal (k-) series, the latter then becoming sibilants,
62 II. Grammar
in the dialect area in which the safam-languages originated. The conditions for such a split are not entirely clear, but must be at least partly connected with the sound following the stop, since the appearance of plain velars in con- trast to palatals is especially frequent before a and r. In dealing with Italic dia- lects, as with cen/«/n-languages generally, the question of k and k is of relatively little importance. In the present work, whether a given pIE reconstruction ac- cording to traditional practice calls for k or k, g or g, gh or gh we write simply J<, 9, gh.
^ The basis of the familiar designation "cen/um-languages" (Greek, Italic, Celtic, Germanic) and "sa^3m-languages" (Indo-Iranian, Armenian, Phrygian, Balto-Slavic, Albanian).
2 Hirt, B.B., XXIV (1899), pp. 218-291; Meillet, Introduction d. V etude comparative des lan- gues indo-europeennes, pp. 66-67; Lehmann, pp. 100-102.
^ Sturtevant-Hahn, A Comparative Grammar of the Hittite Language (New Haven, 1951), p. 55.
* Pedersen, Tocharisch vom Gesichtspunkt der indoeuropaischen Sprachvergleichung (Co- penhagen, 1941), p, 234.
46. pIE k.
a. k, if not followed by e or i, is represented by c k: kanetu: cf. L. canito; com, -co -ku: L. cum, com-; kabru kapru (with cabriner): L. caper, ON hafr.
b. k followed by an e- or i-vowel or by y became 5 s through a process of pal- atalization not shared by Oscan or by classical Latin but doubtless resembling the sound-change observed in It. cento, or in Fr. cent, from L. centum. Often s appears in place of s in the tables written in the Latin alphabet. The following is a partial list of examples: Serfer, Serfe: 0. dat. Kerri, L. Ceres; girau, give: L. cis, citra; sihitu, sihitir: L. cinctos, -is; tases tagez: L. tacitus; fagia: 0. fa- kiiad, L. faciat; tigit: L. decet; desenduf: cf. L. duodecim; curnase: cf. L. cornice; pesetom: cf. L. peccatum.
c. In the following forms with 9 s followed by / the palatalization of k was caused by e in the suffix -elo-, -kelo-, etc., the e being subsequently lost by syn- cope: arglataf: L. arculatas (L. -u-< -e- before I); tiglu, kurglasiu, preuislatu, struhgla. In uesclir the palatalization of k appears to have been prevented by the s.
d. Certain apparent cases of 9 s before back vowels result from the occasional failure to write the i which has caused the palatalization, probably indicating a partial or complete absorption of the i by the preceding consonant: vestiga beside vestigia; fagu beside fagiu; pimigate beside ponisiater; purdinsus pur- dinsust, purtingus, beside purdinsiust; tribfigu < *-ki6n beside abl. tribrisine with i.
e. Of the instances in which c k appears unchanged before front vowels most can be explained as restorations through the analogy of closely related forms
A. Phonology 63
not subject to change; so especially the forms fratreci, Pupfike, todceir, Nahar- cer, Tuscer, labuscer, all belonging to o-stems. That the absence of palataliza- tion is due to the secondary character of the i, e in these dat. sg. and dat. -abl. pi. forms (e < oi) is less likely, especially in view of the palatalized Pupfige, etc., beside Puprike. In pacer also the e is secondary (pacer < ^-pakrsK *pakris), and in any case the c can be explained by the analogy of forms where k still stood unchanged before r.
f. In several other forms, however, the explanation of the unaltered c k is not clear. The evident relationship of Acesoniam to 0. Akudunniad suggests that the e may have denoted a vowel of obscure quality without effect on the c, but the four Umbrian occurrences of the word are consistent in the use of e e. kebu, apparently = L. cibo, is under suspicion of being a loan, not only because of the k before e but also because of the medial b (39). In cehefi and kukehes the c k has been explained on the ground of secondary e< ai or through the analogy of forms with k before a (: L. candeo) or as an archaism preserved through ritualistic use, but no explanation is certain.
g. There are a few forms in which i appears in place of an expected g, but all can be explained by special causes without the need of formulating a general phonetic law. peia, which is always used in contrast with rofu, rufru 'red' of animals, must itself designate a color, presumably 'black', in which case it would be cognate with L. piceus. If contamination of '^pikyos with *nigros (L. niger) is admitted, peia may be derived from ace. pi. f. -^pigyans by 47 a. feia, semaiitically equivalent to fagia, is sometimes taken from stem dhe- (cf. ri-dr]-/Lii) as distinguished from dhe-k- (cf. L. fe-ci), but there is no positive evidence for unextended dhe- in Italic except in compounds of the type of L. condo, 0. pruffed. feia, however, may have been remodeled after imv. feitu (with normal loss of k by 46 j), according to the correspondence */era (implied by ferar): fertu, etc. usaie in lb 45 is almost certainly an erroneous spelUng of usage which occurs in the same formula in I la 44, the lateral loop of the character d being omitted.^
h. It is not certain how the group kw (as distinct from the single labiovelar phoneme k^) developed in the Italic dialects. Even if ekvine means (Via) E- quina or in Circo Equino (MuUer, p. 159), it is most unlikely that it is a Latin borrowing. We must then admit that kw was kept distinct from A:'", which became p by 49 a, or else reject the connection of ekvine with L. equinus.^ kvestur, like the official title 0. aidil, is probably a borrowing, from L. quaestor, and moreover it is uncertain whether the qu of L. quaero, quaestor, etc., is from a labiovelar or from k -\- w. In tekvias the k originally was almost certainly followed by a full u-vowel (< ^dekuwias).
i. The cluster kt (including kt from gt by very early assimilation, but exclud- ing kt resulting from syncope of an intervening vowel) became ht in Umbrian,
64 II. Grammar
as in Oscan, but not in Latin: rehte: L. rede; apehtre: cf. 0. ehtrad (both from *ek-tr- without the s contained in L. extra); uhtur: L. audor; ahtim, Ahtu, if connected with L. ago, adio, adus. That the h in such forms as these was very weakly sounded, or lost with compensatory lengthening of the pre- ceding vowel, is shown both by the frequent omission of h in the writing and by its use as a mark of vowel-length in forms where it has no etymological value. Among the forms where h is omitted are satam beside sahta: 0. saahtum, L. sandus; Speture: cf. L. spedo; petenata: L. pedinatus. For A as a mere mark of vowel-length see 26 c.
j. kt occurring after syncope of the intervening vowel has a different history in both Umbrian and Oscan from the kt of i above, implying that the syncope did not take place until after the change of the earlier kt had been completed. In Umbrian i appears in place of the k, while in Oscan the d kt remains unal- tered: aitu: 0. actud, L. agito; deitu teitu: L. didto; feitu: cf. 0. fadud, L. fa- dto.^ The phonology of ars-ueitu, ku-veitu is more complicated. Most scholars equate them with L. ad-, con-vehito: cf. Skt. vahati, Goth, ga-wigan, Lith. vezii, all from root wegh-. If we assume a development *weghet6d > *wexet6d > *we- hetod > ^wehtod, we should not expect to find ei ei in all twenty occurrences of the two forms (never with e or eh). But if the syncope occurred whi e the prior consonant was still x it is possible that -xt- may have developed as -kt-: thus, *wexet6d > *wextdd > *wekt6d > ueitu.
k. For the voicing of k after n see 60 c. For the assiimlation of k before s see 61 d. For the treatment of final k see 62 a.
^ For the three forms here discussed see R. G. Kent, C.P., XV (1920), pp. 365-6.
* In Latin kw fell together with A-'" (equus prosodically like sequor), but of course neither became p. If we hold that kii; failed to become p in Umbrian, the gentile name Epidius found on Latin inscriptions of Campania must either contain a p of foreign (perhaps Celtic) origin or else must not belong to the etymological group of equus, since it is quite unhkely that Um- brian and Oscan show a different development of kw.
3 The fact that fetu fetu (once feetu) is nearly four times as frequent as feitu feitu may be explained by the fact that the raised e of *fek{e)tdd (the same ablaut-grade as in L. feci) easily contracted with i < k before /, whereas the a of aitu and the e of the other forms did not.
47. pIE g.
a. g, if not followed by e or i, is represented by g- k (2a, 3): gomia kumiaf: L. gomia; mugatu: cf. L. mugio; agre: L. ager, dygog.
b. g followed by an e- or i-vowel or by y was converted by palatahzation to a sound represented by i i: muieto, pf. pcpl. to mugatu. In eveietu also the i is most satisfactorily explained as from g (< *e-weget6d < *e-weigd6d) but the etymology of the verb is not entirely sure.
c. A similar palatalization of g after i is sometimes suspected but cannot be
A. Phonology 65
proven: niru lib 15 may be equivalent to L. nigrum, but the exact sense of the single occurrence is not known; the adjective from the name of the city of Iguvium is written liuvina -as 7 times beside Ikuv- 14 times, and in the later tables consistently appears as liou- or lou-, yet the evidence of the Latin form Iguvium and of mediaeval Eugubium and modern Gubbio is against such a change. It has been suggested that the spelhngs liuv- liou- lou- represent a "pious fraud" on the part of the priests, with a view toward connecting the name of the town with the name of the god Jove.
d. g is lost before n in natine, L. natione: cf. yvrjaiog, Skt. jndtis; naratu, naraklum: cf. L. gnarus,yva)QiCoj. A similar loss in medial position is perhaps to be seen in conegos kunikaz if from *kon-gnigd-, cf. OL gnixus.
48. pIE gh.
a. gh normally becomes h h initially and medially: homonus: 0. humuns (nom. pi.), OL homones; hondomu, hondra, hondu: cf. 0. huntrus, L. humi, Xajjiai; her, herd, heries: cf. 0. herest, OL, ho rior, xf^^Qf^, Germ, gern; hostatu: L. hastatos, cf.h. hasta, Goth, gazds; e-hiato: cf. 0. ee-hiianasum, L. hiare, xaaycoi; pron. erihont with -ho- from the same stem as L. hoc; mehe: L. mihi, Skt. mahyam.
b. Of the representation of pIE gh by / found in L. fundo (or ferus with / < ghw) there are no sure examples in Umbrian. felsva, generally taken to be cognate with L. holera, may be a dialect borrowing.^ The / in Fondlir-e, with L. fans, is probably from dh by 43. The / in erafont is better taken as originating through reanalysis of such forms as if-ont than as a phonological variant of the h in erihont.
^ Cf. Paul, e Fest., p. 74 Li. ... antiqui dicebant ...folus pro holere...; Ernout, Les elements dia- lectaux dans le vocabulaire latin (Paris, 1928), pp. 69-70; Sciirijnen, Neophilologus, VII (1922), pp. 222 (map), 234-6.
49. The Labiovelars. The conversion of the labiovelar stops to labials has already been mentioned in 37 a as one of the most important features in which Umbrian and Oscan agree against Latin.
a. A:"' regularly becomes p: forms derived from the interrogative-indefinite- relative stem: pis: 0. pis, L. quis, rig; pir-e: 0. pid, L. quid; porse pure: cf. 0. pud, L. quod; pane: L. quamde; prepa: L. *prae-quam; panta: L. quanta; puse, as if L. *ut-s-l;^ putrespe: cf. L. utriusque.^ Particle cognate with L. -que, re, Skt. -ca: a-pe: L. at-que; nei-p: cf. L. ne-qiie. The indefinite stem and -pe are both contained in pisipumpe: L. quicumque and in {sei)podruhpei: L. utroque. Forms derived from the stem of the numeral '4': peturpursus: L. quadrupedibus, cf. 0. petora petiro-pert. Among forms derived from the stem of the numeral '5' the second p of pumperias is derived as here described;
66 II. Grammar
the initial p of pumpefias and of puntes are from k^ which came from earlier p through asimilation to k^ in the second syllable (pIE '^penk^e, Skt. panca, Gk. TievTE, Aeol. nsfjine, L. quinque, Olr. coic; cf. 105 n. 3).
b. g^ becomes b in Umbrian and Oscan, as against v, gu (the latter after n) in Latin: benus: cf. 0. kumbened, L. venit, ^aivco; berus: L. verubus; bum: /Sow, L. bovem (with dialectal b for L. v); habina: L. agnlna, cf. d/nvog < *d^v6g. umen, which is equivalent to L. unguen, must be derived from ^umben, the b having caused assimilation of n (rj) to m and having then itself disappeared.
c. Of the rather rare phoneme g^h there are no sure examples in the Itahc dialects serving to show the development in initial position. In medial position g^h appears as / in Umbrian in a single group of forms: vufru <^wog^h-{e)ro-m, vvifetes < *wog^heto- as equivalent to L. votis according to the usual interpre- tation, but see on Ila 31 ; Uofione Vufiune as a god in some way associated with vows, according to the usual interpretation, but see on VIb 19. *eugh-, the root of ev^ofiai, may be regarded as a variant of *weg^h-, with dissimilatory loss of labialization. On ninctu, whose root originally contained g^h, see d below.
d. Loss of labialization. Under certain conditions stops which are known by comparison with Greek, etc., to have been originally labio velar lost their hp-rounding in Oscan-Umbrian or in Latin or in both and show the same de- velopment as Italic k, g, i. Analogical processes have worked in both directions, sometimes producing unlabiahzed forms where p, etc., might be expected, and causing the retention or restoration of p, etc., where loss of labialization might be expected. Loss of labialization before / is common to Latin and the dialects: L. coqu.0 but coctus, inseque but insectio, U. puntes, cf. 0. Puntiis, against U. pumpefias, 0. pumperiais.^ U. prusikurent < ^-sekusent, sukatu < *so- kdtod, both regarded as cognates of evvene, eviane, L. insece, may owe their k to analogy with -/cMorms similar to L. insectio above; prusikurent may also owe its k to loss of labialization before u. subocau, suboco, whatever their precise value, are related to L. vocare, vox, fejiog, etc.; the loss of the labial element in L. vocare and in the oblique cases of vox is generally explained by analogy with vox = voks < ^'wok^'s (loss of the "^ before s), and the Umbrian forms must be similarly explained. The cluster k^t arising from syncope of an intervening vowel (g'^, g^h then being assimilated before the voiceless stop /) is treated in the same manner as earlier k^t > A:/:^iiktu< *flk"'t6d< ^dhlg"^et6d: cf. L, fivo {and figo); ninctu < *mnk^t6d< *sning^het6d: cf. L. ninguit, veicpsi, Av. snae- zaiti.
e. On the other hand the labial was preserved in puze, puse, putrespe (see a above); if we admit loss of labialization before u, then these forms must owe their p to the analogy of related forms in which p is normal. In umtu, where we should expect k as in fiktu, ninctu, the preservation of the labial (impUed by
A. Phonology
67
the m) must result from analogy with unsyncopated forms of the present stem, such as *umbo: L. unguo.
^ Loss of the initial consonant (c < A:*" before u) in Latin was caused by reanalysis of such forms as si-cut, ne-cuter, etc.
2 It is not likely that the development was *ponk'^t- > *pompt- > *pomt- > *pont-, for if we do not admit an early loss of the labial element, we should expect either mft < mpt or per- haps mt as in umtu.
3 The syncope then must have been earlier than the conversion of labiovelars to labials, for if the latter change had preceded the syncope, we should expect to find the same development as with original labial stops. But the kt did not develop early enough to be affected by the change kt > ht.
50. Continuants. The following table is designed to show in the simplest form the history of consonants other than stops; that is, of the nasals, semi- vowels, liquids, and sibilants, all of which may be grouped together under the comprehensive term " continuants ", in several of the principal IE languages, in order that the etymological comparisons in the succeeding paragraphs may be more readily understood.
pIE
Umb.
Osc.
Lat.
Gk.
Eng. Lith.
OCS
Skt.
|
m |
m |
m |
m |
A*^ |
/ni |
m |
m |
m |
|
n |
n |
n |
n |
V |
n |
n |
n |
n |
|
y |
i |
i |
i |
' , c, o ^ |
y |
}' |
i |
y |
|
w |
V u |
V u |
V |
f, o * |
IV ^ |
V |
V |
V |
|
I |
V u, l,r |
I |
I |
;. |
I |
I |
I |
r(ir |
|
r |
r |
r |
r |
q' |
r |
r |
r |
r(ir |
|
s |
s '', r ^ |
s\z^ |
s, r |
', o « |
s, r ^ |
s |
s |
s |
The characters used in the Umbrian and Oscan columns are those of the La- tin alphabet, except that v is included along with u, and f is used to denote the sound derived from intervocalic I under the conditions described in 55 b. The sign / is to be understood in the value of German /, English y. The circle O indicates that the sound under certain conditions is lost.
The table shows the sounds as they appear in the initial position, or in the medial position between vowels. Their history in the medial position postcon- sonantal-prevocalic or postvocalic-preconsonantal, or in the final position, is to a considerable extent the same, but it is not possible to present all the special changes in the table. A few of the most important are given in the notes.
1 Final m becomes v in Greek, and, in part, n in Germanic.
2 y in Greek is represented initially by ' {= h) or by C> the reason for the variation being disputed; medially between vowels it is lost; medially in conjunction with preceding conson- ants it undergoes complex changes.
68 11. Grammar
3 In Gothic and Old Norse intervocalic y and w, through the change known under the name of Holtzmann's Law, underwent strengthened articulation whereby y became Goth, -ddj-, ON -ggj-, and w became Goth., ON -ggw-.
* The labial semivowel w, written / (digamma), was preserved in many mainland Greek dia- lects, but lost in Attic and Ionic.
5 pIE r and / result in r regularly in Avestan and Old Persian and generally in Sanskrit, but I sometimes appears in Skt. for pIE I and sometimes even for pIE r. The irregularity is com- monly attributed to dialect-mixture.
* pIE forms with initial r regularly appear in Greek with a vowel before the q.
' Initial s was lost in some forms when followed immediately by p. A:, m, n, I, or r.
8 Medial intervocalic s appears as r in Umbrian, Latin, and (in part) English; as z (= voiced s) in Oscan and (in part) Gothic; in Greek it became h and was then lost, frequently with result- ing vowel-contraction.
51. Nasals. pIE m.
a. m is regularly preserved initially and medially between vowels: mani: 0. manim (ace), L. manu; Marte: L. Marti; mehe: L. mihi; nome, nomner: L. nomen, nominis, a vowel having originally stood between m and n in the Um- brian form of the genitive.
b. A case of elision of vowel -f /n, similar to that in L. animadverto < anim- (um) adverto, must be admitted in eitipes if we accept the usual derivation from *eitom hepens (see Lexicon).
c. Medial m preceded by a vowel and followed by a consonant is preserved in kuxnne < *kom-nei (not the result of syncope as in nomner in a above).
d. Medial m preceded by a vowel and followed by a consonant is partly lost (especially in the tables written in the native alphabet), partly preserved: am- pentu apentu, combifiatu kumpifiatu kupifiatu, etc.
e. Final m is partly preserved, partly lost. Some typical examples are: ace. sg. puplum puplu poplom poplo, ocrem ocre; gen. pi. fratrum fratru; infini- tive aierum afero; supine anzeriatu anseriato; postposition -kum -com rarely, -ku -co frequently; temporal adverb enom ennom enu eno enno. The 1 sg. of secondary forms of verbs must have ended in -m (cf. L. eram, 0. manaium), but the only Umbrian example is pres. subj. aseriaia.
i. For assimilation of m see 61 a.
52. pIE n.
a. n is regularly preserved initially and medially between vowels: neip: cf. L. neque; nerf: 0. gen.\)l.nerum. Gk. dv^Q, Olr. nert 'strength' ; nertru: vegreQcg, Eng, north; nome: L. nomen; nuvis: cf. L. novem; plener: L. plenis; cabriner: L. caprini.
b. Initial clusters consisting of consonant + n may lose the prior consonant but the n is preserved: natine: L. natione, cf. gnatus, gens, yvriaioQ; naratu.
A. Phonology 69
naraklum: cf. L. gnarus, yvcoQiCco; ninctu: L. ninguito, cf. Eng. snow, Lith. 3 sg. sniega.
c. Medial n preceded by a consonant and followed by a vowel is normally preserved: sesna: 0. kersnu, L. cena; tremnu abl. sg. < *trebndd: cf. 0. triibum.
d. Medial n (including n < m before dentals) preceded by a vowel and fol- lowed by a stop or by g s < A: or by s^ was sufficiently weakened in pronunciation to be frequently, but not uniformly, omitted in writing. The fact that this omis- sion of n is more frequent in the earlier than in the later tables, where it occurs only before s, must be attributed to the influence of Latin orthography, since it is hardly conceivable that the n was actually restored in speech. Among the chief examples are: iveka, iuenga: L. iuvenca; Sagi, Sansi: cf. L. Sancus; hutra hondra < *hom-terad (with n by 61 a); aterafust andersafust; ampentu ampetu < '^am-penn{e)tod < ^an-pendetod with nd < nn (61 e) and then syncope; 3 pi. forms of secondary tenses: etaians etaias, dirsans dirsas, sis sins; n before s after syncope: fons fos < '-^faunis; n by assimilation of m before s after syncope: onse uze: L. umerus with rhotacism, which in Umbrian was here prevented by the syncope. For further examples of omission of preconsonantal n see 2 i.
e. The cluster -nkt- in both Umbrian and Oscan participated in the change -kt- > -ht- and then lost the n, at least in so far as its full consonantal value was concerned: sahata sahatam: 0. saaMum, L. sanctum; sihitir: L. cinctis.
f. The cluster -ndl- lost both n and d, apparently in the sequence ndl > nnl > nl > II: apelus < *am-pendlus; similarly entelus.
g. n in the 3 pi. primary ending -nt < -nti is normally not omitted: furfant, dersicurent, pepurkurent, ■prusikwcent, furent, haburent, sent, etc.; furfaO is the only exception, fefure is best not taken as fut. pf. 3 pi. of futu, standing for *fefurent, but as a distinct verb: see Lexicon and note on I la 4. eruhu lib 22 appears without the -nt of erihont and other forms of the same paradigm, sururo VIb 48, sometimes taken as a similar shortening, is more probably for ^sururor, a dittographic form of suror, just as suront is haplologic for sururont.^
h. Final n is preserved in umen (: L. unguen), lost in nome (: L. nomen; 13 examples). In nuraem (twice in lb 17) final m probably results from assimila- tion after the medial m. The postposition -en appears occasionally with the n preserved, but mostly as -e -e, and occasionally as -em -em, the last form hav- ing apparently originated in the ace. sg. by assimilation as in numem above, and then having spread to other situations.
i. For the history of ns, original and secondary, see 58.
^ ns in Umbrian is in large part converted to /, but escapes this change when original and non-final or when n stands before s as the result of syncope (58 a, e).
2 See 112 d and R. G. Kent, C. P., XV (1920), p. 361, where it is pointed out that both sur- uront against suront and sururo against surur are favored in VIb 48 - Vila 2, which forms a passage complete in itself.
70 II. Grammar
53. Semivowels. pIE y was actually an allophone of the vowel i, which it replaced initially before vowels and medially between vowels, but its separate treatment among the consonants is justified on grounds of convenience.
a. Initial y is preserved as i i in Umbrian as in Oscan and Latin: ioiiies: cf. L. iuvenis, Skt. yuvan-; iuka: L. iocus, cf. 0. iuklei. (But i i in loui luvi is from dy-: 0. Diuvei, 0. Aiovfsi, as well as luvei, L. Diespiter, Diovis beside the forms with lov-.)
b. Medial y between vowels was lost, and if the vowels were similar they un- derwent contraction: nom. pi. of i-stems puntes, pacrer, foner: cf. 0. tris, L. hostes, all from -eyes; adjectives in-eyo-: farsio (i < e by 8 b), spantea. The presence of i i in kuraia, portaia, etc., may be explained by the analogy of other present subjunctive forms where the i is preserved after consonants (126 a).
c. Intervocalic y is not subject to loss if the prior vowel is i or an i-diphthong. Among instances of the latter type may be cited pernaiaf, pustnaiaf, Teteies.^ The instances of y retained, as a glide, after vocalic i are sometimes difficult to distinguish, because a single character may stand for iy or for the conson- ant y alone. In pIE iy and y appear to have stood in complementary distribu- tion,2 but in Italic the old distribution has been effaced. However, in those forms written in the native alphabet with ii (triia, heriiei, etc. ; see 2 h) the value of the second i is unquestionably consonantal.
d. y preceded by a consonant and followed by a vowel is partly preserved, partly lost after imparting to the consonant a palatal quality. In forms like fagiu fagu, Rupinie Rubine, spinia spina, vestigia vestigam the coexistence of spelhngs with and without i i shows that the sound was consonantal and was at least partly absorbed by the preceding consonant which it had palatalized.
e. In kurglasiu, plenasier, urnasier, sestentasiaru the failure of the s to change to r (57 c, d) is often taken as evidence of a consonantal rather than vocahc value for the i.
i. If y is preceded by a consonant and the vowel following suffers syncope, the y itself may become i by samprasarana. For details and examples see 32 includ- ing a. In iuenga iveka: L. iuvenca also the i i must have become vocahc; at least this is more plausible than the assumption that iv stands for yuw-.
^ It might conceivably be argued that y has been lost in these forms and that the written i belongs in the diphthong. But in Oscan spellings of the type of Pumpaiians, vereiiai are common, and examples of L. Maiia, eiius, etc., are known. In the treatment of intervocalic i in general Umbrian, Oscan, and Latin agree. It is therefore better to assume that i in pernaiaf, pustnaiaf does duty for the y and also for the second element of the diphthong ai, which before y then escapes the usual change to e (19).
2 Probably under the conditions familiarly known as " Sievers' Law": iy if preceded by two consonants or by a syllable containing a long vowel or diphthong, otherwise y.
A. Phonology 71
54. pIE w was an allophone of u and replaced it precisely as y replaced i. Moreover the old distribution of uw and u has been effaced and it is not always possible in a given instance to determine whether u indicates the consonant or the vowel followed by a glide, but in the native alphabet uv is employed in a manner which parallels the use of ii.
a. Initial w is preserved as v ii in Umbrian as in Oscan and Latin: uiro: L. vir, cf. Skt. vlras, Lith. vyras; uerfale: L. verbale, cf. Eng. word; uirseto: cf. L. videre, (f)idsiv; vurtus: cf. L. verto, vorti, Skt. vartate; etc.
b. Intervocalic w is preserved as v u in Umbrian as in Oscan and Latin : avif auif: L. avis; uve oui: L. ovem, Skt. avim. If w is preserved as a glide after u, the fact is indicated in the native alphabet by means of uv: tuver duir: cf. L. duo; kastruvu castruo.
c. Medial w preceded by a consonant and followed by a vowel is to a consi- derable extent preserved: sve sue: 0. svai; arvia aruio (in the older tables 19 instances of arv-, with aruv- only once); felsva; tesvam dersua; saluo (24 times salu-, only twice saluu-; not represented in the native alphabet).
d. But w after certain consonants, especially labials, is lost: subocau < *sub- wok-; prufe < *pro-bhwed: L. probe, cf. Skt. prabhu-; perhaps also sopa, supa: L. suppus, if from *supwo-.^
e. w after d does not cause alteration of the d, in contrast to Latin, where dw > b: difue: cf. dtcpvrjq, L. bis, bi-; purditom < *-dwitom with zero-grade as in L. duint; probably also dia < '^dwiydt.
f. Initial swe- suffers loss of w but with alteration of the e, as in Latin: sonitu: cf. L. sonare, Skt. svanati, Olr. sennaim. (In contrast to the Umbrian and Latin development cf. 0. sverrunei, Eng. swear.)
g. seritu with anseriato (cf. L. servare and servire) and caterahamo (cf. L. caterva) appear to show loss of w after r, but it is not possible to state the condi- tions with precision.
^ Since gemination is scarcely ever expressed in writing in tlie Iguvine Tables, it might be supposed that w was assimilated to the preceding consonant, actually producing -pp-, etc., but the evidence of O. fufans amprufid, L. probus, aperio, etc., whose history is parallel, rather favors the view that w has been simply lost, and this despite L. suppus.
55. Liquids. pIE /.
a. Initial /, which is preserved in Oscan and Latin as in the majority of IE languages, is nowhere written on the Iguvine Tables.^ On the other hand there are several virtually sure instances of v- u- < /-: vapefe, napef-e: L. lapide, (in) lapides; vutu < *low{i)t6d: cf. L. lavare; possibly also Vuvgis: L. Lucius. Other alleged examples are quite uncertain: see uocu-com and anderuomu in the Lexicon.^
72 II. Grammar
b. Intervocalic I is partly preserved, partly changed to f ts. The conditions for the latter change are somewhat uncertain because of the scarcity of data, but apparently before e, y, and iy I became d by a sort of palatalization, and then participated in the change d > r rs (42 b). Before back vowels and before e and preconsonantal i the I was preserved. Examples without change are: ulo< *dl6 (or double I as in L. olle, ilk'?); uerfale, i-stem with -e< -f: L. verbale. Examples with r rs < I: karetu carsitu, where e i = e as if to L. *calere for ca- lare; uretu: L. ad-oleto; famefias: L. familiae, cf. 0. famelo. The preservation of I in presoliaf-e and feliuf filiu probably results from the analogy of related forms with / before back vowels.^
c. I preceded by a consonant and followed by a vowel is preserved: plener: L. plenis; dupla: L. duplus.
d. I preceded by a vowel and followed by a consonant is partly preserved: alfir: L. albis; pelmner < ^pelpmnneis (?). But / is lost before t: muta, motar: L. multa, < ^molktd, with secondary -It- after loss of k; comatir pcpl., but imv. comoltu with preservation of I, which was not in contact with the t until after syncope.
e. Of original final I there are no true cases: tigel is from *-k{e)los; sumel is from *someli.
^ There are a few occurrences on minor Umbrian inscriptions : La (abbreviation of a praeno- men) in 353 d Co., L (abbreviation of a praenomen) in 354 Co.
^ The surest examples of the change all have a back vowel after the v- u-, and the quality of the vowel may have induced a back pronunciation of the I, leading to w. It is by no means cer- tain that the change took place before e and i, but there is no good evidence, positive or nega- tive. The alleged cases are treated by von Planta, I, pp. 287-9; add Muller, p. 237 (veskla < -*les-tlom).
3 See Lang., XXV (1949), pp. 395-401, for an attempt to formulate the conditions for the change I > r rs.
56. pIE r.
a. r is preserved initially and medially before vowels: rehte: L. rede; rufru: L. rubrum, eqvOqov; uhturu: L. auctorem; trif: L. tris; agre: L. agri; etc.
b. Medial r before consonants is in general preserved: porca: L. porcas; term- nom-e: L. (in) terminum, cf. 0. teremenniu (with anaptyctic e after the r). But the omission of r before s in both alphabets is frequent, though irregular, indicating a weakening or partial loss in pronunciation: sesna, but gersnatur: 0. kersnu, L. cena, cenati; farsio fasio: L. farrea. See 2 i.
c. Final r is preserved in r-stem nouns, perhaps through the influence of the intervocalic r of the oblique cases: pir: nvq; uhtur: L. aiictor; arsfertur; etc. But loss of -r is frequent in passive verb forms: emantu (but also emantur); ostensendi: cf, L. ostendentur; tefte: cf. L. datur; etc.
A. Phonology 73
d. For r which becomes final as the result of syncope followed by samprasa- rana see 32 d. This r is normally preserved, but against over 100 cases of the postposition -per -per (< pro) there are 3 occurrences of -pe.
57. The Sibilant. pIE s.
a. Initial s is preserved when followed by a vowel and, at least in part, when followed by a consonant: sacra: L. sacras; sersi: L. sede, cf. sdog; sim: vv, cf. L. suem; screhto: L. scriptum; statita: L. statuta.
b. s before /, m, n both initially and medially was normally preserved in Umbrian and Oscan in contrast to Latin, where it was lost with compensatory lengthening of the preceding vowel, if medial. There are no examples of initial si- in Umbrian, but Oscan has slagim 'boundary-road', cf. Olr. slige 'street'; medially, disleralinsust if < ^dis-leisdlinkyust; cf.L. dl-ligo< ^dis-l-; smursim-el (etymology uncertain); pusme, esmei: Skt. kasmai, asmai, cf. Pael. prismu, L. primus; snata: cf. L. no,nare, Skt. sndti; fesnaf-e: 0. fiisnam, cf. L. fdnum < *fdsnom. But s- is lost in ninctu: L. ninguito, Eng. snow, with s indirectly attested in Gk. dydvvKpog and in Epic lengthening before vlcpa, vELcpsi. The treatment of s in these initial groups is believed to depend partly on the general- ization of doublets arising under different conditions of juncture; cf. reyoq, areyog, jjllkqoq: a/nixgog.^
c. Intervocalic s in Umbrian as in Latin underwent the change to r familiar- ly known as rhotacism. The change must have been preceded by a voicing of the intervocalic s, perhaps during the period of Italic unity or in any case pro- bably very early. Oscan never passed beyond this first stage, and the voiced s is indicated by z in those Oscan inscriptions which are written in the Latin al- phabet.2 Umbrian examples of medial rhotacism include: kuratu: L. curato, cf. Pael. pf. 3 pi. coisatens; gen. pi. forms pracatarum, etc.: cf. L. feminarum but 0. egmazum; fut. and fut. pf. forms staheren, furent, etc.: cf. 0. censazet, tri- barakattuset, all from vowel-stems + -sent; pari, nom. pi., < *p6s + enclitic i; demonstrative pronoun ere{c): 0. izic.
d. But the Tables show a number of occurrences of s 5 not changed to r. Some of these exceptions are more apparent than real, since the s s may stand for ss, which is not subject to rhotacism; so, for example, demonstrative pron. in abl. sg. essu esu with ss < ks: 0. eksuk; intensive pron. esuf: 0. essuf, as- sumed to be cognate with L. ipse, ss < ps; osatu: cf. Pael. impf. subj.pass. 3 sg. upsaseter, L. operari, with syncope in the dialect forms before s could become r; uasor, unless s is preserved after the analogy of nom. sg. *uas, may be from a stem *wates- with ts assimilated after syncope: cf. L. vassa in Plant., Merc, 781 in Cod. Ambr.; in pisi the preservation of s before the vocalic enclitic, in contrast to puri (c above), is evidently induced by the simple pis; in kurglasiu and other forms with the suffix -dsyo- the preservation of the s may be explained
74 11. Grammar
by the supposition that i is the consonant y rather than the vowel i: of. 0. kers- snasias (not *kerssnasiias) '^cenariae', but L. denarius; asa: 0. aasai, aasas, in contrast to the rhotacized L. ara, is especially difficult, but the preservation of the Umbrian s may be a ritualistic archaism; so possibly also in eikvasese.
e. In contrast to both Oscan and Latin the later Umbrian of Tables V, VI, and VII has rhotacism in final as well as medial position: tuder (s-stem, as shown by etufstamu < ^ek-tudestdmod); gen. sg. totar (cf. L. familias), popler, ocrer; dat.— abl. pi. Atiiefier, aseriater; 2 sg. verb form sir: L. sis. But final rhota- cism does not occur in the dat.-abl. pi. of i-, u-, and consonant-stems, which have -s < -fs < -fos < -bhos: homonus: cf. L. hominibus, and 0. luisarifs with -fs preserved ; nor does it occur in the 3 pi. secondary ending of verb forms such as etaias (with etaians), sis (with sins).
f. Before enclitics beginning with a vowel -s in dat.-abl. pi. forms of a- and o-stems is rhotacized in the older as well as the later tables: Funtlere Fondlire, fesnere, but before enclitics beginning with a consonant the s is preserved: esunesku. In other words, the juncture is so close that the rule for medial rhotacism applies.
g. A change sr > fr must be assumed in order to explain several Italic forms. In Latin the final result is fr initially, br medially: frigus: glyog, Lett, stregele 'icicle'; L. membrum < *mems-ro-: cf. Skt. mdmsam, Goth. mimz. The only probable Umbrian example is tefru 'burnt-offering' (?), if from ^tep{e)s-ro-: 0. tefurum (-uru- < -ro- by Oscan anaptyxis), cf. L. tepor, Skt. tapas.
h. Final s in Tables I- IV and the -r -r of Tables V-VII which results from rhotacism of the earlier -s are occasionally omitted in writing: gen. sg. agre, Fi- sie, nom. pi. prinuatu, dat.-abl. pi. snate, antakre, kumate, 2 sg. verb forms heri heri, si, etc. Among nouns and adjectives the cases of omission are very few in proportion to the total number of forms in -s, -r, -r, but for verbs the total number of 2 sg. indicative and subjunctive forms attested is low. Forms ending in -s from -fs or -ns are not subject to dropping of the s, just as they are not sub- ject to rhotacism (see e above). ^
i. For the treatment of ns, both original and of secondary origin, see 58. For the treatment of rs, both original and of secondary origin, see 59.
^ For a different view, whereby s- represents a remnant of a prefix, see Hoenigswald, Lang., XXVIII (1952), pp. 182-5.
^ For example, in the Tabula Bantina (no. 28 Co.). In the native Oscan alphabet z has the value ts, while s is used for both the unvoiced and the voiced s. Paelignian, like Oscan, escaped rhotacism, while for the other minor dialects the evidence is inconclusive. Italic rhotacism has recently been treated with special attention to chronology and geographical distribution by F. Altheim, Studies in Honor of David M. Robinson, vol. II (St. Louis, 1953), pp. 459-68.
^ Abl. pi. sevakne (third declension) in IV 9 stands at the end of the line, with no room in the margin for the final s.
A. Phonology 75
58. NS. In treating the history of the cluster ns it is necessary to make a dis- tinction between its occurrence in medial and in final position and between those cases in which ns is original and those in which it arises from some other consonant cluster or from the syncope of a vowel which originally stood between the n and the s.
a. Original medial ns became nts, represented by nz (z = ts), ns, or simply s: anzeriatu aseriatu, anseriato, anstiplatu, all with the prefix an- (= dvd),^ before forms having initial s; menzne (< *mens-en-i^), antermenzaru, both cognate with L. mensis.
b. Original final ns, which is principally represented by the ace. pi. of masc. and fem. nouns and adjectives, became /; the /, however, is sometimes omitted, rarely in the older tables, but very frequently in polysyllabic forms in the later tables: vitlaf, purka porca, apruf abrof, uiro, uvef oui, etc.
c. Medial ns arising from n + the dental clusters tt, dt (through nss by 44 a, b) results in / and is illustrated by the passive participles spefa < *spenssd < *spend-ta: cf. L. pensa; spafu: L. pdnsiis, both < *(s)pand-to-;^ probably also mefa with L. mensa; there is no verb root in -nd to which this last pair can be referred, but they may be derived from *mens-sd analogically formed after those participles which had ss by regular development of dental clusters.
d. Final ns from -nts, as in the nom. sg. of present participles, becomes /, which in the few quotable examples is omitted in the later tables: zefef serse: L. sedens; restef reste; kutef;* probably also prep, tra traf traha trahaf, if we regard it as originally pres. pcpl. of a verb equivalent to L. *trdre.
e. ns resulting from syncope of the intervening vowel is represented in medial position by uze onseK'^omesei: L. umerus;^ in final position by fons fos< *faunis. ostensendi, fut. pass. 3 pi. of ostendu, must also be considered as an example show- ing medial ns after syncope of the short vowel which formerly preceded the s (see 123), but it is not clear whether the development was *-tendesenter > *-tend- senter, etc., or whether it was '^-tendesenter > "^-tennesenter > *-ten(n)senter, etc. ; in other words, whether the d was assimilated to the n before the syncope of the thematic vowel or not.
f. In the nom. sg. of masc. and fem. /i-stems it is possible that we must recog- nize an ending /< ns of secondary origin ; at least this view is necessary in order to explain the Oscan forms fruktatiuf, uittiuf (: L. usio), tribarakkiuf. To the -n (which was either inherited, as in dychv in contrast to L. legio, or added by analogy with the oblique cases) there was attached an -s after the analogy of the other classes of animate nouns with -s in the nom. sg., and the resulting -ns ultimately became -/. The Umbrian examples are karu (dat. karne, abl. karne, cf. L. card, carnis), tribf igu (abl. tribrisine), and esuf: 0. essuf, both equivalent to L. ipse and commonly taken as from *eps6ns. Since final f is seldom
76 II. Grammar
omitted in the older tables — which alone provide examples of the forms under discussion — one is tempted to take karu and tribfigu as exact equivalents of L. card, etc., without the -ns, but esuf cannot be so explained, and it is also perhaps safer to assume a form based on -ns because of the usual close agreement of Umbrian with Oscan.
g. The secondary tenses of verbs, including the present subjunctive, have a 3 pi. ending -ns, -s, the explanation of which is difficult. The fact that in addition to U. etaians, etaias, dirsans, sins, sis, etc., several examples are known from Oscan and some of the minor Italic dialects, might lead one to assume that the formation was developed during the period of Oscan-Umbrian unity, but the fact that the -ns never changes to -/ favors a very recent origin. For the prob- able origin of the formation see 115 d.
h. It is advisable at this point to consider the question of relative chronology among the foregoing sound-changes. In general Umbrian and Oscan agree: both have -nts- from original medial ns, both have -ns where syncope has oc- curred in the final position (as in fons under e) and both have / in the nom. sg. of /i-stems, as in f, unless we here adopt the alternative explanation of U. karu and tribfigu. The forms in a must have undergone a development -ns- > -nts- (written nz in the native alphabet) before the changes resulting in second- ary -ns- had been completed, for otherwise menzne, etc., should have partici- pated in the change ns > /. On the other hand forms like uze onse, fons must not have suffered syncope until after the change ns> f had ceased to be effective. The discrepancy between Umbrian and Oscan in the treatment of original final ns (U. vitlaf, etc., as in b, but 0. viass) is a difficulty, but it is possible that doublets arose under different conditions of juncture, after which -/ was gener- alized in U. and -ss in 0.
^ Some refer the prefix of anzeriatu to am- (: L. amb-, afifp-); in tliis case we merely have to assume an early assimilation, whereby m becomes n and the cluster follows the normal de- velopment of original medial ns.
^ Blumenthal, p. 70, prefers a stem mens-ni-: less probable because -e is a very rare spelling in abl. sg. of i-stems, but the development of the -ns- is not affected.
^ For the variation sp-jp- see 57 a, b. The question of the relation between spahatu, ando), L. pando, pateo, spatium is complicated, but the equation spafu: L. pdnsus is quite safe.
* Etymology uncertain (see Lexicon), but all from Biicheler on make it an active pcpl. For the abnormal spelling kutep in lb 3 see 2 k.
^ Actually with ns by assimilation of ms after the syncope was completed. Observe also the omission of n in uze.
59. RS. For rs as for ns it is necessary to take account of whether the com- bination occurred in medial or final position and whether it was original or arose from the alteration of some other consonant cluster or after the syncope of an intervening vowel. Yet the history of rs is by no means closely parallel to that of ns, especially in regard to the situations in which / can develop. Moreover,
A. Phonology 77
among the reflexes of ns the n was never preserved, at least in writing, before /, but r was regularly preserved in rf resulting from any combination contain- ing rs.
a. Original medial rs is preserved, or else the s alone is written, the r being omitted, in Umbrian, in contrast to Latin, where it becomes rr, and to Oscan, where it ultimately becomes r with lengthening of the preceding vowel: U. tur- situ tusetu and the name of the goddess dat. Turse Tuse, all cognate with L. terreo, terror, sregosv ' eq)6^r]aEv Hesych.; farsio fasio fasiu: L. farrea.^ pars before the enclitic est shows the same preservation of rs as in medial position generally (for other examples of treatment of word-finals before enclitics see 4, 57 f).
b. Original final rs becomes r: far: L. far, 0. far. See a above in regard to -rs before enclitics.
c. rs in which the s has resulted from a dental cluster (44 a) becomes rf in Umbrian, being preserved as rs in Oscan and Latin: trahuorfi: L. transverse, both from *-worssed < *-wrtted: cf. L. versus, 0. feqaoQei.
d. There are several other forms originally containing clusters of three or more consonants in which a stop stood between r and s. In all of the following the rs is preserved after loss of the stop, or else the s remains, the r being dropped, as in a above: sesna, gersnatur: 0. kersnu, cf. kerssnasias, L. cena, cendti, all from *kert~snd-: cf. Skt. kdrtati, krntdti;^ persclo persklum, pesclu, persnimu pesnimu, all apparently from *per/c + sk- or other s-extension: L. posco < *pork- sko; for details see Lexicon; probably also perstu pestu, peperscust: L. parco, compesco (?). The etymology of persondru and the origin of its rs are unknown.
e. Medial rs resulting after syncope of an intervening vowel becomes rf in Umbrian, rr in Oscan and Latin: Serfe ^erfe: 0. Kerri, OL Cerus (= *Cerrus) and unsyncopated Cereri, with the second r from s; parfa: L. parra; the equa- tion cannot be questioned, and the two can only be reconciled on the basis of *par{e)sd, for original rs would not yield rf in Umbrian, and pIE *parbhd would not yield parra in Latin.
f. Final rs resulting after syncope of an intervening vowel becomes simply r: f rater frateer < *f raters < ^frdteres; the original vocalism of the second syl- lable is uncertain, but the length indicated by frateer is probably compensatory; ager (minor inscription n. 355 Co.) < ^agers < ^'agrs < *agros: dygog, L. ager.
g. The fact that original medial rs did not become rf may be explained on the supposition that the s became voiced (-rs- > -rz-) before the development of secondary -rs- had been completed, while in this latter -rs- (< -rss- < -rtt-) the s must have remained unvoiced until the conversion to / occurred. The change of -rs- to -rf- after syncope, as illustrated by parfa, Serfe, is surprising in view of the failure of Oscan Kerri, sverrunei to show the same change, and also
78 II. Grammar
because in corresponding forms with -ns- after syncope (58 e) the change to / did not occur, but the material is too hmited to allow any sure explanation.
1 But farer, gen. sg. of noun far, has r in place of rs by analogy with the nom.-acc. far.
2 A root containing t is necessary in order to account for L. cena, OL cesna, since rsn would yield L. rn, as in cernuus < *kersnowos, cf. Slct. sirsan-. But the root *kert- is evidently an extension of *ker- found in karu and kar-tu.
60. Conversion of Voiceless Stops and Spirants to Voiced Stops. Voice- less stops, and also the Italic voiceless spirants derived from pIE aspirated voiced stops, are liable in Umbrian to become voiced stops when preceded by nasals, and the voiceless stops are similarly altered before r. This latter change does not take place in the initial position, but the position after verbal prefixes such as en-, com-, etc., is generally treated as medial. Since in the tables written in the native alphabet t and k are regularly, and p occasionally, used for d, g, and b respectively, we must depend to a large extent on the later tables for clear examples of the consonantal changes in question.
a. The change t > d after n is illustrated by: anter- ander-: 0. anter, cf. L. inter; hutra hondra < *ghom-t(e)rdd with suffix as in L. extra, contra; hondomu < * g horn- tmmod with superlative suffix as in L. ul-timus; Funtler-e Fondlir-e: L. *(in) Foniulis; mantraklu mandraclo<^man(u)trdgklo-:^ ci.h.manutergium;-per- suntru persondru: the etymology is unknown, and the spelling persontru in VIb 28 is surprising with its / in the Latin alphabet against 8 instances with d; yet in all probability the word is formed with the suffix -tro- (76 b); pres. subj. pass. 3 pi. tursiandu: L. terreantur; fut. pass. 3 pi. ostensendi < *ops-tendesenter; en- tentu endendu: L. in-tendito; the second d of the Umbrian form is from t of the imv. ending -tod, changed to d when the original d had been assimilated to the n (61 e) and the thematic vowel had been syncopated, leaving n in contact with /; similarly ustentu ostendu: L. ostendito. The same explanation can be appUed to imv. hondu 'cast down' if from *ghom + root do of dldco/Ai, L. do, dare, in which case the stages would be *ghon2-ddt6d > ^hon-det6d>^honnet6d > *hont6d > hondu; but if the root is dhe- (cf. ridrj/xi)^ the development must be different- ly reconstructed: ^ghom-dhdtod > ^honpetod > *honpt6d > *honttdd, etc., the syncope having occurred before the possibility of a change np > nd (see d be- low). On final nt see 41 c.
b. A change / > d before n may possibly be exemplified by Padellar, if it is from *padenlds < *padnlds < *patnolds.
c. k becomes g when preceded by n : iveka iuegna : L. iuvenca. But the change k> g in the cluster nkl must have occurred earlier, since it is shared by Umbrian with Oscan and Latin: anglom-e: 0. ungulus 'amilus' (Fest., p. 514 Li.; the variation of the initial vowel in these forms is a/o-ablaut; the original voice- less character of the consonant is attested by dyxdh], L. uncus, etc., yet there
A. Phonology 79
is some evidence for variation in the final guttural of the root even in pIE); angla: cf. L. clamare; of the 8 occurrences, all in the later tables, 6 have g, 2 have c.
d. The pIE voiced aspirates when preceded by nasals are represented not by voiceless spirants (/, h) but by voiced stops.^ Labials: amboltu, with amb-
< '^amf- < ^ambh-: dfx(p-, L. amb-; amprehtu, ambretuto, with ambr-, an ex- tension of amb-,'^ 4- verb equivalent to L. eo, ire; kumpifiatu combifiatu: cf. L. fido, Tisido), root bheidh-; even if we were to adopt the less probable connection with Tiwddvojuai, Skt. bodhate (root bheudh-), the law under discussion is still applicable. Of the effect of n on the dental spirant there are no sure examples (see a above). The voicing of a guttural (Italic ;(; < pIE gh) is illustrated by kren- katrum cringatro: OCS krogii, OHG hring, which together point to original gh in the second syllable. It should be noted finally that in afero, aferener< '^am-jer-
< *amf-fer- the analogy of the simplex fer- has prevented the action of the nasal upon the initial consonant of the verb in the compound.
e. Medial p becomes b when followed by r:^ abrof apruf : L. aper, apri: kabru, cabriner: L, caprum, caprini; tribfigu, tribrisine: the voicing is caused by con- tact with f (in the latter form simply r !) from d if we derive from *tri-pediky6n.^ In supru: L. supra, cf. 0. supruis, the p is in all probability used with the value of b.
t. Medial / becomes d before r in: atru adro: L. ater, atri (consonantism not certain, since there are no sure etyma outside of Italic); seipodruhpei: cf. 0. putereipid (anaptyctic vowel after t), L. utroqiie. But cringatro, fratrom, and other forms escape the change, and it is not possible to formulate any rule.
^ u is believed not to be subject to syncope (29, n. 1); yet it is likely that mandraclo, like late L. manutergium, had the full stem manu- as first member. Probably medial u was actually subject to syncope, though u in final syllables, on the evidence of ems, was not.
2 Compare the fusion of the two roots exemplified by condo, reddo, etc., in Latin, where me- dial d may represent either d or dh. In the dialects, where dh gives / medially, the two sets of verbs would not fall together except when the consonants are obscured by secondary assim- ilatory changes. The thematic vowel e instead of a (the normal Italic reflex of a) is written here because of the transfer of these compounds to the third conjugation. The second of the two explanations here offered for hondu is also applicable to ahauendu, preuendu: Eng. wind (verb), etc., the voiced aspirate being supported by Skt. vandhuram, vivadhas.
' In Latin the voiced aspirates in general became unaspirated voiced stops in medial position regardless of their environment; only in the case of gh does the development appear to be con- ditioned by a preceding nasal (or r): veho, mihi, but lingo (cf. Xeixoi), jingo (cf. xslxoz), spargo. We may assume then that in Latin a preceding nasal produced voicing if the medial spirant was not already voiced under the rule covering the development of voiced aspirates in Latin. In Oscan on the other hand a nasal does not cause a following stop or spirant to become voiced. It seems best therefore to regard the change exemplified by ambretuto (cf. O. amfret), L. lingo, etc., as an independent development of Umbrian and Latin or a tendency which spread after the separation of dialects and which failed to affect Oscan.
* For amb-: ambr- compare L.sub: super, vno: vneQ, Skt. upa: upari, where the enlargement
80 II. Grammar
was of course pro-ethnic. For the prefix ambr- compare also O. amfr-et, L. anfractus = am.fr- actus, the traditional connection with frango, fractus being a popular etymology.
^ The voicing of p before r does not take place in Latin, nor in Oscan apart from the forms embratur, etc. {' imperator') on several coins. O. Abella is best talien as containing original b: cf. Olr. ubull, Eng. apple, and Vergil's maliferae... Abellae (Aen., VII, 740).
6 "We cannot attribute the change p > ft to assimilation before the earlier d since the vowel cannot have been syncopated until after the d had become f.
61. Assimilation of Consonants includes changes which must partly have occurred in pIE itself and which are recognized chiefly through the obvious etymological relationships of the forms with other unassimilated forms. Such, for example, is the change of a voiced to a voiceless stop when the stop imme- diately following is voiceless, as in L. scrip tus beside scribo, and Oscan and Um- brian must have inherited these p-forms before making further changes of their own which resulted in 0. serif tas, U. screihtor. But new instances of assimila- tion occurred as consonants were newly brought into contact, and the Italic languages were especially susceptible to these changes because of the vowel syncope which was so prevalent in them. Unfortunately, however, in Umbrian full assimilation, whereby a sound becomes identical with the following sound (regressive assimilation being the usual type) is difficult to recognize because doubling or "length" of consonants is scarcely ever shown in writing. Some- times the doubling of a consonant in an Oscan form makes it highly probable that a closely related Umbrian form was similarly pronounced, but often it is a matter of conjecture whether a consonant was assimilated or merely lost, es- pecially in clusters originally consisting of three or four consonants.
a. Assimilation involving change in the place of articulation is shown in ape < *at-pe < *at-que: L. atque, with the single instance of appei in Vllb 3 consti- tuting possible evidence of consonantal length despite the 30 instances of spel- ling with single p p. Similar is the transfer of nasals to the position of the fol- lowing stop in etantu, panta (: L. tanta, quanta) if from ^tdm-td, quam-td accor- ding to the prevailing view. uze onse: L. umerus < *omesos shows a change of the labial m to the dental n before s after syncope has taken place. ^ This as- similation of m, however, did not take place in the imperative uintu< *umbet6d: L. unguito, nor in sumtu.
b. A dental or guttural stop before /, the ace. pi. ending of masc. and fem. consonant-stems, suffers complete assimilation, followed apparently by total loss; at least there is no evidence that the change results in -//, and in some cases even the f / is not written. Examples are: capif {-if 3 times, -i twice, -if once by analogy with other cases): cf. L. capides; frif {-if 5 times, -i 6 times): cf. L. fruges; uapef-e {-i -/ 3 times before postposition -em -e): cf. L. lapides.
c. Changes of order, that is, of the manner of articulation, were usual in the parent speech whenever voiced and voiceless consonants came into contact through etymological derivation. Reference may be made to the first paragraph
A. Phonology 81
of the present section, and also to 60, in which the changes described are really of this type. Somewhat different, however, and more complicated is the develop- ment seen in sistu in III 8 from '^'sistt6d< *sizd-t6d < *sizdet6d: L. sidito.
d. k, p, and / are assimilated to a following s, with some evidence that a ge- minate results, although in the triconsonantal clusters it is possible that actual loss occurred. With ks: pron. essu esii: 0. eksuk. With ps: esuf: 0. essuf esuf, cf. L. ipse; osatu: cf. 0. upsannam, L. operari, all from *opesd-. With kst: desire: 0. destrst (= L. dextra est); sestentasiaru : L. sextantariarum. With pst: ostendu. ustentu < *ops-t-: L. ostendito. / is assimilated to final s in -/o5, the dat.-abl. pi. ending of i-, u-, and consonant-stems, after syncope of the o: aueis: L. avihiis; berus: L. verubus; homonus: L. hominibus; etc. The possi- bility that the s was long is shown by the fact that it is never rhotacized and scarcely ever dropped as in the dat.-abl. pi. of a- and o-stems (on sevakne IV 9 see 57 n. 3).
Oscan shows a single example of preservation of / in the ending: luisarifs, beside Anafriss, sakriss with double s.
e. In the following instances stops standing immediately before (in the case of nd, after) nasals are themselves changed to nasals. bn > mn: tremnu< *treb- nod: cf. trebeit,0.triihvira. pm > m{m): somo: h.summum, cf. super, L. super, dn > n(n.): une < *udne, abl. sg. of utur 'water', an r/77-stem: vScoq, vdarog (84). nd > n{n): ponne (twice, and pune puni pone 16 times) < *pon-de < *k'^om-de, cf. L. quamde; gerundives peihaner, etc.: cf. L. piandi, 0. upsannam, sakrannas; also as an intermediate stage in ostendu, endendu, ampentu, before syncope of the thematic vowel and change of nt to nd (60 a).
^ It is possible that the whole consonant cluster was voiced after syncope had taken place since the intervocalic s which became rhotacized in Latin and, when syncope did not interfere, in Umbrian also, is assumed to have been a voiced s in proto-Italic, as in Oscan. The phonetic spelling then would be *ondze; the z of uze shows that the dental stop was preserved between n and z, and the n of onse shows that the nasal was not lost.
62. Final Consonants. Juncture. The preservation, alteration, or loss of final consonants has for the most part been discussed in the sections devoted to the history of the individual consonants, so that the present section is merely a summary, with references to the appropriate sections. The consonants with which we are concerned are partly inherited finals (e.g. -s in nom. sg. of the second declension, -m in various ace. sg. forms, etc.), partly the products of consonant- al changes (e.g. -/ from -ns), partly consonants left in final position after syn- cope of a vowel in the final syllable and alteration of the resulting consonant cluster. To some extent the retention or loss of final consonants depends on the form-categories in which they occur (e.g. final r is retained in nom. sg. of nouns of the type of uhtur, L. auctor, but largely lost in 3 sg. and pi. passive verb forms).
82 II. Grammar
a. Of inherited final labial stops, either preserved or lost, there are no instan- ces. Final t and d are regularly lost, -t having passed through the intermediate stage -d (41 b, c, 42 e). Final k seems to be preserved in huntak, tuplak. For final m see 51 e. For final n see 52 h. Of inherited final / there are no cases; in those instances where final / actually appears its final position is secondary (55 e). For original final r see 56 c. r which becomes final after completion of the change -ros > f s > -er or the change -ro > -f > -er (a consonant having ori- ginally preceded the r) is regularly preserved, the instances of loss being extreme- ly rare (32 d, 56 d). For final s and for the r which arises from it in the later tables see 57 e, h. Original final s is normally preserved in forms where a pre- ceding vowel has been syncopated : fons fos (58 e), pelsans (29 i, 91 b), dat.- abl. pi. forms fratrus, aueis, etc. {-s < -fs < -fos < -bhos, 57 h, 101 c), but not so in forms which originally ended in -los (e. g. tigel, 55 e) or -ros, -res, -ris (e.g. ager, pacer, 32 d, 59 f). -s is lost in original rs (59 b, but observe also a). In the 3 pi. active secondary ending of verbs, whatever the origin of the -ns, the s is regularly preserved (58 g). Final / from -ns, -nts is largely preserved in the earlier, largely lost in the later tables (58 b, d, f).
b. For the treatment of final vowels see 31 a-g, 34.
c. Juncture has to do with the changes which occur (or fail to occur) at the point of contact between morphemes; that is, where stem-suffixes, case- and person-endings, enclitics, prepositional prefixes, etc., are attached to roots or to each other; and also where words follow one another in the sentence. In gen- eral the term "open juncture" or "external sandhi" may be applied to instances of the last-named type, and "close juncture" or "internal sandhi" to those oc- curring within words or within such phrases as form single accentual units. It is not possible, however, to formulate a comprehensive and orderly set of rules for the Iguvine Tables comparable to the rules found in grammars of such languages as Sanskrit. The Tables almost never show variants corresponding to differences in external sandhi; such variations as the presence or omission of final 777, for example, show no consistent relation to the initial of the following word. In part the phenomena of juncture have already been treated; so, for example, the treatment of initial preconsonantal s (57 b) and the assimilation of consonants (61) are largely matters of juncture, although they partly belong to historical stages considerably prior to the existence of Umbrian as a separate dialect. The present section is primarily concerned with the changes which may occur when verbs or other forms are compounded with prepositional prefixes or when postpositions, pronominal enclitics, or the enclitic est are placed after forms of any kind. In such cases two opposing forces may come into conflict: the tendency to allow the free operation of the phonetic law applicable to the particular environment in which the form occurs, and the tendency to maintain the form as it would appear in isolation or in open juncture.
d. A verb compounded with a prefix may undergo changes of such a nature
A. Phonology 83
that the initial of the simplex is lost or altered: subocau with loss of w after b: cf. L. vox, vocare, etc. endendu with change t > d after n (60 a); cf. ostendu with prior t preserved. combifiatu with b < f after m (60 d). anzeriatu with -nz- = -nts- < -ns- in *an-ser- (58 a with n. 1).
e. menes, fut. 2 sg. equivalent to L. venies, may owe its initial m to the in- fluence of a compound *kuni(ni)enes with -m{ni)- < -mb- as in umen < *um- ben < "^ong^en (49 b) or like the assimilation of -nd- to -n{n)- in peihaner, etc. (61 e) ; and this despite the unassimilated Oscan form kumbened.
f. Like 0. kumbened immediately above in b are certain forms in which the influence of the simplex prevented alteration of the initial in the compound : anferener < *am(f)-fer-, escaping the change -mf- > -mb- seen in combifiatu in a above; andersafust, escaping the change -nd- > -n{n)- (61 e). Another in- stance is anderuomu, if any of the etymologies assuming the change I- > u- v- is correct (55 a and note on VIb 41).
g. The change of intervocalic rf to f rs (42 b) may occur when an enclitic with initial vowel is added to a form with final d: pife pirse < *k^id -f i, pure porse < *k^od -f i (vocalism of the enclitic is uncertain; see 106 d). Similarly when an enclitic with initial d follows a form with final vowel: nersa < ne-ddm.
h. Rhotacism of final s, which is characteristic of the later tables, may occur even in the earlier if the form in question is followed by a vocalic postposition: Funtler-e lb 24, fesner-e lib 11 (57 f).
i. The enclitic est in the combination pisest Vib 53 fails to rhotacize the s of the pronoun, yet in sopir habe 54 the s becomes r before the verb habe.
j. Final rs (original or secondary after syncope), which normally loses its s, retains it in the phrase pars est just as medial rs is retained in farsio.
63. Sources of the Umbrian Sounds. The present section is a summary of the etymological values of the Umbrian sounds as they are represented by the characters of the Latin and native alphabets. The treatment is necessarily much condensed and contains references only to the most important sound- changes. For many details the individual sections must be directly consulted. It has seemed necessary, however, to include not only the single vowel-characters but also certain combinations, such as ai, ei, ei, etc., which stand sometimes for simple vowels, sometimes for diphthongs.
a a.: < a 6; < a 7 a;<95 a; aain combination with preceding liquid or nasal
may be from long vocalic liquid or nasal by 18 a. ai ai: < ai 19 c.
b h < 6 39; < g'"' 49 b; < / < bh after m 60 d; < p before r 60 e. c k: < /c 45, 46 a, e, f, (?) h; < k"^ 49 d; < f before I 41 e, before n (?) 41 1 In
regard to k see also under g (2 a), d t (2 a): < d 42 a, d; < / final 41 b; < f after n 60 a; < t before r 60 f.
84 II. Grammar
e e: < e 8 a; < e 9 a; < f 10 b, c; < f 11 b; < ai 19 a; < ez 20 a, b, c; < oi
final 21 d; < di 25 a; < ei 25 d; < oz 25 b. ei ei: < e 9 c; < f 11 c; < az (?) 19 b; < oz 25 b. t i: < bh 40; < dh 43 ; < r^/z ( ?) 48 b; < ^*^7z 49 c; < s before r 57 g; < original
71S 58 b; < secondary ns 58 c, d, e, f ; in the combinations rf rf 59 c, e. g- k (2 a): < g- 45, 47 a; < k after neo c; < gh after n 60 d. h h: < gh 45, 48 a; < A', p (original or resulting from assimilation) before t38 b,
46 i. As mark of vowel length 11 a, 15 c, 26 a, b, c. As mark of hiatus
33 a. I i: < no a; < f 11 with a; < e 8 b, c, d, e, f ; < e 9 b; < u 15 a, b; < ez 20 d;
< ei 25 d; < 6i 25 b; < z/ 53 a, c, d, e, f, n. 1; < k, g, gh 46 g, j, 47
b, (?) c. Z 1: < / £5 b, c, d, e.
m m: < 777 51 a, c, d, e; < n by progressive distant assimilation 52 h. 71 n: < 77 52 a, b, c, d, g, h; < 777 by assimilation 61 a. 0 u (2 a): < 0 12 a, d; < 0 13 with a; < a 7 b, c; < e 8 g; < iz 14 b, c, d; < 01
21 a; < az7 22; < ezz 23; < ozz 24; < eu 25 c; < 6u 25 e. ol, or: < ol, or, or < I, r. See 17 a.
p p: < p 38 a; < /c*" 49 a, e. In regard to p see also under b (2 b). q: see under c k and 2 a. r r: < 7-56 a, b, c, d; < s 57 c, e, f. f rs: < d42 b, c; < / 55 b. For rs, and for rs when not equivalent to f, see 59.
For the same as components of larger consonant clusters see 59 e, s s: < s 57 a, b, d, f; < ks 61 d; < ps 61 d; < /s < bh{o)s 61 d; as a variant of
rs rs 2 i; < U. 5 9, q.v. s 9 : < A' 46 b, c, d.
/ t: < Z 41 a. In regard to t see also under d (2 a), u u: < 17 14 a, 15 c; < 0 12 b, c, 13 c; < 0 13 with b, c, d; < ozz 25 e. In regard
to u see also 0 u (2 a). til, ur: < ol, or or /, f. See 17 a. 17 v: < zz; 54 a, b, c: < Z initial 55 a. z: < Zs 2 a, 58 a.
B. Formation of nouns and adjectives
64. Of the various Indo-European noun and adjective stem-classes the following are represented in the Iguvine Tables: root-stems, a-stems, o-stems, i-stems, u-stems, n-stems, r-stems, one r/77-stem, s-stems, and stems in stops, including U. f rs from intervocalic d.
a. The gender-system is essentially the same as in Latin: a-stems are femin- ine, o-stems masculine or neuter, and the three genders are variously distributed among the other stem-classes, but the gender of the nouns can be positively known only from the forms of the adjectives used in agreement with them. As in Latin, adjectives are either o-stems with a-stem forms serving as femin- ines, or they are i-stems or consonant-stems without change of stem for different genders.
B. Formation of Substantives 85
b. In the following paragraphs an attempt is made to classify the majority of the nouns and adjectives found in the Iguvine Tables on the basis of their stem-formation. A few, however, are so obscure that discussion regarding them is reserved for the commentary or the Lexicon. The a- and o-stems are treated concurrently, for any attempt to separate them would do violence to many of the classes using the same suffix with only a difference of gender indicated by -a- or -o-. Moreover it is sometimes difficult in Umbrian, especially with divine names attested only in the dative singular, to know whether a given form is a masculine o-stem or a feminine a-stem.
65, Root-stems. In these the root serves as stem. There is no suffix of any kind other than the case-endings.^ The variety of roots, however, is such that the forms fall into many different declensional classes with vocahc, diphthongal, or consonantal stems.
a. a-stem: angla < *an-kld, with the same root found in L. cld-rus, cldmo, denominative to *cld-md.
b. u-stem: sim: vv, cf. L. suem (= *suwem).
c. Diphthongal stems, bum, bue 'ox' is a stem in 6u varying with ou and o: cf. L. bovem, hove (with b of dialectal origin, from g^), ^ovv, dat. j5ot, Doric ace. ^(x)v, Skt. gam, loc. gavi. re(-per), ri is a stem in eije: cf. L. res, rel, Skt. rds, ace. rdyam, etc. Di Dei (voc), Dei (ace) 'Jupiter' is a disyllabic base in eu/e: cf. L. dies, Diespiter, diem, Gk. ace. Zfjva (ace. '^Z7]v extended by further addi- tion of ace. ending -a), Skt. Dydus, ace. Dydm; U. dat. luve, voc. lu (both in combination with forms of pater) are built on the grade *Di/e«->pIt. (D)iou-: L. lovis, etc.
d. Z-stem: salu: L. salem, aXa.
e. r-stems: nerf: cf. avdg-ag, Skt. rirn; pin nvq; vepurus abl. pi., if we analyze ve-pur-us 'fire-less.'
f. Stems in stops or in Umbrian spirants derived from stops: pase: L. pace; praco (?), stem-class not certain, but the gender is fem., and a fem. o-stem is unlikely; frif < ^frilg-f: cf. L. fruges; persi peri: L. pede, nodi; dupursus, pe- turpursus: L. bi-, quadripedibiis; pruzuf'e< *pr6{d)-sod-i (?) with ablaut- variant of sj sed-: cf. L. praeside, etc.; «e/acc. pi. < '^'weif-f <*weidh-ns: cf. L. di-vido{1).
* " Case-endings " here may be understood to include elements properly belonging to vowel- stems but secondarily extended to consonant-stems and forming for all practical purposes a part of their case-endings: the ace. sg. of consonant-stems, for example, has the ending -om, consisting of o, the stem-vowel of o-stcms, -f- m, the original ace. sg. ending; and the dat. -abl. pi. has -us, consisting of u, the stem-vowel of u-stems, + s < -fs < -bhos, the original dat. -abl. ending for i-, u-, and consonant-stems.
86 II. Grammar
66. a-stems. In the following forms a is attached directly to the root, or at least the stem cannot be clearly broken down into any other components than the root + a: asa; uia; totar; peica; parfa< '^paresd (f mother analysis not clear); porca; paca (abl. sg. functioning as "improper" preposition, like L. causa); Tursar; probably also gen. pi. antermenzaru; Hule (generally taken as fem.). Cf. also angla, a root-stem in a, listed above in 65 a.
67. o-stems. In the following nouns and adjectives 6 (or its ablaut-variant ^ is attached directly to the root, with no clear possibihty of further analysis: abrof; kabru; furo: L. forum; iuku iuka; numer: L. nummis; onse (loc): L. umerus; peico; perso perum : jiedov; Titis (gen.): L. Titus, -i; toru: L. tamos; uerof-e (ace. pL); uocu-com: either foixog or L. lucus; uinu; probably also an- deruomu (abl.); carsom; kebu; $erfer, but with the possibility of a suffix -es- before the o. Adjectives: alfir: L. albis; rofu; semu; possibly also dersecor.
68. Stems in -io- and -id-. Actually this category includes stems in yojd and iyo/d, but the distribution of the two types in the Italic dialects is secondary and not a survival of the conditions assumed to have existed in proto-IE (see 53 c with n. 2 on "Sievers' Law"). Moreover the two types frequently cannot be distinguished from one another as they appear in our texts, and no attempt is made to separate them in the following sub-sections, although in general forms written with ii in the native alphabet belong to the type with suffix iyo/d. The suffixes treated in the present section furnish a large proportion of all the ad- jectives used in Umbrian, or indeed in the earlier Indo-European languages generally. Many of the nouns, which include personal and ethnic names, etc., are clearly adjectival in origin, others much less clearly so. The suffixes may be attached either to roots directly or to stems already containing extensions, and the analysis of some of the forms is so uncertain that a rigid classification is impossible. In all cases the Lexicon should be consulted for further etymolo- gical details.
a. The following forms may with reasonable certainty be regarded as prim- ary derivatives formed by the addition of yo/d or iyojd directly to the root. Nouns or adjectives used substantively: afkani, as if = L. "^ad-cinium (see 32 a; an i-stem here is far less probable); pruse^ia; Vuvgis: L. Lucius. Ad- jectives: deueia (Via 9, 10; ei probably represents i); Fisier; gomia kumiaf; loui: cf. L. lovius; Martier; peia (cf. 46 g); Sansie; Tlatie: L. Latium (?); Uehier Vehiies; observe also Vugiia beside Vuvgis listed above as masc. noun.
b. The following adjectives are formed with the suffix -eo- < -eyo- with loss of intervocalic y (53 b) as in L. igneus, aureus, etc.: farsio: L. farrea; spantea: cf. spanti; possibly also afepes, if = L. adipeis.
c. The following adjectives, designating situation, are formed with the suf- fix -aio-< aiyo- (cf. 0. kersnaiias beside kersnu, Gk. dixaiog beside dixi-j, Ved.
B. Formation of Substantives 87
sabheyas beside sabha): pernaiaf (to perne), pustnaiaf (to postne), persae pefae (to perso pefum).
d. The following adjectives or substantives of adjectival origin are formed with the suffix -asio- or -ario- < -dsyo- (perhaps originally from gen. sg. in -as + yo-; cf. 0. kerssnasias beside kersnu, L. herbarius beside herba, and 57 d on the lack of rhotacism in most of these forms) : kurglasiu; plenasier; ur- nasier: L. ordinariis (?); sestentasiaru : L. sextantariarum; but ezariaf with rhotacism.
e. The following is a list of miscellaneous nouns and adjectives formed cer- tainly, or at least very probably, with the suffix -{i)yo-. The suffix is mostly secondary, although in a few cases it appears to be added directly to the root. A large proportion of the words, especially of the personal, divine, and ethnic names, are etymologically obscure, and the very brief comments below must be supplemented by reference to the Lexicon. Forms used with substantive value: Acesonia/nAkefuniam, a place-name; cf. 0. Akudunniad, etc.; aiu (?), stem- formation uncertain; am-peria; arsier;armo, derived from the -wo-stemaruo; Kas- trugiie, a gentile name ; cf. Oscan names of the type of Klawatiis; Clauerniur, nom. pi., name of a decuvia; Kluviier, a gentile-name; Coredier Kureties, name of a god ; kvestretie ' term of office as kvestur ', but in formation like L. duritia, ava- ritia, etc. ; famerias : L. familiae ; Fisoiiie, name of a god ; Piquier ; Huntia, Hoier, Noniar, N ur pier, Salter, names oi persons or deities in the enumeration of places given in Via 12-14 in order to define the limits of the augural templum; Pe- raznanie, pL, name of a decuvia; presoliaf, one of the places listed in Via 12; pumpefias; Rubinam-e, Rupinie, name of a place (for phonology see 53 d); sehmenier sememes; spinia; sufafias; gersiaru; tekvias; tehterim (a stem in -di- or -li- is much less probabe); Teteies, a gentile name with stem in -eiyo-; uhtretie, to uhtur as kvestretie to kvestur; uestisiar 'libation' and Uestisier, name of a god, both apparently identical in formation except for the variation -d/o-. Forms used principally with adjectival value: aplenia; arsmatia; Atiersiur from a stem in -dio-, cf. Coredier; filiu feliuf 'suckling' < *fe-liyo-, cf. L. filius; Grabouie, a divine epithet, in formation like Fisouie; Petrunia, name of a gens, in formation like Acesoniam Akefuniam; speturie: L. *spectoriae; $erfiar, a divine epithet; dequrier tekuries.
69. Stems in -uo- and -ud-. These are based on a suffix -(u)wo/d similar to the suffix {i)yold described in 68, and what was said there is to a large extent apphcable here also. No attempt is made to separate the nouns and adjectives, since most of the forms are adjectival in origin and several are used sometimes with one value, sometimes with the other; but the adverbs in -e are given sep- arately. Nouns and adjectives: arvam < *ar-wdm, cf. L. aro, arva (f.), Naev. and Pacuv. apud Non., 192, 30; dersua tesvam; felsva: mersuva; prever: L. privus; saluo: L. salvum; seuom sevum; sopa supa < *sup-wo-; for loss of w
88 II. Grammar
see 54 d. Adverbs from stems in -wo-: give < *ki-wed: cf. simo Qimu, L. ci-tra; priife < *pro-fwed < ^pro-bhwed, cf, L. probus, superbus, Skt. abhvas, and 54 d.
70. Stems in -ko- and -ka-. The c k is sometimes visibly preceded by the "connecting-vowel" i, which spread analogically from stems ending in i; in other cases comparison with cognate forms shows that the i was present before its loss through syncope. The original value of stems in -koja- was predominantly ad- jectival, but a considerable proportion have become substantives. The Umbrian forms include the following nouns: fratrexs fratreks (but fratreca Vllb 2 as adj. abl.sg.fem.): L. *fratr-icus; iuengar: L. iuvencae (on U. ng < nk see 60 c); perca: L. pertica. Adjectives: todceir < *tout-iko-: 0. tuvtiks; the ethnics labuscom lapuzkum, Naharcom, Tuscom Turskum; Tesenocir Tesenakes, name of one of the gates of Iguvium; Pupfiges Pupfges, Pupfige Puprike, epithet of the god Puemunes. A /c-suffix is also to be recognized within the forms aviekate, fratrecate, and iiesticatu.
71. Stems in -to- and -td-. The most productive class of these stems of course consists of the perfect passive participles of verbs (131), There are, however, in Umbrian several examples of the closely related type represented in Latin by armdtus, barbdtus, with the sense 'equipped with, characterized by, resem- bling so-and-so, ' as well as a number of other forms which for various reasons are not suitable for inclusion among the participles. As in the case of the stems in -{i)yold, -{ii)wo]d, -kojd-, the original adjectival (or participial) value is often replaced by substantive value. Substantives: arglataf: cf, L, arculata beside arculum; somewhat similar is ponisiater ■punigate-.h.'^puniceatusheside puniceus, cf. praetextatus beside praetexta; loc. sg. fratrecate 'in term of office as fratri- cus, ' a probable transfer from the u-declension, as in OL senatus, gen. senatl; the divine names Honde Hunte (for -nd-< -nt- see 60 a) and Prestote Prestate; motar, muta mutu: L. multa 'fine'; pertom and tettom, terms of uncertain meaning but probably participial in origin, found in the list of landmarks in Via 12-14; prinuatur, title of certain officials, but of obscure etymology; vaputu 'incense' (?), perhaps related to L. vapor as arbutiim to arbor; urieta:L,. orbit am; pracatarum, although it may be classified as a true participle, especially if we take the preceding praco as gen, pi. Adjectives: hostatir: L. hastatis; anhostatir; petenata: L. pectinatam; etantu: cf. L. tanta < *tam-td; similarly panta: L. quanta; apparently mersta, although the etymology is not clear. The adverb rehte: L. recte < ^reg-ted must also be included.
72. Stems in -do-, U. -rso- -fu-: calersu kalefuf: L. calidos ' 2.evxo/usT(x)7iovg,' cf. L. ccdlgo, Skt. kdla-, from which the suffixal character of the Itahc -do- may be inferred;^ possibly also sorsom sufum, although it is not certain that rs f here is from d rather than /.
> The Latin adjectives in -idus are phonologically ambiguous, since the medial d can represent IE d or dh, but the rs of U. calersu must be from d, since dh would become / (43). In view of
B. Formation of Substantives 89
the semantic value of calersu, therefore, it is likely that the Latin color-terms albidus, rubidus, etc., have original d rather than dh.
73. Stems in -mo-. The o-suffix -mo- is found in the noun arsmor < *ad-mos, cf. also the derived adj. arsmatia and vb. arsmahamo; perhaps also in stahmei < *sta-m6i, although here the possibility of a stem in -mi- cannot be ruled out. Of feminine nouns in -md- (cf. 0. egmo 'thing, ' L. fama) there are no Umbrian examples.
a. -mo- is also used in forming superlative adjectives and ordinal numerals, together with adverbs corresponding to both of these types. Adjectives: hon- domu < *ghom-tm-m6d, with superlative suffix -tmmo- as in L. op-tumus, etc.; soma: L. summum < ^sup-mom (14 b, 61 e); semu, in partitive attribution with pe{r)sclu, ' middle of, half-way through ' : for the formation cf . L. ordinals pri- mus, Septimus. Adverbs: nesimei < ^'nedh-tmmed, with the same superlative value as in L. proxime; nuvime, according to the usual view 'for the ninth time,' as adv. in -e{d) from ordinal stem in -mo-, but perhaps better taken as adv. from superlative *nowf mo- ' latest, last ' ; see on I la 26; simo gimu '(back) to this spot, ' cf. give, L. citra, citimus.
74. Stems in -no- and -nd-. The nouns and adjectives, together with several adverbs, formed with -nojd- as a suffix of primary or secondary derivation are sufficiently numerous and varied to require some division into sub-classes.
a. In the following nouns -no- or -nd- is added directly to the root: kumne loc. sg. < *kom-nei: 0. comenei (with anaptyctic e) ; fesnaf-e ace. pi.: 0. fiisnu, cf. L. fdnum; acnu: 0. akenei, acunum, L. annus, < ^at-no- ? (see 41 f); tremnu < *treb-ndd.
b. The following adjectives are formed with -no- as primary or secondary suf- fix: plener: L. plenis, cf. ple-re; sanes: L. sanis; ahesnes: L. aenis ahenis < * ayes-no-; iifestne (see 44 d).
c. A suffix -snd must probably be recognized in the noun sesna: 0. kersnu, L. cena, OL cesna, all from *kert-snd.
d. A suffix -dno- (cf. L. Romdnus, urbdnus) occurs in several forms, mostly ethnic or topographical names: Treblanir abl. pi., name of one of the gates of Iguvium; luieskanes and Satanes, names of decuviae. asiane is of uncertain etymology and meaning.
e. The similar suffix -ino- (cf. L. caninus, Sablnus) is used chiefly in making ethnics and adjectives from names of animals or gods: liouinar: L. Iguvinae; cabriner: L. caprini; ekvine loc. sg. : L. equinus (?); habina: L. agninas; Fi- souina, related to Fisouie as L. Antoninus to Antonius; Miletinar.
f. The divine names appearing chiefly as dat. sg. in -une -one are somewhat
90 II. Grammar
different and more difficult to analyze. Although most of them occur only in the native alphabet, the form Uofione together with L. Bellona, Pomona, etc., make it plain that the vowel is 6, not u; yet for Vesune see on IV 3. The mas- culine forms Afmune, Puemurie, Uofione Vufiune can be consonant-stems in -077- as well as stems in -ono- (like L. patronus), since no unambiguous case- forms of them occur. In any case the whole group should be regarded as stems in -on- or extensions of them with -o/d-, their treatment among stems in -no/d- being justified merely by convenience.
g. Similarly stems in -m(e)no- are o-extensions of men-stems but may conven- iently be treated with stems in -no-: termnom: L. terminum, cf. OL termen, Skt. tarati, all from the root ter-.
75. Stems in -lo- and -Id-. Here are included stems in -{e)lold-, -tlojd- (Italic -klojd-), -k{e)lold-, and -dhlojd-.
a. The semantic value of the stems in -lo/d- is partly diminutive, partly that of names of instruments, but is in general not sharply defined. Forms based on -lo- and on -elo- are difficult to distinguish because of the loss of vowels through syncope in Umbrian and the development of anaptyctic vowels in Latin; but the suffix -elo- was common in Greek (e.g. axojiEXog, etxEXog), and its use in Umbrian may be inferred from the palatalization of k in arglataf < ^ark-eld- tans, cf. L. arc-ulum, arc-ulata; preuislatu < *prai-wink-eld-t6d, cf. L. vinc- ulum. On the other hand the lack of palatalization in anglom (not ^anslom) is evidence of -lo- rather than -elo-, the unvoiced stop being supported by com- parison with dyxdjv, dyxdXf], L. uncus, etc. The Umbrian forms with -(e)lo/d- are: anglom-e (but not angla: see 65 a); probably Casiler; katel: L, catulus; klavlaf: ecla (?) adj. 'every,' of uncertain etymology; fondlir-e; Padellar (?), apparently name of a goddess, < "^Pateno-lds, cf. 0. Patanai and 32 e; persclo, where however the suffix may have been -tlo- as in b below; sep-les; scapla; uitla and uitlu.
b. Stems in -tlojd- (cf. avrXoQ, xvtXov) became Italic stems in -klojd- by 41 e. The Umbrian examples in general are nouns designating articles or places having a specific function (cf. L. oraculum, poculum): kumnakle; ehvelklu; ficla < * fig-kid; mandraclo; muneklu; naraklum; ooserclom; pihaclu: L. pia- culo; evidently also sufefaklu, although the exact sense is uncertain. On the adjective auiecla, the formation of which is not clear, see Lexicon.
c. Stems in -kelojd- (cf. L, diminutives articulus, etc.) arose originally from the addition of -lo- as secondary suffix to stems with the suffix -ke-, which in Umbrian suffered the change k> g s (46 b, c) before the syncope of e: ereglum; strusla; uasirslom-e; uesclir, with c k unpalatalized after the s. With the above forms belong also kurglasiu (< ^kfkeldsiod), the second k not being an original part of the root, and tigel, tiglu, if the interpretation followed in Ila 15, etc., is correct.
B. Formation of Substantives 91
d. A stem in -dhloja- (cf, ysvedXov, L. fabula, with b < dh before /) is appa- rently represented by tafle: L. tabula < *tld-dhla (?); contained also within staflarem: cf. L. stabularius with stem in -drio-, not -dri-.
76. Stems in -ro- and -rd-. Here are included stems in -rojd-, in the suffix -trojd- used for nouns designating instruments, and in -irojd- < -terofd-, the suffix used for adjectives of comparison and contrast.
a. The following nouns and adjectives are made with -rojd-, generally as a primary suffix added directly to the root: adro atru ace. pi. neut. : L. atra (root and suffix uncertain, since there are no sure cognates outside of Italic, and the word may be of Etr. origin); agre: L. agri < *ag-rei, cf. ag-o (?); abrof: even though L. aper and most of the other cognates have r, the ro must be considered a suffix and not a true part of the root; so also with Kabru, L. caper; antakres: L. integris, \ tag-, cf. L. tango; ebetraf-e< ^e-bad-rd-: cf. L. baeto, but our form could also be derived from *e-baet{e)-trd with suffix as in b below; rufru: L. rubrum, eqvQqov, cf. rofu without suffix; sacra: L. sacra, cf. sancio, sanctiis; tefru < ^tep(e)s-ro- (57 g), -ro- being here a secondary suffix added to tep-es-; uiro: L. viros, cf. Skt. vlra-, all from *wi-ro-, if connection with L. vl-s, vi-m, etc., is to be maintained. naraklum and naratu imply an adj. *(g)nd-ro-:h.gndrus, cf. co-gno-sco, yvcogijuog, etc. Finally the adv. subra (with supru sese IV 3) < *sup-erd, cf. L. super, superus, supra, the suffix -ero- here having compara- tive value like that of -tero- in c.
b. The suffix -tro/d-, used chiefly for making names of instruments or articles having a specific function (cf. L. ara-trum, aQo-rgov, QYj-rQa) is clearly recog- nizable in the Umbrian nouns kletra, krematra, cringatro krikatru, fef ehtru, persondru together with the adjective venpersuntra (partly substantive in use). Discussion of these forms, most of which are partly obscure in etymology and meaning, is reserved for the Lexicon.
c. The suffix -terojd- is used in forming adjectives, together with some ad- verbs, designating position, ownership, etc., where a contrast of opposites is possible (cf. L. dexter, vester, deilrsQog, vpiereQoi; etc.). Syncope of the e is regular in the Umbrian forms. Comparative adjective: mestru: cf. L. magis, magister. Adjectives of position: destre testre: L. dexter, -tra, SsiiteQog; ner- tru: veQXEQOQ, ON nordr; postra: L. posterus; pretra; hondra hutra (on d < t see 60 a). Pronominal adjectives involving contrast of identity: etru: for sense cf. L. alter, including its value as ordinal numeral 'second'; putrespe, seipo- druhpei: L. utriusque, utroque. Pronominal adjective of possession: uestra: L. vestra. Adverb of place whence: apehtre ' extrinsecus ' : source of final e not clear, but eh = ek, and tr is undoubtedly from the suffix -terojd-, cf. L. ex-tra.
77. Stems in -i-. Here are included stems in i not a component of any longer suffix, and stems in -ti- (with -dti-), -ni-, -ri- (with -dri-), -li- (with -dli- and -feli-).
92 II. Grammar
a. In several of the following forms the i has the appearance of a primary suffix attached to the root, but it is not always possible to make a clear analysis into root and suffix: angif ace. pi. < *ank-i-ns (?); auif avif: L. avis, Skt. vi- 'bird'; out ace. pi., uvem ace. sg. : L. ovis, ofiq; sersi, where -i is evidence of an I'-stem in abl. sg., against the possibility of taking as a consonant-stem or a locative (cf. L. sedes partly inflected as an i-stem); spantim-af (?), uncertain and possibly to be analyzed as a stem in -ti-; see on III 33.
b. The IE suffix -ti- was productive in the formation of nouns of action, but the examples in the Iguvine Tables are too few to allow any generalization with regard to its normal function in Umbrian: ahtim-em < ^ag-tim-en: cf. L. ago, actio, the common Latin type in -tion- being an extension in -on- of the older type in -ti-, for which cf. f^d-aig, Skt. ga-tis, etc.; anderuacose antervakaze < *anter-wakat{i)s sit {*wakdtis : L. vacatio as ahtim : L. actionem); puntes nom. pi. 'groups of five' < '^ponk{'^)teijes, cf. 49 d with note 2. A stem ^sta-ii- must be recognized within imv. statitatu, an iterative verb made from the pass, pcpl. of a 4th-conjugation denominative to stati- (cf. ardaLg, L. stati-o), unless we derive i from H (15 b) and explain through a pcpl. equivalent to L. statutus.
c. Stems in -dti- include the dative forms Atiiefiate, Kaselate (beside nom. Casilos < -dt(i)s), Kureiate, Museiate, Peieriate, Talenate, all names of decuviae at Iguvium, and Tarsinater, -te, -tern, name of a hostile state. The evi- dence for the i-stem character of these forms is found in the closely cognate Latin forms Arpinas, etc., with I'-stem forms nom. sg. Arpinatis (Cato apud Prise, II, 337, 21 Keil), gen. pi. Arpinatium, optimatium, etc.
d. Stems in -ni- include: noun poni puni (abl. sg.), literally 'the drink' < *p6-ni- (?); adj. fons: cf. L. Fones, gloss, 'dei silvestres,' Faunas, faveo; perak- nem: etymology and meaning uncertain, but by any reasonable analysis the n belongs to the suffix rather than to the root; the I'-stem, however, may have replaced an o-stem by the same principle of formation as in L. triremis: remus; similarly in sevakni, which is also uncertain. A stem *persni- K'^persk-ni- must evidently be assumed to account for the denominative verb persnimu 'pray.'
e. Stems in -ri- (apart from those in -dri-, for which see f below) include: ocar ukar, ace. ocrem, abl. ocri: okqlq, L. ocris, root ak-, ok-; pacer nom, sg. < *pdc-ri-s (cf. 32 d), nom. pi. pacrer; peracri abl. < *per-ak-rld: L. acri; sakre ace. sg. beside the a-stem form sakra sacra, precisely like OL sacrem, etc., be- side standard -o/a-forms.
f. Stems in -dri- include Staflarem, and also sehemeniar if we take it as an adj. from the noun equivalent to L. semen and admit loss of the final i, but see on Vila 52, where an altogether different interpretation is favored.^
g. A possible example of a stem in -li- similar to L. facilis, habilis may be seen in abl. pi, ise^eles: L. ^insicilibus, unless we follow those editors who emend to isegetes, producing a pass. pcpl.
B. Formation of Substantives 93
h. Stems in -ali- include: uerfale: L. verbale; sorsalir gen., sorsalem ace: cf. sorsom sufum; Tefrali abl. sg.: cf. Tefrei. A stem *-lerdli- < *leisali- must be assumed in the formation of disleralinsust if we adopt the usual analysis, taking it as a 4th-conjugation denominative verb cognate with L. lira, delirus.
i. Adjectives in -feli- represent Italic stems in -fli- resulting from alteration of the IE stems in -clhlo- in line with the Italic tendency to inflect adjectives as i-stems. Since the anaptyxis which explains the i before / in L. credibilis, etc., cannot be assumed for Umbrian, the vowel must have been extended from the nom. sg. *-fel, in which the e could arise by samprasarana (32): fagefele^: -L. *facibilem (in sense *sacrificabilem); similarly purtifele: cf. purtitu purdi- tom, pass. pcpl. to purdouitu.
^ Apart from a few rare forms such as extaris, coquinaris, pecuniaris, paegniaris, the suffix -ari- exists in Latin only as a variant to -ali- produced by dissimilation in forms already con- taining / (e.g. mililaris in contrast to animalis). O. dekkviarim would be an exception to this rule, and also U. sehemeniar if taken as an adjective in -dri-. The dialect evidence is too scanty to permit the formulation of a rule, but the contrast between Staflarem and Tefrali seems to indicate that the distribution of -ali- and -ari- followed the same principle as in Latin.
2 Practically all editors thus emend the actual reading fagefete in lib 9, the only passage where the word occurs.
78. Stems in -u-.
a. Apart from stems with the suffix -tu-, which in Latin and several other languages form a productive class, the Umbrian u-stems include the following substantives: berva ace. pi., berus abl. pi.: L. verua, verubus; mani: L. manu {i < u by 15 a), loc. manuv-e; trifor gen. sg., trifo dat., etc.: L. tribus, etc.; pe- quo ace. pi.: L. pecua; castruo kastruvuf ace. pi.: u-stem on evidence of 0. gen. sg. castrous, in contrast to o-stem L. castra. The divine names Fiso, Trebo (both dat. sg.) must be regarded as «-stems, but both have in the earlier tables dat. sg. forms Fise, Trebe of the o-stem type.
b. The suffix -tu-, used in Latin in the formation of nouns of action and the result of action, is clearly represented in afputrati (i < u as in mani, a above): L. arbitratu; apparently also in Ahtu dat. sg. used in Ila 10, 11 of a divine power identified with Jupiter and Mars. Whether the etymologically obscure uaiuo vatuva is a stem in -tu- or in -tuo- cannot be positively determined.
79. Stems in -ie- ("Fifth Declension"). The origin of this type in Latin and the Italic dialects is partly obscure because of the lack of clear correspondences in other IE languages, but it is reasonably certain that it rests ultimately on a stem-class in -(i)f/e/r- varying in part with -(i)yall- (cf. L. durities: duritia). The Umbrian examples include : auie (contained also within the participle aviekate and the adjective auiecla); iouies. The fifth-declension noun re {-per), ri (:L. res) has already been Usted in 65 c among the root-stems.
94 II. Grammar
80. Stems in stops. Here are included stems in k, t, and d (Umbrian f rs). Several forms from roots ending in these and other stops, but containing no suffixes, have already been included among the root-stems (65 f) and are given only passing mention here.
a. Stems in -k- (including -g- -s- from k by 46 b) include: curnaco ace. sg., curnase abl. : cf. L. cornicem, -e; scalse-to, skalge-ta: cf. L. calice; neuters ace. sg. huntak, tuplak. For the root-stems pase, praco, and frif (^-stem *frug-) see 65 f.
b. Stems in / include: erietu: L. arietem, cf. sQicpoQ with suffix -tpo- < -bho-; Marte: L. Marti; frite < *fret-e, like L. dot-e according to Muller, p. 512, but the stem in -/- is only one of several possibihties. For the present participles, whose stems end in -nt-, see 130 a.
c. Stems in -d- (U.-f--rs-) include: capirse kapife: L. capidi; vapefe, uapef-e: L. lapide, (in) lapides. For the root-stems persi peri, dupursus, peturpursus, pruzure, and also for the root-stem uef (originally in -dh-) see 65 f.
81. Stems in n. Here are included masculines and feminines in -on- without gradation or with zero-grade in the obhque cases, feminines (and masculines?) in -ion- with -In- in the oblique cases, and neuters in -men-.
a. Stems in -on- representing (in part) the type of L. sermo, sermonis, include: abrunu ace. sg. : L. *apr6nem, cf. L. Apronius; homonus dat. pi. : OL homo- nes-} karu, karne : L. ca/o, carni, with zero-grade in the obhque cases. menzne < '^mens-en-i may belong here, but no case other than the ablative occurs, and from the attribute kurglasiu Ila 17 we cannot determine whether menzne is m., like L. mensis, or neut.
b. To L. legio, -onis, ordtio, -onis, etc., without gradation, corresponds an Oscan-Umbrian type based on the gradation -(i)y6n- : -In- (35 d): 0. nom. sg. tribarakkiuf, fruktatiuf, ace. medicatinom, abl. tanginud; U. natine: L. na- tione; ferine as if corresponding to L. *ferione from root of fero; tribfigu, tri- brisine.
c. Neuters in -men include: /lo/ne numem (sic) ace, nomner gen., nomne dat.: L. nomen, etc.; umen ace, umne abl.: L. unguen, unguine; pelmner gen. sg.
1 Similar in formation to abrunu, homonus, is probably esuf: O. essuf 'ipse,' as if = L. *eps6ns. On the evidence of this form and of O. fruktatiuf, etc., we must probably assume that U. karu, tribf igu originally had n by analogy with the oblique cases plus s by analogy with other nominatives of the third declension, the resulting -/ (< -ns) being then dropped.
82. Stems in r. Here are included nouns of agent and nouns of family-relation- ship. The root-nouns nerf, pir, and vepurus, whose stems end in r, have already been treated in 65 e.
B. Formation of Substantives 95
a. Nouns of agent (cf. L. orator, victor) are formed by means of the suffix -tor- without gradation in the paradigm: arsfertur, dat. arsferture, ace. arsferturo; kvestur (probably a Latin loan; see 46 h); Speture, name of a god, as if L. *Spect6ri; uhtur, ace. uhturu: L. auclor, -torem. The abstract nouns kves- tretie, uhtretie contain -tr-, the zero-grade of the suffix -tor-.
b. Nouns of family-relationship are formed by means of the suffix -ter/tr- in ablaut-variation. The Umbrian examples are: f rater frateer frater nom. pL, < '^frateres (26 h), fratrom gen., fratrus dat.; luvepatre dat., lupater voc.
83. Stems in s. Here are included stems in which a suffix ending in s has been attached as a primary suffix to the root. The vowel before the s has been in many instances lost by syncope, and the s has partly been changed by rhota- cism to r. From the identity of nom, and ace. sg., from adjective-agreement (e.g. in uirseto auirseto uas est Via 28, etc.), and from comparison with Latin and other languages it is apparent that the gender is prevailingly neuter. Um- brian examples are : ace. far < ^fars, gen. farer (in place of *farser by analogy with nom.-acc. far): cf. L,. far, f arris; nom. sg. mefs, abl. pi. mersus, < *med- (o)s, med{e)s-ufs; ace. sg. tuder, nom. pi. tuderor, etc. ;i ace. sg. eras < ^-aisus: 0. abl. pi. aisusis (?); ose if < ^•op-(e)s-i: L. opere, cf. U. osatu, but the etymo- logy of ose and the interpretation of the phrase in which it occurs are uncer- tain; pars: L. par (?); uas 'flaw' < *wak(o)s (?); nom. pi. uasor, ace. uaso, abl. vasur: L. uas, vasis, etc. (for lack of rhotaeism see 57 d).
^ Evidence that tuder is an s-stem is found in the denominative verb eheturstahamu etuf sta- mu, which when taken in connection with tader points to an original *-tudes-td- with syncope of e after the change d '^rs r. Whether nom.-acc. tuder has r as the result of final rhotaeism or by the analogy of the oblique cases (cf. L. honor in place of early hands by analogy with ho- norem, etc.) is uncertain, but the latter alternative seems preferable, since the nom.-acc. which we should expect would be *tudos or its reflex *turs (cf. mers).
84. Stems in r/n. This type of neuter noun, in which -n- of the oblique cases alternates with -r- of the nom.- ace, is believed on the basis of Hittite evidence to have constituted a fairly large class in Indo-Hittite or early proto-IE, but it usually survives in the recorded languages in scattered examples with secondary alterations of the stem (e.g. Skt. yakr-t, gen. yaknas, '^Jiag, ijjzarog, L. iecur, iecinoris, early iecoris; femur, feminis, etc.). There are no Oscan examples. The sole^ Umbrian example is ace. utur (t = d) 'water,' abl. une < *ud-n-: cf. vdcoQ, vdarog, Goth, wato, ivatins, Eng. water. Germ. Wasser, the n-stem being generahzed in this word in East annd North Germanic, the r-stem in West Germanic,
1 From Goth, fan ' fire, ' gen. funins we may infer that the stem represented by U. pir, ace. purom, Gk. nvg, nvQog was originally of the -r/n-type, but since no trace of this variation has survived in this word in Greek or Italic, it has been treated here as a root-stem in -r. ie-
95 IL Grammar
pru Ila 32, understood in the present work as ace. sg. 'liver', cannot be equated directly with L. iecur but must represent a transfer to the o-stem class: *yepr-o{m), unless, as Muller, p. 218, suggests, it is a misspelling of *iepur.
85, Nominal Compounds. Umbrian and Oscan in general resemble Latin in the types of compounds formed and in the extent to which compounds are used. The first member is an adverb of the class capable of being used prepositionally, or an inseparable prefix (e.g. a- an-, sei-, ve- ven-), or a numeral stem, or, rarely, the stem of a declinable noun or adjective. The second member is a noun or adjective stem, or, in the case of seipodruhpei, a pronoun. Apart from combi- nations of a- an- with the passive participle, compounds containing parts of the verb proper are not included here (see 134).
86. The compounds treated in this section are formed with prefixes which may be used as prepositions, or which at least have cognates in Latin which are capable of being so used.
a. The following compounds are hypostases; that is, the relation of the pre- fix to the second member is that of a preposition toward its object: antermen- zaru: cf. L. intermenstruanim; peraknem, if equivalent to L. perennem, literally 'through the year,' but see on Ila 5. traha Sahata, name of the place of one of the sacrifices in the lustration, has become a hypostasis in which the case is determined not by traha but by the construction of the whole expression in the sentence.
b. In peracri 'perfect' (semantic value of -ac- as in axfirj, dxjualog) the pre- fix has intensive force; cf. nsQiKaXXriQ, L. peroptimus.
c. In the following compounds the second member contains a verbal root and the whole is a noun of action or of the place of action : antervakaze anderuacose L. *intervacat{i)s sit; arkani: L. '^ac-ciniiim; afputrati; ebetraf-e: cf. L. (in) exitus; ehvelklu; ooserclom-e: L. ^obser{va)culum (?).
d. The following compounds are best regarded as agent-nouns formed direct- ly from the compound verbs to which they correspond:^ arsfertur: cf. L. ad-few, Prestote Prestate, name of a goddess: cf. L. prae-sto.
e. In the following compounds the second member contains a verbal root, and the whole constitutes an adjective similar in sense to an active or passive participle: abl. pi. ise^eles < ^en-sek-elif{o)s; prusegia: cf. proseseto, pruse- katu; pruzufe (65 f).
f. kumne loc. sg. ' in comitio ' is formed by attaching the suffix -no- directly to the prefix kom-, unless we adopt the more dubious view whereby ^komno is from *kom-ben-o- with prefix and root as in L. conventio. kumnahkle may be explained either as *komnd- (stem of a denominative verb to '^komno-) + -klo- (75 b) or as ^komn- + -dklo- after the analogy of other nouns in -dklo-.
1 Not only here but also in c and e we must consider the possibility that the compounding took place first in the verb, from which the verbal noun or adjective was subsequently formed.
C. Declension of Substantives 97
87. The compounds treated in this section are formed with prefixes which, as far as we can determine, appear only in composition and not as prepositions governing objects.
a. The following compounds contain the negative a- a- an- an- -|- a passive participle (in the case of antakres an adjective in -to-): aanfehtal (26 d): L. infectas; anhostatir: cf. hostatir; ansihitir: cf. sihitir; ase^eta; asuata: cf. snata; auirseto: cf. uirseto; antakres: L. integris.
b. sei- (cf. L. se-cerno, sed-itio, etc.) occurs in composition with the indef- inite pronoun stem in seipodruhpei 'separately for each' < *se{d)k^oter6k^e: L. sedutraque, Plaut., Stich., 106.
c. ve- ven- (cf. L. ve-cors, ve-sanus, and Lexicon for the form ven-) occurs in vepurus abl. pi. 'fireless' (?), to which the verb vepuratu appears to be a denominative; and also in the obscure venpersuntra vepesutra: cf. persondro persuntru.
88. The following compounds contain numeral stems as first members: dupla, tupler: L. duplus, second member being apparently from root *pe/- 'fold'; similarly tripler; tuplak 'fork': L. duplex; difue ace. sg. neut.: di(pvsg; dupur- sus and peturpursus, compounds of the bahuvrihi or possessive type: L. bipedi- bus, quadripedibus ; tribfigu, tribrisine < *tri-pediky6n-, -In-.
89. Compounds having noun or adjective stems as first members include: mandraclo mantraklu 'hand-wiper, towel, maniple': cf. L. manus and tergeo; seuacne sevakni, if the first member is identical with the stem in seuom sevum, but see on Ila 21.
C. Declension of nouns and adjectives
90. For the stem-classes of nouns and adjectives and for the gender-system see 64. The declensional system in the Italic dialects is sufficiently like that in Latin to justify the familiar arrangement under five declensions: the first declen- sion then includes a-stems, the second o-stems, the third consonant- and i- and diphthongal stems, the fourth u-stems, while the fifth includes the few scattered examples whose stems end in -e (79). But the difference between o-stems and consonant-stems is less strongly marked than in Latin (see 92, 94, 100 below on the gen. sg., ace. sg., and gen. pi.), while on the other hand the original distinc- tion between consonant-stems and i-stems is much better preserved than in Latin, where it has been largely effaced by analogical levehng (see 94, 96, 99, 101 on the ace. sg., abl. sg., nom. pi., dat.-abl. pi,).
98
II. Grammar
First Declension Second Declension
a-stems o-stems yo-stems
Sg. Nom.
Gen.
Dat.
Ace.
Abl.
Loc.
Voc.
Nom. Ace.
Neut.
PI. Nom.
Gen.
Dat.- Abl.
Ace.
Nom.-
Acc.
Neut.
|
xnuta mutu |
tases tagez pelsans fratrexs fratreks ager katel tig el |
Vuvgis Teteies aTsir( ?) |
|
to tar tutas |
popler kapres agre ponisiater ptmigate katles katle |
Coredier Kureties Fisie Kastrugiie Titis |
|
tote tute |
Tefrei Tefre Tefri $erfe Qerfe |
loui louie luvi luvie Grabouei Krapuvi Sansi Sagi |
|
to tarn iota tuta uestisiam uestisia |
pop lorn puplum kaprum pop to puplu vinu |
Graboui Graboue |
|
asa asa uestisia vestigia |
uinu vinu pihaclu tremnu somo |
Fisiu Fisiu kurglasiu |
|
tote tafle |
pople destre testre onse uze |
Atiiefie |
|
Prestota |
Grabouie Sage |
|
|
or torn or to esono esunu persclo persklum |
arkani tertim terti |
|
|
anclar iuengar famefias pumpefias |
liouinur Ikuvinus Ikuvinu gersnatur |
Clauerniur Atiiefiur Ikuvinus |
|
pracatarum urnasiaru |
pihaclo pihaklu |
Atiiefiu |
|
dequrier tekuries aseriater anzeriates klavles plenasier urnasier |
uerir veres vaputis seueir |
sehmenier aplenies Clauerni Klaverniie Uehier Vehiies |
|
merstaf mersta gomia kumiaf porca purka |
abrof apruf uitlu rufru ueiro |
filiu feliuf |
|
adro atru veskla vesklu |
aruio arvia arviu farsio fasiu |
C. Declension of Substantives
99
Third Declension Consonant-stems i-stems
Fourth
Declension
u-stems
Fifth Declension e-stems
Nom.
Ace.
Neut.
|
arsfertur affertur tribfigu karu serse zefef |
ocar ukar fons Casilos pacer |
||
|
farer nomner |
ocrer punes sorsalir |
trifor |
|
|
capirse kapife nomne |
ocre ukre |
trifo Ahtu |
ri |
|
curnaco uhturu |
ocrem ocre uvem uve perakni |
trifo trifu |
|
|
per si peri nomne pure |
ocri-per ukri-per sevakni sevakne |
mani mani |
re-per ri |
|
ocre ukre |
manuv-e |
auie |
|
|
lupater |
|||
|
pir pir nomne nuniem(sic) umen |
u erf ale |
||
|
f rater frateer frater |
foner puntes |
||
|
fratrom fratrum buo fratru |
peracrio |
||
|
fratrus kapirus |
aueis aves avis sevaknis |
berus |
iouies |
|
cap if kapi uapef-e vapef-e |
auif aueif avif avef uvef oui |
iouie |
|
|
uasor tuderor |
sakreu porakneu |
pequo berva |
100 n. Grammar
91. Nominative Singular Masculine and Feminine. The IE case-ending in a-stems was O, in o-, i-, and u-stems -s, in consonant-stems sometimes O, sometimes -s.
a. In a-stems in Umbrian, as in the IE languages generally, the stem is used without ending; for -a. > -o -u see 7 b.
b. In o-stems the stem-vowel is syncopated (29 i), but -s normally remains: pelsans < *pelsann(o)s < *pelsandos; fratreks, etc. In -yo-stems the y becomes i after syncope of o (32 a) : Vuvgis (on Teteies see 32 b) ; arsir with -r < -s by 57 e, if from "^alyos (: L. alius), but see on Via 6. Stems in -ro-, -lo- lose -s after syncope of the o: ager <*agros (32 d), katel, tigel.
c. Among consonant-stems there are no examples of masc. or fem. stems in stops, apart from such present participles as zef ef serse. Stems in -r- have leng- thened grade of the suffix, -tur < -tor, with zero-ending: arsfertur, kvestur: of. L. orator, etc. Stems in -on- (karu, tribfigu, pronoun esuf) may have had -s, in contrast to L. caro, legio, etc. See the discussion in 58 f.
d. i-stems in general retain -s after syncope of the i: Casilos < *-ldtis; fons. But -s is lost in -ri-stems: pacer < ^pakris (32 d), ocar (see Lexicon).
e. Nom. sg. forms of the fourth and fifth declensions are not attested.
92. Genitive Singular. All O.-U. forms rest ultimately on IE forms with ending -s; there are no traces of the ending -i seen in L. virl, or of -osyo seen in tnnoio, Skt. asvasya. The -s is rhotacized to -r in Tables V, VI, and VII (57 e), but forms without final consonant occur sporadically in both the earlier and the later tables (57 h). In a-stems the s is added directly to the stem: cf. L. (pater) familias in contrast to the usual -ae < -dl. o-stems and consonant-stems use the ending originally peculiar to i-stems: -es -er -er -e -e, all from -eis (cf. 0. aeteis, sakarakleis). The ending in Titis is anomalous but cannot have a different origin from that of o-stems in general. The ending for i-stems is based on the full ("normal") grade of the stem- vowel (-eis) and is inherited: cf. Skt. agnes, Lith. antes. The u-stem trifor has -or < proto-It. -ous < IE -eu-s or -ou-s (23). No gen. sg. forms of the fifth declension are attested.
93. Dative Singular. The history of the forms is complicated, but, with the exception of those which were originally locative, all are derived ultimately from an ending -ei or -ai,^ which in combination with the stem-vowels a, o pro- duced long diphthongs.
a. In d-stems the -di became 0. -ai, U. -e -e through shortening of the first element (25 a).
b. In o-stems the -oi became 0. -ui, U. -e -e through -oi (25 b): cf. OL Nu- masioi in contrast to the standard form with -6 derived by loss of the i of the long diphthong.
C. Declension of Substantives 101
Stems in -(i)yo- may show contraction, especially in the later tables: Sa^i Sansi, Fisi, etc., and even plain o-stems occasionally show forms in -i -i or -ei: Tefri Tefrei, fratreci.
c. In the third declension, consonant-stems and i-stems alike end in 0. -ei, U. -e -e, but the precise history of the endings is obscure. It is not necessary to assume the same origin for the ending in the two stem-types, but it is ex- tremely probable that the development was the same in the two dialects. The forms have been explained as original locatives with the ending belonging to i-stems, but it is possible to refer them all to a dative ending -ei added to con- sonant-stems or to i-stems with full grade of the stem-vowel (thus ocre < *okreei < *okreyei, 53 b), the ultimate result being the same for both stem-types.
d. The ending -u -o of iz-stems is from -ou < -eu or -ou (23), originally a locative formation with the bare stem in full grade or possibly in lengthened grade -eu or -ou; cf. Skt. sunau, and for the Itahc formation cf. L. curru, cornu, etc., in contrast to the standard -ui < -(u)wei with proper dative ending. There are no Oscan dat. sg. forms of u-stems.
e. The fifth declension form ri can be analyzed as either original dative or locative: the case-ending dat. -ei or loc. -i was added to the stem in -ey, and after loss of the y (53 b) the vowels contracted.
^ The difficulty in reconstructing the IE dat. sg. arises from the fact that forms like Aiei- (TQsq)rjQ), Cypr. Gk. Aifeiicpikoq) etc., point to -ei, while xafiai and the infinitives in -ai (e.g. donsvai point to -ai. Whatmough, C.P., L (1955), p. 151, suggests postulating an ending hi as well as -ei, -j. This ending-6i, if I understand it correctly, would be the reduced ablaut- grade of -ei and would provide a source for the Gk. forms in -ai.
94. Accusative Singular Masculine and Feminine. The original ending was -m for all vowel-stems and its allophone m for consonant-stems. In Umbrian the same system in general prevails, but in consonant-stems the -om of o-stems has been adopted; cf. the similar Oscan forms leginum, tanginom. In -yo-stems the y becomes i after syncope of the o (32 a). No forms of the fifth declension are quotable. In all stem-classes the -m -m is subject to loss (2 j, 51 e).
95. Nominative-accusative Neuter. From proto-IE the Italic branch in- herited neuter o-, i-, u-, and consonant-stems, but no nom.-acc. sg. forms of u-stems are quotable from the dialects. The history of the o-stems, including -r/o-stems, is identical with that of the accusative of masculines (94). Conson- ant-stems use the bare stem without ending; for the stem-variation in utur, abl. une, see 84. i-stems use the bare stem with final e < i (so uerfale: 10 b).
96. Ablative Singular. The ending -d, which in proto-IE was peculiar to o-stems, was extended in Italic to vowel-stems in general. The stem-vowel is
102 II. Grammar
regularly long after the analogy of the -6d in o-stems, probably with support from the -dd of a-stems. The -d is regularly lost in Umbrian (42 e) as in classical Latin, but is found preserved in 0. toutad, dolud, slaagid, etc. Consonant- stems use -e < -i (10 b), originally a locative ending; cf. the similar ending in L. rege, etc., in contrast to Oscan tangimid, ligud, with o-stem ending.
97. Locative Singular. All forms can be referred to original endings -i or O- a-stems have -e -e, 0. ai -ae< -a -\- i (25 a), the development after contraction being identical with that of the dative, o-stems have -e -e, 0. -ei -ei, from -e, the ablaut-variant of o, + i.^ Consonant-stems have -e -e < -i and are thus iden- tical with ablative forms, which have adopted the locative ending (96). i-stems have -e, which may be from full-grade -ei + U through -eyi > -ei; no Oscan examples. For the u-stem form see 93 d, the dative having been originally lo- cative. Locative forms sometimes are followed by the postposition -en -e, written separately or together. In scalsie (consonant-stem) the -e prevented the change of the preceding i to e. In manuve the glide v shows that the diphthong ou was preserved before the -e, in contrast to the otherwise similar dat. trifo with monophthong -o < -ou. It is possible that many locative forms are actually based on contraction of the ending with the postpositive -en and omission of the final n.
1 Greek gives evidence for both -ei and-oz (e.g the adverbs itcsl, olxoi), while L. -i and Slvt. -e are ambiguous, but the Umbrian forms probably agree with the Oscan, which clearly point to -ei. Moreover the -i -i < -oi, which alternates with -e -e in the spelling of dative forms of o-stems, is not found in the locative.
98. Vocative Singular. In IE languages generally the voc. sg., in so far as it is distinct from the nom., is regularly without case-ending, although the stem-vowel frequently shows ablaut-variation. The Tables show voc. forms of the first, second, and third declensions.
a. a-stems regularly have -a, never -o as in the nom. Since there are nearly fifty examples, it is not possible to regard the voc. ending as identical in origin with the nom. ; it must therefore represent -a < -d, the reduced grade of -a, as in Homeric vvfi(pa, xo^ora.
b. o-stems have -e, ablaut-variant of o. In -(i)j/o-stems this vowel remains uncontracted, in contrast to L. Tulll, etc.
c. The consonant-stem lupater shows a voc. form in both members; cf. Zev TidrsQ, L. luppiter. The e is short, as may be seen by comparison with nom. 0. patir, where i (in place of the expected i) stands for e, cf. nartjQ.
99. Nominative Plural Masculine and Feminine. Umbrian and Oscan have preserved the inherited endings based ultimately on -es, which contracted with
C. Declension of Substantives 103
the stem-vowel in proto-IE; similarly in Sanskrit and Gothic, in contrast to %MQai, Innoi, etc., L. terrae, equi, etc., which have the pronominal ending. a-, 0-, i-, and consonant-stem forms are attested.
a. a-stems have -as in the older tables, -ar (57 e) in the later. Cf . Skt. senas, Goth, gibos.
b. o-stems have -us (occasionally -u, 57 h) < -6s < -oes in the older tables, -ur in the later, including Va. Cf. Skt. vfkas, Goth, wulfos.
c. i-stems have -es in the older tables, -er in the later. The IE case-ending -es was added to -ei, the full grade of the stem-vowel, giving -eyes > -ees (53 b) > U. -es -er, 0. is -is, L. -es.
d. Consonant-stems must originally have had -^s (cf. (pv?.aK-S(;, Q'^roQ-sg, etc.) in contrast to Latin, where the ending -es of i-stems was adopted also by consonant-stems. O.-U. -es lost the e by syncope: 0. humuns, 0. jLieddeii, meddiss. The Umbrian examples all belong to r-stems and show not only syn- cope but subsequent loss of the r with some evidence of compensatory lengthen- ing (59 f).
100. Genitive Plural. The original ending was -6m,^ or a combination con- taining it, for all stem-classes. Italic dialect forms are attested in the first, second, and third declensions.
a. a-stems have -arum -aru: 0. -azum -asum, L. -arum, Epic Gk. -acov, all from -dsom borrowed from pronouns. The -um (not -om) of pracatarum is common- ly taken as evidence that the original vowel-length of the ending is preserved in this and other gen. pi. forms (27 b).
b. o-stems, including -yo-stems, have -u -o, -iu -io from -6m, -y6m with loss of -m: 0. -um -om, -ium, L. -um in archaic virum, etc. Standard L. -drum, modeled after -arum of a-stems, has no parallel in the dialects.
c. Consonant-stems have -um -u -om -o, i-stems -io, all formed by addition of the ending -6m to the stem as in Oscan and Latin.
1 With some evidence of an ablaut-variant -em in Gotliic.
101. Dative- Ablative Plural. There are two endings, one for the first and second, the other for the third and fourth declensions, as in Latin. No fifth-declension forms are quotable.
a. a-stems have -es, -er -er: 0. -ais -ais -ais, L. -is, with early -eis -es, all from -ais in imitation of *-ois, the ending for o-stems, or perhaps rather from
ais in imitation of -ois, the earlier form of the o-stem ending (25 a).
b. o-stems have -es, -er -e -ir -eir -er, once -i: 0. -liis -ois, L. -Is with early
104 II. Grammar
-eis -es -oes, all from -ois < -ois (25 b), an ending originally instrumental: cf. Skt. -dis.
c. i-stems have -is -es -is, once -eis (aueis): 0. -ifs -iss, L. -ibus, all from -i + -fos < -bhos with O.-U. syncope of the vowel in the final syllable. 0. luisarifs, the sole form with / preserved, gives evidence of the survival of the bhos-ending in the Italic dialects, and the analysis commonly given receives support, where the Umbrian forms are concerned, through the failure of the -s to suffer rhota- cism (57 e) or to be omitted in writing.^
d. Consonant-stems have -us -us adopted from u-stems, in contrast to both Oscan -is -is and L. -ibus adopted from i-stems.
e. u-stems have -us < -u + -fos with derivation as in c above.
1 sevakne abl. pi. IV 9 is not a normal case of omission of final -s but the result of lack of space at the end of the line.
102. Accusative Plural Masculine and Feminine. The IE ending was -ns for vowel-stems or its allophone -ns for consonants-stems. Umbrian forms are quotable in all declensions, if we count kastruvuf, neuter u-stem with ending as explained in 103 below.
a. a-stems have -af -a -af -a (58 b) : 0. -ass -as, L. -as. Skt. -as, Goth, -os reflect an early loss of n in the ending -dns, but the O.-U. forms require the as- sumption of an -ns in proto-Italic. The -ns may have been adopted from o-stems, where n was not subject to early loss after the short o, or it may represent an inherited variant -ans with shortening of a and preservation of n.
b. o-stems have -uf -u -uf -u -o (58 b; on the vowel-length indicated by -u see 26 e; on turup see 2 k) : 0. -uss, L. -os, Cretan -ovg, Goth, -ans, all from -ons with compensatory lengthening in Latin after loss of n.
c. i-stems have -if -i -ef -e -if -i -eif (58 b): Oscan not attested, Latin -Is (the rival form -es being taken from consonant-stems). These endings are from -ins, with resulting long vowel inferred from spelling -eif (usually, though not always, an indication of length) and from the parallel development -ons > -of > -uf.
d. Consonant-stems have -f -/ or O, with loss of d before / in vapef (cf. L. lapid-es), capif (cf. L. capid-es), of g before -/ in frif (cf . L. fruges), of / < dh in uef (see Lexicon). The absence of a vowel before -/ in these forms and in 0. inalaks,^ usurs^ is difficult to explain. It could not have been the result of syn- cope, since proto-Ital. -ens < IE -ns should have lengthened its e (cf. the treat- ment of -ons, -ins in b, c above^. It is possible, however, that the nom. pi., where the e in -es was syncopated, may have exerted an analogical influence which could be expressed by the proportion *vitlas: vitleif = * vitlus; vitluf == pTin- tes: *puutef = *vapes: uapef.
D. Numerals 105
e. The diphthongal stem buf buf is readily explained as from bo (cf. Dor. /5a>g with monophthongization of the long diphthong in ^g'^ou-) + -ns. Similarly sif sif si from the long-vowel stem su- (i < u by 15 a).
f. On the u-stem kastruvuf see 103 a below.
^ We might expect -ss as in the o-stem forms feihuss, etc., with -iiss<; -ons, but the two consonant-stems here cited occur on the Curse of Vibia (Co. 130), where gemination is never expressed in writing.
^ It is altogether unlikely that -ons -Ins were inherited directly from proto-IE; the evidence of Cretan eXevdeqovq, Goth, wulfans, etc., is against such an assumption.
103. Nominative-Accusative Plural Neuter, o-, i-, u-, and consonant-stems are quotable. The ending is, in part, -a -a -u -o: 0. -u -o, L. -a, all from original -a. The forms in -u (0. -u) -o are evidence that the vowel-length was preserved in the dialects in contrast to Latin, where a is shortened to -a (cf. the similar treatment of -a. in nom. sg. of a-stems).
a. But in Umbrian, without parallel in Oscan, some neuter forms were al- tered by the addition of nom. pi. -r, ace. -f to the neuter endings in -o -u. Nom. : o-stems: totcor, screihtor, arsmor, dersecor, subator; consonant-stems: uasor, tu- deror. Ace. : o-stem : veruf : u-stem : kastruvuf . The classification of these forms as neuter, despite their endings, depends in part on the coexistence of such un- mistakably neuter forms as krematra, ace. sg. tuder.
D. Numerals
104. The remains of the Italic dialects are not extensive enough to allow more than a fragmentary account of even the cardinal and ordinal numerals. Yet the stems of the numerals from 1 through 6, and also of 9 and 10, are all repre- sented in the Iguvine Tables in one form or another, and the numerals for 2 and 3 show several different inflectional forms.
a. The numerical symbols IIII, V, VI, VII, X, XV occur in Table Vb 8-18 (Latin alphabet), CCC in Vllb 4 (Latin alphabet); XII in lib 2 (native alphabet).
b. The form-classes in which numeral stems occur are: cardinals, ordinals, distributives, multiplicative adverbs, compounds of several types with numeral stems as first member, and miscellaneous noun and adjective formations.
105. The numerals and their various derivatives are treated in numerical order in the following paragraphs.
a. 1. The cardinal is unu, ace. sg. Ila 6, 8, <*oinoni: L. unum, OL oino, oivrj ' one on dice, ' Goth. ains. The ordinal is prumum promom ace. sg. with ad- verbial value: cf. tiqojuoq 'chief. '^ The multipUcative adverb is sumel 'once': cf, L. simul, semel, and 28 b.
106 II. Grammar
b. 2. The vestigial dual form represented by L. duo has been replaced in Um- brian by a nom. pi. form dur, ace. tuf, ace. neut. tuva, dat.-abl. duir tuves. The nom. dur < *duw6s resulted from vowel-contraction after the 6 had been raised in the direction of u (33 c). On u and uv see 2 h. The ordinal is etru, etre, etc., with normal -o/a-stem inflection, from *e-tero- (stem 109 c, suffix 76 c), its ordinal use being precisely similar to that of L. alter. The adverb duti 'for the second time' is apparently an analogical formation after tertim. dupla ace. pi. f., tupler abl. pi. m. are identical in formation with L. duplus but are used with the distributive value of L. binas, -is. tuplak ace. sg. neut. '(two- pronged) fork'(?) is similar to L. duplex, du- appears as first member of the compound dupursus (88). The compound dlfue ace. sg. (88) has as first member di- < dwi: dupvrjQ, cf. L. bi-fidus, Skt. dvi-jas.
c. 3. The cardinal is trif tref ace. pi. m.-f., triia ace. neut., tris abl., with stem tri/trei- and regular I'-stem inflection (58 b, 102 c, 103). The ordinal is tertie dat. sg., tertiam-e ace. sg. f., etc., with regular o/a-stem inflection: L. tertius. tripler abl. pi. is precisely similar in formation to dupla, tupler (b above). The multiplicative adverb trioper triuper 'three times' is formed by the addition of -per < -pert to the neut. pi. ^triya: cf. 0. petiropert, L. semper for suffix. For the compound tribrisine see 88 and Lexicon.
d. 4. This numeral appears as first member of the compound peturpursus^: cf. du-pursus in b above, and for the form of the stem cf. Skt. catur-; for p < k'^ (L. quattuor, etc.) see 49 a.
e. 5. The cardinal, which is not attested in any dialect of the Oscan-Umbrian group, was apparently O.-U. ^pompe < ^penk^e with o < e (8 g): nevre, Aeol. nipiTiB, Skt. panca, L.^um^ue with initial ^u < p. ^ puntes nom. pi., puntis abl. pi., ' pentads, groups of five, ' is a -^i-stem noun (77 b and for phonology 49 d with n. 2). pumpefias nom. pi. lib 2 is of uncertain formation and meaning (for discussion see introductory note on lib), but is undoubtedly a derivative of the numeral for 'five'.
f. 6. The only form is sestentasiaru gen. pi. f. 'bi-monthly' from *seksto- (: L. ordinal sextus) + -ent-dsioja, the -ent- in contrast to L. sextantarius being possibly the result of analogy with an e-form corresponding to L. trientarius ; see 61 d, 68 d.
g. 9. nuvis 'nine times' is apparently made from *now- (cf. L. novem, novie(n)s) with -is after *dwis, ^^tris.'^ nuvime 'for the ninth time' is an adverb in -e(d) from an ordinal stem ^nowimo- like L. septimus, decimus, according to the usual view, but see on lla 26. The proper name Noniar gen. sg. Via 14 may belong to the root of the numeral 'nine'.
h. 10. There are two derivatives of the numeral 'ten.' tekvias nom. pi. lib 1, signifying originally tenths of the population, is an -lya-stem derivative
E. Pronouns 107
of *dekw-, which may be assumed as an ordinal stem built after the analogy of some form corresponding to L. oddviis, oydofog. tekuries dequrier abl. pi. 'of the decuviae ' is an adjective in -uryo-, a suffix derived by reanalysis of * k'^etur-yo-, in which *k^etur- is an ablaut-variant of *k^etwer-o- 'four'.^
i. 12. desenduf ace. pi. occurs in Vllb 2; from *dekm-duwo remodeled with pi. -/ < -ns in place of dual ending (see b above).
^ L. primus is from *pri-is-mos: Pael. prismu.
^ Possibly also in the proper name Petrunia: L. Petronius, cf. Pomponhis, if from the nu- meral ' 5 ', the Latin forms being necessarily dialectal (p << k^") if these etymologies are ad- mitted. Oscan forms include petora 'four', and the more recently discovered form pettiur on an inscription from Samnium (Pi. 35 D = Vet. 141 = Bott. 5 b).
^ Since the assimilatory change which produced L. quinque<C *penk'"'e was very early, having affected Celtic as well as Latin, we must assume that the proto- Italic form was *k^enk'^e, whence O.-U. *pompe with both stops reconverted to p (49 a).
* Not like L. noviens, since -ns would become / by 58 b.
^ For the etymology and interpretation of these numeral derivatives, which have been much debated, I follow Devoto, Atti del reale istituto veneto, LXXXIX (1929-30), pp. 927-36.
E. Pronouns
106. For our purposes, the pronouns may be divided into personal, including the possessive pronominal adjectives; demonstrative; and interrogative, indefinite and relative. As in Latin, the personal pronouns show no distinction of gender, the interrogative-indefinite has distinct neuter forms in the nominative and accusative, while the remaining pronouns show the gender of the substantives to which they refer. Several of the pronouns in oblique cases are capable of being used with adverbial or conjunctional value, and the line of demarcation between such uses and the true pronominal uses is not always clear. For this reason and in order to fill gaps in the paradigms several of these adverbial forms are included in the table and discussion.
a. Case-endings are for the most part the same as in -o/a-stem adjectives. Final s is rhotacized in the later tables and also before enclitics beginning with vowels (e.g. nom. pi. m. eur-ont < *ey6s-hont). Nom.-acc. neut. forms were made with the ending -d (cf. 0. id-ic, L. id, Skt. tad, etc.) with loss by 42 e in este and r rs < d (42 b) in ef ek erse, pife pirse, etc. On the ending in esmei see 109 c.
The following enclitics are attached to certain pronominal forms:
b. -k -c (cf. L. hi-c, hun-c, tun-c, illi-c, etc.) in abl. sg. eruk, f. erak, etc. The nom. sg. m. erek erec and nom.-acc. neut. erek contain an enclitic combination from -ik (cf. 0. izic, idic). Many forms not showing -k -c probably had it and subsequently lost it (62 a; observe also ere ere beside erek erec). -k -c should not be confused with the postposition -ku(in) -co{Tn), which does not lose its vowel.
c. -hunt -hont after vowels, -unt -ont after consonants, in the pronoun of
108 II. Grammar
identity erihont, and also in the adverb isunt. This enclitic is in origin probably an ace. sg. *hom cognate with L. hi-c hun-c, etc., further extended by means of -t (cf. per-t). On the abl. sg. f. erafont with / in place of the normal h see 48 b. d. -i -i -e -e -ei in many forms of the interrogative and relative pronouns and related conjunctions: poi poe poei, pure porse, pisi pisi, pire pirse, puri, pafe, pifi. i (cf. ovroG-l-) must be assumed in order to account for the forms written with -i -i, since final i normally became e (10 b). The relation between this enclitic and the i of erihont is uncertain; it is not possible to bring it into connection with the second vowel of erek erec, 0. idik, without assuming a variant -i- as a possible basis for the vocalism of these forms.
107. The extant forms of personal pronouns, including the reflexive, are:
a. First person. Dat. sg. mehe, from *meghei: L. mihi with vowel weakening e > i in first syllable because of unaccented use, and (for the most part) iambic shortening in second syllable. Cf. also Skt. mahyam < *meghi + enclitic par- ticle -am.
b. Second person. Dat. sg. tefe tefe, from *tebhei: 0. tfei, L. tibi with vocalism as in mihi above. Cf. also Skt. tubhyam with formation as in mahyam above. Ace. sg. tiu tiom tio teio, from te + om of doubtful explanation ; possibly adopted from ace. sg. of o-stems, or possibly enclitic as in Skt. nom. sg. tvam, dat. tubhyam, etc. Cf. also 0. nom. ( !) sg. tiium.
c. Third person reflexive. Dat. sg. seso, generally analyzed as from *s{w)oi (: ol) + -so, an enclitic of unknown origin^
1 The final syllable in svesu lb 45, Ila 44, sueso Vllb 1, all loc. sg. of the pronominal adj. equivalent to L. suus, must be of the same origin. The most natural supposition is that it is itself a form of some pronoun used with intensive value, a single grammatical case being used in all situations, with the first member alone inflected, as in early L. eapse, eumpse, eopse, eaepse, etc. Our forms may be explained on the basis of an original ace. sg.if we admit loss of -m in the dat. sg. pronoun seso (VIb 51 only) and in the three examples of the adjective svesu sueso. In any explanation it is necessary to account for the lack of rhotacism of the second s. If as the second element we take ace. sg. *esso{m) *epsom (: L. ipsum), we may derive seso from *soi essom, with *soi as a variant of *swoi (Gk. ol, oi) and -e < -oi by 21 d, the syllable being unaccented; or possibly the first component suffered elision before *esso{m). The fact that esui, O. essuf has an extended stem *esson (58 f) is not a serious obstacle; although common to Umbrian and Oscan it is probably of relatively recent origin, while the simpler *esso- may have been used in the combinations here under discussion.
108. The forms of possessive pronominal adjectives are:
a. For the second person singular. Gen. sg. tuer touer, abl. sg. tua tuua: from *towold- <*tewo/a- (23): Hom. reo'g, L. tuus (with vowel-weakening).
b. For the second person plural. Abl. sg. uestra: from *wes-, ablaut-variant of *wos, wos (cf. Skt. vas, L. vos) + -tero-, suffix of comparison and contrast (29 a, 76 c).
E. Pronouns
109
c. For the third person reflexive, Loc. sg. svesu sueso: from swo/a- (: Gk. possessive og) + fossihzed ace. sg. *esso{ni) generalized first in the reflexive pronoun and then in the pronominal adjective, according to the suggestion in 107, n. 1.
109. The demonstrative pronouns are shown in the following table. The sep- arate pronominal stems are discussed in the paragraphs following the table. The forms used in the paragraph headings are the same as those in the Lexicon. Forms used adverbially have been omitted, even when, as in the case of ulo ulu, they contain recognizable case-endings.
erec erek erihont esmei este, estu
esuf orer ures
Sg. Nom.
Gen.
Dat.
Ace.
N.-A. Neut.
Abl.
Loc.
PI. Nom. Gen.
DaL- Abl.
Ace.
N.-A. Neut.
|
erek ere erec ere |
erihont eront |
esuf |
||||
|
erer irer erar |
erarunt |
orer ures |
||||
|
esmik esmei |
||||||
|
earn |
estu |
|||||
|
efek erse |
este este |
|||||
|
eruk erak eru-ku eru-com |
eruhu erahunt erafont |
esu essu esu-ku (eh)esu esa |
||||
|
esme |
||||||
|
euron t |
||||||
|
eru erom ero |
esumek esome |
|||||
|
ererunt eriront |
esir isir esis-co |
|||||
|
eaf eaf |
estu esto |
|||||
|
eu eo |
110 11. Grammar
a. The anaphoric erek erec: 0. izic, cf. L. is, ea, id. A comparison of the O.-U. and Latin forms shows that originally the nom. sg. m. and nom.-acc. sg. neut. had i-, the other cases ey- ; the e- < ey- in the Umbrian forms seems to have spread through the whole paradigm by analogical leveling. The Umbrian r r, Oscan s z in gen., abl., and (0.) loc. forms was generalized in these cases through reanal- ysis of forms where the case-ending contained s, thereby producing a stem U. erojd- < *eisojd-. The most probable source of the s was the pronominal gen. pi. ending -som : cf . Skt. tesdm, f . tdsdm, etc. The enclitic -ek -ec is used in nom. and ace. sg. and gen. sg. ererek, and -k in abl. sg.
b. The pronoun erihont 'the same.' The first part is the same as in erec but without the enchtic -ec -e: cf. L. is and the first part of Idem, 0. isidum. On the vocalism see a above; on the i of erihont see 106 d. On the enchtics -(h)unt -{h)ont, -font see 106 c, 48 b.
c. The pronoun represented by dat. sg. esmik esmei, loc. esme, ' this. ' These forms are from the IE demonstrative stem *e- which appears also in etantu (cf. L. tanta), etru (105 b), L. e-quidem, i-xei, e-xelvog, and probably also in the Greek verbal augment. These forms are cognate, as regards both stem and case-ending, with Skt. asmai. For -k in esmik see 106 b.
d. The pronoun este, estu, etc. 'this', with reference sometimes to that which has just been mentioned (e.g. Via 15, b 62, 63, Vila 51), sometimes to that which is newly introduced (e.g. lib 23, 24). Although este has not the special nuance of association with the second person, it is generally regarded as related to L. iste; yet the etymology of both is uncertain. The first part may be the stem e~ of c above, altered in Latin to i after is, id; the s is of uncertain source but may be connected with the s in esmei esmik; the second part is probably the same stem as in Skt. tad, tasya, the Greek article to, rev, etc. Among the inflec- tional forms ace. sg. neut. este (in contrast to L. istud) probably has -e < -id by analogy with *id, *pid reflected in er-ek, pir-e.
e. The pronoun esu, etc., 'this', and the derived adverb eso, etc., 'thus'. The Umbrian forms have s throughout, ss in abl. sg. essu and adv. issoc. Like 0. eksiik, exac, etc., they have the same semantic value as L. hie, haec, but the Oscan forms have ks x (= U. s[s]) in dat., abl., loc, k in nom. and ace. The initial e is probably the pronominal stem discussed in c above. The next part may be from *ke-: cf. L. ec-ce, huius-ce, etc., perhaps also cis, citra. The origin of the s is uncertain but its distribution in the Oscan paradigm is parallel to that of s in 0. izic, U. erec, and it may be from the same source (see a above). The derivation then may be from *ekso- < *e-ke-s-o-, the medial vowel being necessary because otherwise -ks- would give 0. -ss-. For addition of the encli- tics -c -k, -ek see Lexicon under eso and esu.
f. The pronoun esuf: 0. essuf esuf 'he, he himself, the intensive meaning being clearest in the Oscan Tabula Bantina (Co. 28), line 19. The etymology is
E. Pronouns HI
as uncertain as that of L. ipse, to which it is commonly assumed to be related. The -uf appears to be from -6ns, the nom. sg. ending -s being added to a stem extended by means of -on, as in Latin agent-nouns praedo, etc. (for -/ < -ns see 58 f).
g. The adverb ulo ulu 'there, to that place: cf. 0. gen. sg. f. ulas, L. olim. The vowels of both syllables are probably from 6 (on the development of o in Umbrian see 13; on the ending see 113 c).
h. The pronoun orer ures, etc. The use is mostly equivalent to that of L. hie 'this, ' but in lb 18 = VIb 55 it has rather the value of the anaphoric pronoun L. eo ' (with) him. ' There are no cognate forms in Oscan and the etymology is wholly uncertain, since the initial vowel may be from 6, 6, oi, or an u-diphthong, and the r may be original or from s.
i. The adverbs enem ene eno enu enumek, etc., 'then, next, after that'. All these forms are related to 0. inim 'and,' L. enim, Skt. instr. sg. anena, Russ. on, ona, ono 'he, she, it,' the various Italic meanings being derived ultimately from a demonstrative value similar to that of L. ilk. If L. nam, nempe are also related, as seems likely, the initial vowel must be of separate origin and is probably the same as e- in c above, but 0. i- can only be from i or e, not from e. It may be possible to reconcile the various cognate forms by assuming several ablaut-variants, but no explanation is certain.
j. The stem so- found in seso (107 c, with n. 1), 0. sifei, L. sibi, etc., possibly but not certainly related to *so, *sa, Skt. sa, sd, 6, rj. Here belongs probably surur < *s6-s6-s (?), suront, sururont. From "^swo-, a variant of so-, comes sue sve as well as the pronominal adjective sueso (108 c).
k. The stem to/d-: rov, rijv, ro, etc., Skt. tarn, tdm, tad, Goth, pata, etc. This stem is found in the second part of the adjective etantu (: L. tanta) from *e-tdm-td. For the e see c above. The origin of the final syllable is uncertain.
110, Interrogative, Indefinite, and Relative Pronouns. All forms are derived from a group of stems with initial /c"', which in Umbrian and Oscan be- came p, but in Latin remained as qu (49 a). The stem /c^i- (O.-U. pi-) in the Italic languages was partly specialized for interrogative-indefinite use, the stem k^o/d- (O.-U. pojd-) for relative use, but there are no Umbrian examples of the interrogative use (interr. pis once in Oscan, in no. 164 Co.). The ace. pi. form pifi as relative with definite antecedent occurs in Vllb 2. The stem k'"u-, which might have been expected to appear as cu- ku (49 e) regularly appears as pu, apparently through the analogical influence of related forms. In addition to the pronouns a variety of adverbs and conjunctions are derived from the stems pi-, pojd-, pu-, and it is not always easy to distinguish these uses from true pro- nominal uses.
112
II. Grammar
Forms with Stem pi- Masc.-Fem. Neuter
Forms with Stem po/a Masc. Fem. Neuter
|
pis pis -pir pisi |
pirse pife pere |
poi poei porse porsei |
(sue) po (sve) pu |
pure |
|
pusme |
||||
|
pora |
||||
|
puri pure |
porsi |
|||
|
pifi |
pafe |
porse |
Sg. Nom.
Gen. Dat. Ace. Abl. Loc.
PI. Nom.
Gen.
Dat.- Abl.
Ace.
a. The enclitic -i -i -e -e is extensively used in both the indefinite and the relative pronouns (see 106 d). The -s of nominative case-endings is partly rho- tacized before the enclitic (57 c), but pisi preserves its s through the influence of pis. On the other hand pir (in sopir) with r from final s occurs in VIb 54,
b. In addition to the forms described in a above, the following combinations of pis occur: svepis, svepu suepo < *swai k^a, cf. L. si qua; pisher like L. qui- libet 'anyone at all,' the second part being probably from *herit 'wishes' with enclitic use causing syncope of i and loss of /; pisest 'whoever is,' with enclitic use of est; pisi pumpe 'whoever' (112 e, 113 1).
c. The forms parse, porsi, porsei, pure may contain an enchtic from -di (42 b) arising through reanalysis of such forms as nom.-acc. sg. neut. *k^od-l, abl. *k^dd-l. The nom. sg. m. forms may then be from *k^oi- di (o u for o < oi, 21 a) and the nom.-acc. pi. neut. forms from *k^d-di with a <: a either by normal development (7 b, c) or analogically. The nom. pi. m. forms are more difficult; since d would not change to rs after a consonant, the rs must be analogical (42 c), unless the loss of s in the nom. pi. ending occurred sufficiently early to leave *p6-dl.^
F. Indeclinables 113
d. Dat. sg. pusme corresponds closely to Skt. kasmai. For the ending cf. esmei 109 c.
e. Abl. sg. f. pora is from *k^oisad (: 0. poizad) with o < oi (21 a) and r < s as in the oblique cases of erec (106 a).
^ The analysis of the forms in this paragraph is based for the most part on von Planta, II, p. 228. Buck, §199 f., explains these forms as examples of an indeclinable relative identical in formation with the conjunction pure 'when' in Ila 26. Such a use would have a parallel in Modern Greek ttov who, whom', but von Planta's explanation has the advantage of conform- ing to normal use of cases, even though the phonological details are somewhat troublesome.
111. The following forms not included in the table in 110 are derived from the interrogative-indefinite or the relative stem:
a. panta nom. sg. f, Vb 2, ace. Vb 3 'how great': L. quanta; analogous in formation to etantu (109 k), but without the initial e-.
b. putrespe, gen. sg. IV 14 'of each'; from *k^o-tereis-k^e: cf. seipodruhpei, L. utriusque, norsgog. For -tero-, the suffix of contrast, see 76 c; for -pe: L. -que, re, see 113 1.
F. Indeclinables
112. Indeclinables include adverbs, prepositions, and conjunctions. They are de- rived occasionally from nouns, much more frequently from adjectives, numer- als, or pronouns, sometimes from other indeclinables, and very rarely from verbs (e.g. heri, conj. 'or,' from pres. ind. second (?) sg. of verb her, herti, etc., 'wish'). A considerable proportion of indeclinable forms, especially of the adverbs, have clearly recognizable case-endings, and if their forms are identical with those of stems actually declinable, it is not always possible to distinguish between adverbs and oblique cases of adjectives and pronouns; so, for example, pirse as neut. of pronoun pis and as conditional conjunction 'if.' In addition to fossilized case-endings several suffixes not belonging to the declensional system of the language are used, and a number of encUtics used in the formation of pronouns recur among the indecUnables.
The following classification is based on the stem-types from which the indeclin- ables are derived. Since a clear division of categories is not always possible and the arrangement is sometimes arbitrary, cross-references are freely used; but for etymological details it is necessary to consult the Lexicon.
a. The following adverbs are derived from adjectives, including perfect pas- sive participles but excluding numerals, whose derivatives are listed in b below: preve, prufe, rebte, sarsite, sepse, trahuorfi, nesimei. These forms are in general adverbs of manner, except nesimei and trahuorfi, whose value is local, nuvime
114 II. Grammar
also belongs in the present paragraph, if derived from the superlative of *nowo-; see on Ila 26.
b. The following adverbs are derived from numeral stems: duti (see 105 b), trioper (105 c), nuvis. For promom, which serves as the ordinal 'first', see 113 a.
c. The following adverbs are derived from local prepositions, or from decUn- able adjectives derived from prepositions by means of suffixes of comparison: apehtre (cf. L. adj. exterus), perne, postne, postro, superne, subra, Qive, gimu (cf. L. prepos. cis).
d. The following adverbs of time, place, and manner are derived from the stems of demonstrative pronouns and are in certain instances scarcely distin- guishable in use from oblique cases of the pronouns themselves: from *f-, stem of the anaphoric erec (109 a): ife, itek. From *esso-, stem of the demonstrative esu, etc. (109 e): eso, esuk, isec, isunt. From a pronominal stem represented in Oscan by gen. sg. fem. ulas and in Latin by olle, adv. olim (109 g): ulo. From a pronominal stem '^eno- represented by Skt, instr. sg. anena, etc. (109 i) : enem, eno, enumek, etc. From a pronominal stem *so- represented by 6, i^, Skt. sa, sd (109 j): surur, suront, sururont, and also the conjunction sue.
e. The following local, temporal, conditional, and final conjunctions are de- rived from the stems of interrogative-indefinite and relative pronouns: from pam- (: L. quam): pane < *k^dm-de, prepa < '^prai-k^dm. From po-, pu- < *k^o-, '^ku (see 49 e), with various case-endings or other suffixes: pue, pufe, pone ponne pune, pumpe, pufe, pusi puze. From pi-< *k^i-: perse pefe. The stem po- is also contained in the adverb of direction seipodruhpei (111 b).
f. The conjunction heri heris etc. 'or' is a fossilized form, apparently 2 sg. fnd., of the verb her, herti; for the semantic development cf. L. vel 'or' from a iorm of volo 'wish' (2 sg. *welsi ?).
g. The following indeclinables do not fall into any of the etymological groups treated in the preceding paragraphs: the negatives nei, neip; the temporal conjunctions arnipo, nersa, which contain the negative *ne; the conjunction nosue < "^noi-swai, but for the -sue see 109 j; the conjunction ote ute; the tem- poral adverb sumel; the conjunction et; the conjunction ape etc, < *at-k^e.
113. The following classification is based on the case-endings, enclitics, or other elements attached as suffixes to the stems used in the formation of adverbs and conjunctions.
a. The following adverbs and conjunctions are in origin accusative singular forms, the adverbial use being derived in part through use of the accusative to express the inner object or effect of the action of the verb: promom, duti < *dutim < ^du-tyo-m, tertim terti; the temporal conjunctions pone ponne pune < ^k^om-de, pane < *k^dm-de, nersa if from *ne-ddm, cf. L. quondam.
F, Indeclinables 115
h. A neut. ace. pi. ending is found in the multiplicative adverb trioper triiuper < *triyd-per(t) (105 c).
c. The following adverbs are based on -od, the abl. sg. ending of o-stems, the adverbial use being developed largely through the instrumental value of the ablative case:^ tertio tertiu, gimu, eruk, eso etc. 'thus,' ulo ula,surur,seipodruh- pei. The same formation appears perhaps also in postro Vila 43, 44 = pustru lb 34, 36, taken by Buck, § 190. 6 a rather as derived from a predicative use of the neut. pi. like that of pustra Ila 32, pustru b 19, etc. The question cannot be conclusively answered, but an adverb of direction from a neut, pi. form is by no means a usual or inherently probable formation; in Ila 32, b 19, VIb 5, Vila 8 pustra, pustru, postro is in predicative agreement. The conjunction pue 'where', if from '^k^"6-l, must have received the enclitic -I after loss of the ablatival -d.
d. The following adverbs are based on -ed, in origin an ablaut-variant to the -6d of c above, and are thus analogous to L. longe, OL facilumed, 0. amprufid, etc.: preve, prufe, rehte, sepse, sarsite, trahuorfi, nesimei, nuvime; similarly apehtre, unless with a suffix from -im (cf. L. illim), which is preferred by von Planta, II, pp. 190, 194, 454, not only for this form but for several others as well. give probably originated not as an abl. but as a loc. form (see on lib 11).
e. The following adverbs are based on -dd, the abl. sg. ending of a-stems: hondra huntra < *ghom-terdd, subra: L. supra.
i. The locatival suffix -fe < *dhe appears in ife: L. ibi,^ Skt. iha, Prakrit idha; pufe: L. ubi. Here must also be included ifont < ^i-f-{e)-hont.
g. The enclitic -i (106 d) is used in the local conjunction pue < *k^o-i, the con- ditional conjunction perse < *k^id-l and the temporal conjunction pufe < k'^od-i.
h. The enclitic -k -c (106 b) is used in the adverbial forms eruk, itek, esuk, isec, enuk, enumek.
i. The enclitic -{h)ont used in pronouns of identity (106 c) is found also in the adverbs ifont (see f above), isunt, suront, sururont.
j. An enclitic -ne appears in the adverbs perne, postne, superne: L. superne. The origin of the enclitic and the quantity of its e are uncertain, but it may be related in some way to the demonstrative stem found in L. nam, nempe, etc. (109 i).
k. A group of enclitics -de, -do, -dam related to L. dum, donee, etc., appears in the temporal conjunctions pane < *k^dm-de, pone < *k^om-de, panupei: L. quandoque, nersa < *ne-ddm.
1. The enclitic -pe < *-A:^e appears in the temporal conjunction ape< *at-k^e: L. atque, and in the negative neip neip nep with loss of final vowel (see 31 b and cf. L. nee beside neque). So probably in (pisi) pumpe, putrespe: L. utriusque; seipodruhpei, panupei; there seems to be no compelling reason for deriving the
116 II. Grammar
enclitic from *-k^id (: Skt. -cid) rather than from *-A:"'e beyond the desire to equate it with the enchtic in 0. pukkapid. The spelhng -ei in some of these forms may signify that the enclitic -i (see g above) was further added at the end. m. An enchtic -ti is commonly assumed in the formation of puse puze, etc. < *A:"'iz-/(0-s-i ( ?) : cf. L. ut, uti and see 49 a with note 1. The -ti in posti is doubt- less of the same origin.
^ It is, for example, from this use of the abl. that ulo, L. illuc, acquire their value as indi- cating direction toward a goal, the intermediate stage being the route traveled. In a formal sense, however, there is no justification for setting up a series of original instrumental as dis- tinct from ablative forms, for the fusion of the two cases into a single Italic case with ablative -d took place very early; the absence of -d in the Umbrian forms represents a normal phonetic loss, and the distinction of forms in -o and -u has no value as evidence for originally distinct case-endings.
* with b for d after ubi, where dh yielded b instead of d because of the preceding u. For the absence of initial consonant in ubi see 49 a with note 1.
G. The verb
114. General Survey. The extant monuments of the Itahc dialects, being mostly prescriptive rather than narrative in content, give us an unbalanced picture of the verb system. Among tenses w^e have examples of the present, perfect, future, and future perfect, the last being especially frequent. Of the imperfect there are no examples except 0. fufans 'erant,' as if equivalent to L. ^fubant, and there are no pluperfect forms at all; yet it is altogether probable that this tense existed in the dialects as well as in Latin.
a. The voices are the active and the passive; as in Latin the difference in form is largely one of personal endings, and there are several deponent verbs. Pas- sive inflection, however, has spread into the perfect system in a manner not paralleled in Latin, as exemplified by such forms as pf. subj. pass, pihafi, heriii, 0. sakrafir, lamatir, fut. pf. pass. U. henuso, couortuso, 0. comparascuster.
b. The subjunctive mood is known from examples in the present and perfect tenses, with a few forms of the imperfect in Oscan and Paelignian. In the im- perative mood the present is rare, but forms corresponding to the Latin future imperative are extremely frequent.
c. The forms of the non-finite verb include a present active infinitive, a per- fect passive infinitive of periphrastic formation as in Latin; a supine similar to the Latin supine in -turn; a present active participle, a perfect passive parti- ciple, and a gerundive, these last three formations being fully analogous to those in Latin.
d. The classification of verbs into four conjugational types is applicable to Umbrian and Oscan, as to Latin, since the present stem-classes are to a large extent the same throughout the Italic group.
G. The Verb
117
e. The following tables do not contain all the verbal forms in the Iguvine Tables, but are intended to show the forms which best represent the categories to which they belong. Since the four conjugations in the Italic dialects, as in Latin, are, strictly speaking, present stem classes, the use of the traditional scheme for the verb as a whole sometimes involves difficulties, and in some instances the practice followed may appear inconsistent. For example, fut. pf. portust, despite the absence of the characteristic a, has been placed in the First Conjugation along with its pres. subj. form portaia, but the pres. ind. pass, forms herter herte herti have been placed in the Third Conjugation, in contrast to the Fourth Conjugation forms heris, heriiei, on the theory that the presence or absence of (long) i in forms of the same tense marks a genuine varia- tion of conjugational class. In the table of irregular verbs only the verbs 'be* and 'go' have been included. Although afero and the equivalents of L. facio, capio, and habeo are treated in 122 c, d, e as irregular, their irregularities are not so extreme as to make it impracticable to include their forms under the appropriate headings in the table of regular verbs.
EXAMPLES OF REGULAR VERBS
Pres. Ind.
Pres. Subj.
Fut. Ind.
Perf. Ind.
Perf. Subj.
sg.
pi. 3
sg.
pi. 3
sg.
sg. pi.
sg.
Conjug. I
ACTIVE FORMS
Conjug. II Conjug. Ill
Conjug. IV
|
subocau (?) furfant furfae |
trebeit (or IV?) tigit |
sestu |
heris |
|
aseriaia kuraia portaia kupifiaia |
prehabia |
dersa tefa dirsans (nei)fh.abas |
fagia |
|
prupehast |
menes anpenes ferest |
heries purtuvies fuiest staheren |
|
|
eitipes (?) |
fefure (?) usage (?) |
||
|
combifiansi |
heriiei |
118
II. Grammar
Conjug. I
Conjug. II
Conjug. Ill
Conjug. rv
|
Fut. Perf. |
||||
|
Ind. sg. 2 |
apelus entelus benus |
purtiius purtingus |
||
|
3 |
andirsafust |
apelust |
purdinsiust |
|
|
atefafust |
entelust |
disleralinsust |
||
|
partus t |
benust |
|||
|
combifiansiust |
dirsust terust peperscust |
|||
|
pi. 3 |
haburent |
facurent benurent procanurent eiscurent pepurkurent prusikurent |
||
|
Imv. Pres. |
||||
|
sg. 2 |
stiplo aserio |
|||
|
Imv. Fut. 2, 3 |
||||
|
combifiatu |
car situ karetu |
aitu |
purdouitu |
|
|
kumpifiatu |
habitu habetu |
fertu fertu |
purtuvitu |
|
|
' |
naratu |
sersiiu |
comoltu |
amparitu |
|
portafu |
tursitu tusetu |
kumaltu |
seritu |
|
|
purtatu |
upetu |
fetu fetu |
stahitu |
|
|
slip lata |
ditu titu teftu endendu ententu |
|||
|
pi. 2, 3 |
etato etatu |
habituto habetutu tursituto tusetutu |
aituta fertuta ustentuta |
stahituto |
|
Inf. Pres. |
afero aferum fagiu fagu (or IV?) |
|||
|
Supine |
aseriato anzeriatu |
|||
|
Pcpl. Pres. |
[traf] |
kutef serse zeref |
restef |
G. The Verb
119
PASSIVE FORMS
Con jug. I
Conjug. II
Con jug. Ill
Conjug. IV
Pres. Ind.
sg. 3
Pres.
Subj. sg. 3 pi. 3
Fut. Ind. Perf. Ind.
pi. 3
sg. 3
pi. 3 Perf.
Subj. sg. 3 Fut. Perf.
Ind. sg. 3
pi. 3
Imv. Fut.
sg. 2, 3
pi. 2, 3
Inf. Perf.
Pcpl. Perf.
Gerundive
|
habe (?) |
herter herte |
||
|
habe (?) |
herti |
||
|
ferar |
|||
|
tursiandu |
emantur terkantur |
||
|
ostensendi |
|||
|
tuderato est |
screhto est |
||
|
stakaz est |
screihtor sent |
||
|
pihafi |
herifi |
||
|
pihos fust |
benuso |
persnis fust |
|
|
couortuso |
purdito fust |
||
|
gersnatur |
|||
|
furent |
|||
|
spaJiamu |
persnihmu |
||
|
eturstahmu |
persnimu axnparihmu anouihimu |
||
|
arsmahamo |
|||
|
caterahamo |
persnimumo |
||
|
kateramu |
|||
|
kuratu eru |
|||
|
erom ehiato |
|||
|
conegos |
tases tagez |
screhto |
persnis |
|
kunikaz |
opeter |
comatir |
purditom |
|
kuratu |
uirseto |
kumates |
purtitu |
|
snata |
vufetes |
spafu |
heritu |
|
hostatu |
spefa |
||
|
gersnatur |
|||
|
peihaner |
anferener |
||
|
pelsans |
|||
|
pelsana |
120
II. Grammar
IRREGULAR VERBS
THE VERB BE
THE VERB GO
|
Pres. Ind. Act. sg. 3 pi. 3 |
est est sent |
|
|
Pres. Subj. Act. sg. 2 3 pi. 3 |
sir si sei si se si se sins sis |
|
|
Fut. Ind. Act. sg. 3 pi. 3 |
fust fus fust furent |
est eest |
|
Perf. Subj. Pass. sg. 3 |
ier (?) |
|
|
Fut. Perf. Ind. Act. sg. 3 pi. 3 |
iust ambrefurent |
|
|
Imv. Fut. Act. sg. 2,3 pi. 2,3 |
futu futu fututo |
etu eetu etu etuto etuta etutu |
|
Inf. Pres. Act. |
erom eru |
|
|
Pcpl. Perf. Pass. |
daetom, peretom |
G. The Verb 121
115. Personal Endings. It is not possible to give a full account of the per- sonal endings in the Italic dialects; first person plural forms, for example, are not attested, and there are no second plural forms except in the imperative. The distinction of primary and secondary endings, however, is preserved to a greater extent in the dialects than in Latin; in the third person it can be ob- served in both singular and plural, and it is found in the passive as well as the active voice. The primary endings are used in the present, future, and future perfect indicative and the secondary in the perfect (as well as in impf. 0. fufans) and throughout the subjunctive.
The following table shows the personal endings actually quotable in indicative and subjunctive forms. The term "ending" is here understood in the narrowest sense and excludes such elements as the stem-vowel or the s of the future. The symbol O shows a "zero-ending" which in indicative and subjunctive forms has resulted from phonetic loss of the final consonant which constituted the ending. Endings which are found with identical spelling in both alphabets are shown in the Latin alphabet only.
For the endings of imperative forms see 127 a, b, c, d.
a. Active first Singular. The primary -u < -o occurs in sestu, stahu (: L. 5/0; from minor Umbrian inscription no. 355 Co.), and in subocau subocauu, if it is present, as seems likely, rather than perfect. On the failure of unlike vowels to contract see 33 a. The secondary ending -m is attested in 0. pf. mana- fum, but the -m has been lost in the sole Umbrian example, pres. subj. aseriaia (126 a).
b. Active Second Singular. The original distinction between primary -51* and secondary -s was effaced in the Italic languages as a result of the early loss of-i in the primary ending. The difference which appears in the table is therefore coincidental and results merely from the distribution of quotable forms among the earlier and later tables (-s > -r, 57 e). The form heris used as a conjunction in la 4, b 6, is in origin a 2 sg. pres. ind. verb form. In fut. and fut. pf. forms the tense-sign s and the ending -s are written as one. Pres. subj. sir (: L. sis) shows rhotacism of the s, and omission of the final consonant occurs in several forms, including pres. ind. heri, subj. se si si sei.
c. Active Third Singular. The primary -ti became t through proto-Italic loss of the I, and this / is partly preserved, partly lost: e. g. tigit but heri, fust but fus. A large proportion of both 2 sg. and 3 sg. forms are fut. or fut. pf. in subordinate clauses depending on principal clauses with imperatives in -tu. Since these imperatives are themselves ambiguous, it is not always possible to distinguish fut. and fut. pf. 2 sg. forms from 3 sg. forms with loss of -t, but an examination of the usage in the tables as a whole shows that in general the rit- ual instructions are in the second person in Tables I, II, and III, in the third person in V, VI, and VII. Table IV is poor in applicable examples, but fakust
122
II. Grammar
PERSONAL ENDINGS
ACTIVE PA
Singular Plural Singular
Plural
Oh
'a a o
C/2
|
-u |
|||
|
-s O |
|||
|
-t O |
-nt -n -0 -ent |
-ter -/i -te -tei |
-Tidz |
|
O i<-m) |
|||
|
-r (<-s 57 e) O |
|||
|
o(<-rf <-t, 41 b) |
-ns -s -es |
-r O |
-nrfu -ntur |
31 and prupehast 32 are third person. — The secondary ending -t became -d (41 b), and this -d was regularly lost in Umbrian, though it is preserved in Oscan and in such Old Latin forms as sied, which antedated the generalization of the primary -t.
d. Active Third Plural. Primary -nti became -nt through loss of -i. -ent -ent < *nti is in origin a variant of -nti used in unthematic forms and exempli- fied in fut. staheren, fut. pf. facurent, etc. Final nt is normally preserved, but furfaS lb 1 (= furfant VIb 43) and staheren lb 19 (before a word beginning with t) are exceptional. The secondary ending is -ns, -ens} with frequent omis- sion of the n. The origin of the ending, which is known in Oscan, Paelignian, and Volscian, but not in Latin, is difficult to explain. It cannot be a normal phonetic development of -nt, and any discussion of it must take account of the fact that this -ns escaped the usual change to -/ (58). According to the view pre- ferred here -n < -nn < -nd (61 e) < -nt was strengthened by the addition of -s through analogy with the 1 pi. and 2 pi. endings (cf. L. -mus, -tis).^
e. Passive Third Singular. The primary ending is -ter or -te -tei -ti (8 c, 56 c): herter herte hertei herti. For the future perfect forms in -so see g below.
G. The Verb 123
The secondary ending is -r, which is subject to loss (56 c): ferar, but pihafi, he- rifi.
f. Passive Third Plural. The primary ending is -ndi in ostensendi, the sole example; *nter may be assumed as the original ending (60 a; 8 c, 56 c). The secondary ending is -ntur -ndu, both for *-ntur (60 a, 56 c),
g. Remarks on the Passive Endings, r-endings are known to have existed in Sanskrit, Armenian, the Italic and Celtic languages, Hittite, and Tocharian, primarily though not exclusively in deponent and passive forms. Since their history is complicated and requires some revision, especially as a result of the recent study of the two last-named languages, it is not intended here to give a full discussion of the origin of theUmbrian formations, but rather to emphasize those features in which they differ from the corresponding Latin forms. As in the active voice the distinction between primary and secondary endings in the third sg. and pi. is maintained in Umbrian, not lost, as in Latin. The O.-U. 3 pi. ending *-nter is best explained as from a combination of the two IE medio- passive endings -nto -ro, with syncope of the medial o and -er < -r < -ro, as in per< pro (32 d). The L.and U. secondary 3 pi. ending ^-ntur may be from -nto-r, though several other origins through combination of nt- and r-endings are pos- sible. In any case the 3 sg. O.-U. -ter and L. -tur must have been derived from -nter, -ntur analogically after the correspondence between active 3 sg. -t{i) and 3 pi. -nt{i). U. forms of the type of ferar, in which -r alone serves as the 3 sg. ending, must have resulted from simple replacement of the active ending -d
< -t) by -r, as in Latin the 1 sg. ferar, ferebar resulted from replacement of the -m of active feram, ferebam. The fut. pf. pass. 3 sg. forms benuso, couortuso are the most difficult of all. Since there are only four examples, all from Vila 64 through VI lb 2, it is fully possible that an -r has been lost. They would then be closely related to the ferar-type, and this close relationship would receive addi- tional support from the fact that the short passive forms of the /erar-type, as well as benuso and couortuso, are predominantly impersonal in use. The o, how- ever, remains unexplained, and the whole formation is the more obscure be- cause of the uncertain origin of the future perfect itself.
^ -es in pf. eitipes, the one actual Umbrian example with the secondary ending of unthema- tic forms. Cf. O. uupsens. The relation of -ens to -ns is like that of -ent to -nt. 2 So R. G. Kent, I.F., LIII (1935), pp. 41-4.
116. Present Tense Formation and Conjugational Classes. The distinc- tion of thematic and unthematic present stems (" o-verbs and mi-verbs "), which is the primary basis of classification in Sanskrit and Greek, has little significance in the Italic group. A few unthematic verbs survive as irregulars (e.g. L. sum, volo, esse 'to eat'), while those from roots ending in vowels have mostly become indistinguishable from vowel-stems of other types and have joined with them
124 II. Grammar
to form the classes familiarly known as the first and second conjugations. Most of the inherited primary thematic verbs, including those of the reduplicating, nasal-infix, and -s/co-classes, fall into the third conjugation, while primary verbs with present stems in -yejo- are included within the fourth conjugation or as a subdivision within the third (the capio-type). Denominatives, which form a very large and productive class, are in origin -f/o-stems; most of them belong to the first conjugation, but some to the fourth and a few to the second. The classification here briefly outlined applies equally to Latin and to Oscan-Um- brian.
a. Although the four conjugations in principle are present stem-classes, their special characteristics persist to a large extent throughout the non-present ten- ses, including the perfect passive participle, so that the validity of the division into four classes is all the greater. Yet some deviations from the system occur : e.g. L. domo, -are, domui, domitiis; cupio, -ere, ciiplvi, cupitus. There is some ground for believing that these deviations are rather more frequent in the dia- lects than in Latin, if we may generalize from the limited material : e.g. mugatu but pf. pcpl. miiieto; portatu but fut. pf. portust; osaiu but pf. pcpl. oseto (minor inscr. no. 354 Co.), in contrast to L. operari, operatus; pesetom with no pres. attested, but cf. L. peccatum. Not infrequently Umbrian and Latin show a dif- ference of stem-class in otherwise equivalent forms : e.g. dirstu <*did{e)t6d like didcofii in contrast to L. do ; amboliu but L. ambulato; iiirseio as if L. ^videtum instead of visum; conegos kunikaz < -dtiis but L. conixus; neifhabas as if L. *ne adbibas, like advenat, instead of adhibeas.
b. Sections 117-120 below contain an account of the four conjugations with lists of those verbs which can with reasonable probability be assigned to them. Although the classification is mainly applicable to present stems, some non-pre- sent forms are included when they clearly show the characteristics of their res- pective classes. The discussion of individual verbs is held to a bare minimum; for etymological details the Lexicon should be consulted.
117. The First Conjugation includes a few primary verbs and a large num- ber of denominatives. As in Latin, the primary verbs are partly derived from root-stems with unthematic inflection, partly from present stems with the suffix -z/e/o-. The denominatives, whether from nouns or adjectives with stems in -a, -0, or consonants, are all derived by addition of -j/e/o- to a stem in -a.
a. Possible examples of root-stems, with original unthematic inflection are: spahaiu (root spa- ?); restatu < ^re-std-tod, but it is also possible to assume a present in -ye/o-.^ The pf. pcpl. snata gives indirect evidence of a present stem snd-: cf. Skt. sndti, but also sndyate, L. no, nare.
h. Other primary verbs of the first conjugation are derived from disyllabic bases in -a or from stems in -a built analogically after such bases. The original
G. The Verb 125
inflection was partly unthematic, but we may infer from Gk. primary presents like ondo), yaXdco, that the -yo-type of inflection came into early use. Among Latin examples of this class may be mentioned dico, oc-ciipo, with persistent a, and seco, domo, sono, with a. restricted to the present system. The Umbrian forms are: mugatu (but pf. pcpl. muieto); prusekatu (but pf. pcpl. proseseto); sukatu, cf. Lith. sakau; a present of this type is indirectly attested by pf. pcpl. ehiato: cf. L. hio, -are, -atus; fut. pf. andersafiist < '^an-didd- through transfer of ^dide/o- (cf. dirstu, didcofii) to the first conjugation, probably in connection with its speci- alization of meaning. The pf. pcpl. conegos kunikaz, with no present attested, shows the a. of the first conjugation in contrast to L. co-nitor, conixus.
c. The following present forms can with reasonable certainty be regarded as denominatives: arsmahamo, caterahamo, kuraia, naratu, osatu, pihatu with fut. prupehast, {an-)stiplatu, ahatripursatu, eheturstahamu, vepuratu, uesticatu, preuislatu, subocau. Similar denominatives are indirectly attested by the pf. pcpls. gersnatur, tuderato.^
d. The iterative or frequentative type represented by L. cantare, etc., is in origin probably a denominative formation from the pf. pcpl. Umbrian has several examples, although without the characteristic meaning: etato to the primary etu; portatu, with L. porto formed from a pcpl. to the root seen in jtsiQco, noQog; statitatu, a denominative to a pf. pcpl., whatever view we take of the stem before the participial suffix.
e. The analysis of the following forms is less clear, but all are more probably denominative than primary verbs: combifiatii, aseriatu, preplotatu, furfant. The pf. pcpls. aviekate and stakaz presuppose denominative stems in -a.
f. The present tense of verbs of the first conjugation was originally formed by the addition of the personal endings directly to the a. of the root or disyllabic base in verbs of the unthematic type, or to the thematic vowel e/o in the denom- inatives and other verbs with stems in -dye/o-, but the two types implied by this description have been replaced in Oscan-Umbrian and Latin by a sing e type of inflection combining features of both earlier types. The first person sing- ular is regularly of the thematic type : e.g. subocau^ < ^sub-wokdyo without con- traction after loss of y; cf. L. no < ^(s)ndy6 in contrast to Skt. sndmi which preserves unthematic inflection. In other forms of the present indicative the loss of intervocalic y was followed by vowel-contraction;* the resulting vowel was d, whereby thematic and unthematic forms became identical, or if a dif- ferent vowel resulted, it was analogically replaced by d. For the shortening of a in Latin 3 sg. -at there is no evidence in Oscan-Umbrian, and in fact the evi- dence of spelling in second-conjugation forms is against such shortening. For the future see 123. For the present subjunctive see 126 a.
^ Such a stem is often assumed for L. sto, stare, which is intransitive and of durative aspect. Because of the transitive meaning of restatu, Buck, § 213, 4 a, regards it as a reduplicated
126 n. Grammar
pres. like L. sisto, with transfer to the first conjugation, but as support for an unreduplicated form with transitive value we may compare L. praesto ' offer, ' unless praesto and U. restatu may both be of the same class as i-ar7j/j,t with loss of reduplication through syncope after the prefix, an unlikely supposition in view of the absence of any L. *sistdre.
* Not so, however, with hoslatir and ponisiater, which belong to the same type as L. barbatus. Such formations are older than the corresponding finite verbs, when the latter existed at all.
^ Some take subocau as perfect rather than present, but this involves greater difficulties. See on Via 22.
* In the proto- Italic period a + e would yield a, and a + o would normally yield 6 (e.g. sol <; *sdivol < *sdwel), with replacement by d in forms of the first conjugation.
lis. The Second Conjugation includes verbs derived from roots or disyllabic bases in -e, and from stems in -eyejo- of several different types. It is not possible in all cases to separate the primary verbs with original unthematic inflection from those based on stems in -eye/o-. Moreover because of the fluctuation in the use of e and i on the tables it is sometimes difficult to determine whether a given verb should be assigned to the second conjugation or to the fourth. The question is usually decided by reference to the semantic value of the verb (as intransitive or as causative) or by comparison with related Latin forms.
a. The primary verbs which undoubtedly belong to the second conjugation include: habitu habetu: L. habeto; sersitu: L. sedeto; tenitu: L. teneto; carsitu kafetu in contrast to L. calato. Indirect evidence of present forms correspond- ing to L. taceo, video is given by the pf. pass, pcpls. tases tagez, uirseto.
b. The Umbrian and Latin second conjugation includes a class of verbs with o-vowel in the root and a stem in -eyejo-. The meaning is predominantly intensive or causative: e.g. L. moneo 'remind' beside memini 'remember'. The clearest Umbrian example is tursitu tusetu < ^torscyetod (12 b) with normal o-grade in contrast to L. teneto. upetu, with pf. pcpl. gen. sg. opeter, may represent the same type.^ kanetu may also be an intensive; it is generally taken as a third conjugation form equivalent to L. canito, but it is difficult to believe that the e would escape syncope; see on IV 29. The pf. pass. pcpl. vufetes, if taken with the majority of editors as equivalent in sense to L. votis, would give evidence of a present indicative like L. voveo, of the type treated in this sub-section; see however on Ila 31, where a different interpretation is proposed, making it equi- valent to L. libitis, pcpl. of libet. If this view is adopted, viifetes still belongs in the second conjugation, but probably with the forms in a above rather than in b.2
c. To the denominative type represented by L. albeo beside adj. albus may belong U. kutef if it is pres.pcpl. of *kauteyo, a denominative to *kautos — L. cautus; but several other etymologies have been proposed.
d. Several other verbs are commonly assigned to the second conjugation, but with some uncertainty, which arises in part from the ambiguity of i in the Latin alphabet and the lack of forms in the native alphabet: iremitu with causative
G. The Verb 127
sense but without the characteristic o-vowel; possibly nepitu, sauitu, sonitu; probably trebeit because of its intransitive meaning, although the etymological value of the ei is not clear; probably also eveietu, tigit.
e. The treatment of tlie stem before the endings is in general analogous to that in the first conjugation (117 f): in denominatives and other formations in -eye/o- the intervocalic y was lost and similar vowels contracted, producing par- tial identity with the unthematic forms. Quotable indicative forms of this con- jugation, however, are limited to the 3 sg.
^ The normal e-grade may be contained in Hittite epmi 'je saisis, ' as suggested by E.-M.', p. 824, who however take the o-vowel of the Italic forms as evidence of an old unthematic pres- ent. But upetu is identical in sense with L. optato, and it is possible to see an original intensive value expressed in Umbrian by ablaut-variation in the root and stem in -eyo- and in Latin by the use of the iterative stem in -ta-. The semantic evolution may be expressed by the propor- tion: root *ep- 'seize': *opey6 (or *optdyd) 'wish for, choose ' = L. capio 'seize': capto 'strive for'.
2 Those who prefer to adhere to the traditional interpretation may compare evxofiai as a form illustrating the normal ablaut-grade; the connection of voveo with evxo/iai is favored by E.-M. and Boisacq. The intensive notion in voveo has faded to about the same extent as in spondeo, a verb of similar meaning.
119. The Third Conjugation contains a large majority of the inherited pri- mary verbs with thematic inflection, whether the thematic vowel is added di- rectly to the simple root or whether the root is strengthened by reduplication, nasal infix, or any one of several suffixes.
a. The majority of verbs of the third conjugation form the present stem by the addition of the thematic vowel to the root, which is usually, but not invaria- bly, in the normal e-grade. In the imperatives, which are the most frequent forms, the e before the / has been syncopated, but its former presence can be inferred through comparison with Latin and through the fact that the treatment of the final consonant of the root often presupposes the presence of a vowel before the t (cf., for example, original -kt- and -kt- arising after syncope, 46 i, j). The Um- brian forms are: aitii; menes for *benes, fut. 2 sg. of verb corresponding to L. venio but without /, as in OL pres. subj. advenat; kartu; deitu teitu; pres. subj. terkantur and emantur; sumtn; fertu and its compound represented by inf. afero; fiktu; probably holtu, the etymology of which is, however, uncertain; comoltu kumultu;^ amboltu (in contrast to L. ambulato); ampentu; afpeltu;^ the series of compounds andendii,^ endendu, ostendu, pertentu, sutentu; umtu; arsueitu and kuveitu; veltu and ehueltu; ahaiiendii and preuendu; couertu; vetu; vutu.
b. Several verbs are made from reduplicated stems. The original stem was probably unthematic as in didco/Lii, iarr]fii, but the intrusion of thematic forms began very early (cf., for example, Skt. tisthati in contrast to unthematic laTrjf^i),
128 II. Grammar
and the Italic forms have become fully thematic. The forms are: ditu titu < *did{e)tdd (42 c), with pres. subj. dirsa tefa < ^diddt; sestu < *sist(e)t6d; sistu: L. sldito, from ^sizd(e)-t6d; probably also restef < *re-sistens with loss of the reduplication through syncope of its vowel; comparison with L. favors a reduplicated stem in the third-conjugation verb, but an unredupUcated stem in the first-conjugation restatu (117, n. 1).
c. hondu ' cast down ' (?) is a compound of horn- + a verb similar to that found in L. condo, reddo, etc. Since proto-It. p became d after n, hondu has the same ambiguity as the Latin series; that is, the second element may be either from *dd- 'give 'or ^dhe- 'place,' but the similarity in meaning to L.pessum dare favors the former alternative. It is to be assumed that the stem here is unreduplicated, but the verb has adopted the thematic type of inflection exactly as in b above. For phonological details see 60 a, d.
d. The type of present stem with nasal infix (cf. L. vinco, vici) is represented by nindu (49 d): L. ninguito.
e. The verb-class characterized by the inchoative suffix -sko- is represented by perstu < ^ perk-sk{e)-t6d, fut. pf. peperscmt; eiscurent fut. pf. of a verb com- monly referred to root ^ais- 'seek'. The two fut. pf. forms, with 0. comparas- custer, show the -sko- extended into non-present forms of the verb, in contrast to Latin cresco, crevi, etc., where the suffix is restricted to the present system.
f. revestu is apparently formed with a suffix -so- and is thus equivalent in formation to L. revisito < ^re-weid-se-tod.
g. In the third conjugation the thematic vowel was regularly syncopated in the fut. imv. but not in the 2 sg. and 3 sg. ind., where, however, the evidence is very meagre: 2 sg. seste (lib 22, taken by some as 3 sg. pass.), 3 sg. Marruc. feret. Vest, didet. Identity between the true thematic forms and those with a < 3^ originated in forms where either e or a suffered syncope, after which the un- thematic forms changed fully to the thematic type.
^ These forms, with L. molo, -ere, are from a disyllabic base, but the 3 which stood as the re- duced grade in the second syllable has been replaced by the ordinary thematic vowel.
2 An ordinary thematic present if we follow Devoto's interpretation 'circumilo,' whereby the verb is from root *k^el-, cf. L. ac-colo, nsKo^ai, Skt. carati. If we equate with L. appellere, with most editors, we have a third-conjugation verb of the -no-class.
^ In this series the root is actually *ten- and the suffix -do- (the unstrengthened form of the root being found in L. con-ien-lus), but this does not seem a sufficient ground for setting up a separate class.
* 9 is the reduced grade of pIE a, e, 6 (35 c, and b, c above). What is said here applies parti- cularly to the verb ' give ', including Vest, didet here cited.
120. The Fourth Conjugation includes some primary verbs and some de- nominatives, the latter class being derived originally from i-stems and then by analogy from nouns and adjectives of other types (e.g. L. finio to finis, then
G. The Verb 129
custodio to cusfos, servio to servus). The present stem must originally have been made for primary verbs by adding -iyejo-, or its ablaut-variant -i-, to the root and for denominatives by adding -{i)ye'o- to the noun-stem fini-, custod-, serv-, etc., but thematic and unthematic forms do not coexist in parallel series; rather they supplement one another in a single composite paradigm, as in the first and second conjugations.
a. The primary verbs include: amparitu; anouihimu; pres. subj. fuia with fut. ind. fuiest, the length of the i being attested by comparison with L. fio, fiunt and 0. 3 pi. fiiet; purdouitu purtuvitu, fut. pf. purtiius and purdinsus, etc., pf. pcpl. purditom purtitu. The persistence of i in all these forms (against e in purtuvetu purtuetu once each) plainly points to a verb of the fourth con- jugation derived from *dowi/dwl, an extension of root *d6(w) 'give'. Pres. subj. dia is apparently from ^dwlydt made from the same extension of the same root. stahitu, fut, staheren, and the cognate Oscan forms stait, stahint stahint, etc., are usually placed in the fourth conjugation, the classification being partly dependent on the evidence of -yo-stem derivatives of the root "^std- in Indo- Iranian, Balto-Slavic, and Germanic, since the vocahsm of the Italic forms is not entirely clear.
b. The denominatives include: persnimu persnihimu persnihmu, etc., ap- parently from a noun ^persk-ni--} fut. -pi.disleralinsust with -nky-pi. stem (124 f) to ^dis-lerdli-; pf. pcpl. stahmito, stahmeitei gives evidence of a denominative to a stem *std-mi- or ^std-mo- represented by dat, sg. stahmei.
c. The following verbs appear to belong to the fourth conjugation but their origin is in some respects uncertain: seritu probably denominative like L. servio, if we admit loss of w after r (54 g); statita pf. pcpl. of a denominative ^statiyo to *sta-ti- or ^statuyo to *sta-tu-, in which case our form would be equivalent to L. statuta; the ambiguity arises from the fact that I can be from u (15 b) as well as original; uestis uesteis, of uncertain etymology, but probably pf. pcpl. from *westlt{o)s, almost certainly belongs to the fourth conjugation.
d. Reference should also be made to the discussion of fagia and heri and their cognate forms in 121 below.
e. The vocalism of the stem before the endings and the relation between thematic and unthematic inflection have been partly covered in the first para- graph of the present section. Among forms actually preserving the thematic vowel may be cited 1 sg. stahu < '^stayo (minor inscr. no. 355 Co.), fut. 3 pi. staheren < *sta{y)escnt.
^ Since the noun is not actually attested, it may possibly have been *persk-no-, the denom- inative formation then being like L. servlre to serous; cf. also Skt. denominative prasnayati ' ask ' to o-stem noun pras-na- ' question '. In stahmei also the stem-class is uncertain, even though here the noun itself is attested.
130 11. Grammar
121. Third-Conjugation Verbs of the Type of L. capio. The reason for recognizing these as a special class in Latin is, of course, the fact that in some of their forms they resemble normal verbs of the third conjugation but in others those of the fourth. It is to be expected that such verbs in the dialects should suffer syncope in the f ut. imv. and should have i or ii in certain of the forms re- corded in the native alphabet (see 53 c, 68). Actual examples showing this va- riation are: imv. 0. factud, but pres. subj. fakiiad, U. fagia, inf. fagiu fagu; pres. pass. 3 sg. herter herte herti hertei, pres. act. 3 sg. her as enclitic in pisher, but heri after svepis IV 26, fut. heries, heriest, pres. subj. 0. heriiad; the verb 'take,' which appears to be a contamination of the cognates of L. capio and habeo (122 e) shows imv. hahtu hatu hatu, but pres. subj. habia, where, however, i represents e (cf . L. habeat). In the forms outside the present system, which have some indirect bearing on the question of assignment of the verbs to conjuga- tional classes, we find fut. pf. fakust, fakurent facurent; pf. subj. act. heriiei, pass, herifi, pf. pcpl. heritu hereitu eretu; fut. pf. habus, haburent.
a. The origin of the capfo-type in Latin and in the dialects must naturally be treated as a single problem.
According to the prevailing view L. capio and audio both represent inherited types, with corresponding classes in Balto-Slavic, and the original distribution of verbs between the two classes depended on the length of the radical syllable.^ If we adopt this view, the dialect forms corresponding to L. facio and capio (with the h of habeo) probably belong, like capio and facio themselves, in the third conjugation. 0. factud and U. hatu etc. then resulted from syncope of i in con- trast to the syncope of e in normal third-conjugation verbs, and 0. fakiiad may simply represent a transfer to the fourth conjugation.^ The pres. pass, forms herter etc. suggest a similar classification for this verb, but the non-present forms resemble those of verbs of the fourth conjugation.^ In its inflection this verb may be compared with the semantically similar L. cupio, -ere, -Ivi, -Itus.
b. According to the alternative explanation of the capio-type, it developed within Latin from the audto-type through iambic shortening,* *cdpis, ^capit, etc. thus becoming cdpis, cdpit, etc. Naturally this explanation will not hold good for the dialect forms, and if we admit it for Latin, the dialect forms must properly belong to the fourth conjugation, and, since i is not subject to syncope, hatu herter, 0. factud, etc., must represent an unstrengthened type like OL advenat beside venio, -Ire, either inherited or secondary. On the whole the former explanation seems somewhat preferable, but in any case the three verbs under discussion show fluctuation between the third and fourth conjugations and cannot be fitted perfectly into either class.
^ By Sievers' Law (53, n. 2). Cf. Brugmann, Gdr., II, in, i, p. 189; Leumann-Hofmann, p. 321, with extensive bibliography.
* U. pres. subj. fagia is less certain. It is difficult to separate it from O. fakiiad, although
G. The Verb 131
in the native Umbrian alphabet ii is normally written where -iy-, in contrast to -y-, is intended (see 2 h, where some exceptions are shown). On the other hand inf. fagiu with its variant fagu probably represents *fakyom. It should be observed that the distinction of -iy- and -y- has been effaced in Latin, capio, capiunt thus being like audio, audiunl, but is preserved in the dialects.
* heris hefi heri is less decisive. On syntactical grounds we cannot be sure whether it is ind. or subj. in origin, and it is not impossible to take it as like L. capis, since the evidence in regard to syncope in the 2 sg. of third-conjugation verbs is not conclusive. But it is probable that these verbs belong to the fourth conjugation.
4 Cf. F. Skutsch, Archiv fur lat. Lex. und Gram., XII (1902), pp. 210-13; GL, II (1910), pp. 367-9; Kent, Forms of Latin, p. 99; Sommer, Hdb. d. lat. Laut- und Formenlehre, pp. 505-6.
122. Irregular Verbs. Irregularity in the inflection of verbs in the Italic languages may result from fluctuation between two conjugational classes, or from the survival of unthematic conjugation in parts of the verb which are nor- mally thematic, or from suppletion, whereby different tense-systems of the verb are made from different roots, or, in at least one very probable case, from con- tamination of roots.
a. The verb 'be' is made, as in Latin, from the root *es/s- and from the di- syllabic base *bhewd- with its ablaut-variants, especially *bhu.-. *es- appears in the pres. ind. est, sent, subj. sir, si, etc., inf. erom, and pres. pcpl. 0. praesentid. The 3 pi. sent, 0. sent set, like Goth, sind, preserves the ending characteristic of unthematic forms, in contrast to Latin sunt < *sonti, which has been altered after the analogy of thematic verbs. The root *bhil, which in Latin is mostly restricted to the perfect system, the fut. pcpl., and fore, forem, etc., furnishes the imv. U. futu, impf. ind. 0. fufans, impf. subj. 0. fusid (= h.foret), and the forms fust fust, etc., which appear in Oscan partly as fut. pf . but in Umbrian regularly with future value.
b. The verb ' go ' is from the root *ei/i-, and, as in Latin, the development of *ez/- before vowels produces partial resemblance to verbs of the second conju- gation. The normal grade ei may be assumed for imv. etu, etc., fut. 3 sg. est eest, fut. pf. amprefuus, ambrefurent. The zero-grade is found in fut. pf. iust, pf. subj. pass, ier, both with stem like that of L. ii, and in pf. pcpl. da-etom, per-etom with e < i as in 10 c; cf. L. red-itum, etc.
c. The compound verb afero < *am{f)-fer- 'perform a lustration' supplies its perfect system from '^am{f)-dojdd-, as shown in fut. pf. an-dersafust a-tefafust. It is not known what root supplied the perfect of the simplex fer- but in view of L. few, tuli, etc., Gk. (pegco, oiaco, ijvsyxa, etc., it is probable that this verb resorted to suppletion in the Italic dialects also.
d. The verb corresponding to L. facio and the classification of its forms in the third and fourth conjugations have been discussed 121 in. The forms without i include fut. pf. fakust, fakurent facurent with a in contrast to the e of L. fe- cerit, etc. The vocalism of pres. subj. £eia and imv. fetu (52 times) feitu (5 times)
132 II. Grammar
feetu (once) fetu (48 times including the erroneous feiu) feitu (20 times), however, presents a problem. These forms have been referred sometimes to the full grade of the root ^dhe- as seen in ridi^fxi, etc., sometimes to the extended *dhe-k- of L. feci. In favor of the second analysis are the spelling with ei ei in some of the examples^ and the fact that this root without the /c-extension is not posi- tively known to exist in the Italic languages except in compounds of the series represented by L. condo, etc. \
e. The cognates of L. capio and habeo in the dialects present a complicated situation. All the forms in question have initial h, never c k, but the meaning is partly that of L. capio: so, e.g., in pres. subj. habia, fut. pf. haburent, and in some of the imv. forms, especially the third-conjugation forms hatu, etc. More- over 0. fut. pf. hipust, pf. subj. hipid, etc., show p in place of b, and 0. fut. ha- Jiest, pres. subj, pass. haf[iar^ show /. This / cannot correspond to L. b as in 0. tfei: U. tefe: L. tibi (40), since the Umbrian and Latin forms with b plainly point to original b in the verb (39). The facts seem to be approximately as follows: the verbs corresponding to capio and habeo, which are closely related in meaning, suffered contamination,^ perhaps in the period of Oscan-Umbrian unity, and developed a series of forms all having the initial h of habeo, while the second consonant is partly b, partly p. The Oscan forms with / must have spread analog- ically from forms in which / could have come from p before / (38 b). U. hatu etc. give no evidence of whether they originally had b or p, since after the syn- cope had once taken place the history would be the same for b as for p. The meaning in Oscan is that of habeo but in Umbrian to a large extent that of capio,
^ *dhe-tod would probably give a form spelled uniformly with e e. *dhe-k- would presumably lead to forms with thematic inflection and phonological development as in 46 j with note 3. In favor of derivation from */e- cf. PI., II, pp. 257-8; in favor of *fe-k- cf. Buck, p. 168.
^ On no. IS'^, Buck, p. 365, published subsequent to the time of the collections in Conway, von Planta, and Buck's earlier edition. The form lia,f[ia,r, although partly dependent on res- toration, gives some support to the / in hafiest, previously suspected of being an error.
» So Buck Oscan-Umbrian Verb-System, p. 165; Kent, T.A.P.A., LVII (1926), pp. 51-2; Buck, 0.- U. Grammar, p. 167.
123. Future Tense. There are no forms corresponding to either of the nor- mal Latin types; all are sigmatic in origin and are therefore closely related to the Greek future and to Latin futures of the type of faxo.^
Although the future, or its prototype, was originally independent of the pre- sent stem, it has become in Italic a part of the present system, sharing in such characteristics as reduplication in verbs like ditu < *did{e)tdd, 0. fut. didest, or the -z/o-suffix in purtuvies. The tense-sign s is added to the a of first-conjugation verbs (e.g. prupehast) and to the e or ije of third- or fourth-conjugation verbs. There are no clear examples from the second, since in habiest the i is not necessari-
G. The Verb 133
ly from e by 8 b, but may be the mark of the fourth conjugation. The e before the s is syncopated in pass. 3 pi. ostensendi < '-^-tendesenter (60 a), but in general this vowel remains, while the short vowel between the s and the ending suffers syncope: so in the 2 sg. and 3 sg. forms heries < '^'heriyeses, menes < *beneses, ampenes, purtuvies, ferest < ^fereset, eest, fust, fuiest, prupehast. 3 pi. furent and staheren < *stayeseni show no syncope in either syllable, and con- sequently the s is rhotacized.
1 All these formations were formerly regarded as being of the same origin as the Homeric short-vowel aorist subjunctive; cf., for example, Bucl<, § 221. More recently they have been equated with Sanskrit desideratives of the type of pipdsdmi and with Old Irish sigmatic futures; so, for example. Buck, Comparative Grammar of Greek and Latin, §§ 388-9. But perhaps the distinction is more apparent than real, since it appears now that the s-future and the s-aorist subjunctive were originally a single formation with future value, from which the subjunctive value was a secondary development; cf. E. A. Hahn, Subjunctive and Optative (New York, 1953), pp. 59-65. At the same time it should be observed that the Skt. desiderative and the Irish sigmatic future differ from the Italic and Greek formations in being reduphcated.
124. Perfect Tense and Perfect System. In describing the perfect system in the Italic dialects it is necessary to a considerable extent to depend on future perfect forms, since the perfect tense itself is so rare in the extant inscriptions that it is not possible to find examples representing every one of the various types. Since the origin of the future perfect is an unsettled problem, some facts in the history of the perfect stem are uncertain, and the treatment of it must be mainly descriptive. The perfect formations without distinctive suffix, though partly with reduplication or gradation of the root vowel (a, b, c) were undoub- tedly inherited from proto-IE perfect (and perhaps in part aorist) forms and are represented in Latin as well as Oscan-Umbrian. But the other formations, the Latin perfects in -si and -vijui, the O.-U. /-perfect, the Oscan ^/-perfect, and the Umbrian I- and n/cy-perfects have a more limited distribution; not one of them appears in all three languages, and, while allowance must be made for the meagerness of the dialect texts, it is safe to say that these formations, in their function as perfect tense-forms, were developed in pro to- Italic or even later.
a. The forms of the perfect system with reduplication are : dede (minor inscr. no. 352 Co., with d represented by the character q); fut. pf. dirsust tefust (42 b); dersicust; pepurkurent ; fefure, of very doubtful etymology and inter- pretation but in any case a reduplicated form of the perfect system. The vowel of the reduplication is regularly e as in Greek and in OL memordi, cecurri, etc.^
b. Unre duplicated forms of the perfect system are derived partly from ori- ginal perfects, partly from aorists corresponding to the Greek " second " aorist. The principal forms are: fut. pf. fakust, portiist, vurtus with coiiortiis, benus, eiscurent, iust, purtiius, pf. subj. heriiei, pf. subj. pass, ier; probably also pf. 3 sg. usa^e usaie, for which see on la 45. The Italic verb has suffered such ex-
134 II. Grammar
tensive alteration through loss of ablaut-distinctions and loss of distinction of meaning between perfect and aorist that it is not possible to classify the forms on the basis of these two tenses.
Moreover proto-lE apparently had some reduplicated aorists as well as some unreduplicated perfects. Some of the verbs above may have lost reduplication through composition, like L. concidi in contrast to simplex cecidi. There are no clear examples of the o-grade of the radical vowel characteristic of Greek "second" perfects; the vowel of vurtus couortus may be from f as well as or (35 d). Of the conjugational classes the third and fourth are chiefly represented among the foregoing forms; portiist beside imv. portatu shows a simplification of stem (116 a), eiscurent shows the extension of the present tense-suffix -ske/o- into the perfect system (119 e). The stem of iust and ier is similar to that of L. il, and it probably served as the model for the stem in heriiei and purtiius, both of which have rival perfect stems of other classes (d, f below).
c. Another type of unreduplicated perfect without special tense-suffix is that in which the radical vowel is lengthened, as in L. vidi to video, veni to venio, feci to facio, etc. The long vowels are of diverse origin, but in a large proportion of cases they represent ablaut-variants, as lengthened grade to normal grade e, or as normal grade to reduced grade 9 (35). Since the number of Umbrian exam- ples is small, the Oscan examples are added as an aid toward their clarification: 0. upsed, uupsens, 0. ovnaevg with u uu ov for o, in contrast to O.upsannam, U. osatu, L. operari, all present-system forms with o; 0. sipiis 'knowing' < *sepus or '^sepuwos in ablaut-variation with L. sdpio; U. prusikurent with *sek- in variation with L. insece; 0. hipid, hipust, U. eitipes, if from ^eit{om) hepens, all in variation with U. habitu, L. habeo, etc., as if L. had *hebi as pf. to habeo (see 34 on eitipes and 122 e on the consonantal variation).
d. The /-perfect is represented by the following forms : f ut. pf . 2 sg. amprefuus, 3 pi. ambrefurent; 3 sg. andersafust atefafust; pf. subj. pass. 3 sg. pihafi, herifi, perhaps also cehefi, although not according to the view preferred here (see on Via 20). To these may be added the following Oscan examples: pf. aikdafed 'decre- vit\ of which -da- belongs to the root of L. do, dare; fufens 'fuerunt'; pf. subj. sakrafir ' sacrato\^ Although the origin of the formation is not entirely clear, there is little doubt that the / is of the same origin as that in 0. impf. fufans, or the tense-signs of the L. impf. in -bam and f ut. in -bo ; in other words it belongs to a verb-stem from the root *bhewa/bhil- used as an auxihary in combination with some form which cannot be precisely identified. Now among the foregoing verbs fufens belongs to the same root as U. imv. futu, L. fui, etc., and ampre- fuus to that of L. eo, ire; herifi is of the fourth conjugation, and the classifica- tion of cehefi is unknown; all the others are of the first conjugation. This distrib- ution therefore is roughly analogous to that of the Latin futures in -bo, which belong to eo: ibo, to verbs of the first and second conjugations, and in some early instances to verbs of the fourth. The two formations appear then to be based
G. The Verb 135
on the same second element, a thematic stem ^bhwejo-, reflected in Skt. aor. subj. bhuvat, etc.; but it must be assumed that this stem could be used both as an indicative and as a subjunctive similar in formation to the "short-vowel" subjunctive of Vedic Sanskrit and Homeric Greek, and that the Latin 6o-future is derived from the latter use and the O.-U. /-perfect from the former. This twofold development has a close analogy in the history of the stem *bhwa-, which furnished on the one hand Lith. biivo 'he was' and the tense-suffix of the Latin impf. ind. and on the other the L. subj. fuat, etc.
e. The /-perfect is found only in Umbrin and only in the fut. pf. forms ape- lus apelust beside imv. ampentu < *a/i-pe/id-e/orf, entelus entelust beside imv. endendu < *en-tend-et6d. The suffix -lo-, which is attached directly to the root, is apparently of the same origin as that in L. credulus, pendulus, etc., together with the suffix used in the formation of the participle in -lo/a- which furnishes the preterit in the Slavic languages. The phonetic development is : *pend-lo- > *penn-lo (61 e) > ^pen-lo- > *pel-lo.
i. The -n/cy-perfect is found only in Umbrian, where it is represented by: fut. pf. combifiansiust combifiansust combifiansiust, pf. subj. combifiansi; purtin- gus, purdinsiust purdinsust purdinsus; disleralinsust. The origin of this forma- tion cannot be determined with any certainty. It is quite possible, however, that it arose from a combination of an accusative noun in -am, -im + *ke -\- iiist and other forms of the perfect system of the verb 'go'.^ The use of verbs of going as future auxiliaries, as, for example, in L. factum itur, factum iri, etc., is favorable to such a view. The noun which forms the first part of the periphra- sis (with -n < -m before k) could be an a-stem like L. cura, fuga (whence denom- inatives and -n/cy-perfects of the first conjugation) or an i-stem like L. mors < *mort(i)s (whence denominatives and n/cy-perfects of the fourth conjugation) or a stem in (i)yo- with ace. sg. in -im (32 a). The enclitic k(e) may serve to indic- ate direction, like -de in olxovds, (pvydde.^ The whole formation, therefore, may have spread analogically from such periphrases as ^fugansiusl < ^fugdm- k{e)-iust 'he will have gone in flight,' *meddiki{o)m-k-iust 'he will have gone to the pronouncement of justice. '
g. Indicative forms of the perfect tense are so rare that their description must be supplemented by forms from Oscan and other dialects. The 1 sg. is re- presented by 0. manafum (Co. 130, the Curse of Vibia), in which the secondary ending -m is added to the thematic vowel.^ The 3 sg. is represented by 0. aik- dafed, deded, etc., U. dede (minor inscr. no. 352 Co.), all based on the secondary ending -d < -t (41 b). The 3 pi. is represented by 0. uupsens, etc., U. eitipes < *eitom hepens (?), all based on -ens (115 d, with n. 1).
1 Cited by Aulus Gellius, VI, 9, 2-8. The classical reduplication exemplified by momordi, cucurri, etc., arose secondarily upon the principle of imitation of the radical syllable, with forms like tetendi serving as a model.
136 II. Grammar
2 O. aamanaffed, manafum are omitted here. They were taken as forms of the /-perfect to a verb coresponding to L. mandare by Buck, Oscan- Umbrian Verb System, pp. 151, 171, and by PI., II, pp. 338-9, but as reduplicated pf. < *-man-fefom, -ed (root *dhelg-, second a being anaptyctic) by Buck, O.-U. Grammar, § 223.
2 Tliis explanation is substantially that of PI., II, pp. 352-3, with certain alterations and additions.
* Observe the close similarity in the use of -ce and -ds in hi-c, huius-ce, hun-c, o-de, rov-de; observe also O. cebnust 'venerit', where ce- is prefixed to a fut. pf. form of the verb 'come'.
* In this particular instance the thematic vowel may have replaced the vowel of the root. The analysis of the form is uncertain; see n. 2 above.
125. Future Perfect Tense. Examples are numerous, and perfect stems of all the types described in 124 are represented. The following list is not complete, but is designed to show some examples of each type: pepurkurent, dersicust (124 a); vurtus, portust, faciirent (124 b) prusikurent (124 c); andersafust, am- brefurent (124 d); entelust (124 e); combifiansiust (124 f).
a. The origin of the future perfect is disputed,^ but it has been plausibly ex- plained as having arisen through analogy with the forms fust, etc., which appear in the Iguvine Tables with future value (see 122 a and Lexicon under futu), but which sometimes serve as future perfect in Oscan. The development of the new formation from perfect stems would be sufficient to establish the future perfect as distinct from the future value. For the explanation of the endings -us, -ust, -urent reference may be made to 123.
1 Buck, Verb-System, p. 176, and O.-U. Grammar'^, § 230, favored derivation as a periphras- tic formation from a pf. act. pcpl. with suffix -us- -}- a future of subjunctive origin equivalent to L. ero, etc., thus giving a combination which would be analogous to ysygacpax; eaofxai. Von Planta, II, pp. 373-6, with detailed discussion, favors connection with the L. yi-perfect. Buck, O.-U. Grammar^, in correction of 230, upholds the view here adopted.
126. Subjunctive Mood. Subjunctive forms of the present and perfect tenses are found in both Umbrian and Oscan. There are no Umbrian forms of the im- perfect subjunctive, but Oscan has fusid: L. foret and several other forms ana- logous in origin with the Latin imperfect subjunctive.
a. All present subjunctive forms from regular verbs are of the a-type, even in the first conjugation: 1 sg. aseriaia, 2 sg. kupifiaia, 3 sg. 'kuraia, portaia; second conjugation: 3 sg. prehabia, 3 pi. pass, tursiandu (8 b); third conjuga- tion: 3 pi. (or 2 sg. ?) neifhabas (116 a), 3 sg. dirsa tefa, 3 pi. dirsans, 3 sg. pass, ferar, 3 pi. pass, emantur, terkantur; fourth conjugation: 3 sg. iagia, dia. The forms of the second, third, and fourth conjugations are precisely like the corresponding Latin forms, but the forms of the first conjugation fail to agree with Oscan and Latin, which agree with one another in the use of a modal suffix -e- (either subjunctive e added to denominative stem in -ay- or old optative -ye- added to unthematic stem in -a-, in either case producing -aye >
G. The Verb 137
-ae- > 0. deiuaid, L. curet, etc.). U. kuraia, etc. cannot be derived by addition of a to a stem in -ay-, for y would be lost and the vowels would contract (53 b) ; the formation must therefore have developed secondarily by analogy with forms like fa9ia, etc. with normal i.
la. The subjunctive formation characteristic of unthematic verbs, illustrated by L. sim, velim, etc., is found only in forms of the verb 'be': 2 sg. sir si sei, 3 sg. si se si se, 3 pi. sis sins. This formation was originally optative, and the mood-sign was -yell-, which was added to the root. The original gradation shown by L. 2 sg. sies < ^s-ye-s contrasting with 1 pi. slmus < *s-r-/nos has been effaced as in classical L. sim, sis, etc., by the generalization of the grade I throughout.
c. The perfect subjunctive is illustrated by the following examples: 3 sg. combifiansi, probably also heriiei; 3 sg. pass, herifi, pihafi, pihafei, probably also ier. It is not certain whether the perfect subjunctive is in origin an optative made by adding the sign -i- to the perfect stem or a subjunctive made by adding -e- to the same stem. In favor of the former view is the optative origin of the Latin pf. subj., in which the I is added to a stem in -is- of aoristic origin and iden- tical with the -is- in the 2 sg. pf . ending -istl (e.g. 1 pi. fecerlmus < '^dhek-is-l-mos). But 0, 3 pi. tribarakattins^ and 3 sg. pass, sakrafir^ present an obstacle, since i may be from e or f but not from I. The question must be regarded as still unset- tled.3
^ Line 48 of the Cippus Abellanus (no. 95 Co.); 3 pi. subj. of t-pf. of a first-conjugation verb equivalent in sense to L. aedijico.
^ No. 114 Co.; 3 sg. subj. pass, of /-pf. of a denominative verb identical in origin with L. sacrare. The inscription is one of the group known as "lovilae dedications." Forms occurring in texts which have no character for i (e.g. the Tabula Bantina = no. 28 Co., in the Latin al- phabet, or the Curse of Vibia = no. 130 Co.) are useless for our problem, since their i i may stand for either I or e.
* Von Planta, II, pp. 369-71, with discussion, favors optative origin. Buck, § 234, favors derivation from an e-subjunctive. E. A. Hahn, Subjunctive and Optative (New York, 1953), pp. 71-2, n. 156, prefers to take heriiei, heriei herie as ye-optative forms like OL stem, pihafi pihafei as from an Z-optative.
127. Imperative forms are found in both voices, in both singular and plural, and with both second-person and third-person value; but there are no instances of formal distinction between forms of the second and third persons. As in Latin, the imperative nowhere has a special stem with a special mood-sign, but depends solely on distinctive personal endings.
a. Of forms corresponding to the Latin "present imperative" there are no absolutely sure examples, but stiplo Via 2 and aserio Via 4 probably belong in this category. As in Latin observd, etc., the stem alone is used without ending (o < a as in 7 b). The less probable alternative to taking these forms as imv. is to take them as infinitives.
138 II. Grammar
b. Future imperative active forms of the 2-3 sg. are very numerous, and the following is merely a list of selected examples : naratu naratu, portatu purtatu, habitu habetu, tursitu tusetu, fertu fertu, endendu ententu (60 a), purdouitu purtuvitu, fiitu, etu, fetii, hatu. The ending is from -tod (13 b): 0. -tud -tud, OL -tod, L. -to, Gk. -rco. For the stem before the ending see 117 i, 118 e, 119 a, 120. On the value of the forms as 2 sg. or as 3 sg. see 115 c.
c. The forms of the imv. active 2-3 pi. include: habituto habetutu, tursituto tusetutu, fertuta, ustentuta, fututo, etc. The ending -tuta -tutu -tuto points to an original -tota with -a partly changed to -u in the older tables and fully to -0 in the later tables (7 b), but the origin of the ending is uncertain, etatu etato are apparently 2 pi. of a frequentative verb (117 d, as if = L. *itare beside ire) with loss of a syllable through haplology in *etatutu *etatuto, but etuta etutu etuto are normal pi. imperatives of the primary verb etu.
d. The forms of the imv. passive include: 2-3 sg. eturstahmu, spahmu, anouihi- mu, persnimu persnimu; 2-3 pi. arsmahmo armamu (written arraanu), ca- terahamo kateramu, persnimumo. These forms are in some way related to the rare and archaic Latin passive imv. forms progredimino, fruimino, but it is im- probable that they actually result from a change -m{i)no > -mo; rather the U. and L, forms show inherited suffix-variants, for which the Balto-Slavic pass, pcpls., with suffix -mo- in contrast to Gk. -jxevoq, Skt. -manas, provide some evi- dence. U. -mu < -mod may then represent a passive imv. formation modeled after the active imv. in -tod, while 0. imv. pass. 3 sg. censamur shows a further remodeling with -r from other passive forms. The pi. imv. persnimumo has the same relation to sg. persnimu as act. pi. etuta etutu etuto to etu. The plural value of arsmahamo, caterahamo is clearly shown by their association with the pi. louinur Ikuvinu in VIb 56, lb 20; the syllabic loss in arsmahamo can be explained by haplology (<*arsma/iamii7no) and in caterahamo by its juxtaposition with arsmahamo.
128. Infinitive. The only simple infinitive preserved in the Italic dialects is the present active. The Umbrian examples are: afero aferum, fagiu, fagu, erom eru, these last belonging to the verb 'be' and used in some instances as auxiliaries in the formation of the pf. pass. inf. The Oscan pres. inf., which is precisely similar, may be illustrated by the following examples: moltaum, tri- barakavum,! fatium, deikum, deicum, ezum. The origin of the formation is uncertain, but it is difficult to avoid the belief that it is an ace. sg. of a verbal noun, with the ending -om which belongs properly to o-stems. The Oscan forms have u u < 0 before final m. Although the verbal nouns which developed into infinitives in the various IE languages were made from the root and not from separate tense-stems, nevertheless the O.-U. infinitives, like the Latin forms in -re, are made from the present stem and share in its formal characteristics.
G. The Verb 139
a. For the perfect passive infinitive see 133,
1 Twice on the Cippus Abellanus (lines b 2, 10 = no. 95 Co.), but in the former instance the last four letters depend on restoration and in the latter the u is not clearly legible.
129. Supine. The only certain example of this formation is aseriato anzeriatu. Other alleged examples, including aso VIb 50 and vagetum lb 8, are better taken as pf. pass, pcpls. The supine in Umbrian is precisely like that in Latin, both in formation and in use: it is in origin the ace. sg. of a verbal noun with stem in -tu-, as is also the classical Skt. infinitive in -turn. For -o < -urn see 14 b. There are no examples of a supine in Oscan.
130. Active Participles.
a. The present active participle is represented by restef (119 b), serse zefef : L. sedens, and probably kutef (22, 118 c). Although there are no oblique case- forms and no sure Oscan examples at all, there is no reason to doubt that the present participle, as in Latin, Greek, etc., was formed by the addition of -nt- to the present stem. For final / < -nts, and also for the possibility of deriving the preposition traf tra from a pres. pcpl., see 58 d.
b. The pf, act. pcpl. made with the suffix -wes/wos/us- and exemplified by sidcbg, Idvla, Skt. vidvams, vidusl, and Balto-Slavic forms, is not preserved as a productive category in the Italic languages, yet several forms are explained as remnants of it: L. prepos. apud as neut. nom.-acc. sg. to verb-root found in apiscor;^ 0. sipus (124 c) < ^sepus or, with lengthened grade of the suffix, *sep- wos, although it may be from o-stem ^sepwos with samprasarana as in 32. The pf. act. pcpl. is sometimes taken as a component of the O.-U. fut. pf. tense. A different explanation is adopted here, but see 125 a with n. 1.
1 See R. G. Kent, Forms of Latin, § 423, II.
131. Perfect Passive Participle. Examples are very numerous, and most of the conjugational classes are represented. The following is a selected list: kuratu: L. curatum; gersnatur: L. cenati; kunikaz conegos; tases tagez: L. tacitus; screihtor: L. scripta; spafu if < ^spand-tom (58 c with n. 3); similarly spefa; persnis; statita; purtitu purditom; daetom; peretom. See also 71 for adjectives of the type of L. barbdtus, which in a formal sense are precisely like pf . pcpls. Both these and the true participles in Umbrian, as in Oscan and Latin, are formed by means of the suffix -to/d-. In verbs of the first and fourth conju- gations the suffix is normally added to the -a-, -i- of the verb stem. In verbs of the third conjugation it is added directly to the root, causing secondary changes in some of the resulting consonant-clusters, of which the most striking is that illustrated by spafu, spefa above. There are no sure examples of participles
140 II. Grammar
preserving the characteristic e of the second conjugation, since even uirseto may have unsyncopated e. On the difficult question of the e here and in tases, muieto, prusegetu, daetom, peretom, frosetom, pesetom, uasetom, some of which belong to verbs of the first conjugation, see 30 c with n. 2.
132. Gerundive. The examples are: gen. sg. m. anferener, pihaner pehaner peihaner; nom. sg. m. pelsans, ace. sg. m. pelsanu, ace. pi. f. pelsana pelsana. 0. upsannam, sakrannas give evidence of etymological gemination of the n. The 0. and U. forms are therefore to be equated with L. gerundives piandus, ferendus, etc., on the basis of the O.-U. change -nd- > -nn- (61 e), the ultimate origin of the formation remaining obscure.
133. Periphrastic Forms. As in Latin, the passive forms of the perfect system are normally made by means of the pf. pass. pcpl. with appropriate forms of the verb 'be, ' including forms of the root /u- as well as es-. The perfect indicative forms are: uaseto est, pesetom est, peretom est, frosetom est, daetom est, orto est, screhto est, pi. screihtor sent, subator sent. The only pf. subj. form of this type is kuratu si. The only pf. imv. form is purtitu futu. The pf. inf. forms are erom ehiato and kuratu eru. The fut. pf. forms are muieto fust, persnis fust, pihaz fust = pihos fust, purtitu fust = purdito{m) fust, spafu fust, pi, gers- natur furent. For pass. pf. subj. forms of the type of pihafi see 114 a, 126 c. For fut. pf. forms of the type of benuso see 115 g.
134. Verbal Composition. The general principles for compounding verbs with prefixes are the same as in Latin. The effect of the prefix on the meaning of the verb is to define more precisely the location, direction, etc., of the action. The compound then often shows the attainment of a goal where the simplex shows no such attainment (cf., for example, L. consequor beside sequor. Germ. ersteigen beside steigen), and the distinction becomes similar to that of the " per- fective " and " imperfective " aspects of the Slavic languages, but not to such an extent as to be reduced to an orderly system. Consequently no attempt is made to distinguish between those compounds in which the prefix has local value and those in which it merely shows action brought to completion.
Many of the verbal prefixes are also found as prepositions (or postpositions) with noun objects, and among the rest the lack of similar examples of preposi- tional use may be accidental; the Latin equivalents of dis- and re-, however, are used only as prefixes, and there is no reason to beheve that they were ever otherwise used. For those prefixes which also serve as prepositions with nouns cross-references are given below, but detailed discussion of etymology and nuan- ces of meaning is avoided here.
a. aha- ah- a- a-: L. a-, ahatripmsatu ahtrepufatu etc., ahauendu.
b. am- am- an- a- a- 'around, about': df^(pi-, L. amb-. This and an- immedi-
G. The Verb 141
ately following are sometimes difficult to distinguish, as a result of the assimila- tion or loss of the nasal to which both are subject. The chief criterion for sep- arating the two prefixes is the sense of the compounds in which they occur. The following are to be referred either certainly or probably to am- : afero af erum, anferener, and its suppletive fut. pr. an-dersafust aterafust; amboltu.
c. ambr- ampr- apr- 'around, about': 0. amfr-. Related to am{b)- as L. super to sub. ambretuto, amprehtu apretu.
d. an- an- am- a- a- 'up, upon', etc.: dva-, L. an- (in anhelo). The forms which are to be assigned with certainty or reasonable probability to this prefix, in contrast to am- above, are: afiktu, amparitu, ampentu apentu, andendu atentu, anouihimu; anseriato aseriatu, etc.; anstintu astintu, anstiplatu.
e. ander-: 0. anter, L. inter. andersistu.
f. ars- af- ar- ar- a-: L. ad-, (a)fliabas in crasis neifhabas, afpeltu, ar- sueitu etc. Cf. 147 a.
g. com- kum- co- ku-: L. com- con-, combifiatu kumpifiatu, comohota, co- moltu kumultu, conegos kunikaz, couertu kuvertu, kukeh.es (?), kuveitu. Cf. 147 b.
h. da-: 0. dat (and more remotely L. de). daetom.
i. dis-: L. dis-. disleralinsust (?).
j. ehe- eh- e- e-: 0. eh- ee-, L. e. efurfatu, eheturstahamu etufstamu, ehia- io, ehueltu, eveietu. Cf. 147 c.
k. en- en-: 0. en, L. in-, iv-. endendu ententu, enetu. Cf. 147 d.
1. OS- us-: 0. o{s)- (in ossi[ns = 'L.*ob-sint), L. ob- obs- os-. ostendu ustentu.
m. per- per-: 0. per-, L. per-, peretom, pertentu. Cf. 147h on prep. pert.
n. pre- pre-: 0. prae- prai, L. prae-. prehabia, prepesnimu, preplotatu, preuendu, preuislaiu. Cf. 147 k.
o. pro- pru-: 0. pru-, L. pro-, procanurent, proseseto, prusekatu, prupe- hast, prusikurent. Cf. 147 g on -per.
p. pur- pur-: L. por-. purdouitu purtuvitu.
q. re- re-: L. re-. restatu reste, revestu.
r. sub- sub- su-: L. sub-. subator, subahtu subotu, subocau, sumtu < "^subemetod, with assimilatory loss of b after syncope of the radical e; sutentu.
s. irah-'. L. trans- tra-. trahuorfi (adv. from pf. pass. pcpl. of verb corres- ponding to L. trans-verio). Cf. 147 p.
142 II. Grammar
H. Syntax
135. General Remarks. The syntax of the Iguvine Tables, like that of the other Italic dialect inscriptions, resembles Latin so closely in most respects that a detailed account of it is unnecessary. In the following brief outline the chief emphasis is on those features of the syntax which are unusual (although many of them have parallels in Plautus or in early Latin prose) and on those principles which are important for the solution of problems of interpretation.
136. Ellipsis. The omission of the subject, verb, or direct object of the sen- tence is moderately frequent, but it is not possible in all cases to make a sharp distinction between clear cases of ellipsis and mere omission of words which are unnecessary because they can easily be supplied from the context.
a. The subject of the imperative forms used in the ritual instructions is omit- ted far more frequently than it is included. See Introduction, § 16.
b. Omission of the copula is so widespread in many IE languages as hardly to deserve notice among cases of ellipsis. In lib 1-2 tekvias / famefias pum- perias XII, however, the 3 pi. sent is probably to be supplied if the interpre- tation adopted in this edition is correct. See translation and notes.
c. Most of the following passages contain an accusative direct object but omit the verb on which the object depends: — I la 25 tiu puni tiu vinu: it is not certain what verb is to be supplied, but cf. Via 25 Di Grabouie, tio esu hue peracrei pihaclu ocreper Fisiu totaper loiiina irer nomneper erar nomneper, where the verb would be subocau, as appears from the almost identical 33-4. Cf. also VIb 9-10, Vila 9-10. — la 18 kapif purtitaf sakref etraf purtitaf etraf sa- kref tutaper Ikuvina: the verb may be inferred from the corresponding VIb 18 capif purdita dupla aitu sacra dupla aitu.- — IV 14 putrespe erus: cf. erus.../ teftu 27-8. — Via 17-8 parfa dersua curnaco dersua peico mersto peica meersta, etc.: the accusatives cannot depend on carsitu which immediately precedes, for the sentence is oratio recta, as tefe in 18 shows. Some verb equivalent to ' I see, I have seen ' must therefore be supplied. — Via 27, 37, 47, b 29 pus{e)i neip her(e)itu: the full form is shown in Ila 4 fetu puze neip eretu.
d. Ellipsis of the direct object: — Va 17 ape apelust: the object of the verb would designate the sacrificial victim. — III 20-1 ap / vuku kiikehes iepi pers- klumar kafitu: the object of kafitu would designate the persons called to the ceremony. The forms comoltu VIb 17, 41, Vila 39, 44, 45, kumaltu Ila, 9, IV 28, kumultu la 34 are without direct object, kumaltu Ila 41 is preceded by the series of partitive genitives struhglas fiklas sufafias, and in every instance it is probably the cakes of the type used in each particular sacrifice which are to be ground up.
H. Syntax 143
137. Asyndeton. The omission of the conjunction 'and,' like the omission of the copula, was a widespread and well-established IE phenomenon. In the Iguvine Tables it is so frequent as to be the rule rather than the exception. For the following pairs line references are unnecessary: arepes arves 13 times, antakre(s) kuniate(s) 3 times, snata asnata (-u, -es, -e) 4 times, dupursus peturpursus, sihitu ansihitu {-ir) 7 times, hostatu an{h)ostatu (-ir) 7 times, atru alfu; nome nerf arsmo ii{e)iro pequo castruo frif 11 times; Puemune Puprike Vesune Puemunes Pupfikes IV 10-11, but Puemune Puprike et Vesune Puemunes Pupfikes 12-13. etu veltu IV 21 is asyndetic, but the verbs are not coordinate in the full sense. Via 1 parfa curnase dersua peiqu peica merstu shows two asyndetic pairs, but the connective is not necessarily et, since the re- quirement may be for a parra or a crow dersua, a woodpecker or a magpie merstu. See note on passage.
138. Agreement. The fundamental rules are the same as in Latin : an attrib- utive or predicative adjective agrees with its noun in number, gender, and case; a verb agrees with its subject in person and number; a relative agrees with its antecedent in gender and number, while its case is determined within its own clause. But deviations occur as a result of case-attraction, anacoluthic sentence- structure, or the replacement of strictly formal agreement by agreement ac- cording to sense.
a. A singular subject which for any reason is plural in sense may have its verb in the plural : Va 24-6 sve mestru karu / fratru Atiief iu pure ulu benurent / prusikurent rehte kuratu eru... The verb prusikurent takes its number not only from the collective sense in karu but also from the influence of the intervening relative clause, whose verb benurent receives its plural number from fratru... pure; similarly 27-9; similarly b 3-6, where pepurkurent is the verb showing attraction of number. In Vila 52 hondra furo sehemeniar hatuto totar pisi heriest, despite the singular clause pisi heriest it is understood that the persons catching the heifers will be more than one; hence plural hatuto. In VIb 56 the adfertor with the two prinuatur make a pi. subject: eno com prin- uatir peracris sacris ambretuto. Cf. Livy, XXI, 60, 7 ipse dux cum aliquot prin- cipibus capiuntur.
b. Via 19-20 shows an example of attraction of the antecedent into the case of the relative; or, more strictly, the antecedent in the nominative case introdu- ces the sentence, followed by the relative clause, after which the antecedent is resumed by a demonstrative in the accusative as object of the principal verb, the sentence as a whole being anacoluthic: uasor... porsi ocrer / pehaner paca os- tensendi eo iso ostendu pusi...: uasor is nom., since ace. would end in -/ (103 a), but eo, object of ostendu, must be ace.
c. The following examples show a shift in number between a pronoun and the
144 II. Grammar
noun or pronoun to which it refers: Vila 13-14 futu fons I ... pople totar liouinar tote liouine erom nomne erar nomne. erom refers to pople, being plural by con- struction according to sense, but in the similar passages 17-18, 31, the sg. erer is used. In III 33-IV 3 the sg. erek is used three times with reference to tuva (triia) tefra; see note on III 34. Va 7-8 shows a shift in number between re- lative and antecedent: revestu pure tefte eru emantur herte 'revisito, quae dantur [num] ex iis accipiantur oporteaf (Buck): the usage of pure resembles that of efek in III 33-IV 3 above, while eru is a pi. partitive subject, unless we prefer to take eru as abl. sg. with Conway, Dial. Ital. Ex. Sel., who translates 'ex eo quod datur.'
139. Interchange of Predicate adjective and Adverb or Preposition- al Phrase. The following constructions are striking chiefly because they show an adverb or prepositional phrase where English usage leads us to expect an adjective, or vice versa. Parallels can sometimes be found in Latin, and to some extent the usage varies between Umbrian and Oscan, or even within the Iguvine Tables themselves.
a. An adverb is used where we should expect a predicate adjective: — ■ Va 24-27 sve mestru karu... prusikurent rehte kuratu eru, erek prufe si (as if L. used id probe sit in place of id probum sit). — Via 8-10 angluto / hondomu porsei nesimei asa deueia est anglome somo porsei nesimei uapersus auiehcleir j est: with the adverb nesimei (-ei < -ed) contrast the predicate adjective in Oscan no. 109 Co. nessimas staiet veruis luvkei ' proximae slant portae in luco.'
b. An adjective is used where we should expect an adverb or prepositional phrase, or the adjective may replace the preposition: — tases persnimu, kutef pesnimu, these two expressions with their variations making 29 examples; cf. Livy, XXXV, 19, 1 contumeliam tacitus tulit. — Ila 10-11 uve... / peraem fetu 'sacrifice the sheep on the ground,' with variations 11 examples in all; similarly Ila 21-22 esunu / perae futu 'the sacrifice shall be on the ground. ' — Ila 30-31 supa spantea pertentu ' at the side. ' — 32 sup a pustra perstu ' at the back ' ; cf. b 19, VI b 5, Vila 8, the forms of postro- in all these examples being accusa- tive in agreement with the things which are placed at the back; similarly IV 17-18 persuntru supu eregle... purtuvitu 'at the bottom of the statue'; but 19 has super eregle with a preposition, — VIb 17 eno mefa uestisia sopa(m) purome efurfatu, subra spahmu 'down into the fire'; sopa{m) agrees with ues- tisia or mefa, but subra is best regarded as an adverb, as appears plainly in Va 20, Via 15 (first instance), Vllb 3. Vila 38-39 is nearly equivalent to VIb 17.
140. The Cases. The syntactical differences between the Italic dialects and Latin in regard to the cases are considerably less striking than the morphological differences. As in Latin the ablative case combines the functions of the IE abla- tive, instrumental, and (in part) locative; but just as the locative has maintained
H. Syntax 145
itself as a distinct formal category to a greater extent than in Latin, so it is ac- tually used to a greater extent.
141. Nominative. The uses as subject and as predicate noun or adjective need no explanation. The Iguvine Tables show no clear examples of vocative use where distinct vocative forms are available. There are two possible instan- ces of an unconstrued nom. used as a heading for what is to follow: — Via 12 Tuderor totcor: uapersusto auieclir ebetrafe, etc., unless we prefer to consider tuderor totcor the subject of a sentence in which the copula is omitted. — Ila 15 Huntia may be a heading for the instructions on the dog-sacrifice which follow, if the suggestion contained in the note on the passage is correct.
142. Genitive. Most of the uses, especially the possessive and other adnom- inal uses, need no explanation. Uses which seem to deserve some notice in- clude the following:
a. Genitive of origin: Vb 9 farer opeter p. Ill I agre Tlatie Piquier Martier, The Ager Tlatius Picuvius Martins is the source of the spelt.
b. Among partitive uses some are obvious; among those which are more unusual, at least from a Latin standpoint, are the following: — Ila 41 struhglas fiklas sufafias kumaltu: the three gen. sg. forms are partitive objects of ku- maltu. — Via 13 tertiame praco pracatarum: the third in a certain series of objects is used as one of the landmarks in defining the augural templum. praco may be ace. sg. modified by tertiam or gen. pi, modified by pracatarum, but the last is clearly partitive gen. See note on the passage. — Va 8 eru emantur: a partitive subject; but see 138 c above, with the alternative explanation.
c. Judicial genitive: a probable instance is Vllb 4 fratreci motar sins a. CCC. For discussion see note on the passage.
d. Gerundive in the genitive case expressing purpose: Via 8 Uerfale pufe arsfertur trebeit ocrer peihaner. b 48 sururo stiplatu pusi ocrer pihaner. Cf. Tac, Ann., II, 59, 1 Aegyptum proficiscitur cognoscendae antiqiiitatis.
In Via 18-19 the gerundive construction may depend loosely on the noun esoneir: esisco esoneir seueir j popler anferener et ocrer pihaner perca arsmatia habitu. In 19-20, immediately following, paca functions as an improper preposi- tion governing the genitive: uasor uerisco Treblanir porsi ocrer / pehaner paca ostensendi...; so most construe the sentence (otherwise Dev. and Pis., taking paca as abl, of manner ' ritu' or 'ex pacto').
143. Dative. Most uses require only the briefest mention. The dative of in- direct object is illustrated by Via 22 laue Grabouei buf treif fetu, with similar expressions at the beginning of the instructions for other sacrifices. Dative with adjectives: Vib 7-8, etc., fons sir pacer sir ocre Fisi tote liouine erer nomne erar
146 II. Grammar
nomne. Dative with compound verbs: Ila 19-20 = III 21-22 pir ase / antentu; IV 21 erek persuntre antentu; Via 56 prosesetir mefa spefa ficla arsueitu and other examples with arsueitu; la 31 esmik vestiga afiktu. The following uses are less familiar or less obvious:
a. Dative of possessor used with a form of the verb ' be ' : — Vb 6-7 etantu mutu afferture / si. — Vllb 1 Pisi panupei fratrexs fratrus Atiersier fust 'Who- ever at any time shall be brother-superior among (or for) the Atiedian Brothers. '
h. Dative with a form of the verb kuratu: Va 4-5 ri esune / kuraia. Cf. the rare Latin use which appears in Plant., Trin., 1057 sed ego sum insipientior qui rebus curem publicis.
c. With the impersonal verb herter: lla 40 esuf pusme herter. Cf. Ter., Ad., 928 ita nobis decet.
d. Dative of advantage or disadvantage: — Vila 11-3 Prestota Serfia... preuendu uia ecla atero tote Tarsinate, etc., in the curse against Iguvium's ene- mies. With opposite effect 26-7 Prestota I Serfia... ahauendu uia ecla atero pople totar liouinar tote liouine, etc. — Via 17-8 parfa dersua curnaco dersua peico mersto peica meersta, J mersta aueif mersta ancla eesona tefe, tote liouine esmei stahmei stahmitei: the dative construes less easily with dersua... mersta as 'west... east' than as ' prosper am... iustas' with earlier editors, yet the notion of the omens serving the interests of the individual and the state is quite clear.
144. Accusative. The most frequent uses are as direct object of verbs and as object of prepositions (or postpositions). For elHptical expressions with omis- sion of the verb governing the object see 136 c. For the uses with prepositions see 147 a, d, e, h, j, 1, n.
a. An example of the ace. with a verb of putting on (passive with middle sense) occurs in Vib 49 perca arsmatiam anouihimu. cringatro hatu, destrame scapla anouihimu. For the probable sense of perca and the semantic evolution of anouihimu see on Via 19. For the case-construction in general cf. Verg., Aen. II, 392-3 Androgei galeam clipeique insigne decorum / induitur.
145. Ablative. The uses fall into three main categories : " true " ablative constructions involving source and separation; instrumental-comitative con- structions; and constructions of place and time in which the ablative competes with the locative. The ablative absolute is treated among participial construc- tions; see 153 h. For prepositional uses see 147 b, c, g, i, k, o.
a. The ablative of place whence, unaccompanied by a preposition or postposi- tion, is rare: — Via 8-9 angluto j hondomu porsei nesimei asa deueia est ' proxume ab ara divina.' — III 23 = IV 15-16 testru sese asa; sese as a preposition governs testru, and the resulting combination governs asa. So probably also
H. Syntax 147
in Via 2 tremnu serse arsferture ehueltu, 16-17 tremnu serse / combifiatii. For reasons of word-order it would be easier to take tremnu with serse than with ehueltu, combifiatu, but in this case we should expect the locative tremne.
b. Instrumental and comitative constructions include the following: — ^ Ab- lative of accompaniment: lib 13 persutru vaputis mefa vistiga feta fertu: probably persutru is ace. object of fertu, while vaputis and the succeeding nouns are abl. ; see note on passage. — Ablative of means: VIb 52 uia auiecla esonome etuto; cf. 65, lb 14, etc.; Ill 12-13 sakre uvem / kletra fertuta. A variety of the instrumental use is the frequent ablative with fetu: Via 56-7 este j esono heri uinu heri poni fetu; similarly b 1-2, etc. Cf. Acta fratrum arvalium, pp. 11, 12, 24, 26, etc. (Henzen) ture et vino fecerunt. Somewhat different and probably to be referred to the " true " ablative is VIb 55 fetu uru pirse mers est 'do with him what is the law. ' Cf . L. quid hoc homine facias ? — Instrumental ablative with persnihmu: VIb 9 poni pesnimu, and frequent similar expressions. — -Ablative of accordance: Va 11-12 felsva/afputratifratruAtiiefiu prehubia 'at the dis- cretion of the Atiedian Brothers'; III 4-6 uhturu... / ... ustentuta pure / fra- tru mersus fust. See note on passage. — Ablative of price: Va 17-18 ape apelust, muneklu habia numer / prever pusti kastruvuf 'shall receive a donation [at the rate of] a single nummus per head'; similarly 19, 21.
c. Locative uses: These include both local and temporal constructions. In the plural there is of course no formal distinction between the ablative and locative cases, and even in the singular the formal distinction is not always clear. Local uses: — Via 5 sersi pirsi sesust; sersi is probably abl., since the loc. sg. of both consonant- and i-stems in the 3rd declension ends in -e -e. — Via 2 trem- nu serse arsferture ehueltu. — • Via 11 todceir tuderus seipodruhpei seritu 'within the city boundaries.' Temporal uses: Ila 17 menzne kurglasiu fagia tigit. The ending of the adjective shows that the case is abl., not loc. The justification for the case is that the construction is very close to being an abl. of attendant circumstance or abl. abs. rather than a purely temporal construction.
146. Locative. The constructions are partly local, partly temporal. For uses with prepositions and with the postposition -en -em -e see 147 d, m. This post- position frequently contracts with -e, the loc. sg. ending in the first and second declensions, to which most of the attested loc. sg. forms belong; and since final nasals are frequently not written, it is possible that many or even all loc. sg. forms, if not accompanied by some preposition, actually contain -e{n). Examples not preceded by prepositions and not clearly followed by -en -em -e include the following :
a. Local uses: — Via 57, etc., ua/uo ferine fetu. — VIb 50 aso destre onse fertu (but in lib 27, 28 testre e uze habetu the postposition is clearly present). — Va 15-16 kumnah/kle Atiierie ukre: the first noun, with its adjective, is loc-
148 II. Grammar
ative, and so is ukre; eikvasese Atiiefier, which follows, has postpositive -e. — VIb 54-5 sopir habe / esme pople 'if anyone is caught among this people.'
b. Temporal uses: — lb 45 = Ila 44 kvestretie... svesu (svesu is inflected before the -su; the case-ending is -e-); similarly Vllb 1 sueso fratrecate. — Ila 15, III 2 sume ustite 'in the final period' (?). — Va 2, 14-15 plenasier ur- nasier.
147. Prepositions and Postpositions. The cases used are the accusative, ablative, and locative. -ar -a, hondra, pert, posti pusti, subra, superne have only the accusative. com -co -ku, ehe e, -per, post pus, pre, -fo -ta -tu have only the ablative, super has the locative (but cf. superne with ace), -en -e -em and traf have the accusative and locative, paca used in Via 20 with a preceding gerun- dive construction resembles Latin causa used as an improper postpositive pre- position with the same construction. For the use of prepositions as prefixes in combination with verbs see 134.
The follov/ing account of the prepositions is merely intended to illustrate the most important and typical uses. For a full listing as well as for etymological data see the Lexicon.
a. -ar -a with ace, 'to, toward': spinamar etu, persklumaf kafitu; 'in the direction of, at ' : tuva tefra spantimar prusekatu, spiniama persnihinu vestikatu ahtrepuratu.
b. -co -ku with abl., 'with': etuto com peracris sacris, com prinuatir; in post- positive use ' at ' : asaku, destruco persi, nertruco persi, etc. ; only with pronouns is the postpositive used in sense 'with': VIb 50-1 erucom prinuatur dur j etuto.
c. ehe e with abl., 'out of, from among': VIb 54 eetu ehesu poplu; 'away from': Ila 38-9 manf easa / vutu.
d. -en -e -em. With ace, 'into, to, upon': pune fesnafe benus, uapefe auieclu, (cringatro) destrame scapla anouihimu 'place over his right shoulder'; showing result, vasetome fust; cf. L. in irritum. With loc, 'in, upon': fesnere purtuetu, ocrem Fisiem ' (if) on the Fisian Mount (fire has occurred). ' Occasionally -e{n) is attached both to the noun and to a qualifying adjective: with ace, lb 14 vapefem avieklufe kumpifiatu (but not in the corresponding VI b 51 uapefe auieclu); with loc, ocrem Fisiem, toteme louinem. Buck, § 169, 7a, compares 0. hurtin Kerriiin (Agnone Dedication = Co. no. 175, line 1-2) and suggests that the extension of -m to the adjective, as if it were a case-ending, was favored by the parallelism of locative forms with and without -m and of accusative forms with and without -m, the -m of the ace being of course quite natural in the ad- jective, toteme actually contains a double loc. ending, induced perhaps by such ace forms as totame. Ill 20 vukumen esunumen etu is probably not an in- stance of repetition of the -en; see note on passage.
H. Syntax 149
e. hondra hutra with ace, 'below': hondra furo sehemeniar 'below the Forum Seminarium, ' hondra esto tudero ' below these limits. '
f. paca with gen., 'for the purpose of : ocrer pehaner paca 'for the purification of the Mount. ' See introductory remarks in the present section.
g. -per with abl., 'for, on behalf of: teio subocau... ocriper Fisiu totaper liouU na, erer nomneper erar nomneper (frequent).
h. pert with ace, 'beyond, on the other side of: pert spinia (but -pert Ila 35 is an error for -per).
i. post pus with abl., 'behind': post uerir.
j. posti pusti(n) with ace, with distributive force like >card, Eng. per, L. in (singulos, etc.): pustin angif 'at each turning'; pusti kastruviif 'per head' (?); posti acnu 'each year.'
k. pre pre with abl., 'before, in front of: pre uerir.
1. subra with ace, 'above': Via 15-16 subra esto / tudero; elsewhere only as adv.
m. super with loe, 'above, at the top of: lb 41 super kumne 'above the place of assembly'; IV 19 super eregle 'at the top of the statue. '
n. superne with ace, 'above': superne adro 'above the black (vessels).'
o. -to -ta -tu with abl., 'from': uapersusto auieclir 'from the augural seats* (in defining the limits of the templum); scalseto skalgeta (purtuvitu or dirstu) 'from a cup.'
p. traf traha{f) tra 'across.' With the exception of Ila 13 tra ekvine 'across the Via Equina '(?), all examples govern forms of sahatam, making a place-name similar to Trastevere, and the case-syntax then depends not on traf but on the relation of the resulting hypostasis to the rest of the sentence. See on Vila 5.
148. General Remarks on the Syntax of the Verb. The treatment of the syntax of the verb is brief, partly because in most respects the resemblance to Latin is so close as to make a detailed description superfluous, and partly be- cause the extent and nature of the material have limited both the number of morphological categories and the range of uses of those categories which are actually represented. In the indicative only the present, future, perfect, and fu- ture perfect are found, the last being the most frequent; in the subjunctive only the present and perfect, with a few imperfect forms in Oscan and Paelignian. With the exception of stiplo and aserio all imperative forms are of the "future", type. The infinitive is limited to the present active and perfect passive, the par- ticiples to the present active, perfect passive, and gerundive.
149. Impersonal Use of Verbs. The use of 3 sg. forms, especially passives
150 II. Grammar
without definite subject is moderately frequent. The chief examples are: — Via 27-8 = 37 = 47 persei tuer perscler uasetom est pesetom est peretom est frosetom est daetom est; Vllb 3 pusei subra screhto est 'as it is written above'; but Via 15 has screihtor sent with definite pi. subject; VIb 42 purdito fust, Va 18-19 ape purtitu / fust; yet lb 38-9 esunu / purtitu fust has esunu as subject, and we may have to explain the equivalent Vila 45 purditom fust as an ellipsis; similarly IV 31-2 purtitu / futu, but esunu / purtitu futu in Ila 43. — Va 20 ape subra spafu fust; see on Va 17-22. — Va 23-9 ehvelklu feia... / sve rehte kuratu si. sve... / ... / prusikurent relite kuratu eru. ../... sve ... / ... prusikurent / kuratu rehte neip eru... — • Via 7 sue muieto fust 'if a noise is made.' — VIb 63-4 ape termnome j couortuso ... ape termnome benuso 'when the boundary has been reached'; 65 = Vila 1-2 erafont uia pora benuso. — Ila 17 fagia tigit; cf. L. decet -\- inf. — Va 5-6 pire uraku ri esuna / si herte, like L. oportet + subjunctive; cf. 8, 10, b 8, 11, 13, 16, Vllb 2, III 1. — Ila 40 esuf pusme herter ' he whose duty it is. '
150. The Voices. Passive forms are found in normal passive use, in imper- sonal use, and as deponents. It is unsafe in specific cases to affirm the non-exis- tence of active forms in languages as poorly preserved as the Italic dialects, and yet it is reasonably certain that deponents must have existed in Umbrian as in Latin and Greek.
a. For the normal contrast of active and passive the following examples will suffice: — Via 29 Di Grabouie, pihatu acre Fisei, pihatu tota louina; b 46-7 eno ocar / pihos fust. — Via 19-20 uasor... porsi ocrer / pehaner paca ostensendi, eo iso ostendu... 'the vessels which are to be exhibited... he shall so exhi- bit... '
h. For the impersonal use of passive forms see 149 above.
c. Among deponent verbs should probably be included the following: persnimu: cf. L. precor with different stem but same root and meaning; terkantur: degxo- fj,ai (?); eheturstahamu. ^ersnatur furent, the sense of which is active, may, like L, ceno, have possessed active forms, but no such forms are quotable.
d. When active and passive forms coexist, the difference may be one of tran- sitive and intransitive meaning: III 13-14 kletram / amparitu 'set up the Utter'; Ila 42 amparihmu stand up'; cf. tarrj/ui : tarajuaL. It is not clear, however, why the active subra spahatu is used in VIb 41 but the passive subra spah(a)mu in 17 = Vila 39.
e. No active forms are quotable beside passive arsmahamo caterahamo, but the active would presumably have been used of the commander who arranged the companies, the passive being a middle used of the persons arranging themselves. So probably ^anouihitu would mean 'put (clothing) on (another)' as anouihimu is ' put (clothing) on (oneself). '
H. Syntax 151
151. The Tenses. There are few, if any, usages which are pecuhar by con- trast with Latin.
a. The present indicative is occasionally used in subordinate clauses where we might expect the future: VIb 43 (= lb 1) ponne oui furfant; Vllb 2 ponne iuengar tursiandu hertei, but fut. in VIb 48 (= lb 10) pone poplo afero heries; VIb 54-5 sopir habe / esme pople 'if anyone is caught among this people'; the form habe is obscure but undoubtedly of the present tense.
b. The perfect indicative is used of past action with lasting effect in Va 1-2 (= 14) esuk frater Atiieriur / eitipes; the last word is partly obscure but al- most without doubt a pf. ind. verb.
c. The future and future perfect differ as in Latin. In subordinate clauses, where the majority of our examples of both tenses are found, the fut. shows the action as in progress, the fut. pf. as completed, when the main action occurs; so, for example, in lib 27-8 pune anpenes krikatru testre e uze habetu. ape apel/us, mefe atentu 'while you are slaying... when you have slain.' But there is some fluctuation in usage: lb 15-16 pune menes (= *benes, L. venies) I Akefuniamem, but fut. pf. in the corresponding VIb 52-3 ape Acesoniame/ ... benust; lib 16 pune fesnafe benus.
d. Among the tenses of the subjunctive only present and perfect forms are quotable in Umbrian. The distinction between the two tenses is not purely one of relative time, or rather most of the few examples do not occur in situations where such a distinction can be observed. Like the distinction of the Gk. present and aorist subjunctive, it appears to be partly a matter of aspect. Via 28-9 (=z 38, 48, b 31) esu bue / peracrei pihaclu pihafei; the subj. is pf. pass., and the purification is contemplated as a completed act. VIb 51-2 neip / amboltu prepa desua combifiansi 'until (the augur) has announced.' Vb 3-7 panta muta... / ... / ... afferture eru pepurkure/nt herifi, etantu mutu afferture / si 'what- ever fine they demand it be pleasing to be (imposed) on the adfertor, so great a fine shall be (imposed) on the adfertor. ' For the sequence of tenses it is impos- sible to find adequate examples, but the few imperfect subjunctive forms in Oscan and Paelignian occur in subordinate clauses depending on perfect indicative forms; in other words with the same tense sequence as in Latin (cf. Cippus Abellanus = Co. 95, lines 19, 50, 51, 54; Pael., Co. 239).
e. The imperative forms, being used in permanent instructions for the perfor- mance of the ritual, are almost all of the "future" type: naratu = L. narrato, habitu = L. habeto, etc. The two present imv. forms stiplo Via 2, aserio Via 4 are in oratio recta, in the dialogue between the augur and adfertor during the taking of the auspices at the beginning of the purification of the mount.
f . The participles have the same tense-values as in Latin, the pres. act. showing action contemporaneous with that of the main verb, the pf. pass, showing action
152 II. Grammar
prior to that of the main verb. For the time-value of the pcpl. in the ablative absolute aueis aseriater Via 1, see introductory note on Via 1-18.
152. The Moods. The uses of the indicative and subjunctive in subordinate clauses are treated in 1E4. The present section is concerned with the subjunc- tive and imperative in expressions of command, exhortation, etc., and with dependent uses of the subjunctive where no subordinating conjunction is present.
a. In the following passages 2nd and 3rd person forms of the subjunctive are used with a value scarcely different from that of the imperative: Va 3-5 affertur ... / ... ri esune / kuraia, prehabia pife... 10-12 affertur... / ... esunesku vepurus felsva / ... prehubia, 17-23 ape apelust, muneklu habia numer / prever pusti kastruvuf,... (19) habia... (21) habia... (23) ehvelkki feia fra- treks ute kvestur. Cf. 27 prufe si, b 1 ehvelklu feia, 6-7 etantu mutu affer- ture / si. In the above passages the use of the subjunctive in place of the imv. may have been partly induced by eitipes 2, 14, which introduces the decrees; in other words, the subjunctive may be quasi-subordinate. Not so, however, in Vllb 1-2 ...fratrexs ... portaia seuacne fratrom / Atiersio desenduf. 4 fratreci motar sins a. CCC. lb 35-6 ene tra Sahta kupifiaia / erus tefa, but imv. kupifiatu earlier in 35, combifiatii Vila 43, 44, the latter instance correspon- ding to the subjunctive in lb 35.
b. A prohibitive use of the subjunctive occurs in IV 32-3 efek / ures punes neifhabas '... ne adhibeant.' In Via 6, b 51-2, however, prohibitions are ex- pressed by ne{i)p + imv.
c. "With verbs of necessity, obligation, and certain other ideas a dependent subjunctive may appear with no subordinating conjunction. With herter: III 1-2 esunu fuia herter sume / ustite... Cf. Va 6 si herte, 8, 10 emantur herte, b 8, 11, 13, 16 dirsas {-ans, -a) herti; Vllb 2 ponne iuengar tursiandii hertei; but in Vb 3-6 panta muta fratru / Atiiefiu raestru karu... / ... afferture eru pepurkure/nt herifi the infinitive eru depends on herifi, which itself de- pends as a pf. subj. on pepurkurent. With ti^it: Ila 16-17 affertur... / ... fagia tigit. With stiplo: Via 2 stiplo aseriaia parfa dersua... With combifiatu kupifiatu: Vila 43-4 combifiatu Rubiname erus / dersa (= lb 35), Vila 44 (= lb 35-6) practically the same. With deitu: VIb 64 etaians deitu, 65 = Vila 1 deitu efaias. In Via 22-3 the presence of subocau seems to be the occasion for the use of the subjunctive: teio subocau suboco / Dei Graboui... fos sei pacer sei ocre Fisei; cf. b 6-7, 26. In all other instances of the formula subocau is absent from the context, and the imv. futu, fututo is used. Cf. also a above for the verb eitipes as an influencing factor in the use of the subjunctives in V.
d. Indirect questions have partly the indicative, partly the subjunctive: Va 7-8 revestu pufe tefte / eru emantur herte 'whatever are offered, he shall examine whether any of them should be accepted.' herte is ind.; the form cannot
H. Syntax 153
be explained as subj., and the use of the ind. is no more surprising than in early Latin, where it is frequent. Cf. 10 emantu herte. The subjunctive is used in 23-4 ehvelklu feia fratreks ute kvestur / sve rehte kuratu si; b 1-3 ehvelklu feia... / ... panta muta / afferture si. In the last example the subj. can be explained by the deliberative nature of the clause, but not so in the prior exam- ple, which is an indirect question of mere fact.
e. The use of the imperative in the ritual instructions of the Tables is too frequent to require illustration. Certain special cases, however, deserve mention: Ila 42-3 esunu / purtitu futu; IV 31-2 purtitu / futu. Contrast the indicative in VIb 42 purdito fust; similarly Vila 45, lb 39. In each case the formula occurs at the end of the description of a particular ceremony.
f. The prohibitive use of the imv. occurs in the following passages: Via 6 neip mugatu nep arsir andersistu; b 51-2 neip / amboltu. See also b above.
153. The Non-finite Forms of the Verb. The verbal nouns include the present active and perfect passive infinitive and the supine in -turn, while the verbal adjectives include the present active and perfect passive participles and the gerundive.
a. Examples of the complementary infinitive are found only with forms of the verb heries: — Vib 48 pone poplo afero heries (= lb 10). — Ila 16-17 heriiei fagiu, affertur... / ... fa§ia tigit. — ^ lib 21-2 vitlu vufru pune heries / fagu.
b. The infinitive accompanied by an accusative subject and depending on an impersonal verb is exemplified by: — Vb 3-6 panta muta ... / ... / ... affer- ture eru pepurkure/nt herifi. The inf. eru depends on the subjunctive herifi, which in turn depends on pepurkurent. — Vllb 1-2 seuacne ... / ... pifi reper fratreca pars est erom ehiato. For the form ehiato see the note on the passage.
c. The infinitive in oratio obliqua is found in the following passages: Va 24-6 sve ... / ... / prusikurent rehte kuratu eru. Similarly 27-9, but with negative neip.
d. Clear examples of the supine occur only in the verb anseriato, and all de- pend on forms of etu: lb 10 avef anzeriatu etu = VIb 48 auif aseriato etu. Cf. Via 1, 6, (twice).
e. All examples of the present participle show the "predicative " use: — VIb 22 serse fetu = la 25 zefef fetu, and so with the other examples of serse zefef ; see Lexicon.— VIb 47 reste esono feitu = lb 9 restef esunu feitu. — la 6 kutef pesnimu and similarly with all other instances of kutef.
f. The attributive use of the pf. pass. pcpl. may be seen in the following exam- ples: — VIb 18 capif purdita dupla aitu, sacra dupla aitu (= la 18), where pcpl. purdita and adj. sacra designate two different classes of vessels. Similarly veskia snata asnata Ila 19, etc. — Ila 42 antakres kumates persnihmu, where
154 II. Grammar
kumates are the ground in contrast to the unground grains (or cakes ?) ; simil- arly lb 36, 37. — VIb 59-60, etc., nerf sihitu ansihitu iouie hostatu / anhostatu.
g. Examples of the predicative use of the pf. pass. pcpl. are: — Via 55, etc., tases persnimu. — VIb 5 uestisia et mefa spefa scalsie conegos fetu, and similarly in all the other instances of conegos kunikaz.
h. Examples of the ablative absolute are: Via 1 aueis aseriater (= la 1, Ila 16-17). Possibly also VIb 17 comoltu, comatir persnihimu, unless comatir is rather to be taken in an instrumental sense 'with the ground grains' as in the passages with antakres kumates cited in f above.
i. The examples of the gerundive in general involve the notion of purpose: — Via 18-19 esisco esoneir seiieir I popler anferener et ocrer pihaner, where the gerundive phrase depends, at least to some extent, on esoneir. — Via 19-20 ocrer / pehaner paca, where paca is generally taken as equivalent in sense to L. causa. — Via 8 uerfale piife arsfertur trebeit ocrer peihaner, where the gerun- dive phrase has no direct dependence on any word; cf. Tac, Ann., II, 59 Aegyp- tum proficiscitiir cognoscendae antiquitatis. — VIb 22 (= la 26) pelsana fetu 'offer for (ultimate) burial'; cf. Ila 6, III 32. — Ila 43 katel asaku pelsans futu is more nearly like the obligational use in L. Carthago delenda est.
154. Subordinate Clauses. The use of dependent subjunctives without a subordinating conjunction has already been treated in 152 c above, and indirect questions in 152 d. Relative clauses and clauses introduced by the local con- junctions pue and pufe present no problem in regard to the use of moods and tenses. The present section is therefore limited to substantive clauses with puse puze, temporal clauses, and conditional clauses.
a. The following passages show puse puze + subj. corresponding to L. ut -\- subj.: — Vila 43 carsitu... puse erus dersa =■ lb 33-4 kafetu ... /... puze erus tefa. — Via 20 eo iso ostendu pusi pir pureto cehefi dia; dia is subj., but see note on the passage.
h. The temporal conjunctions are arnipo, nersa, prepa 'until,' ape, pane, pone pune 'when.' VIb 41 sersitu arnipo comatir pesnis fust (fut. pf.); in 25 the verbal form uesticos is obscure; see note. Via 6 neip mugatu nep arsir ander- sistu nersa courtust. b 51-2 neip / amboltu prepa desua combifiansi (pf. subj.). VIb 5 ape sopo postro peperscust, and fut. pf. with all other instances except lib 28 ape purtuvies, where the verb is fut., and possibly III 20-1 ap / vuku kuke- hes, where the verb is obscure; see note. Vila 46 (= lb 40) postertio pane poplo andirsafust. In contrast to ape and pane, which introduce action prior to the main action, pone may introduce contemporaneous as well as prior action, and is used with the pres., fut., and fut. pf. ind., and once with the pres. subj.: VIb 43 (= lb 1) ponne oui furfant, and pres. also in Vllb 2, and in lib 22, where seste is better taken as pres. than as fut. Va 8-9 pihaklu pune / tribfigu fuiest,
H. Syntax 155
and similarly with fut. in lb 10 = VIb 48, lb 15, 19, lib 21, 27. Fut. pf. in lb 11 pune kuvurtus and in 20, 33, Ila, 1, 7, b 16. Pres. subj. in VIb 49-50 pone esonome ferar.
c. The conditional conjunction is sue sve, including its combinations with enclitic pronominal forms. Examples are found with the fut. pf. ind., with the pres. subj., and in two probable instances with the pres. ind. Va 24-6 = 27-8 sve mestru karu... prusikurent rehte kuratu eru (kuratu rehte neip eru), and fut. pf. also in Via 7, 16, Vllb 3. Pres. subj. in VIb 47 (= lb 8) saepo esome esono anderuaco se. In Va 24 the verb with the first sve 'whether' is a subjunc- tive of indirect question; see 152 d. The present indicative, if the forms are correctly explained, is found in lb 18 svepis habe 'if anyone is caught,' IV 26 svepis heri. The apodoses of the above conditional sentences, generally being instructions rather than statements of fact, show imperative (lb 18, IV 27, Via 17) or pres. subj. forms (Va 27 si, b 1 feia, Vllb 4 sins). The fut. pf. ind., however, is found in Via 7, where disleralinsiist is fut. pf. by any reading or interpretation. In lb 8 = VIb 47 the forms ise, {e)fust are uncertain; see on VIb 47.
155. Word-Order. As in early Latin prose the normal arrangement of words is seldom altered for the sake of artistic effect, and yet it is not so rigid that we can set up invariable rules. Even in two different versions of the same formula the words may be differently arranged. A full account of the word-order cannot be attempted here; it must suffice to show the normal arrangement in sentences and phrases and to point out a few of the more striking exceptions.
a. The verb commonly stands at the end of its clause: Via 3 arfertur eso an- stiplatu; 22 pre uereir Treblaneir luue Grabouei huf treif fetii; 34 Di Grabouie, tio siibocau; etc. But there are many exceptions; e.g., lb 40-1 iveka perakre tusetu / super kumne arfertur, with subject last, perhaps in emphatic contrast to prinuvatu tuf tusetutu which follows; Ila 9 ape purtiius sufu; lb 15-6 pune menes / Akef uniamem ; etc.
b. The complementary infinitive precedes the verb on which it depends in VIb 48 afero heries (= lb 10) but follows in Ila 16 heriiei fagiu and similarly in b 21-2. In all the instances of anseriato with forms of the verb etu (lb 10, Via 1-2, 6, 6-7, b 48) the supine precedes.
c. As in Latin, the conjunction which introduces a subordinate clause some- times takes second place within the clause: Va 8-9 pihaklu pune / tribf igu fuiest; Via 5 sersi pirsi sesust, poi angla...
d. A dependent genitive sometimes precedes its noun, sometimes follows it: Via 29-30, etc., ocrer / Fisier totar louinar nome nerf arsmo ueiro pequo castruo fri pihatii; Via 23, etc., erer nomneper erar nomneper. With the genitive in sec- ond place: VIb 51 perca ponisiater habituto — lb 15 perkaf habetutu punigate; VIb 57-8 Prestota Serfia Serfer j Martier.
156 II. Grammar
e. In placing the genitive of the fathers' praenomen between the son's praeno' men and nomen Umbrian differs from Latin and Oscan, which place the father's praenomen after the son's nomen: lb 45 Vuvgis Titis Teteies; similarly Ila 44; Va 3 T. T. Kastrugiie, on which see note.
f. A qualifying adjective normally follows its noun: Via 58 si gomia = la 7 sif kumiaf ; VIb 3 sif filiu = la 14 sif feliuf ; VIb 19 buf trif calersu = la 20 tref hvd kalefuf. In the frequent expressions f rater Atiersiur, ocre Fisi, tote louine, uerir Treblanir {Tesenocir, Uehier) the adj. invariably follows. The position of the adj. with persondru, however, varies: VIb 28 sorsu persontru and similarly in 31, 35, la 27, 30, Ila 8; pesondro sorsom in VIb 24 and similarly in 37, 39, 40.
g. The arrangement of adj. + prepos. + noun, as in L. magna cum cura, is fairly common: — VIb 24 destruco persi, and so in 38, 49, la 29, lib 27, 28; VIb 37 neriruco persi, and so in 39, la 32, but VIb 24-5 mani / nertru. — Ila 21, 35 Petruniaper natine, b 26 Vugiiaper natine — Via 10 todcome tuder, but 12 tuderor totcor.
h. Numeral adjectives precede the noun in the earlier tables but follow the noun in the later tables: la 3 tre buf fetu, and so in 7, 11, 14, 20, 24, b I, 4, 24, 27, 31, but Via 22 buf treif fetu, and so in 58, b 1, 3, 19, 22, 43, 45, Vila, 3, 6, 41. Similarly the numeral '2' precedes in Ila 27, III 32, 34, but follows in Vb 10, 15, VIb 50, Vila 46. Ill 17-19 seples / ahesnes tris ... etres tris / ahesnes... tuves ahesnes is a special case in which we must take account of the failure to repeat seples and of the contrast between 'three' and 'two'.
III. TEXT, TRANSLATION, AND COMMENTARY
la 1 - b 9 (= Via 1 - b 47). Purifica- tion of the Fisian Mount. For a de- tailed analysis see the outline of the ta- bles, above, Introduction, § 13. As in Table VI, sacrifices are to be performed before and behind the three gates, at the grove of Jupiter, and at the grove of Coredius. The deities honored and the victims offered are regularly the same in the corresponding sections of I and VI, but there are some differences in details and in the order in which the instructions are presented.
la 1 - 2 (= Via 1 - 21). Taking of the auspices. In contrast to the expanded version in Via, with the names of the birds to be observed, the formulas to be repeated, and the landmarks to be used in defining the augural templum, we have here a single short sentence. — 2 pernaies pusnaes: literally ' anlicis posticis,' 'before and behind.' For an attempt to equate them with merstu, dersua and make both pairs refer to east and west, see on Via l,and also the sum- mary at the end of Via 1 - 18.
la 2 - 6 ( = Via 22 - 57). Sacrifice of three oxen to Jupiter Grabovius before the Trebulan Gate. The present version differs from that in VI in its failure to include the text of the prayers, which is merely hinted at in the expression
ukriper Fisiu tutaper Ikuvina; in the use of kutef instead of iases; in the in- clusion of the phrase afepes arves and the absence of any distinct reference to calies; and in the order in which the various instructions are arranged. — 3 arvia ustentu: this instruction appears regularly in the sacrifices performed for the purification of the Mount and the lustration of the people, in the later as well as in the earlier tables. In the former, however, the verb feta takes the place of ustentu, and the expression afepes arves is omitted. Devoto, p. 200, sug- gests that the fetu of VI and VII covers the whole ceremony with the aruia, which in I is described in two steps, arvia ustentu and pesnimu afepes arves. He is probably correct in understanding aruia of vegetable offerings, with Bii., PL, Bk., Bott., against K. O. Muller, Hu., Blum., Vet., who translate ' exta' (see on Ila 18), afepes is more difficult. Most connect it with L. adipeis, which provides a sound basis for the ending (direct connection with adipibiis would lead us to expect *afepus; cf. 101 d). It is then the first member of an asyn- detic pair. Dev., on the other hand, translates ' tostis granis, ' deriving from *ad-eph-, with the root found in erpco, Armenian ephem 'cook'. The idea of
158
III. TEXT AND TRANSLATION
(la 1) este : persklum : aves : anzeriates : enetu : / (2) pernaies : pusnaes : preveres : treplanes : / (3) iuve : krapuvi : tre hvd : fetu : arvia ustentu : / (4) vatuva ferine : feitu : heris : vinu : heri puni : / (5) ukriper : fisiu : tutaper : ikuvina : feitu : sevum : / (6) kutef : pesnimu afepes arves : /
(7) pusveres : treplanes : tref sif : kumiai : feitu : / (8) trebe : iuvie : ukriper : fisiu : tutaper : ikuvina : / (9) sup a sumtu : arvia ustentu : puni fetu : / (10) kutef pesnimu : are arves /
(11) preveres : tesenakes : tre buf : fetu : marte : krapuvi / (12) fetu : ukripe : fisiu : tutaper : ikuvina : arviu : ustentu / (13) vatuva ferine : fetu : puni fetu : kutef : pesnimu : arpes : arves : /
(14) pusveres : tesenakes : tref sif : feliuf : fetu : / (15) fise sagi ukriper : fisiu : tutaper : ikuvina / (16) puni : fetu : supa : sumtu arviu : ustentu : mefa : / (17) vestiga : ustetu : fiiuvi : fetu : xikriper :
Commence this ceremony by observ- ing the birds, those from in front, and those from behind. Before the Trebu- lan Gate sacrifice three oxen to Jupiter Grabovius. Present grain-offerings, place the ribs on a tray, sacrifice either with wine or with mead, for the Fisian Mount, for the state of Iguvium. Pray each (portion) in a murmur with (of- ferings of) fat and grain.
Behind the Trebulan Gate sacrifice three pregnant sows to Trebus Jovius for the Fisian Mount, for the state of Iguvium. Take up the under-parts, pre- sent grain-offerings, sacrifice with mead, and pray in a murmur with (offerings of) fat and grain.
Before the Tesenacan Gate sacrifice three oxen; sacrifice to Mars Grabovius for the Fisian Mount, for the state of Iguvium. Present grain-offerings, place the ribs on a tray, sacrifice with mead, and pray in a murmur with (offerings of) fat and grain.
Behind the Tesenacan Gate sacrifice three sucking pigs to Fisus Sancius for the Fisian Mount, for the state of Igu- vium. Sacrifice with mead, take up the under-parts, present grain-offerings, pre- sent a mefa cake and a libation. Sacri- fice to Fisovius, sacrifice for the Fisian
10 are: for arepes. An erabure occupying the space of one letter stands after are and also after arv. 17 fiiuvi: for fisuvi.
the roasting of the grain as a part of the ritual is most attractive, but the etymo- logical support is weak in comparison with the easy equation of afepes with L. adipeis. I have therefore decided, though hesitantly, to maintain the usual interpretation. It is curious that at the end of the series of instructions for each
sacrifice in I we regularly find afepes arves but no mention of cakes, while in VI and VII, usually near the end of the series, a ficla and one other type of cake are prescribed. Only for the sacrifice at the grove of Coredius does I call for a special type of cake (tenzitim lb 6 = tesedi Vlb 46). Possibly there was a change in
la 3
COMMENTARY
159
the ritual whereby cakes in the later stage took the place of simple offerings of grain and fat in the earlier. — 4 va- tuva ferine feitu: see on Via 57. — heris vinu heri puni: on Via 57. — sevum: on Via 18 (near end). — 6 ku- tef: found only in la and b, where it corresponds to tases in Via and b; but tagez is used in la 26 (sacrifice behind the Veian Gate) and regularly in the instructions for the sacrifices to be per- formed during the lustration of the people (lb 26, 30, 32, 44). kutef is almost certainly a present participle with -ef < -ens, but there is no general agreement in regard to its etymology. It is alto- gether probable, however, that its sense is similar and not opposite to that of tases; in other words, that between the earlier and later tables there was a change in the phraseology but not in the mean- ing of the instructions.
la 7 - 10 (= Via 58 - 59). Sacrifice of three pregnant sows to Trebus Jovius be- hind the Trebulan Gate. For the char- acter and sex of this deity and the de- clensional class to which the name be- longs see on Via 58. The instructions here are given in substantially the same or- der as in VI, but differ, as in other similar instances, in the use of kutef for tases, arvia ustentu for aruio fetu, and are- pes arves where the later version calls for two type of cakes (see on 3 above). Moreover la omits persae fetu, but has supa sumtu. References to the sopa, however, are regularly associated with victims which are slain persaia (cf. la 16 = VIb 3, Ila 22, VIb 3 with 5, Vila 7 with 8; note on VIb 5). supa sumtu must then refer to the removal of the entrails from the victim which has been slain lying on its back. — 7 sif kumiaf : see on Via 58. — 8 Trebe luvie: on Via 58, — 9 supa sumtu : on sopo VIb 5. la 11 -13 (= VIb 1 - 2). Sacrifice of three oxen to Mars Grabovius before the Tesenacan Gate. The differences of de- tail between this and the later version shown in VI consist in the contrast be-
tween arviu ustentu and aruio fetu, kutef and tases, arepes arves and pro- sesetir farsio ficla arsueitu, much as in the case of the second sacrifice.
la 14 - 19 (= VIb 3 - 18). Sacrifice of three sucking pigs to Fisus Sancius be- hind the Tesenacan Gate. The instruc- tions for the supplementary offerings to Fisovius which follow the sacrifice proper are not introduced by a temporal clause commencing with api as in VIb 5, or as in both the earlier and later versions of the sacrifice behind the Veian Gate (la 27 ^ VIb 23). Moreover the words kutef pesnirau arepes arves appear at the very end of the present passage, in contrast to the more orderly arrange- ment in 24-34. Yet the provision for the supplementary offerings in la as in VIb is made clear by the use of the name Fisuvi (cf. VIb 5, etc.), the repetition of ukriper Fisiu, the requirement of the raefa (nowhere else mentioned in Table I; cf. VIb 5), and the instruction with regard to the cups (cf. VIb 18). — ■ 14 sif feliuf : see on VIb 3. — Fise Sagi: on VIb 3. — 16 mefa: on Via 56. — 17 vestiga: on VIb 5. — ustetu: this verb is used in the passive with uasor as subject in Via 20 and in the active, with reference to the electing of the uh- tur in III 5; in all other passages apart from its use here with mefa vestiga, it is restricted to the expression arvia us- tentu. In regard to the variation be- tween ustentu and fetu see on 3 above; Devoto's theory may be applicable to the present passage as well. — 18-19 kapif purtitaf sakref, etraf purtitaf, etraf / sakref: the general sense must be the same as in VIb 18 (q.v,), but there are certain differences in the language, of which the most striking is the lack of a verb on which the accusative might depend. Since ellipses of this sort are not uncommon in the Tables (see 136 c), we may assume that the aitu of VIb 18 is not absolutely indispensable in the abbreviated instructions given here. The important point is how many vessels of
160
III. TEXT AND TRANSLATION
fisiu : fetu : / (18) kapir : purtitaf : sakref : etraf : purtitaf : etraf : / (19) sakref : tutaper : ikuvina : kutef : pesnimu : afepes : arves : /
(20) preveres : vehiies : tref : buf : kaleruf : fetu : vufiune / (21) krapuvi : ukriper : fisiu : tutaper : ikuvina : / (22) vatuva : ferine : fetu : heri : vinu : heri : puni : / (23) arviu : ustentu : kutef : pesnimu : arepes : arves : /
(24) pusveres : vehiies : tref : hapinaf : fetu : tefre : iuvie : / (25) ukriper : fisiu : tutaper : ikuvina : puste : asiane : fetu : zef ef : fetu : / (26) pelsana : fetu : arvia : ustentu : puni fetu : tagez : pesnim / u : afiper : arvis : api : habina : purtiius : sufum : pesuntru / (28) fetu : esmik : vestigam : preve : fiktu : tefri : iuvi : fetu : ukri /per fisiu : tutaper : ikuvina : testruku
each kind are to be used, and etraf clearly takes the place of the dupla of VIb IS. The only other plural forms of etru in the Tables are dat. etre accom- panying Kiaverniie, Satane, luieska- nes in lib 3, 4, 6 (second Clavernii, ' etc.). The use of etraf, or of plural forms of L,. alter, with the value of cardinal numerals would be most unnatural, yet there would be no point in saying ' some loaned, otliers sacred (property)' as a mere amplification of purtitaf sakref immediately preceding, etraf must tbere- fore have approximately the same sense as the distributive dupla in VIb 18. A remote parellel to this use of an ordinal with enumerative value may perhaps be seen in such Greek expressions as nifimog avToQ 'with four others.'
Mount. (Move) the loaned and the con- secrated bowls, two loaned, two conse- crated, (in the ceremony) for the state of Iguvium. Pray in a murmur with (offerings of) fat and grain.
Before the Veian Gate sacrifice three oxen with white foreheads to Vofionus Grabovius for the Fisian Mount, for the state of Iguvium. Place the ribs on a tray, (sacrifice) either with wine or with mead, present grain-offerings, and pray in a murmur with (offerings of) fat and grain.
Behind the Veian Gate sacrifice three lambs to Tefer Jovius for the Fisian Mount, for the state of Iguvium. Sa- crifice at the rear of the altar, sacrifice seated, sacrifice (the victims) for burial, present grain-offerings, sacrifice with mead, and pray silently with (offer- ings of) fat and grain. When you have made the presentation of the lambs, offer a pig-persondro, and to it add a libation separately. Sacrifice to Tefer Jovius for the Fisian Mount, for the state of Igu-
la 20 - 23 (- VIb 19 - 21). Sacrifice of three oxen to Vofionus Grabovius be- fore the Veian Gate. The earlier and later versions agree very closely, the only differences being tliose already observed in the instructions for the third sacrifice (la 11 - 13 = VIb 1 - 2). The arrange- ment of the instructions is consistent between the two versions, but is not quite the same as in the third sacrifice. — 20 kaleruf: see on VIb 19. — Vufiune Krapuvi: on VIb 19.
la 24 - 34 (= VIb 22 - 42). Sacrifice of three ewe-lambs to Tefer Jovius be- hind the Veian Gate, followed by sup- plementary offerings to Tefer Jovius. The sacrifice proper is contained in 24-27. Tlie differences between this and the later version in VIb include the usual varia-
la 18
COMMENTARY
161
tion of arvia ustentu with aruio feitu and of afiper arvis with the two types of cakes in VIb 23. Moreover tlie ob- scure puste asiane fetu is peculiar to la 25. On the other hand in place of kutef pesnimu, which is regularly used in la, b in describing the sacrifices connected with the purification of the Mount, la 26 has tagez pssnimu, as in the instruc- tions for the lustration of the people and throughout Via, b, Vila. — 24 ha- pinaf: see on habina VIb 22. — Tefre luvie: on VIb 22. — 25 puste asiane fetu: very uncertain; the most satisfac- tory interpretation is probably that of Devoto, who translates 'in parte postica arae, ' taking puste as loc. of a stem posto- derived from preposition post, and asiane as in some way connected with asa, Lat. ara, though stem and suffix are not quite clear. For other interpre- tations see L. H. Gray, B.B., XXVII (1902), pp. 306-7; Blum., p. 59; Pisani, p. 159; Vetter, p. 177. — zeref: for the sitting position see on VIb 17. — pel- sana: on VIb 22.
la 27 - 34 api habina purtiius, etc. : as in VIb 23 ff. the porreclio is followed by special offerings to Tefer Jovius. The earlier and later versions agree in most of their principal features: the priest of- fers near his right foot a sorsom persondro and a libation to Tefer Jovius, and near his left foot a persondro and libation to Stabilis Jovius (?), and both versions conclude with the grinding of the cereal offerings and accompanying prayers. However, there are numerous discrepan- cies of detail; in addition to the failure of la to include the text of the prayers, the identification of the priest by the clause eront poi habina purdinsiist (VIb 24) is also omitted, and similarly the instruction earn (sc. capirso) mani nertru tenitu, etc. (VIb 24-5). Moreover la only once mentions an erus, that of the ewe-lambs (33), while VIb calls for an erus in 25 just after the offering of the persondro and libation at the right foot and for the erus of the prosecta in 38,
and the erus of each of the libations in 38 and 39. The instructions pesondro... endendu pelsatu (39-40) and uaso... subra spahalu... anderuomu sersitu (40-41) are also without any precise equivalent in la. On the other hand VIb has nothing cor- responding to the expressions esmik ves- tigam preve fiktu la 28, esmik ves- tiga aiiktu 31, used with reference to the manner of offering the libation. — su- fum pesuntru: see on VIb 24, — 28 esmik vestigam preve fiktu : in contrast with 31 esmik vestiga afiktu, where the verb is compounded and the adverb preve is wanting. The sense is obscure chiefly because the composition of both the vestiga and the pesuntru is unknown. For preve Dev., p. 246, suggests ' gut- tatim. ' It seems not impossible, however, that the present passage may be explained by comparison with VIb 24-5 earn (sc. capirso) mani / nertru tenitu arnipo uesti- sia uesticos; that is, if he holds the li- bation-bowl in his left hand until he has poured the libation, he does not pour it upon the persondro but in a separate spot, even though near his right foot {destruco persi). The absence of a prefix to the verb in preve fiktu should be noted, in contrast to afiktu in 31 just below, where preve does not appear. The dative esmik is a difficulty, but here it may merely signify that the vestiga be- longs to the pesuntru, even though it is not placed upon it as in 31, where the compound afiktu governs esmik quite naturally. So far as the translation of fiktu and afiktu is concerned, nothing can be affirmed with certainty, since we do not know the exact nature of the vestiga which depends on it. On the whole, however, connection is easier with L. figo than with fingo, and if the ves- tiga is solid (or possibly melted fat which solidifies when it falls?) there is no se- rious difficulty ; if it is a liquid, we must assume a semantic evolution 'attach > add > pour (over).' — 29 kapife: see on VIb 18 capif. — pefum: on VIb 24 perso. — 30 Staf /li luve : the bronze here
162
III. TEXT AND TRANSLATION
: peri : kapife : pefum : feit / u api : ef el purtiius : enuk : suf um : pesuntrum : feitu : staf / li iuvesmik : vestiga : aiiktu : ukriper : fisiu tutaper : ikuvina / a : feitu nertruku : peri : kapire : pefum feitu : puni : feitu : / (33) api sufuf purtitius : enuk : hapinaru erus : titu : zef ef / (34) kumultu zefef : kumats : pesnimu : /
(lb 1) vukukum : iuviu : pune : uvef : furfaO : tref : vitluf : turuf : / (2) marte : hufie : fetu : pupluper : tutas ; iiuvinas tutaper ikuvina : / (3) vatuva : ferine : fetu : puni : fetu : arvia : ustentu : kutep : pesnimu : / (4) afepes : arves : vukukumi kureties : tref : vitlup : turup hunte ge / fi : feitu : pupluper tutas : iiuvinas : tutaper : iiuvina vatuva : / (6) ferine fetu arvia ustentu : tenzitim : arveitu : heris vinu : heris : / (7) puni : feitu kutef : persnimu : afipes : arvis inuk ukar : pihaz : fust : / (8)
vium. At your right foot make a mound for the bowl. When you have presented this (persondro), then offer a pig-person- dro, to Stabihs' Jovius, upon it add a hba- tion, for the Fisian Mount, for the state of Iguvium. At your left foot make a mound for the bowl. Sacrifice with mead. When you have presented the pig-persondra, then distribute the eras of the lambs. Grind (the grain) sitting and pray sitting, with the ground (grain).
At the Grove of Jupiter, while they are placing on a platter (the remains of) the sheep, sacrifice three bull-calves to Mars Hodius for the people of the state of Iguvium, for the state of Igu- vium. Place the ribs on a tray, sacrifice with mead, present grain-offerings, and pray in a murmur with (offerings of) fat and grain. At the Grove of Coredius sacrifice three bull-calves to Hondus Ser- fius for the people of the state of Igu- vium, for the state of Iguvium. Place the ribs on a tray, present grain-offer- ings, add a twisted cake, sacrifice either with wine or with mead, and pray in a murmur with (offerings of) fat and grain. Then the mound will have been
30 erel: for erek. 31 ikuvina: final a not completed on 31 and so repeated on 32.
33 purtitius: for purtiius. 34 kumats: for kximates. lb 4-5 ge/ fi: for ge/rfi,
with space for the r unused.
reads stafii ;iuvesinik with the word- divider between the first and second i very faint. There have been many at- tempts to read and Interpret the pas- sage correctly, with editors in general taking stafli(i) as in some way connected with L. stabulum and in some cases as- suming wrong word-division, so as to read the second word as uve 'sheep' ('of the stable'). For a detailed discus- sion of the problem see R. G. Kent, C.P.,
XV (1920), pp. 354-6, who argued in favor of the word-division actually in- scribed on the bronze. For the sense of the passage, however, there are good rea- sons for following G. B. Pighi, Latinitas, II (1954), pp. 22-3, R.F.I.C, XXXII (1954), p. 234, who sees in stafii a divine name which may be rendered etymolo- gically '(il dio) per opera del quale si stabilisce (il rito)'. The parallelism be- tween 27-29 and 30-32 is remarkably
la 30
COMMENTARY
163
close, and if we admit Stafli luv, as name of a god to balance Tefri luvi in 28, the symmetry is nearly perfect. We may assume that Tefri luvi should have followed fetu at the beginning of 28 but was omitted, then added after fiktu with a new fetu, while the correct arrangement was followed in 30-31. The reading iuvesmik may be explained as an instance of haplography resulting from the resemblance of v and e, which differ in the native alphabet only in the presence of the cross-bar in the middle of the e: the scribe, having engraved the v, saw on his copy the e of iuvie, confused it with the V which he had just written, and continued with esmik (Kent, ibid., p. 354). See also on VIb 37, where the possi- bility of contrasting Staflarem with Tefrali 28, 35, is taken as support for the inter- pretation of stafli adopted in the present passage. — afiktu: see on 28. — 33 erus: on VIb 16. —34 kumultu: on VIb 17.
lb 1 - 4 (= VIb 43 - 44). Sacrifice of three bull-calves to Mars Hodius at the grove of Jupiter. The agreement be- tween the earlier and later versions is very close, the differences of detail and the arrangement of the instructions being the same as in the case of the fifth sacri- fice (la 20-23 = VIb 19-21). The clause ponne oui furfant = pune uvef furfaS, however, is found only here. — 1 vu- kukum: for the translation 'grove' in preference to 'temple' see on VIb 43. — furfae: on VIb 43. The character o used for the final letter occurs also in purtuviSu IV 20. — 2 Huf ie : on Horse VIb 43. — 3 vatuva ferine fetu: on Via 57. — puni: on poni Via 57. — arvia ustentu: on la 3. — kutep: on la 6; for the final p in place of the normal f cf. vitlup turup in 4 below. ■ — 4 afepes arves : on la 3.
lb 4 - 7 (= VIb 45 - 46). Sacrifice of three bull-calves to Hondus Serfius at the grove of Coredius. The instructions in general agree with those in VIb 45-6, but the ficla prescribed in the latter pas- sage is not mentioned with the tenzitim
in lb 6, and the cakes are called for at a. different place in the series of instruc- tions in the two tables. — 4 Kureties : see on VIb 45. — Hunte Qerfi: on VIb 45. — 6 tenzitim: see on tesedi in the equivalent VIb 46, the only other passage in which this type of cake is prescribed.
lb 7 - 9 (= VIb 46 - 47). Instruction for repeating the whole purification in case there has been any omission. The similarity between the two versions is extremely close, the principal difference being between anter: vakazevagetum- iseavif and ander.uacose.uasetome.fust. auif, as the words appear in the actual texts. For discussion of the word-divi- sion and interpretation see on VIb 46-7.
lb 10 - 45 ( = VIb 48 - Vila 54). Lus- tration of the people. The present ver- sion contains the instructions for the taking of the auspices (10-11), the fire- ceremony (11-12), the second taking of the auspices (13-14), the procession to Acedonia with the prinuati (15-16), the banishment of aliens (16-18), the three circuits around the assembled people (19-23), the sacrifices at Fontuli, at Ru- binia, and at Trans Sanctam, which seem to have been partly in progress at the same time (24-39), the ceremonial heifer- chase (40-44), and finally the authen- tication by the quaestor (45). For the principal discrepancies between the whole passage and the later version, and for the problem of the sequence of ev- ents in both of them, see the introduc- tory portion of the notes on VIb 48 - Vila 54. — 10 aferum: see on Via 19. — avef anzeriatu etu pernaia/f pust- naiaf : on VIb 48-52. — 11 krenkatrum : on VIb 49. — 12 abtimem: probably a /i-stem (+ postposition -em) from the root ag- (cf. L. ago, actio), but the sense is very uncertain. Some, in order to recon- cile it with abtisper III 24, 29, take it to designate the ceremony for which the fire is prepared: 'ad agonium' (Bii.), 'ad caerimoniiim ' (Bk.), ' ad actionem ' (Bott.) ; others to designate the object in which the fire is to be carried: 'in foculam'
164
Hi. TEXT AND TRANSLATION
svepu : esumek : esunu : anter : vakazeva^etumiseavif : azeriatu : / (9) verufe : treplanu : kuvertu : restef : esunu : feitu : /
(10) pune : puplum : aferum heries : avef : anzvriatu : etu pernaia / i : pustnaiaf : pune kuvurtus : krenkatrum : hatu enuraek : / (12) pir : ahtimem : ententu : pune : pir : entelus : ahtimem : / (13) enumek : steplatu parfam : tesvam : tefe : tute : ikuvine : / (14) vapefem : avieklufe : kumpifiatu : vea : aviekla : esunume : etu : / (15) prinuvatu : etutu perkaf : habetutu : punigate pune : menes : / (16) akeruniamem enumek : eturstamu : tuta tari- nate : trifu : / (17) tarinate turskum : nalaarkum : numem iapuzkum : numem : / (18) svepis habe : purtatulu : pue : mefs est : feitu : uru : per e : mers : est / (19) pune : prinuvatus : staheren : termnesku : enumek : armanu : / (20) kateramu : ikuvinu : enumek : apretu : tures : et : pure : puni : amprefu / us : persnimu : enumek : etatu : ikuvinus : tri- iuper : amprehtu : / (22) : triiuper : pesnimu : triiuper : etatu : ikuvinus : enumek : / (23) prinuvatus : gimu : etutu : erahunt : vea : gimu : etutu : prinuvatus : /
(24) funtlere : trif : apruf : rufru : ute : peiu : feitu gerfe : marti : / (25) vatuvu : ferime : fetu : arviu :
purified. If there is any interruption of these rites, they shall be invahd: take an observation of the birds, return to the Trebulan Gate, and perform the rite anew.
When you wish to perform a lustra- tion of the people, go to observe the birds, those from in front and those from behind. When you have returned, take a stole, then place fire in the car- rier. When you have placed the fire in the carrier, then demand a parra in the west for yourself and for the state of Iguvium. (The augur) shall make his announcement in the direction of the augural seats. Go to the sacrifice by the Augural Way. The prinuati shall go (along); they shall hold the wands of a puniceatus. When you come to Ace- donia, then pronounce banishment on the Tadinate state, the Tadinate tribe, the Tuscan, the Narcan name, the lapudic name: "If anyone is caught, bring him to that place to which it is the law, do with him that which is the law." When the prinuati shall be standing at the boundary, then (say): "Arrange your- selves in priestly ranks and military ranks, men of Iguvium." Then make the cir- cuit around (them) with the victims and the fire. When you have made the cir- cuit, pray. Then (say): "Go, men of Iguvium." Make the circuit three times, pray three times, (say): "Go, men of Iguvium" three times. Then the pri- nuati shall return; by the same road shall the prinuati return.
At Fontuli sacrifice three boars, red or black, to Serfus Martins. Place the ribs on a tray, present grain-offerings,
10 anzvriatu: for anzeriatu, the cross-bar being omitted. 20-1 amprefu /us: for amprefus. 25 ferime : for ferine.
19 armanu: for afmamu. f eiu : for fetu or feitu.
lb 12
COMMENTARY
165
(Br., Blum.), 'in ferculum' (PI.), 'vas- culum igniarium ' (Dev., p. 267). From a comparison with the corresponding pas- sage VIb 49-50 it appears that the latter interpretation is preferable. The two pas- sages placed side by side read:
lb 11-13 enu- VIb 49-51 pir endendu .
znek / pir pone / esonome ferar piife
ahtimem en- pir entelust, ere fertu poe
tentu. pune pir perca arsmatiam habiest .
entelus ahti- erihont aso desire onse fer-
mem enumek tu . erucom prinuatur dur
steplatu par- j etulo, perca ponisiater ha-
fam tesvam tefe bituto . ennom sliplalu par-
tute Ikuvine. fa desua seso tote liouine.
It will be noticed that in the later version there is nothing corresponding to ahtimem which stands with pir... en- tentu in the earlier, and also that beside pir entelus ahtimem we find pufe pir entelust. This omission of *alitimem in one place and use of a substitute expres- sion in the other appears to be a clear instance of tabu. Since esonome is 'to the sacrifice, ' in other words the desti- nation of the fire, ahtimen and pufe must in all probability designate that in which it is carried; that is, a sort of fire-pan or brazier. For a noun with suffix -ti- as the name of an instrument or utensil we may compare L. vecfis < *wegh-ti-. For tabu against the mention of the fire-carrier see on VIb 45. — 13 steplatu: on Via 2. — parfam, tesvara: on Via 1, where the parfa is associated with three other birds. "V^^ly it is alone required here cannot be determined. On the possible reason for this second taking of the auspices see on VIb 48-52. — 14 vapefem avieklufe kumpifiatu: on VIb 51. — vea aviekla: on VIb 52. — 15 prinuvatu: on VIb 50, where in contrast to the present passage the mention of the prinuati precedes the instruction sti- platu parfa desua. — perkaf: on Via 19. — punigate: on VIb 51. — menes: the normal spelling would be *benes; menes probably originated through the decompounding of *com(m)enes < *com-
benes. — 16 Akefuniamem: this place in relation to Iguvium proper cannot be positively identified, but see on VIb 52.
— etufstamu, etc.: as in VIb 52-5, when the adfertor has arrived at Acedonia he pronounces banishment against any members of the Tadinate people and tribe and the Tuscan, Naharcan, and lapudic "name" who happen to be present. For trifu, numem, and the possible identi- fication of the nations named, see on VIb 54. — 18 svepis habe, purtatulu pue mefs est, feitu uru pefe mefs est: see on VIb 54-5. In the latter pas- sage the text of the sentence of banish- ment is given more fully, commencing at pisest tolar Tarsinater and including the names of the alien states. — 19-20 pune priniivatus staheren termnesku, enumek 'afmamu / kateramu Iku- vinu': in contrast to VIb 55-6 the older version omits the instruction to pronounce the sentence of banishment three times, and uses pi. termnesku in place of sg. termnuco; the participation of the pri- nuati is differently described, and there is no *d.eitu to introduce the command afmamu kateramu Ikuvinu. For termnesku see on VIb 53, for afmamu kateramu see on 55. — enumek apre- tu tures et pure: VIb 56 reads eno com prinuatir peracris sacris ambretuto. Since bulls are mentioned nowhere else in the description of the lustration of the people, the victims at the three stations Fontuli, etc., being boars, pigs, and heifer-calves, tures must be used here in the general sense of 'victims'; cf. Bii., p. 98, Dev., pp. 278-9. pure may be the same fire which was placed in the carrier in 12.
— 21 etatu Ikuvinus : see on VIb 63.
— 23 prinuvatus gimu etutu, etc. : as in VIb 65 the prinuati depart by the same way by which they came, and as in Vila 51-2, so in lb 40-41 the same (or other?) prinuati participate in the heifer-chase. For the problem involved see the introductory remarks on VIb 48 -Vila 54.
lb 24 - 26 (= Vila 3 - 5). Sacrifice of
166
III. TEXT AND TRANSLATION
ustentu : puni : feiu : / (26) ta^ez : pesnimu : af epe : arves : /
(27) rupinie : e : tre : purka riafra : ute : peia : fetu : prestate / (28) 9erfie : gerfe : marties pefaia : feitu : arviu : ustentu : / (29) kapi : sakra : aitu : vesklu : vetu : atru : alfu : puni : fetu : / (30) tagez : pesnimu : afeper : arves : /
(31) tra : sate : tref : vitlaf : feitu : tuse : gerfie : gerfe : marties : /
(32) pefaia : feitu : arviu : ustetu : puni : fetu : tagez : pesnimu : /
(33) afeper : arves : pune : purtingus : kaf etu : pufe : apruf : /
(34) fakurent : puze : erus : tef a : ape : erus : tefust : pustru : / (35) kupifiatu : rupiname : erus : tefa : ene : tra : sahta : kupifiaia : / (36) erus : tefa : enu : rupiname : pustru : kuvertu : antakre : / (37) kumate : pesnimu : enu : kapi : sakra : aitu : vesklu : vetu : / (38) enu : satame : kuvertu : antakre : kumate : pesnimu : enu : esunu : / (39) purtitu : fust : /
three boars to Serfus Martius at Fontuli. For details see on Vila 3-5. In general the two versions agree, but there is a slight difference in the arrangement of the instructions, as well as the usual dif- ference in the expressions relating to the cereal offerings. Moreover the instruc- tion to distribute the erus upon receiv- ing the order from Trans Sa.ictam is given in Vila 5 in connection with the present sacrifice, but in lb not until
sacrifice with mead, and pray silently with (offerings of) fat and grain.
At Rubinia sacrifice three pigs, red or black, to Prestota Serfia of Serfus Martius. Sacrifice (the victims) upon the ground, present grain-offerings, move the consecrated bowls, divide the black and the white vessels, sacrifice with mead, and pray silently with (offerings of) fat and grain.
At Trans Sanctam sacrifice three heif- er-calves to Tursa Serfia of Serfus Mar- tins. Sacrifice (the victims) upon the ground, present grain-offerings, sacrifice with mead, and pray silently with (of- ferings of) fat and grain. When you have made the presentation, the order shall be called to the place where they have sacrificed the boars, to distribute the erus. When the erus has been dis- tributed, the instruction shall be passed back to Rubinia to distribute the erus. Then the instruction shall be passed to Trans Sanctam to distribute the erus. Then return to Rubinia and pray with the whole and the ground (grain). Then move the consecrated bowls and divide the vessels. Then return to (Trans) Sanc- tam and pray with the whole and the ground (grain). Then the sacrifice will have been completed.
33-4 in connection with the third sacrifice. lb 27 -30 (= Vila 6 - 8). Sacrifice of three pigs to Prestota Serfia of Serfus Martius at Rubinia. In the later version the sacrifice proper is followed in 8-40 by supplementary offerings closely re- sembling those described in VIb 5-18 in connection with the sacrifice behind the Tesenacan Gate. In the present version, however, the only suggestion of such of- ferings is found in the passage kapi
lb 24
COMMENTARY
167
sakra aitu, vesklu vetu atru alfu, on which see below. — 28 pefaia: here for the first time in Table I; in the second and fourth sacrifices of the purification of the Fisian Mount, where Via 58, b 3 have persae felu, la 9, 16 in the corres- ponding position have supa sumtu, but in lb 32, 44 pefaia fe(i)tu is used in con- nection with the sacrifice at Trans Sanc- tam and the sacrifice after the pursuit of the heifers. For its meaning see on Via 58. — 29 kapi sakra aitu: the same instruction occurs in Vila 40 at the conclusion of the ceremony. For the de- tails see on VIb 18, — vetu: of uncertain etymology and meaning. The view pre- ferred here is that of Bii., PL, and Bk., who translate 'dividito.' The phonolog- ical development would then be *weid- hetod > *weif(e)t6d >vetu, with syncope followed by the same treatment of -//- as in 38 b, where -ft- is from -pt-. *veitu might be a more natural spelling, but there are certain difficulties in deriving vetu from *weltu, as Blum., Dev., and Bott. do, for it is not certain that loss of I before /, as in kumates < *kom-mal- tois, would occur here; see 55 d. In fact veltu IV 21, ehueltu Via 2, are against such a view, while at the same time there is not enough similarity in sense be- tween lb 29 and IV 21 to lead to the as- sumption that the verb in the two pas- sages is the same. The meaning then is probably ' divide the black and the white vessels. '
lb 31 -39 (= Vila 41 -45). Sacrifice of three heifer-calves to Tursa Serfia at Trans Sanctam. The details closely re- semble those in the preceding sacrifice at Rubinia, and also agree in most re- spects with Vila 41-45, if allowance is made for those differences which are usual between the ritual instructions of the earlier and later tables. In the earlier as in the later version we must picture sacrifices being performed at three ad- jacent stations partly at the same time, and in each version the adfertor goes from Trans Sanctam to Rubinia for the grind-
ing of the cereal offerings and the cere- mony of the vessels, after which he re- turns to Trans Sanctam. — 31 Tra Sate: for this place-name, which has the form of a prepositional phrase and which we render ' Trans Sanctam, ' see on Vila 5. — Tuse gerfie gerfe Marties: see on VIb 58. — 33 kafetu: for the difference between this verb and kupifiatu see on Vila 43. — pvife apruf fakurent: the same substitute for *Funtluf-e, possibly euphemistic in pur- pose, is used in Vila 43. — 34 erus: for the procedure in distributing the erus at the three stations see introductory remarks on Vila 41-45, and on the erus itself see on VIb 16. —36-7 antakre / kumate pesnimu: the same instruction occurs immediately below in 38 and also in I la 42 in the instructions for the dog- sacrifice. All recent editors take antakres as equivalent to L. integris, no forms in the Latin alphabet being recorded to serve as a test for the value of the k. The priest prays over the ground and un- ground grain or cakes; see on VIb 17 on comoltu, which is usually not accom- panied by a direct object. In Vila 44, 45 comoltu, comatir persni{hi)mu the in- instruction to grind is not merely im- plied but directly given, while on the other hand no distinction is made be- tween unground and ground offerings. — enu kapi sakra aitu, vesklu vetu: precisely as in 29 above, except that the color-names atru alfu are omitted. — 38-9 enu esunu / purtitu fust: not of the porrectio but of the conclusion of the whole sacrifice; similarly in the equi- valent Vila 45 and in VIb 42 at the end of the ceremony behind the Veian Gate. lb 40 -44 (= Vila 51 -54). Ritual pursuit of the heifers. Actually the intro- ductory clause pustertiu pane puplu atefafust corresponds to Vila 46, but the whole passage Vila 46-51, containing the curse against the enemies of Iguvium, is without counterpart in lb. For the chase in general and the differences be- tween the earlier and later accounts of
168
III. TEXT AND TRANSLATION
(40) pustertiu : pane : puplu : atefafust : iveka : perakre : tuseiu / (41) super : kumne : affertur : prinuvatu : tuf : tusetutu : / (42) hutra : furu sehmeniar : hatutu : eai iveka : / (43) tre : akefunie : fetu : tuse iuvie : arviu : ustetu : / (44) puni fetu : per aia fetu : tagez pesnimu : afepe : arves : / (45) kvestre : tie : usaie : svesu vuvgis titis teteies : /
When he has made the circuit around the people for the third time, the ad- fertor shall chase a heifer fit for sacri- fice, above the Comitium, and the prin- uati shall chase two. Below the Forum of Semonia they shall catch them. These three heifers he shall sacrifice at Acedon- ia to Tursa Jovia. He shall present grain-offerings and sacrifice with mead. He shall sacrifice (the victims) upon the ground, and pray silently with (offer- ings of) fat and grain. Lucius Tetteius, son of Titus, approved (the foregoing) in his quaestorship.
40 tuseiu: for tusetu.
it see the Introductory remarks on Vila 51-54. — 40 pustertiu pane puplu atefafust: the pursuit probably takes place, according to the view adopted here, after the completion of the whole lustra- tion Including the sacrifices, and not merely after the third circumambulation. See introductory remarks on VIb 48 - Vila 54. — iveka: the adfertor chases a single heifer and the two prinuafi one each; in Vila 51-2 the adfertor and pri- nuafi chase possibly twelve heifers, though the number is not precisely stated. — perakre: on Via 25. — 41 kumne: on- ly here, being omitted from the instruc- tions in Vila 51-2, but without doubt the same word as O. comenei, comono 'comitium. ' The location of this place of assembly cannot be determined with cer- tainty, but Miss Rosenzweig conjectural- ly places it a short distance southwest of the modern Corso Garibaldi. — 42 hutra furu sehmeniar: on Vila 52. — hatutu: since there is no new subject, the adfertor and prinuati are evidently expected to catch the heifers themselves. In Vila 52, where the number of heifers is larger, any person is authorized to catch them below the Forum.
lb 45. Formula containing the name
of the authenticating official. Nothing equivalent is found at the end of the later version in Vila, but the formula at the end of Ila is substantially identical. In consequence of certain differences in spel- ling and word division, however, it is best to show the exact text of both passages: lb 45 kvestre: tie: usaie: svesuvuv- gistitisteteies : Ila 44 kvestretieusa- gesvesuvuegistiteteies, the last passage being written in the lower left margin commencing a little to the left of esunu and extending upward. That kvestre- tie is to be read as the first word of the sentence is bej^ond question. The last half of the formula, on the other hand, has caused considerable difficulty be- cause of the lack of word-division in both versions, but the recognition of usaie / usage as a verb and of Vuvgis Titis Teteies as the name of the quaestor has finally led to a satisfactory analysis and interpretation of the passage. — kves- tretie: for the kvestur see Va 23, b 2, where the kvestur or the fratreks is directed to take a vote on whether the adfertor has provided satisfactorily for the ceremonial meal and, in case of a negative vote, on what the amount of the penalty should be. For the for-
lb 40
COMMENfTARY
169
mation of the noun kvestretie cf. uhtretie Va 2, 15. In Va 1-3 the epo- nymous magistrate was the uhtur ra- ther than the kvestur; for the relation of this fact to the chronology of the ta- bles see Introduction, 18. — usaie: Ila 44 shows the correct form usage, the er- ror being caused by omission of the loop at the lower left of the vertical hasta in the letter d- Devoto, p. 302, derives from *op-sak-ed as pf. 3 sg. from the root *sa(n)k (formation as in 124 b), partly anticipated by PL, I, p. 555, II, p. 671, and by Linde, GL, XX (1932), p. 292, in so far as the latter took the form for a verb, although his etymology was different (< *opesankiit, pf. of osatu). Vet. and Bott. have followed Dev. ; the former suggests that the authorization was for the cost of preparing the table. — svesu : loc. sg. of the possessive pronominal adj. equivalent in meaning to L. suus; for an attempt at an etymological explanation see 107, n. 1. — Vuvgis Titis Teteies : the praenomen, father's praenomen, and nomen ; for the arrangement see on Va 3. — Vuvgis: taken by Thurneysen, K.Z., XXXII (1893), pp. 559-61, and by Buck (with reservations) and Blum, as equi- valent to L. Lucius. Dev. p. 303, rather than allow uv to stand for a retained diphthong, derives from *Vof(i)kio-, as if corresponding to L. *Vovicius, voveo, etc. Bott. reads Vuvgis here, Vuegis in Ila 44, * Voecius in his Latin transla- tion in both places. In Ila 44 the third letter has a cross-bar making an e, which is visible in the photographic plate in Devoto's edition. BottigUoni's reading is therefore very tempting, since it seems
more natural to assume that the cross-bar was erroneously omitted in lb 45 than erroneously added in Ila 44. Yet there are certain objections. Latin Voesius, which Bott., p. 270, cites as support for * Voecius (cf. Schulze, Lat. Eigenn., pp. 105 f.; Conway, //. Dial., I, p. 328), is a nomen, while the Iguvine Vuvgis, Vue- gis, like L. Lucius, is a praenomen. Moreover forms with -s- from scattered sites in Etruria or from Praeneste are not reliable evidence for Umbrian forms with 9 (from k before front vowel) unless we assume that such palatalization had an extraordinarily wide distribution. It seems best, therefore, to adhere to the old equation of Vuvgis with Lucius, or possibly * Vovicius after Devoto. For e in place of v we have a possible parallel in la 17, where the fifth letter of Fiiuvi (for Fisuvi) has in its center a blur which gives evidence of a not entirely successful attempt to correct an e into a v. As be- tween Lucius and Vovicius the former seems preferable, since * Vovicius is no- where actually attested. The spelling with uv may possibly represent a family tra- dition. — Titis: abbreviated Ti in the other version; etymological equivalence with L. Titus is virtually sure. On the gen. sg. ending see 92. — Teteies: the gentile name Tetteius occurs in C.LL., VI, 5296, 27296 (Rome), X, 1403 (Her- culaneum), 1781 (Puteoli). For the end- ing -ies see 32 b. It is otherwise unknown in Umbrian, but occurs in Paelignian and in certain Oscan inscriptions in the Greek alphabet, and is better understood here as nom. than as gen. sg.
170
III. TEXT AND TRANSLATION
(I la 1) pune : karne : speturie : atiiefie : aviekate : naraklum : /
(2) vurtus : estu esunu : fetu : fratrusper : atiiefie : eu : esum /
(3) esu : naratu : pefe : karne : speturie : atiiefie : aviekate : / (4) aiu : urtu : fefure : fetu : puze neip eretu vestige sage : /
2 esum: for esunu (W for VH)-
Ila. This table falls into two portions quite distinct in content. Lines 1-14 con- tain instructions for sacrifices to be per- formed in case of unfavorable auspices, while 15-43 describe the sacrifice of a dog to Hondus Jovius. The distinct char- acter of the two parts may originally have been shown by a space after 13, which was later filled by the words Agetus perakne fetu (cf. Vetter, p. 189); such a theory is at least plausible in view of the space at the end of 13, which would have been fully adequate for Age- tus. Ila ends with a sentence written in the lower part of the left margin, which, despite certain differences of spelling and word-division, is substantially identical in content with the formula of authenti- cation at the end of lb. It is evident therefore that lb and Ila were written at about the same time, but it is not cer- tain that the auspices in question are those connected with the sacrifices dur- ing the purification of the Fisian Mount and the lustration of the people, as Bii- cheler, p. 123, and Rosenzweig, p. 47, suggest. The expression fratrusper Atiiefie (2), in place of ukriper Fisiu tutaper Ikuvina or pupluper tutas liuvinas tutaper Ikuvina is unfavorable to such a view (cf. Dev., p. 309), but the question must be left undecided.
Ila 1-14. Sacrifices in case of un- favorable auspices: a young pig to Ves-
When for the augural section (of the) Atiedian (Brotherhood) in taking the auspices the response turns out (unfavor- ably), perform the following sacrifices for the Atiedian Brothers. Proclaim these sacrifices thus: "If for the augural section (of the) Atiedian (Brotherhood) in taking the auspices the response hath turned out for ill, bring it to pass that it be as not intended." Offer an addi- tional sacrifice of a young pig to Vesti-
ticius Sancius, an ox to Jupiter, and a victim of uncertain species to Spector, a ram to Jovius, a sheep to Ahtus Jupiter, and a boar to Ahtus Mars. For the Ro- man practice of offering additional sa- crifices if the previous one gave unfavor- able auspices cf. Cic, Div., II, 17, 38 Quid'} cum pluribus dels immolatur, qui tandem evenit, ut litetur aliis, aliis non litetur'^ quae autem inconstantia deorum est, ut primis minentur extis, bene promit- tant secundis^ — 1 karne: it seems better to discard 'carni', the rendering of earlier editors, and, following Blum, and Dev., to adopt 'part, portion, sec- tion ' in all passages w4iere the word oc- curs. In Va 24, 27, b 4 this is unavoidable, and it agrees with the sense in O. maimas carneis senateis tanginud (Tabula Ban- tina = Co. 28, line 3), while the Latin sense ' meat, flesh ' shows a later stage in the semantic evolution. See Blum., p. 68. — speturie: only here and in the similar formula in 3, but the formation is clear: the L. equivalent would be *spectoriae, related to *spector, U. Spe- ture, as amatorius to amator. It is there- fore that section of the Atiedian Brother- hood which is concerned with spectio or the taking of auspices. — aviekate: from a denom. vb. stem avieka-, and pro- bably a deponent pf. pass, pcpl., although analysis as a pres. pcpl. with loss of n before t is also possible. The whole ex-
Ila 1
COMMENTARY
171
pression karne speturie Atiiefie avie- kate, here and in 3, then means ' for the augural section (of the) Atiedian (Brother- hood) in taking the auspices. ' — na- raklum vurtus: the unfavorable nature of the omens is not directly stated, but Bii., p. 124, interprets the passage in the light of the ancient belief that evil omens could be produced by a miraculous change in the organs of the victim at the time of sacrifice; cf. Cic, Div., II, 15, 35 cam immolare quispiam velit, turn fieri ex- torum mutationem, ut aut absii aliquid aut supersit. Pighi, R.F.I.C., XXXII (1954), p. 244, takes vurtus as 2 sg. and translates ' quom parti augurali Atiediae auspicanti narrationem verteris,' under- standing vurtus to denote some error made by the priest in pronouncing the naraklum; this gives good sense in its context, but the construction of the da- tive karne ... aviekate seems slightly more natural by the other interpretation.
— 3-4 pefe karne ... / ... fefure: the formula actually to be used in announc- ing the expiatory sacrifice differs from the clause in 1-2 not only in the use of pefe rather than pune, but also in the words designating the nature of the omens, and it is not unreasonable to sup- pose that the formula in 4 preserves more archaic language. — 4 aiu: found only here and variously interpreted, but best taken in etymological connection with L. aio, prod-igium, as ' response. ' For the suffix, R. G. Kent, C.P., XV (1920), p. 363, admits three possibilities : -id-, -ion- (see 81 a, n. 1), or -io- with declen- sion as neuter; the last is the most likely.
— fefure: taken as fut. pf. 3 pi. of futu by Br., Bk., Blum., and more recently by Vet., but there are several objections: it is unlikely that a fut. pf. of the auxil- iary would be used in making a fut. pf. of the verb as a whole, and even if a fut. pf. were so used, we should not expect a reduplication; the Tables have many in- stances of unreduplicated fust as auxili- ary, whether we prefer to call it fut. or fut. pf. : cf., for example, lb 7, 39, Va 19,
20, 22, Via 7, b 39, etc. Moreover -nt in 3 pi. verbs and other forms is preserved with considerable regularity, the few examples of loss being subject to special explanation, fefure then is probably pf. 3 sg. of some other verb. Connection with cpvQco, 7ieq)vorai is possible, even though Biicheler rejected it in favor of L. forare or furcillare. The semantic evolu- tion then may have been 'mingle > confound > disturb > make ineffective, ' and a literal translation may be 'if ... the response (which has) come forth has made (the sacrifice) ineffective. ' Cf. Kent, C.P., XV (1920), p. 363; Pisani, Gl., XX (1932), pp. 96-7; the latter takes fefure as 3 pi. with an ending -re similar to cer- tain 3 pi. forms in Tocharian B, but aiu and fefure can as easily be sg. as pi. (cf. esunu 2, and 115 c), and a sg. is somewhat easier here in view of narak- lum vurtus in the alternate formula. — - fetu puze neip eretu : cf. the more ellipt- ical expression pusei neip heritu Via 27, etc. — Vestige: this god, who is here grouped in a triad with Jupiter and Spec- tor, is not otherwise known except from the expression carsome Uestisier, which occurs in Via 13-14 in the list of points used in defining the city-limits. There is little doubt, however, that his name is connected with uestis, uesticatu, uestisiar (see on VIb 5), and if the connection of these words with Vesta is valid, Vesticius is probably a god associated with the hearth. Bii., p. 125, hesitantly speculates on the possibility of identifiying Vesticius with the Oscan Vezkei of the Agnone Dedication (Co. 175), A 2, B 3, on the basis of an interchange of z (= ts) and St. — Sage: an epithet of lupater in lib 17, 24, and used alone in b 10; see also on VIb 3. The nasal before 9 s is omitted in all 5 instances in the native alphabet and written in all 12 instances in the Latin alphabet. Yet it is un- likely that we have a distinct non-nasal form replaced in the later tables under Latin influence by a nasal form, as Dev., p. 217, suggests. In the earlier tables n
172
III. TEXT AND TRANSLATION
(5) sakre : iuvepatre bum perakne : speture : perakne : restatu / (6) iuvie : unu erietu sakre : pelsanu fetu : arviu : ustentu : / (7) puni fetu : tagez : pesnimu : afepe : arves : pune purtiius : / (8) unu : sufu pesutru : fetu : tikamne : iuvie : kapife : / (9) pefu : preve
cius Sancius, an ox brought from away to Jupiter, and a victim brought from away to Spector. Offer to Jovius one ram from the temple-property, to be buried. Present grain-offerings, sacrifice with mead, and pray silently with (of- ferings of) fat and grain. When you have made the presentation, present one ^ig-persondro to Dicamnus Jovius. Make a mound in a separate place for the
before g is often omitted, and written only in angif Ila 25, purtingus lb 33. Besides, the wide distribution of the stems Sancu-, Sanco- has received new confir- mation from the appearance of the name San. qus on one of the Itahc inscriptions in the Val Camonica (see Altheim, Ge- sch. d. kit. Spr. [Frankfurt-am-Main, 1951] pp. 100-04). — S sakre: for the sub- stantive use cf. Ill 8, 9, 12 (where sakre is not an epithet of the follo¥/ing uvem), 22, 30, Va 6, VIb 52, 56 ; O. sakrim (Curse of Vibia = Co. 130, hne 11), sakriss (Co. 113), sakrid (Co. 114), sakrim (Co. 117). That the I'-stem sacri- may be used of a young pig as a sacrificial victim is ap- parent from Varro, R.R., II, 1, 20 E quis (sc. porcis) qui iam pari sunt ad sacri- ficium, ut immolentur, olim appellati sa- cres; 4,16 Cum porci depuisi sunt a mam- ma, a quibusdam delict appellantur neque iam lactanfes dicunlur, qui a parlu decimo die habentur puri, et ab eo appellantur ab antiquis sacres, quod tum ad sacrificium idonei dicuntur primum. sacres as attri- bute of porci occurs in Plant., Men., 289-90, Rud., 1208. — perakne: most- ly used as an attribute of sacrificial vic- tims, but of uncertain meaning; common- ly connected with acnu, O.akensi, akun., L. annus, as a precise formal equivalent to L. per-ennis though with a sense more like that of sollemnis (substantially so A.-K., II, p. 355, Buck, § 159a, Blum., p. 45, E.-M.3, p. 63. But this fails to give a satisfactory interpretation of some of the passages where perakn- occurs. — It
is a striking fact that with the exception of Via 54, where we must almost certain- ly emend peracnio to peracrio, the adjec- tive perakn- occurs only in Ila, b, and Va 7, while peracr-, apart from perakre lb 40, occurs only in Via, b, and Vila, and further that both are at times associated with sacr- sakr- in what appears to be a contrast of meanings (e.g. not only the present passage, but also VIb 52 uia auiec- la esonome etuto com peracr is sacr is, with the closely similar VIb 56). Devoto calls attention to the distinction between ani- mals raised for sacrifice (sakri-) and those from secular sources selected as victims (pp. 190-91; cf. his similar classification of vessels, p. 237), and he maintains con- nection with acnu, which, however, he renders by ' excensione, mole' rather than by 'anno. ' In view of O. akenei, akun., it seems best to preserve the meaning ' year ' for U. acnu and abandon connec- tion between it and perakne. Now the k of perakne can stand for g as well as for k (la). It may therefore be possible to derive from *per-ag-ni- 'driven (from) beyond, ' with the value of per- similar to that in L. peregrinus, as suggested by Devoto. -no- as a suffix used in the for- mation of passive verbal adjectives is represented not only by L. planus, etc., but by the pf. pcpls. of numerous Indo- Iranian, Slavic, and Germanic verbs. In the present instance we have in place of -no- an f-stera suffix -ni-, and the change of stem cannot be explained as in the bahuvrihi compounds L. imberbis, trire-
Ila 4
COMMENTARY
173
mis, etc. But the change may have been occasioned by the use of the word as member of a contrasting pair with sacri- sakri-, which outnumber the o/a-stem sacro/d- by 13 to 4. We may then render the sense by 'brought from outside, brought from away. ' See also on Via 25 peracrei, Ila 21, sevakne, — Speture: this god, the third of the present triad, is mentioned only here, but in view of the expression karne speturie Atiierie aviekate in 1, 3 above, and of such tech- nical terms as L. auspicium, extispicium, haruspicium, ignispicium, and speciio 'right to take auspices,' there can be little doubt that Spector was in some v/ay a god of augury. See Rosenzweig, pp. 95-6. — restatn. : cf. lb 9 restef esunu feitu = VIb 47 reste esono feitu in the in- struction for repeating the purification of the Mount in case of any omission in the ceremony.
Ila 6 - 10. Sacrifice of a ram to Jo- vius, followed by the offering of a per- sontru to Dicamnus Jovius. The prin- cipal problem is the identification of Jo- vius, for Ila 6 is the only passage in the tables where a form of the stem lovio- ap- pears alone as a divine name. As a par- allel Rosenzweig, p. 80, mentions the Dea Dia of the Fratres Arvales, and there is no question as to the etymological re- lation of the names, even though the gra- dation and stem-types are different. Yet it is hard to believe that luvie does not involve an ellipsis of some other name, to which it may have stood in the same relation as in the names Tikamne luvie, Hunte luvie, etc. Dev., pp. 315-6, calls attention to the strikingly close resem- blance between the present passage and the description in la 24-34 of the sacrifice of three lambs to Tefer Jovius and the supplementary offerings which follow. The passage in la is more detailed, but both contain the instructions pelsana fetu arvia ustentu puni fetu tagez pes- nimu afiper arvis (with orthographical variations); both call for the offering of a sufura pesvmtru after the porrectio; for
the construction of a mound (per um.) for the bowl; and for the grinding of the cereal offerings, to be followed by prayers. It is possible therefore that in order to balance the repetition of la 24 Tefre luvie ... 26 pelsana fetu ... 27 api habina purtiius, sufum pesuntru / fetu ... Tefri luvi fetu we must under- stand Tikamne to belong in sense with luvie in Ila 6 as well as in 8. — 6 unu erietu: the bronze reads unue- rietusakre, but the word-division is beyond reasonable doubt. For erietu all editors since Aufrecht and Kirchhoff adopt the translation ' arte tern,' with the exception of Breal. unu occurs only here and in 8 below, and it is strange that the numeral should be used at all with re- ference to a single victim. Br. translated ' agnum, ' but with an untenable etym- ology (: dfivoQ, cf. p. 277). Pisani, A.G.I., XXVII (1935), pp. 165-6, translates ' ovillum,' with a better etymology (unu < '^ovino-), but the normal Italic suffix for adjectives from animal-names is not ino- but ino-, as in cabriner, habina, where the unsyncopated / must belong. The interpre- tation of tmu as derived from the stem ovi- also creates difficulties in 8 if suru is to be connected with L. sus, according to the usual view. The word-order also favors taking unu as ' unum, ' since the normal practice in I and II is to place the numeral before the noun (see 155 h). — pelsanu: see on VIb 22 — arviu: on la 3 and below on Ila 18. — 7 puni: on Via 57. — arepe: on la 3. — 8 sufu pesutru: on VIb 24. — Tikam- ne: taken by most earlier editors as abl. of a noun related to L. dicare and signifying a dedicatory formula. Br., p. 278, took it as a divine name, and Dev. and all subsequent editors have so taken it, while retaining the traditional etymological connection. Most recently G. B. Pighi, R.F.I.C., XXXII (1954), p. 247, pointed out its resemblance to L. Vertumnus, Picumnus, of suspected Etruscan origin, and alumnus, femina, from IE medio-passive participles: what-
174
III. TEXT AND TRANSLATION
fetu : ape : purtiiusufu : erus : tetu : enu : kuma/ltu kumate pesnimu ahtu : iuvip : uve
peraknem : / (11) pefaem fetu : arviu : ustentu puni : fetu : ahtu marti : abrunu / (12) perakne : fetu : arviu : ustetu : fasiu : pruse^ete : afveitu : / (13) pefae : fetu : puni fetu : tra : ekvi : ne : fetu : / (14) agetus : peraikne : fetu : /
9 purtiiusufu: for purtiius : sufu.
bowl. When you have presented the pig- persondro, distribute the erus. Then grind (the grain) and pray with the ground (grain). To Ahtus Jupiter sa- crifice upon the ground a sheep brought from away. Present grain-offerings and sacrifice with mead. To Ahtus Mars sa- crifice a young boar brought from away. Present grain-offerings, add spelt-cakes to the parts cut off, sacrifice the victim upon the ground, and sacrifice with mead. Sacrifice across the (Via) Equina. Sa- crifice a victim brought from away, to the Ancites.
ever the actual origin of *Dicamnos, it was popularly associated with *deik-, *dik-a, and had approximately the sense ' he who speaks. ' See further on Ahtu in 10 below. — kapire: on VIb 18.
— 9 pefu: on VIb 24. — preve: found also in la 28, on which see note. In the latter passage, however, preve referred to the manner of offering the uestisia and not to the building of the mound for the bowl. The present passage is best clarified by a comparison with la 29 testruku peri kapire pefum feitu, 32 nertruku peri kapire pefum feitu: in the latter ceremony two bowls were used, while in I la 8-9 a single bowl is used. The function of preve then is similar to that of unu in 6, 8; both appear some- what unnecessary, but may be used in intended contrast to the corresponding instructions in la, as Dev., p. 316, sug- gests. — erus: for a discussion of the nature of this offering see on VIb 16.
— kumal/tu kumate pesnimu: on VIb 17.
Ila 10 - 14. Sacrifice of a sheep to Ahtus Jupiter and of a boar to Ahtus Mars. The word ahtu occurs nowhere except in the present passage, but prac- tically all editors take it as in some way forming a part of the names of the two
deities. The form may be easily analyzed as dat. sg. of a stem in -tu-, and the root may be either that of L. ago or that of L. aio, ad-agium, prod-igium, with -ht- < -kt- < -gt-, as in rehte. Probably the latter alternative is the more satisfactory. Pighi, R.F.I.C., XXXII (1954), pp. 245- 6, in substantial agreement with Dev., p. 314, so takes it, along with aiu < *ag-yo- and Agetus < *ag-yent-, com- paring also the Roman Aius Loquens (Varro apud Aul. Gell., XVI, 17, 2; Cic, Div., II, 32, 69) and Aius Locutius (Livy, V, 50, 5). Both Jupiter and Mars have oracular functions through which they make known their will to men, and it is entirely natural that the oracular aspect of their character should have a prominent place in the first part of Table Ila. The utterance of the gods is deified and the deification is shown not by using adjectives *Iuvie and *Martie but by apposition of the names of the gods them- selves, for Iuvip in 10 can reasonably be taken only as dat. of the noun lu- pater. — 11 pefaem: see on Via 58. — abrunu: in (b 24 = Vila 3 during the lustration of the people of Iguvium three boars are sacrificed to Serfus Martius at Fontuli. The details of the ritual, how- ever, are strikingly different. The in-
IlaS
COMMENTARY
175
structions for the sacrifice at Fontuli in- clude among other matters the expression uaiuo ferine fetu, which is elsewhere used only in connection with oxen (la 4, 13, 22, Via 57, b 1, 19) or bull-calves (lb 3, 5-6, VIb 43-4, 45). The present sacri- fice, on the other hand, contains the in- struction peraem fetu, which is used in Via 58 of pregnant sows, VIb 3 of sucking pigs, lb 28 = Vila 7 of sows, lb 32 - Vila 41 of heifer-calves, lb 44 = Vila 54 of heifers, and III 32 of a sheep; in Ila 21-2 the expression esunu / pefae futu is used in the description of the dog- sacrifice. In other words the two expres- sions uatuo ferine fetu and peraem fetu never occur in the instructions for one and the same sacrifice and never with vic- tims of the same class, unless we equate aprunu with apruf abrof of the sacrifice at Fontuli. It is probable therefore that the stem-variation between aprunu and apruf (as if L. *apr6nem stood beside ace. pi, apros) corresponds to a difference in the victim itself, with aprunu pro- bably standing for ' young boar, ' since the instruction to slaughter the victim stretched upon the ground (peraem fetu) is generally used in connection with smal- ler victims. Cf. A.J.P., LXXVII (1956), pp. 177-80. — 12 fasiu prusegete af- veitu: see on VIb 1-2, where spelt-cakes form a part of the offerings made to Mars Grabovius in the third sacrifice of the purification of the Fisian Mount. The latter passage, together with VIb 43 on the sacrifice to Mars Hodius at the Grove of Jupiter, and the present passage are sufficient to show that spelt-cakes had an especially close association with the cult of Mars. — 13 tra ekvine: the actual reading of the bronze is tra: ekvi:ne. Von Planta, I, pp. 347-8, Dev., and Bott., assume connection with eik- vasatis and eikvasese but are unable to find any entirely satisfactory solution, and moreover the normal Umbrian neg- ative is neip neip, the short form nei being found only in IV 33 in combination with the subjunctive (a)fhabas. It seems
best therefore to read ekvine, taking it as equivalent to L. equinae. For the pho- nological problem see 46 h. For a clear instancp of erroneous use of the word-di- vider cf. kvestre.tie lb 45. Blum., p. 57, compared tra ekvine ' trans equinam viam ' with tra sate lb 43. For the case- usage see on Vila 5. Vetter, p. 191, sug- gests that the sacrifices to the two Ahtus deities were performed in a different place from the sacrifice to Spector. — 14 Agetus: again the interpretation is uncertain, and there is no other instance for purposes of comparison. The form is almost certainly dat. pi. of a ^stem, or of a pres. pcpl. with loss of n before t by 52 d. As a pres. pcpl. it has been referred to ag-, root of ago (Bii.), and to ag-, root of ahtu, L. aio, prod-igium (Dev.), but there is no satisfactory explanation of the 9, for the normal product of palataliza- tion of g would be i by 47 b. It might be possible to assume a variant root-form ak- spreading analogically, as perhaps in O. infinitive acum, from forms where g is subject to unvoicing before t (cf. Brug- mann, Ber. kon. sacks. Ges. Wiss., 1890, pp. 237-8), but a translation ' Agentibus ' or its semantic equivalent ' Actoribus' tells us nothing whatever about the na- ture of the deities in question. Pighi (see above) on 10-14 reaches a better solution by deriving A5e(n)t- < *ag- yent- 'aiens,' which may be tenable from a phonological standpoint if we admit ag- < ak- < ag- as explained above. The Agetus then are apparently Ahtu luvip. and Ahtu Marti themselves, and Agetus is related to Ahtu as the speaker to the word, both being here identified with Jupiter and Mars. But it is rather strange that the instruction to offer vic- tims perakne to the two deities should be repeated here with substitution of the name Agetus for Ahtu luvip. ... Ahtu ... Marti. On the whole it seems safest to follow PL, Bk., Blum., Vet., in tenta- tively identifying Agetus "with L. dis Ancitibus, C.I.L., IX, 3515, from near Furfo in the Vestinian territory. On,
176
III. TEXT AND TRANSLATION
(15) huntia : katle : ti9el : stakaz : est : sume : ustite : / (16) anter : menzaru : gersiaru : heriiei :
phonological grounds the equation is easy, and the Umbrian palatahzation occurred early enough to preclude the change -nk- >-ng-.
Ila 15 - 43. Sacrifice of a dog to Hon- dus Jovius. For a defense of the tradi- tional view that katle means 'dog, pup- py,' not 'young pig,' see below on 15 katle, A number of details of the ritual are obscure, the more so because we have not the advantage of being able to com- pare an earlier and a later version, as in the case of the purification of the Fisian Mount and the lustration of the people. Those features, however, which are clear may be briefly outlined. The time at which the sacrifice is to be performed — sume ustite antermenzaru gersiaru — is given in 15-16, but there is good ground for believing that a distinction was made between the sacrifice perform- ed privately and that performed by the adfertor himself, and tha.t in the latter case a special time, designated by the expression menzne kurglasiu, was pres- cribed. The articles necessary for the sacrifice include cereal-offerings (arvia), the cakes known as struligla and fikla, wine and mead (puni), crushed salt, a maniple, wet and dry vessels, and an unguent (18-19). The victim is slain stretched on the ground, on behalf of the gens Petronia. The dancing of the tripudium takes place when the grain and mead are offered (24-5) after the slaughter of the victim, and again after the porreciio. A certain portion of the ceremony (34-9) takes place near an ob- ject known as the spina, possibly a sort of obehsk; this portion includes libations and prayers, the dancing of the tripudium, the anointing of the spina, and a ceremo- nial washing of the hands. After the adfertor has returned to the altar, he of- fers further prayers, the cakes are ground, and the remains of the dog buried. —
The Hondia: the day for (the sacri- fice of) the dog is appointed for the final period of the Cerealia occurring between
15 Huntia: generally taken as abl. sg., but as ace. sg. by Vetter, pp. 191, 194, who understands it as ace. of goal indi- cating the name of the place where the sacrifice is to be performed. If we so take Huntia, it is difficult to construe it with stakaz est. In 17 with fertu it would give good sense, but the evidence for ordinary accusatives of goal without a preposition in the Italic dialects is ne- gligible; we should rather expect *Hun- tiama(r) or *Huntiame(n),like asamaf , spinama, fesnafe. It is possible that Huntia should be left unconstrued and understood as the title of the body of instructions which follows. This is na- tural in view of the change of subject- matter after 14 and has a partial parallel in Via 12 where tuderor totcor as a gen- eral heading precedes the list of places which follows. Huntia is then best taken as neut. pi. and may be compared with such neut. pi. festival-names as Am- barvalia, Lupercalia, Saturnalia. Huntia in 17 may possibly be a neut. pi. meaning ' the things for the Hondia, ' which are then identified by the long series of nouns following fertu. — katle: this word, which occurs only in the descrip- tion of the sacrifice in the present table, is generally understood to mean 'puppy, dog. ' Vetter, p. 194, takes it to mean ' young pig, ' on the ground that the elaborate instructions which follow would not be devoted to the preparation of dog-meat, and as support for katel = 'young pig' he cites Plant., Epid., 579 aliter catuli longe olent, aliter sues, which, even apart from the mutilated condition of the text, does not seem to give ade- quate support. In Plant., fr. inc. XLII (Goetz and Schoell) et ego te conculcabo ut sus catulos suos it is the presence of sus in the context that establishes the mean- ing of catulos. I prefer to adhere to the usual interpretation of Table Ila 15 ff.
Ila 14
COMMENTARY
177
The sacrifice of dogs to infernal deities and for protection against grain-rust, and also the eating of dog-meat are attested by the following passages: Paul, e Fest., 39 Li. catulinam carnem esitavisse hoc est comedisse Romanos Plautus in Saturione refert; Ov., F., I, 389-90 Exta canum vidi Triviae libare Sapaeos, et quicumque tuas accolit, Haeme, nives; IV, 907-8 Flamen in aniiquae lucum Robiginis ibat, exta canis flammis, exta daturas ovis ; Colum., X, 342 mala robigo viridis ne torreat her- bas, sanguine lactentis catuli placatur et extis; Plut. Quaest. Rom., 52 did ri rfj KakovfXEvi] reveirii Mdvrj nvva Bvovai ; 68 <5td xi xvva dvovaiv ol Aovneqxoi; Plin., AT. H., XXIX, 58 catulos lactentes adeo puros existimabant ad cibum ut etiam placandis numinibus hostiarum vice uterentur his. Genitae Manae catulo res divina fit et in cents deum etiamnum ponitur catulina. If the Atiedian sacrifice included the eat- ing of dog-meat, the elaborate prepara- tions might be all the more necessary to make it fit for human consumption. The case of katle may be gen. (Qerfe lb 28, 31, punigate lb 15 give some support to loss of -s in gen. sg. in the older tables), but is perhaps better taken as dat., since -s is not omitted in katles Ila 22, 27. — tigel : commonly referred to root *dik- of L. dico, etc., and interpreted 'dicatio, dedicaiio, declaratio.' But Ribezzo, II, pp. 85, 93, followed by Vet. and Bott., gained a better interpretation by deriving from *die-kelo-; cf. O. zicolom in the Ta- bula Bantina, where the sense 'day' is practically certain. See also on b 22, III 25. — sume ustite: the expression re- curs in III 1-2. For ustite there is no entirely satisfactory solution, but it is possible to reconstruct a stem *op-stlto-, with sti~ an ablaut-variant of stdi- (so Devoto) and semantic evolution 'that which stands at hand ' > ' time. ' For the temporal use of sume cf. Verg. Aen., II, 324 venit summa dies. — 16 antermenzaru gersiaru: for anter- menzaru the etymological translation ' interlunary ' is easy, and since kur-
glasiu below is best taken as indicating the time for the sacrifice by the adfertor, there need not be any conflict in sense between antermenzaru and kurglasiu. gersiaru is commonly connected with sesna, gersnatur, L. cena, but since this latter group is regularly based on a nd- stem in Italic, it is probably better, with Dev. and others, to associate gersiaru with Ceres, -rs- where we should expect -rf- may be explained by derivation from *ker(e)s-siarum. The sense of sume us- tite anterinenzaru gersiaru then is apparently ' at the final time of the inter- lunary Cerealia. ' There were Roman fes- tivals in honor of Ceres at the beginning of spring (cf. Verg., G., I, 339-40 Sacra refer Cereri, laetis operatus in herbis, ex- tremae sub casum hiemis, iam vere sereno). — heriiei fagiu arfertur ... fagia tigit: it is uncertain whether we should punc- tate after arfertur and make it the sub- ject of heriiei, with the majority of edi- tors, or before arfertur and make it the subject of fagia, with Devoto. In the former case a distinction is supposed to be made between the dog-sacrifice of- fered privately and pubhcly, and the time specified for each is different. Dev., pp. 322-3, argues for the opposite view on the ground that the sacrifice is a single one, offered for the gens Petronia of the Atiedian Brothers, and that the mere occurrence of the adfertor's title is not suf- ficient to indicate a public sacrifice. If we accept Devoto's interpretation, the sub- ject of heriiei must be impersonal, with reference to any person who wants the sacrifice to be performed, and the ap- pearance of arfertur as subject of the following fagia must mean that the ad- fertor alone is authorized to perform it. But menzne kurglasiu leads to serious difficulty, because whether we translate 'at the full moon' or 'in the month Curslasius, ' this new designation of the time for the ceremony is superfluous after sume ustite antermenzaru gersiaru, if not actually contradictory, whereas it is quite acceptable if a distinction is
178
III. TEXT AND TRANSLATION
fagiu : affertur : avis : / (17) anzeriates : menzne : kurglasiu : iagia : tigit : huntia : fertu / (18) katlu : arvia : struhgla : fikla : pune : vinu : salu : maletu : / (19) mantrahklu : veskla : snata : asnata : umen : fertu : pir : ase : / (20) antentu : esunu : puni : feitu :
moons. If he desires to perform it, the adfertor is required, after having ob- served the birds, to perform it at the full moon. He shall bring the things pertaining to the Hondia; he shall bring the dog, grain, a strusla cake, a ficla cake, mead, wine, ground salt, a maniple, wet and dry vessels, and unguent. He shall place fire on the altar; he shall perform the sacrifice with mead. He
made between a sacrifice performed by the adfertor and by some other person. It is not necessary to assume that there were two types of sacrifice, one pubhc and one private, for the ceremony is des- cribed throughout as one, performed for the gens Petronia. The explanation is rather, as Vet., p. 195, suggests, that sometimes the adfertor himself was a member of the gens, in which case a special day was prescribed for the sacri- fice and the taking of the auspices had to be included. — 17 menzne kur- glasiu: menzne as a nasal extension of the same stem which is disguised in /^r;v, L. mensis, can be 'month' (so Bii., PL, Bk.) or 'moon' (so Blum., p. 70, Dev., Vet., Pis., Bott.). In favor of the former is the distribution of meanings in L. mensis and tuna, and also the reasonable- ness of taking kurglasiu as the name of a month, like L. lanuarius, Februarius (in the present case 'that which com- pletes the chcle, ' in other words the last month of the year). But it is natural that the precise date should be named, as Blum, and Dev. observed, and also if the meaning is ' in the month Gircularius, ' we should expect the loc. case (cf. 145 c, 146 b). If we interpret ' when the moon is full, at the full moon, ' menzne kur- ylasiu becomes an ablative absolute or ablative of attendant circumstances and so is quite natural. The use of the name of a round object to designate the fulness of the moon receives some support from Ov., M., VII, 530-1 Dumque quaier iunc-
tis explevit cornibus orbem luna, quater plenum tenuata retezuit orbem. — ti- git: equated with L. decet by nearly all editors except Vetter, who connects with L. licet; the sense is equally good, but for d from initial I before front vowels there is no reliable evidence positive or ne- gative. — Huntia: see on 15, where it was suggested that Huntia in the present passage may refer to the things needed for the sacrifice to Hondus. — fertu: the repetition of fertu again in 19 is a difficulty, but the second fertu may have been placed at the end simply because of the extreme length of the series. — 18 arvia: this passage is of crucial impor- tance for the interpretation of all the oc- currences of arvia aruio. Hu., Blum., Ribezzo, I, pp. 83-4, Vet., p. 195, trans- late ' exta,' in general comparing L. ar- vina fat, lard.' Bu., Co., PL, Bk., Dev., Pis., Bott. on the other hand connect with L. arvum and interpret as grain or similar products of the fields, Bii., p. 62, having pointed out that arvia in the present passage cannot be a part of the victim; so also Br., pp. 101-2, although he translates 'ollas' in the formula aruio fetu, arvia ustentu, and 'extis' in are- pes arves. Assuming that the victim has not been previously slain, I regard the argument of the second group of editors as cogent and take arvia aruio to refer to grain offerings. Moreover there ap- pears to be a sort of supplementary re- lation between the arves of la, b and the cakes prescribed in the corresponding por-
Ila 16
COMMENTARY
179
tions of VI and VII (see on la 3). — struhgla: see on Via 59. — fikla: on Via 56. — salu: the initial s here has the form M, as also in seritu in 24 below. For the use of salt in religious ritual of. Ov., F., I, 337-8 Ante deos homini quod conciliare valeret, far erat et puri lucida mica salts; Plin., N.H., XXXI, 89 ma- xtme tamen tn sacrts tntelltgttur auctort- tas (so. salts) qiiando nulla confictuntur stne mola salsa. Saltworks are known to have existed at Veil, Volaterrae, and in the Vestinian country; see R.-E., 2te Rethe, II, pp. 2078-9. — maletu: the only instance of the simplex of this verb in the Italic dialects. Devoto takes it as imv. of a second-conjugation stem in contrast to the third-conjugation forms kumaltu, kumates and by so breaking the series into two provides full justifica- tion for the second fertu in 19. It is easier, however, not to assume a change of conjugational class, and a pcpl. in -etu is quite natural in view of f rose torn, pesetom, etc. The syncope of the post- radical vowel in kumates comatir may be connected in some way with different accentual conditions in the simplex and in the compound. — 19 mantrahklu: see on VIb 4. — veskla snata asnata: also in 34, 37 below and in IV 9. The two adjectives are clearly of opposite mean- ing and are generally derived from the root snd-: Skt. sndtt 'bathe,' L. ndre, so that their sense should be 'wet and dry' or 'washed and unwashed'; but it is not clear how the vessels are to be used. Dev., p. 325, believes that the two classes of vessels are those which have been used for the unguent (umen) and those which have not, and it may be significant that Ila and IV, the only tables which mention veskla snata as- nata, are also the only ones referring to umen and containing the verb umtu. It is safest, however, simply to translate 'wet and dry,' understanding the snata to be used for liquids or for unguents and the asnata for grain or other substances which do not wet the vessels. —
umen : only in the present table, while the related verb umtu is found in 38 and in IV 13. On etymological grounds the sense is reasonably certain, but the exact nature of the unguent is of course un- known. 19-20 pir ase / antentu: cf. Ill 21-2 vuke pir / ase antentu. — 20 esunu puni feitu: since mead and wine are both called for in 24-5, the use of mead alone here may pertain simply to the opening part of the ceremony, the placing of the fire on the altar. — Hunte luvie: the present table is the only one containing a description of a sacrifice to this god, but Hontus Serfius is mentioned in VIb 45 = lb 4 as the recipient of three bull-calves in the sa- crifice at the Grove of Coredius during the purification of the Fisian Mount. That Hondus was an infernal deity is clear not only from his name, the connection of which with hondra, hondomu, imv. hondu, is fairly sure despite the uncertainty of the suffix, but also because of the literary evidence for the association of the dog- sacrifice with the cult of the underworld and of the dead; see above on 15 katle. — ampentu : practically all editors treat as etymologically equivalent to L. tm- pendtto and translate accordingly, but the exact meaning is not clear. In all passages where forms of this verb occur a word designating some sacrificial vic- tim is present or easily supplied. At first glance 'offer up ' might be a suitable enough translation, and the sense would not be altogether different from that of impendere in Lucan, II, 382-3 patrtaque impendere vttam / nee sibi; Petr., I, 1 "hunc oculum pro vobis impendi". But ampentu indicates a ritual act quite dif- ferent from purtuvitu and performed prior to it, as appears plainly from Va 17-22 ape apelust, muneklu habia nu- mer / prever pusti kastruvuf, et ape purtitu / fust, muneklu habia numer tupler / pusti kastruvu, et ape subra spafu fust, / miineklu habia numer tripler pusti / kastruvu. There is no verb in the Iguvine Tables which un-
180
III. TEXT AND TRANSLATION
hunte : iuvie : ampentu : katlu : /
(21) sakre : sevakne : petruniaper : natine : fratru : atiiefiu : esunu /
(22) perae : futu : katles : supa hahtu : siifafiaf : supaf : hahtu / (23) berus : aplenies : prusegia kartu : krematra : aplenia
shall slay in honor of Hondus Jovius an unblemished dog from the temple- property for the gens Petronia among the Atiedian Brothers. The sacrifice shall be (performed) upon the ground. He shall take the under-parts of the dog; he shall take the lower limbs. He shall roast the sacrificial pieces on a suf- ficient number of spits; he shall place a sufficient number of roasting-pans under
mistakably signifies 'slaughter,' yet it is probable that ampentu implies the slaughter of the victim, whicli must pre- cede the porrectio; cf. on the one hand the figurative expression inter caesa et por- recta in Gic. Att., V, 18, 1, and on the other the imphcation of slaughter in Sen., Tro., 298-9 Quis iste mos est, quando in inferias homo est impensus hominis (of the sacrifice of Polyxena). As L. immo- lare ' sprinlvle with salted meal ' is euphe- mistically employed for ' slay, ' so may ampentu be used in order to avoid direct use of a verb 'slay'; cf. Dev., pp. 326-7. It is not necessary in the passages where forms of ampentu occur to assume that the slaughter was accomplished by hang- ing or strangulation. — 21 sevakne: etymology and sense are both very un- certain. The word is often appUed to victims: the dog here, the sheep in III 26, pig lib 8, goat 10, heifers Vllb 1. But it is also appHed to objects of various kinds: unguent Ila 38, wine 39, roasting- spits Ila 36, vessels 37, IV 9, 25, and to the cakes called arglataf IV 22. What- ever its origin, it must have acquired ap- proximately the sense ' fit for ritual use, ' which may be modified to ' solemn ' or 'perfect' according to the context. It is not fully synonymous with sakre, which is used to distinguish that which is temple-property from that which is sec- ular (see on Ila 5) nor with arsmatia, which is applied to the wand carried by the priest. According to the usual view sevakni is a compound of sev- 'whole'
(cf. seuom, seueir, O. siuom 'omnino') + akni- (cf. perakni, which however is otherwise explained in note on Ila 5). sevakni is then analogous in formation to L. sollemnis, for whose second member Ernout-Meillet ^ suggest connection with O. amnud, while Pisani, p. 187, proposes to derive the second member of sevakni from the same root as angif, L. ancus, uncus, ciyxog, ayytwv, etc. The semantic evolution then would be 'all the way around > complete > perfect, ' and the association of both sollemnis and sevakni with the words for 'year,' O. akenei, U. acnu, L. annus, may be secondary. Dev., p. 307, explains sevakni quite dif- ferently as 'sine vitio, ' so that the prefix is the same as in seipodruhpei, while vak- is the same root as in ander-uacose, uasetom: no less plausible than the view of the majority. — Petruniaper na- tine fratru Atiiefiu: cf. 35 below, where a libation is offered to Hondus Jovius at the spina on behalf of the gens Petronia, and b 26, where a bull-calf is offered to Jupiter Sancius on behalf of the gens Lucia, natine is the formal equivalent of L. natione (for the different vowel- grade see 81 b), but the sense is rather that of the more remote cognate gente. The relation of the gentes Petronia and Lucia to the organization of the Atiedian Bro- therhood as a whole is not clear, but it is to be noted that in the early history of Rome particular cults were assigned to particular gentes, e.g., Sol to the Aurelii, Hercules to the Potitii and Pinarii, etc.,
IIa20
COMMENTARY
181
but were later transferred to the priestly colleges. Bii., p. 131, suggests that the gens Petronia may have become extinct, the implication apparently being that the brothers representing other gentes as- sumed the obligation which it was no longer able to fulfill. — 22 supa: see on VIb 5 sopo. supa hahtu occurs only here, but supa sumtu la 9, 16 must be scarcely different in meaning. The next reference to supa is in 30-31 supa span- tea / pertentu, while 32 includes the in- struction supa pustra perstu ' place the under-parts at the back (of the altar?).' Similarly in the description of the sacri- fice behind the Tesenacan Gate, VIb 3 has persae fetu corresponding to la 16 supa sumtu, while VIb 5 contains the clause ape sopo posfro peperscusL — su- fafiaf : found again in gen. sg. in 41, where it is to be ground up together with the cakes known as struhgla and Sikla. It is therefore probably some bony portion of the victim. The Latin passages com- monly cited in explanation of sufafiaf are: Plant., M.G., 1180 id conexum in umero laevo, exfafillato bracchio (sc. pal- liolum; the reference is to a seaman's dress, and the medial / is a mark of dia- lectal origin, but the reading is not cer- tain); Fest., 199 Li. exerfum, quod scilicet omnes exerto brachio sint exfilati, i.e. extra vestimentum filo contextum; Ps.-Plac. (Gloss. Lat. IV, ed. Lindsay) E 6 effafi- latus exserto humero id est extra filum manum proferens. If the root in sufa- fiaf (< *sub-fafians) is one of 'extend- ing out, ' as would appear from the Latin forms, sufafiaf may mean 'limbs,' and the following supaf may specify the 'lower limbs ' or ' lower parts of the legs. ' The view followed here is essentially that of Dev., p. 329, but nothing can be affir- med with certainty. — 23 berus: re- cognized as the etymological equivalent of L. verubus by Bii., p. 132, whom all subsequent editors follow. — aplenies: found only here and in krematra aple- nia just below. Most translate ' impletis, ' deriving from *am-ple-n(i)yo-, but the
formation is an improbable one, for we should rather expect a stem in either -to- or -no-, and in fact we have in Vila 21, 34 a form plener = L. plenis. Vetter, p. 196, obtains better sense from the two passages where aplenies, -ia occur, by taking as an extension like that in se- memes from a stem equivalent to that of L. amplus. The spits then must be in sufficient quantity, and so must be the roasting-pans which are to be placed below. An alternative possibility is that aplenies may be a gentile name and that the instructions call for a particular type of spits and roasting-pans known by the name of the manufacturers. Schulze, Lat. Eigenn., pp. 152, 346, cites several similar names, of which the most suitable for our purpose are Abulenius (C.I.L., XI, 6056, from Urvinum Mataurense, Augustan period), Abulenus (X, 3766, from Suessula). — prusegia: almost certainly to be equated with L. prosi- ciae. It is not clear just how prusegia differs from proseseto, etc. (L. prosecta), which is used frequently in describing the sacrifices of the purification and lus- tration in VI and VII. The difference, strictly speaking, may have been be- tw^een what is intended for cutting off and what has actually been cut off (other exemples of -r/o-stem adjectives with quasi-gerundive value are L. eximius, in- ferius, ayiOQ, acpdyioq), but there is no basis for such a distinction so far as the present passage is concerned. Cf. the mixture of forms in Non., p. 330 (L. Miiller) Aemilius Macer in Ornifhogonia: Vulcani tosta va.pore / cum virgis prosecta ferunt. Feminino Lucilius lib. XIII: ce- nam, inquit, nullam neque divo prosiciam ullam. Varro Rerum divinarum lib. XI: prosiciem extorum vel in mensam porri- gere. — kartu : generally translated ' di- vidito' on the assumption that it is con- nected with karu 'piece. ' Vet., pp. 196- 7, equates with L. carpito, while Pis., p. 192, translates ' torreto,' with Lith. ku- riii, OCS kurenje, OHG herd, and also L. carbo as support for the sense 'roast';
182
III. TEXT AND TRANSLATION
sutent/u : pefu : seritu : arvia : puni : purtuvitu : vestikatu : ahtrepufa/tu : pustin : angif : vinu : nuvis : ahtrepufatu : tin : puni : tiu : vinu : / (26) teitu : berva : frehtef : fertu : pure : nuvime : ferest : krematruf : / (27) sumel : fertu : vestigia : pefume : persnihmu : katles : tuva : tefra / (28) terti : erus : prusekatu : isunt : krematru prusektu : struhgla : / (29) fikla : afveitu : katlu : purtuvitu amperia : persnihmu : asegeta : /
(the spits). He shall watch the ground. He shall present grain-offerings with mead, pour a libation, and dance the tripudium, offering wine at each turn- ing. Nine times he shall dance the tri- pudium. "(I invoke) thee with mead, (I invoke) thee with wine," he shall say. He shall take the spits and the boiled portions. That which he carries last (with it) he shall carry the fat-drippings at the same time. He shall pray with a libation upon the ground. He shall cut off two (pieces) of the dog as burnt- offerings, and a third as an erus. At the same spot he shall cut the fat-drip- pings. He shall add a strusla cake and a ficla cake. He shall make the presen- tation of the dog. He shall pray with the portion around the (dog's) foot; he
of. also Muller, pp. 74-5. There is no sure solution, but Pisani's interpretation gives perhaps the most satisfactory sense. — krematra: it is plain from sutentu that krematra designates something to be placed under the spits on which the pieces of meat are roasted, and the most obvious connection Is with L. cremare. Most editors translate etymologlcally '*cre- matra,' with varying explanations, but the most probable view is that the kre- matra were roasting pans designed to catch the dripping fat. See also on kre- matruf In 26. — sutentu: clearly 'place underneath (supponito) ' ; cf. andendu, endendu, ostendu, all from *-tendet6d. — 24 pefu seritu: not the pefu called for In 9, which Is a portion of a different ceremony. It Is possible that pefu In the present passage means simply 'ground' and not 'mound,' the meaning adopted in the present edition for most passages (see on VIb 24). The adfertor Is apparently expected to keep watch over the place where the meat is roasted in the fire. In order to see that everything takes place
according to rule. Dev., p. 333, notes the durative aspect of the uncompounded seritu in contrast to anzeriatu; the ef- fect of the prefix might be roughly com- pared with that In Germ, erblicken. — arvia pvuii purtuvitu: the only passage where arvia Is used as object of purtu- vitu, although In IV 22 the same verb governs arglataf, a type of cakes pro- bably of circular form. On the possible relation between arvia and offerings of cakes, see on la 3. — vestikatu : see on VIb 6. — ahtrepufatu : on VIb 16. — 25 pustin angif vinu: nearly all editors place a period after vinu, so that the three words belong in the sentence with vestikatu ahtrepufatu. The phrases pusti kastruvu(f) Va 13, 18, 20, 21, posti acnu Vb 8, 12, 14, 17 have distribu- tive force, and the same force should be admitted for pustin in the present pas- sage, angif Is generally derived from *anki-ns with the same root as In L. uncus, angulus, dyxvXoq, dyxcov, In all of which the fundamental sense Is 'bent, angular. ' The apparent sense then is
Ila 23
COMMENTARY
183
that the priest offers a. libation of wine at each turning in the tripudium. — nuvis ahtrepuf atu : for the number nine in ritual usage cf. Ov., F., V, 435-9 (of the Lemuria) Cumque manus puras fonta- na perluit unda, vertitur et nigras accipit ore fabas, aversusque iacit, sed dum iacit, ' haec ego mitto, his' inquit ' redimo meque meosque fabis. ' Hoc novies dicit, nee res- pieit; Frazer, The Golden Bough, I (Lon- don, 1913), p. 193, mentions the custom in the French Pyrenees of leaping nine times over the midsummer fire in order to be assured of prosperity; cf. the index in Vol. II for further magical associations of the number nine. — tiu puni tiu vinu: for the ellipsis of the verb see 136 c. — 26 frehtef : IV 30-31 contains the instruction pune frehtu habetu, but the sense in both passages is very ob- scure. Formally frehtef can be ace. pi. of a ti-stem, or, with, much less probability, nom. sg. of a pres. pcpl. hke kutef. Sev- eral editors (Bii., Ribezzo, II, pp. 88, 100, Bott.) assume connection with L. frigus, frigeo, etc. : the hot spits are cooled in Ila 26, and the mead is to be cool in IV 31. In this case frehtef must be a pres. pcpl. equivalent in sense to frigidans or refrigerans, but this is troublesome from the viewpoint of morphology, for it is dif- ficult to imagine a denominative verb in the third conjugation, or in the second conjugation with transitive sense (kutef '*cautens' is intransitive), and moreover we should expect Umbrian, like Latin, to use the pf. pcpl. of an action which pre- cedes the main action. The most satis- factory interpretation is probably that of Dev., p. 335, who suggests connection with Skt. bhurajanta 'roast,' while as- suming that in Umbrian the root *freg- replaced the verb represented by L. co- quo in the meaning ' boil. ' The advan- tages of taking frehtef as 'boiled por- tions ' are that it removes the danger of a rivalry in meaning with krematru(f) and that the sense ' roast ' would not suit pune frehtu in IV 31. — pure: the pronominal form pure porse is nowhere
used as a conjuction 'cum' and must here be regarded as a relative without antecedent; cf. Cato ap. Diom. (Gramm. Lat., I, p. 362 Keil) lepus multum somni adfert qui eum edit. — nuvime : taken by nearly all as an adverb from the or- dinal ' ninth, ' but by Vet. as superlative to *nowo- (cf. L. novus) in the sense 'an letzter Stelle, ' on the ground that No- niar Via 14 shows the true ordinal stem 'ninth' and that the carrying of the spits and frehtef is a separate act from the ninefold tripodatio. On the whole his view appears preferable. — ferest: the spits, the frehtef, and finally the kre- matruf are apparently carried from the fire to the altar. — krematruf: since krematru in 28 cannot possibly be 'roasting-pans, ' we may assume that the difference in ending between krematru(f) and krematra corres- ponds to a difference of sense, kre- matru(f) according to Vetter refers to the fat which falls into the pans, then congeals and is cut up; this is better than to refer it to the pieces of roasted meat, which have already been accounted for in the instruction berva ... fertu. — 27 vestigia: see on Vib 5. — tefra: on Vila 46, where, however, tefru-to de- signates the place of burning; here the tefra are the burnt-offerings themselves. Similar offerings are cut from the sheep in III 32, 34, IV 2. — 28 prusektu: not a pcpl., for the normal form of the pcpl. is represented by prusegetu in lib 12, and moreover if krematru and pru- sektu were both accusative they would have to depend on the preceding pruse- katu, giving a tautology, prusektu may represent an alternate stem (cf. L. se- catus: sectus, lavatus: lautus) but is more probably simply an error for prusekatu. — 29 ampef ia : many different solutions have been attempted. On phonological grounds the analysis as from *am(J)-ped- id(d) 'part around the foot' (Bk., Pis.) is free from objection. If this view is cor- rect, a certain portion including a foot of the victim has been cut off and the ad-
184
III. TEXT AND TRANSLATION
(30) karne : persnihmu : venpersuntra : persnihmu : supa spantea : / (31) pertentu : veskles vufetes : persnihniti : vestikatu : ahtrepufatu : / (32) afpeltu
shall pray with the uncut portion; he shall pray with the vempersondro ; he shall place the under-parts over at the side ; he shall pray with vessels of which- ever type he desires. He shall pour a libation, dance the tripudium, approach
fertor holds it or keeps it before him as he prays. Although Biicheler's translation ' pedem servato' for pefura seritu in 24 has not been accepted by later editors, the passages which he cites in support of it are relevant to the present passage: Varro, L.L., V, 23 qui Romanus combustus est, <si> in sepulcrum eius abiecfa gleba non est aut si os exceptum est morfui ad familiam purgandam, donee in purgando humo est opertum ... familia funesta ma- net; Fest., p. 148 Li. membrum abscidi morfuo dicebatur cum digitus eius decide- batur, ad quod servatum iusta fierent reli- quo corpore combusto; ' digito hominis ca- tuli respondet petiolus' (Bii., p. 133). — asegeta: taken by all recent editors as 'non(dum) sec fa.' Apart from the nega- tive prefix the participial stem is the same as in prusegetu. — 30 venper- suntra: of the three other passages where forms of this word occur, lib 18-19 ve- pesutra pesnimu, veskles pesnimu, atrepufatu / afpeltu statitatu is most similar to the present passage. But ven- persuntra cannot be any clearer than persuntru, to which it is obviously re- lated; see on VIb 24, where it is suggested that ven- may reinforce rather than nullify the meaning adopted for persuntru. Ac- cording to Ribezzo II, p. 81, the n in the prefix may have resulted from the assi- milatory influence of the penultimate syl- lable. — 30-31 supa spantea / per- tentu : the number and case of supa, the form and construction of spantea, and the meaning of pertentu are uncertain. Whether supa is ace. sg. fem. or ace. pi. neut. makes little difference for the sense. It is unlikely that it is abl. sg., since it is
improbable that pertentu was normally intransitive, despite IV 8 supes sanes pertentu. spantea from the viewpoint of syntax would be best taken with Vet. as *spante(m)-a(d) 'auf die Seite, ' but there are 13 instances of ace. sg. -m pre- served and none lost before postpositive -a(f). spantea must then be taken as a predicate adj. lil^e pustra in 32, cf. b 19. Just as L. profendo may mean 'extend (the hand)' or 'extend (the object in the hand),' so pertentu here means 'place the under-parts at the side (of the altar, with outstretched hands)' and in IV 8 'reach out with ..." - — 31 veskles vu- fetes persnilimu: the majority of edi- tors derive vufetes from *wog^heto- (cf. Evxof-iai, L. voveo) and translate 'votis' or ' consecratis,' but this does not give a very satisfactory sense, especially if we take account of the other passages in which the expression veskles persnih- mu occurs. Dev. translated ' vacuis, ' that is, not containing the viscera, but gave no etymology to support his interpreta- tion. I propose to equate vufetes with L. libita, pf. pass. pcpl. of libet, lubet, with phonology as in 14, 40, 55 a, and participial stem in -eto corresponding to L. libita from libet as tagez to L. tacitus from taceo. The sense intended may be seen from a comparison with other re- levant passages. In 19 the list of articles to be brought to tlie place of sacrifice included veskla snata asnata, and in 34 these wet and dry vessels, with otlier things, are to be carried to the spina. In 37 the priest is directed to pray at the spina with the wet and dry vessels, ves- kles being this time further accompa-
Ila 29
COMMENTARY
185
nied by the adjective sevaknis. In IV 9 a prayer is offered to Pomonus Poplicus and Vesona of Pomonus Poplicus with the wet and dry vessels, but in 24 to Ve- sona of Pomonus Poplicus with veskles vufetes. In both passages in IV sevak- iie(-is) accompanies veskles, as in Ila 37. The significant contrast appears to be between snates asnates and vufetes; in other words, sometimes the adfertor when he prays must hold or have before him both the dry and the wet vessels, while at other times he may use whichever vessels he wishes. - — 32 af peltu : usual- ly taken as equivalent to L. adpellifo and translated 'adpellifo' or ' admoveto,' but the sense here as well as in b 19, IV 8 calls for an intransitive verb, and Devoto's derivation from the root *k'^el- with neXojJiai, L. colo, which is adopted also by Pisani and Bott., is decidedly preferable. Dev. translates ' circumito,' the idea of circular motion being often inherent in derivatives of this root, and he takes af- as a prefix of perfectivation,but it is more probable that the verb merely signifies the priest's approach toward the altar after dancing the fripudiam. — statitatu: regularly found after arpeltu. Its mean- ing 'halt, stop' is beyond reasonable doubt. — supa pustra perstu : the un- der-parts are here placed on the back part of the altar. Cf. VIb 5, Vila 8, where ape sopo postro peperscust introduces the instructions for libations and supplemen- tary offerings after the main sacrifices behind the Tesenacan Gate and at Ru- binia. — iepru: by some taken as abl. pi. of a demonstrative pronoun + -pru, by others as cognate with ^tzgq, L. iecur. By the former view the pronoun before the enclitic should belong to a stem cor- responding to L, is, ea, id and should be abl. pi., referrring to supa, but the loss of -s before the p would be without pre- cedent, and no other analysis starting from a pronominal stem is entirely satis- factory either for form or for sense. The equation with L. iecur, which seems deci- dedly preferable, was advocated by Pan-
zerbieter, Quaestiones Umbricae (Progr. Meiningen, 1851), p. 12; Savelsberg, K.Z., XXI (1873), p. 213; with especially de- tailed argument and documentation by R. G. Kent, C.P., XV (1929), pp. 367-9; more recently adopted by Pis. and Bott. If we adopt this interpretation, the form iepru is to be derived from *yek^rom either as neut. with the original r-stem extended into an o-stem, or, with change of gender, as masc. r-stem with the usual third-declension ace. sg. ending in -om. That it was written by mistake in place of *iepur, as Muller, p. 218, suggests, is most unlikely. With it may be compared piirom-e VIb 17, Vila 38. So far as the sense is concerned, the liver is here iden- tified as the erus. — mani: purposely inserted, according to Dev., to make it clear that on this occasion the prohibi- tion against touching certain consecrated objects with the hand did not hold good. — kuveitu: the usual verb governing erus is dirstu teftu, although Ila 40 be- low has erus kuveitu tertu. Strictly speaking the actions indicated by -veitu and teftu are not the same: the former is the carrying or conveying of the object to its destination, the latter the placing of it in the possession of those who are to receive it, but the prefix ku- affects the aspect of the verb in such a way as to make it almost synonymous with teftu. Ila 33 - 39. Ceremonies at the Obe- lisk. At this point the adfertor moves to a spot designated by an object called spina, to return to the altar again in 39. The objects which are to be taken to this spot include two bowls of mead, spits, smearing-sticks, wet and dry vessels, and unguent. A libation is offered to Hondus Jovius, prayers are said in conjunction with the spits, with the smearing-sticks, and with the wet and dry vessels. Then follow another libation, the dancing of the iripudium, the anointing of the obe- lisk, a prayer with the unguent, and finally the washing of the hands. — 33 spinamaf : it is of course not clear precisely what object is designated by
186
III. TEXT AND TRANSLATION
statitatu : supa : pustra : perstu : iepru : erus : mani : kuveitu / (33) spinamaf : etu : tuve : re kapifus : pune : fertu : berva : klavlaf : a/anfehtaf : vesklu : snatu asnatu : umen : fertu : kapire : hunte : / (35) iuvie vestikatu : petruniapert : natine : fratru : atiieriu : berus : / (36) sevaknis : persnihmu : pert : spinia : isunt : klavles : persnihmu : / (37) veskles : snate : asnates : sevaknis : spiniama : persnihmu : vestikatu : / (38) ahtrepufatu : spina : umtu : umne : sevakni :
(the altar), and stop. He shall place the under-parts at the back. The liver, (which is) the erus, he shall carry in his hand. He shall go to the ObeUsk. He shall bring mead in two bowls; he shall bring the spits, the smearing-sticks not yet used, the vessels wet and dry, and the unguent. He shall pour a li- bation from a bowl to Hondus Jovius for the gens Petronia among the Atie- dian Brothers. With the ceremonial spits he shall pray on the other side of the Obehsk. At the same spot he shall pray with the smearing-sticks. With the ceremonial vessels wet and dry he shall pray at the Obehsk. He shall pour a libation and dance the tripudium. He shall anoint the Obehsk, pray with the
33 tuve : re : for tuvere. petriiniaper.
33-4 a/anfehtaf: for anfehtaf.
35 petruniapert: for
this word, found only here and in 36, 37, 38 below, but there is no reason to ques- tion the general view that it v/as some sort of column or other monument whose shape had some resemblance to a thorn. Although this particular semantic de- velopment is not attested for L. spina, it offers no difficulty, and it is not out of place to mention the spina in the Roman Circus, a low wall or fence extending longitudinally down the middle, with the track running around it (Cassiod., Var., Ill, 51; Schol. Juv., VI, 588; cf. Bii., p. 136). Compare also "Cleopatra's Needle." • — tuvere kapifus: one bowl is to be used in the libation prescribed below in 34-5, and the other for the purpose of extinguishing the altar-fire (kapire pu- nes vepuratu) in 41; so Dev., but Vet. understands the two bowls in relation to the libations called for by vestikatu in 35 and 37. — klavlaf: the abl. pi. klav- les occurs in 36 and in IV 11, in both instances followed by persnihinu. There are two possible interpretations. Ear-
ner editors, and more recently Vetter, understand in the sense of ' dunes, ' sup- porting their view in part by the fact that Germ. Keule, the semantic equivalent of L. clava, has the additional meaning 'leg, thigh ' (of mutton or veal). The haunches of the dog according to this view were therefore included among the articles to be brought for the ceremony at the obe- hsk. Devoto, however, translates coch- learia, ' taking the klavlaf to be spoons used in applying the unguent to the obehsk (in IV 11-13 to the ereglu or statue), and his view is adopted by Pis. and Bott., although they use 'clauulas' in their translation. Both in I la 35-8 and in IV 9-13 the other objects kept at hand during the prayers are parts of the ceremonial equipment and not of the sacrificial victims : in the former passage the spits, then the klavles, then the wet and dry vessels, in the latter the wet and dry vessels, then the klavles; and in each passage the prayers are followed by the anointing of the obehsk or of the sta-
Ila 33
COMMENTARV
187
tue. Devoto's interpretation therefore seems decidedly preferable, and is adopted here. It is possible, however, that the klavlaf were not actually spoons but mere- ly pieces of wood, club-shaped or with flattened ends, for smearing the unguent on the object to be anointed. The rare L. clavula gives no direct support for either ' haunches ' or ' spoons ' as the meaning of klavlaf, but ' smearing-sticks ' is closer both to the sense which we should expect in a diminutive of *kldva and to the actual use of clavula as a synonym of talea in Varro, R.R., I, 40, 4 with refe- rence to olive-cuttings. — 33-4 a/an- fehtaf : the doubling of the a may repre- sent vowel-lenthening before nf, but it is much more probable that the engraver simply recommenced the word on a new line after realizing that there was not sufficient space on the original line. Near- ly all editors analyze as a compound of negative an- + pcpl. of fagia or fetu. Those who take klavlaf as 'dunes' then understand aanfehtaf to mean 'uncooked.' If klavlaf is ' smearing-sticks ' aanf ehtcif is somewhat more difficult, but Dev. makes satisfactory sense by suggesting ' nondum adhibitum, nondum oblatum' as developments from the literal sense ' non factum. ' The spits, the vessels, and the unguent must be sevakni-; for the smear- ing-sticks it is sufficient that they should not have been used previously. — 36 pert spinia: cf. the Oscan usage in Cip- pus Abellanus (Co. 95), B 33 pert viara 'across the road. ' On the orthographical variation spina: spinia see 53 d. — 37 spiniama: 'on the hither side of the obelisk. ' -a here contrasts with -pert in 36. — 38 spina umtu: parallels to this curious ceremony are moderately abundant. The anointing of the ereglu or statue in IV 13 has already been men- tioned. The Acta fratrum arvalium con- tain, among others, the following referen- ces to the anointing of cult-images : pp. CLVII (Henzen) and CLXXVII signis unctis ; XXIV deas unguentaverunt ; CC XXV deam unguentaver{unt). Pausanias,
X, 24, 6 in reference to a stone near the tomb of Neoptolemus at Delphi says : ijiavafidvTi de anb xov ixvr}fj,aroQ XiQoq sariv ov fidyai; ' rovrov xal eXaiov oarj/ueQai xaraxeovai xal Kara ioQrrjv exdarrjv eoia enixideaai ret ayqa. In Vedic ritual the sacrificial post called the yupah was anointed with butter; cf. Satapathabrah- mana, III, 7, 1, 10-13; Apastamba's Srautasutra, VII, 10, 1-5 (Caland); J. Schwab, Das aliindische Tieropfer (Er- langen, 1886), pp. 68-70. In Genesis, XXVIII, 18, Jacob took the stone on which he had rested his head while he slept, and set it up for a pillar and poured oil over it. — 38-9 manf easa / vutu: the interpretation of the passage in gen- eral receives support from the instruc- tion manus interluito in Cato, R.R., 132, between the prayers with the food and the wine in the offerings to Jupiter Dapalis. The precise meaning of easa, however, is not quite certain. According to Bii. the connection between this passage and what follows is so close that the instruction to wash the hands from the altar is given before the instruction to return from the obelisk to the altar. It seems safer, how- ever, to take the two events in the se- quence in which they are presented, so that easa, ' ex ara, ' means ' away (at some distance) from the altar' (so Dev., who discusses both views).
Ila 39 -43. Ceremonies after the re- turn to the altar. The adfertor prays si- lently with ceremonial wine, the eras is distributed, and also wine and mead; a portion of the cakes and of the sufafia are ground; the fire is extinguished with a bowl of mead; the adfertor prays with the whole and ground cakes, arises and removes whatever is left on the altar. The ceremony is then completed and the remains of the dog are buried at the altar. — 39 asaku : -a(f ) indicates mo- tion, -ku(m) position, although -a in 37 means ' cis, ' contrasting with pert 36 (cf. Dev., p. 347). — 40 esuf pusme herter : the value of esuf as the equivalent of L. ipse receives its best support from
188
III. TEXT AND TRANSLATION
persnihmu : manf : easa : / (39) vutu : asama : kuvertu : asaku : vinu : sevakni : tagez : persnihmu : / (40) esuf : pusme : herter : erus : kuveitu : tertu : vinu : pune : tef tu : / (41) struhglas : fiklas : sufafias : kumaltu : kapife : punes : vepuratu : / (42) antakres : kumates : persmhniu : amparihinu : statita : subahtu : esunu : / (43) purtitu : futu : katel : aseiku : pelsans : futu : / (44) kvestretie usage svesu vuegis ti teteies /
ceremonial unguent, and wash his hands away from the altar. He shall return to the altar; at the altar he shall pray silently with ceremonial wine. He whose duty it is shall carry and distribute the ems. He shall distribute the wine and mead. He shall grind (a portion) of the strusla cake, the ficla cake, and the limbs. He shall extinguish the fire with a bowl of mead. He shall pray with the whole and the ground (grains). He shall stand up and take away the things which remain. The sacrifice shall have been completed. The dog shall be bur- ied at the altar. Lucius Tetteius, son of Titus, approved (the foregoing) in his quaestorship.
42 persmhniu: for persnihmu, with confusion of KM and IH . 44 vuegis: for vuvgis.
O. esuf in Tabula Bantiiia, 19, 21, where it clearly indicates the individual in con- trast to his property and household, es- suf in Conway's Oscan inscription no. 169 is less helpful because of the mutilated state of the text, and the only other Umbrian example, esuf IV 15, is not de- cisive. For pusme there are no other U. or O. examples, but its clear etymological correspondence to Skt. kasmai, dat. sg. of the interrogative pronoun, places its value here as dat. of the relative beyond reasonable doubt, lierter III 1, herte Va 6, 8, 11, herti Vb 8, 11, 13, 16, regu- larly accompanied by subjunctive forms, is clearly a verb of obligation, similar to L. oportet. The adfertor has not been designated by name since 16, but it is to him that esuf must refer. We may pos- sibly have here another instance of eu- phemistic avoidance; see on Via 1 poei angla aseriato eest. — erus kuveitu tertu: see on 32 above. — 41 vepu- ratu: generally regarded as a denomina- tive verb from the stem ve-pur- 'fireless' seen in abl. pi. vepurus Va 11. For the sense cf. Verg., Aen., VI, 226-7 Postquam collapsi cineres et flamma quievit, reli-
qiiias vino et bibulam lavere favillam. — 42 antakres kumates persnihmu: see on lb 36-7. — amparihinu : beside this passive imv. form an active amparitu occurs in III 14 with ace. kletram as object. Nearly all editors derive from the root seen in L. paro, pario, reperio. From an-, am- (Gk. dva-) + *par- 'produce' are developed the active sense ' raise ' and the middle sense 'rise.' It has not been stated that the priest is to sit during the prayer immediately preceding, but the sitting position is prescribed in VIb 17, 41 = lb 33-4 for the grinding of the cakes and the prayers which follow. — statita: generally translated ' statuta,' whether the verb to which the pcpl. be- longs is denominative to * statu- or to *stati-; see 120 c. The correct sense ap- pears to be given by Dev., who trans- lates '[res] exstantes,' understanding it to refer to the remnants of the cakes, etc., left on the altar. ■ — subahtu: appar- ently the same verb as in the expression capirso subotu VIb 25, and generally ana- lyzed as sub- + hahtu. The force of sub- here is similar to that of L. de-; for -hahtu, the value of which is that of L. capio
Ila 40
COMMENTARY
189
rather than of hdbeo, see 122 e. The in- struction is to 'take down' or 'away' whatever remains on the altar. — 42-3 esunu / purtitu futu: the expression purtitu futu occurs in IV 31-2; cf. also VIb 42 purdito fust, Vila 45 enom pur- diiom fust = lb 38-9 enu esunu purtitu fust. — 43 katel asaku pelsans futu: on the arguments in favor of retaining the old interpretation 'bury' for pelsatu, pelsans see on VIb 22, where it is pointed out that the present passage and VIb 39-40 offer serious difficulties to any at- tempt to translate pelsatu by 'pulsato.'
Ila 44. Formula containing the name of the authenticating official. See on lb 45, which is substantially the same. The differences in the two versions are: 1) lb 45 shows word-division in kvestre: tie (erroneously) and before and after usaie, while Ila 44 shows no division at all; 2) lb 45 has usaie in contrast to Ila 44 usage, which is evidently the correct form; 3) la 45 has the praenomen Vuv- Qis, Ila 44 Vuegis; 4) la 45 has the full form of the father's praenomen Titis, Ila 44 the abbreviation Ti.
190
III. TEXT AND TRANSLATION
(lib 1) seme : nies : tekuries sim : kaprum : upetu : tekvias / (2) famefias : pumperias : xii
1 seme : nies: for semenies.
lib 1 - 21. Sacrifice of a pig and a goat at the decurial festival of Semo. The opening hnes present extraordinary difficulties, which involve, among other matters, the meaning of semenies; the meaning and interrelation of tekuries, tekvias, and pumperias, and their signi- ficance for the organization of the citi- zens of Iguvium; the reference of the numeral XII and its relation to what otherwise seems to be a decimal rather than a duodecimal system; the relation of Atiiefiate, Klaverniie, etc., and of the accompanying etre, to the organiza- tion indicated in the passage immediate- ly preceding. lib 1-6 and Vb 8-18 are partly interdependent for their interpre- tation: in both passages the instructions relate to the same festival, the sehmenier dequrier, and the names of two of the ten decuviae listed in lib reappear in Vb, where it is stated that the Glavernii are annually to present 4 p. of barley to the Atiedian Brotherhood and receive in re- turn 10 portions of pork and 5 of goat- meat, the Casilas to present 6 p. of barley and receive 15 portions of pork and 7^4 of goat-meat. It is this ratio of 2 : 3 taken in connection with lib 3 Klaverniie, etre Klaverniie ... 6 Kaselate, etre Kaselate, tertie Kaselate which has led to the notion of an expansion whereby the decuvia Casilas was split into three and all the others, including the Glavernii, into two, the Peraznanii alone remaining undivided. The interpretation adopted here is essentially that of Devoto, who with most editors begins a new sentence with tekvias and who arrives at the sum XII by adding together the ten decuviae and two famefias pumperias, that is two families, or groups similar to gentes, each representing a *pontis or pentad of decuviae. The principal bibliographical references for the opening lines of lib.
At the decurial festival of Semo choose a pig and a he-goat. The decuviae and the quincurial families are twelve. Say:
apart from the editions, are: Schulze, Lat. Eigenn., pp. 543-47; A. Rosenberg, Der Staat der alien Italiker, pp. 118-28; Devoto, Atti del reale istituto veneto di scienze, lettere ed arti, 1930, pp. 927-36; Ribezzo, II, p. 90; Pighi, Umbrica, pp. 20-21. The sacrifice of the pig and of the goat are alike in some respects and dif- ferent in others. Both sets of instructions call for the consecration of the victim (eveietu) and for a formal pronounce- ment as to its suitability for sacrifice, although naratu is used alone of the goat, without a preceding sevakne. Both sacri- fices call for incense, if that is the correct interpretation of vaputu. But the goat- sacrifice differs both in the nature of some of the instructions and in the amount of detail. It is to be performed partly out- side, partly in the temple; the persondro, the cake called mefa, and a libation are all specified; not only wine and mead but also water are to be carried to the temple for the second part of the goat-sacrifice, as are also fire, the maniple, and certain other objects of uncertain nature. The dancing of the tripudium is also prescribed. — 1 semenies tekuries : the prior word is translated ' semes tribus ' by Bii.; ' con- ciliis' by Dev., p. 301, connecting it with Skt. samanya-; 'singulis' by Vet., from *sem-, the IE root indicating iden- tity and individuality. Most others as- sign it to the family of L. semen, in the sense ' (feriis) sementivis' (Bk., Blum.) or 'Semoniis' (Pis., Bott., after Pighi, pp. 20-21 of op. cil. above in introductory remarks to 1-21). This last interpretation appears the most satisfactory in the light of two Latin inscriptions cited by Dev. as support for his interpretation of te- kuries: C.I.L., VI, 567 Semoni Sanco Deo Fidio ... decur(ia) bidentalis donum dedit (from Trastevere); 568 Sanco Sanc- to Semoni Deo Fidio sacrum decuria sacer-
ITb 1
COMMENTARY
191
dotum bidentalium (from the Quirinal). Cf. also Ov., F., VI, 213-8 cited below on VIb 3. The appearance of Sagi in 10, Sagi luvepatre in 17 of the present table, taken in connection with the passages just cited, gives further support to the notion of a festival in honor of Semo, or his Umbrian counterpart. For the ablaut- variation in L. Semo : U. sehmenier cf. L. Nerio: Nerienis, Anio: Anienis. te- kuries is in a formal sense equivalent to L. decuriis 'groups of ten'; yet the sense of the passage here seems to call for an adjective corresponding to tekvias, and the tekvias are not groups of ten but, at least originally, groups based on a divi- sion of the population into tenths; so much is probable from the ten names in the series commencing in 2 with Atiie- riate, and also from the formation of tekvias as an extension of an ordinal stem *deku-. But the multiplicative value which is assumed for tekvias on the ground of its formation may be preserved if we consider that ten tenths make up the whole. The ' decurial festival of Semo ' then is celebrated by the ten decuviae united into one body. — sizxi: in the sacrifices of the purification of the Fisian Mount three pregnant sows are offered behind the Trebulan Gate (la 7 = Via 58) and three sucking pigs behind the Tesenacan Gate (la 14 = VIb 3). — kaprum: the present sacrifice is the on- ly one in the Tables in which a goat is offered, but mention was made in the introductory remarks on 1-21 of the obli- gation of the Atiedian Brothers to fur- nish certain portions of goat-meat to the decuviae Clavernii and Casilas according to the contracts shown in Vb 8-18. — upetu: generally translated 'oplato' or 'deligito' on the basis of etymological connection with L. op tare. The verb re- curs in 8 with si perakne sevakne and in 11 with kapru perakne sevakne as object, and it is difficult to see why, if the victims have already been ' chosen, ' they should have to be ' chosen ' again. Similarly in III 9-10 the 'pentads' of the
Brothers are directed to 'choose' a pig and a sheep (sakre uvem urtas puntes fratrum upetuta), after which again there is an apparent repetition in 22 sakre sevakne upetu, 26 uvem sevakni upetu. Devoto, followed in substance by Vetter, sought to avoid the difficulty by taking upetu as the equivalent of L. ob-lto with the sense of ad-ito or capito. It is difficult, however, to see that much is gained, for the repetition of upetu in lib 8, 11, III 22, 26 is almost as awk- ward as before. It is possible that in these four passages upetu should be tak- en not as imv. but as ace. sg. masc. of the pass. pcpl. The participial stem is known from opeter Vb 12, 17, and the case-ending -u(m) receives support from kapru lib 10, vuku III 21. At the same time there is no serious difficulty in the fact that the concinnity in the series upetu eveietu (sevakne) naratu lib 8, 11 is destroyed, nor in the fact that the new interpretation makes it impossible to end the sentence with upetu in III 22, 26. — tekvias: usually translated 'decuriae, decuriales,' or 'decuviae,' the last being merely an adaptation of the Umbrian word. On the reasons for taking it to refer to divisions into tenths and not groups of ten see above on seme- nies tekuries. The punctuation of the majority of editors is followed here: the first sentence of lib ended with upetu, and tekvias commences a new sentence which ends with XII. Blum, and Pis. make a single sentence, with tekvias famefias as gen. sg. depend- ing on sim. kaprum upetu in the manner of Ar., Eq., 85 (niveiv) axga- rov olvov dyadov dai^ovog (cf. Blum., p. 40). — 2 famefias: here taken as nom. pi. in asyndetic connection with tekvias. It is a clear etymological equi- valent of L. familia and doubtless desig- nated a similar grouping, but whether it was related to *natiuf (abl. natine) pre- cisely as L. familia to gens is not certain. Dev. suggests that gens (U. *natiuf) may have belonged rather to cult, familia
192
III. TEXT AND TRANSLATION
atiiefiate : etre : atiiefiate : / (3) klaverniie : etre : klaverniie : kureiate : etre kureiate / (4) satanes : etre satane : peiefiate : etrep : eiefiate : talenate / (5) etre talenate : museiate : etre museiate : iuieskane : / (6) etre iuieskanes : kaselate : etre kaselate : tertie kaselate : / (7) peraznanie : teitu afmune : iuve : patre : fetu : si : pera : / (8) kne : sevakne : upetue :
"For the (decuvia) Atiedias and for the Second Atiedias, for the Clavernii and for the Second Clavernii, for the Cu- reias and for the Second Cureias, for the Satani and for the Second Satani, for the Peiersias and for the Second Peier- sias, for the Talenas and for the Second Talenas, for the Museias and for the Second Museias, for the Juvescani and for the Second Juvescani, for the Casilas, for the Second Casilas, and for the Third Casilas, and for the Perazna- nii." Sacrifice to Jupiter Arsmo. Con- secrate the pig brought from away and
4 etrep : eieriate: for etre : peiefiate. veietu: for upetu : eveietu.
7-8 pera : kne: for perakne. 8 upetu©:
rather to organization of the citizen-body. Blum, and Pis. understand it as a collec- tive designation of the slaves who are to be included with the decuvia as benefi- ciaries in the ceremony; this view has much in its favor in view of the use of famel in the Italic dialects: Paul, e Fest., p. 77 Li. Famuli origo ab Oscis dependet, apud quos serous famel nominabatur, unde et familia vocata; Paelignian no. 208 Co. famel inim loufir ' servus et liber'; it is not adopted here, however, because it conflicts with the interpretation of the passage as a whole. — pumperias : un- doubtedly derived from the numeral *pom- pe 'five.' Since the closest connection is with O. pumperias, pumperiais, we should expect r, and the f of the Umbrian form may represent an analogical altera- tion of the phrase *famefia pumperia or may simply be an error induced by famerias (cf. PL, II, p. 200). Oscan pumperias, -ais on several Capuan in- scriptions (Co. 105, 107, 114, 115, 116) is generally taken to designate the fifth day of the month (cf. L. nonae from the ordinal 'ninth'), and so Blum, and Pis. take it in our passage. The XII is then the designation of the month December
according to Blum., pp. 39-40, and the festival is similar to the Saturnalia. This interpretation, despite its ingenuity, has certain disadvantages. It is unlikely that the month would be designated by the numeral XII rather than by its name, and pumperias as a genitive of time, which is necessary for this interpretation, is less natural than the abl. pi. which appears in the Oscan inscriptions 114, 115, 116 cited above, pumperias is therefore to be taken as nom. pi. of an adj. agreeing with famerias, and related to the collective puntes as tekuries would be to the collective *dekenti-. A division of the Atiedian Brothers into groups of five is known from the use of puntis III 4, puntes 9, 10, and a similar division of the citizens of Iguvium into two groups of five decuviae each may be assumed, according to Dev., pp. 353-4, with each *ponlis represented by one family, as the puntes of the Brothers appear to have been represented by the gentes Petronia and Lucia. The ten de- cuviae and the two familae *quincuriae together make up the twelve indicated by XII. Whether the families changed per- iodically by a system of rotation or
lib 2
COMMENTARY
193
resembled certain Roman gentes in being permanently associated with a definite cult cannot be determined.
lib 2 - 7 Atiiefiate ... teitu: an ad- monition that every one of the decuviae for which the sacrifices are to be per- formed must be specifically named. Livy, XL I, 16, 1 records an instance in which the sacrifice at the Feriae Latinae was invalid because of failure to mention the Roman people: ... quia in una hostia magistralus Lanuvinus precalus non erat populo Romano Quiritium, religioni fuit. Cf. also Paul, e Fest., p. 59 Li. did mos erat Romanis in omnibus sacrificiis preci- busque 'populo Romano Quiritibusque.' That the dative should be used instead of the ablative -f -per is slightly sur- prising, but the Latin passages just cited (after Bii., p. 141) seem to give adequate support for the view adopted. For the names themselves no attempt is made here to establish connections with local or ethnic names known from other sour- ces, whether Latin, dialectal, or modern Italian. However, certain observations may be made. In the first place the ten names are not all of the same type: six are singular forms similar to L. Arpinas, dat. -ati, while the rest are gentilicia made with suffixes -io- (Klaverniie, Pe- raznanie) or -/io-(Satan0S, luieskanes). Among the individual names Atiieriate is clearly related to the name of the Atie- dian Brotherhood. The appearance of the Clavernii and Casilas in Vb, which states the terms of their contract with the Atiedian Brothers, has already been mentioned in the introductory remarks just above. — 7 afraune: not only are the etymology and meaning uncertain, but it is not even clear whether the word is dative of a divine name belonging with luve patre, or ablative of a common noun. Against the former view is the fact that Armtme luve patre appears to conflict with 10 vaputu Sagi ampetu, 17 vaputu Sagi luvepatre prepesnimu. Devoto renders 'in afmone' (T.G., 'nell' assemblea sacra'), Vetter 'ordine,' Bott.
'in sacerdotum concilia,' all deriving it from the same root as arsmor, arsmahamo, arsmatiam. Earlier editors and also Pis. take it as a divine name, but without agreement as to its origin. The forma- tion of the word favors taking it as a divine name similar to Puemune, Ve- sune, Uofione, and the connection with arsmor is as good as any other. The re- lation of afmune to arsmor then is si- milar to that of L. praedo to praeda; it designates Jupiter as the god having an interest in or presiding over arsmor or sacred ordinances. The conflict with Sa- gi 10, Sagi luvepatre 17 is not irrecon- cilable, or at least it is scarcely more serious than the conflict of these names with luve patre which is left alone in 7 if we take afmune as abl. of a common noun. Read as a divine name it becomes simply an additional epithet of luve patre, who must probably be under- stood as identical with Sagi, Sagi luve- patre of 10 and 17. — perakne: see on I la 5. — 8 sevakne: on Ila 21. — eveietu: usually taken as the second in a series of imperatives, but it was sugges- ted on 1 above that upetu in 8 is best taken as a participle, eveietu then be- comes the prior of two instructions, to be folloved by sevakne naratu. The only other occurrence is in 11 below, and comparison with passages where other verbs are followed by naratu is not of much help: III 26-7, for example, shows the series uvera sevakni upetu Pue- mune Pupfike ampetu tiglu sevakni naratu, but there is no possibility that eveietu in lib 8, 11 can be synonymous with ampetu, for ampetu occurs in 10, 11, occupying a later position in each ser- ies of instructions. The most satisfac- tory etymology and the one enjoying the greatest favor is the derivation from *e-weig-e-i6d, cf. L. vic-tima, Germ, wei- hen (substantially so Osthoff, I.F., VI [1896], pp. 39-47, Bk., Muller, Dev., Pis., Bott.). We must assume a root-form *weig-, since *weik- inferred from the Germanic forms would give -veg-, and we
194
III. TEXT AND TRANSLATION
veietu : sevakne : naratu : arviu /
(9) ustetu : eu naratu : puze : iageiete : sevakne : heri : puni :
(10) heri vinu fetu : vaputu : sagi : ampetu : kapru : pera : kne : seva /
(11) kne : upetu : eveietu : naratu : give : ampetu : fesnere : purtu/etu: iie : fertu : tafle : e pir fer : tu : kapres : prusegetu / (13) ife af veitu : persutru : vaputis : xnefa vistiga : feta fertu : / (14) sviseve iertu : pune : etre : svi : se : ve vinu fertu : tertie : / (15) sviseve utur fertu : pistuniru fertu
10 pera, : kne: for perakne. se : ve: for sviseve.
without blemish, which has been se- lected, and pronounce it free from blem- ish. Present grain-offerings and pro- nounce upon them as upon (a victim) fit for sacrifice, without blemish. Sa- crifice either with mead or with wine. Slay (the victim) in honor of Sancius with incense. Consecrate and pronounce upon the he-goat brought from away and without blemish, which has been selected. Slay it outside and present it in the temple. Carry it there; carry the fire there on a board, and add to it the parts cut off from the goat. Bring the persondro with incense, with a mefa cake, and with the prepared libation. Bring mead in a flask, in a second flask bring wine, in a third flask bring water.
12 epirfer
tu: for e : pir : fertu.
14 svi
must assign eveietu to the second con- jugation because of the unsyncopated e before the ending, unless we choose to make it a pf. pcpl. with Vetter, who com- pares muielo. In view of its probable ety- mology and its position in the series of instructions it is best taken to refer to the consecration of the designated victim.
— sevakne naratu: it has already been stated that the pig chosen must be free from blemish; here the ad fer tor is to make a formal declaration of the fact. For sim- ilar uses of naratu + noun and predi- cate adjective cf. 25 (sc. vitlu) vufru naratu. III 27 tiglu sevakni naratu.
— 8-9 arviu ustetu: see on la 3. — 9 eu naratu puze fagefele sevakne: eu of course refers to the preceding arviu. fagefele is adopted by practically all editors as the correct reading for fage- fete which appears on the actual table, piu'tif eie in 25 gives support for the emen- dation. The sense is as if L. *facibilem were made to show the special ritual sense of facio. The sense then is that the arviu
are to be pronounced upon in the same manner as a victim which is declared to be free from blemish and suitable for sacrifice. However, the passage is not sufficient ground for taking arviu with Blum., p. 59, and others as ' exta' ; see on I la 18. — 9-10 heri pimi / heri vinu: cf. Via 57 with note on poni. — 10 vaputu Sagi ampetu: cf. 17 vaputu Sagi luvepatre prepesnimu, and also 13 persutru vaputis mefa vistiga feta fertu. The best etymology and sense are obtained by connecting vaputu with L. vapor ' incense, ' with the majority of editors. For the correspondence between the s-stem vapor (vapos) and the to-stem vaputu Bii., p. 144, compares L. arbos: arbuta. Blum., p. 86, questions the pro- priety of translating 'tare' on the ground that the introduction of incense into Italy was relatively late, but nothing is im- plied as to the type of incense used, unless he restricts tare to frankincense. On the native materials which preceded the use of oriental incense cf. Ov., F., I, 343-4
lib 8
COMMENTARY
195
Ara dabat fumos herbis confenta Sabinis, et non exiguo laurus adusta sono. For other interpretations of vaputu see Dev. {'laqueo'), Bott. (apparently as a divine name, without discussion). — ampetu: see on Ila 20.
lib 10 -21. Sacrifice of the goat. See the introductory note on the present table. The instructions commence in the same manner as in the case of the pig, except that sevakne is omitted before naratu, whether through oversight or because it is sufficiently clear from 8 and
10 what the pronouncement is to be. —
11 give ampetu, f esnere purtuetu : since the verbs are in contrast, designating res- pectively the slaying and the offering of the victim, it is probable that give and fesnere are also in contrast. The inter- pretation of fesnere as 'in the temple' receives excellent support from O. fiis- nii, ace. fiisnam, fisnam, on the Cippus Abellanus (Co, 95), A 24, B 4, 6, 19; fiis[nai Co. 169; Pael. fesn. Co. 239. The Umbrian use of the plural here and in 16 fesnafe deserves notice, give occurs no- where else, but gives good sense if taken as an adverb from loc. sg. of a stem *ki- W0-; cf. gimu simo, L. cis, citra. The question then arises whether the princi- pal emphasis in give is on a place not in the temple precinct or on a place under the open sky; Devoto decides in favor of the latter, taking account of the Roman practice of standing out of doors while swearing an oath by Dius Fidius and other similar cult practices in connection with heavenly gods. On etymological grounds give seems merely to suggest re- lative nearness to the speaker, and in any case the slaughter of the victim occurs outside of the temple. — 12 tafle e pir fertu: incorrectly divided tafle: epirfer: tu on the bronze, with a dot like the upper point of the word-divider between the r and f, although the space would not be adequate for a whole letter; but the real meaning is clear enough, and all editors translate the prepositional phrase ' in tabula. ' The * tafia is evi- dently a broad fiat carrier of some kind
on which are to be placed not only the fire but the prosecta of the victim as well. In lb 12 in the ceremony between the first and second taking of the aus- pices, before the lustral procession has advanced to Acedonia, the fire is placed in a carrier designated by a different name: pir aMimem ententu, while in the equivalent VIb 49-50 the name of the object for carrying the fire is avoided entirely. — 12-13 kapres prusegetu / ife arveitu: for the prosecta see on Via 56. arveitu in most instances has as direct object that which is added to the prosecta, generally cakes of various kinds. — 13 persutru vaputis mefa vistiga feta fertu: for the persutru see on VIb 24, mefa on Via 56, vistiga on VIb 5. Except for the abl. pi. vaputis the case- endings of all the nouns in the series are ambiguous, but at least one must be the ace. object of fertu. It is probably best to follow Bu., PL, Bk., Pis. in taking persutru as ace, to which the remain- ing nouns are abl. of accompaniment, but the problem is after all one of syntax rather than of sense. The pi. vaputis contrasts with sg. vaputu in 10 and 17, but there is no possibility of taking this form as neut. pi., since prepesnimu in 17 almost certainly calls for an abl. The form feta, which is nowhere else used as an epithet of uestihia, seems to mean that the libation was already mixed, but in this case Devoto can scarcely be right in taking the mead, wine, and water in 14-15 to be the actual ingredients of the libation, for it is made clear that they are to be carried in separate vessels. It is not certain exactly what the uestisia is, and it may not always even have been a liquid substance (see on la 28). — 14 sviseve: clearly a container of some kind, one being used for each of the three liquids, but nothing certain can be af- firmed with regard to the etymology. Several editors suggest possible relation- ship with L. simim; according to Pis. our form may be a loc. (-|- -e) in -eu < -eu analogous to Skt. sundu, etc., from a stem *suit-su- or *suit-iu-, with the Latin cog-
196
111. TEXT AND TRANSLATION
vepesutra : fertu : / (16) mantraklu fertu : pune : fertu : pune : fesnafe benus : / (17) kcibru purtu : vetu vaputu : sa9i : iuvepatre prepesnimu : / (18) vepesutra pesnimu : veskles pesnimu : atre pufatu : / (19) afpeltu : statitatu vesklu : pustru : pestu : ranu : / (20) pesni : mu : puni : pesnimu : vinu : pesnimu : une pesni/mu : enu erus tetu : vitlu : vufru :
17 purtu : vetu: for purtuvetu. xnu: for pesnirau.
Bring dark meal, bring the vempersondra, bring a maniple, and bring (additional) mead. When you have come into the temple, make the presentation of the goat. Offer a preliminary prayer to Ju- piter Sancius with incense. Pray with the vempersondra and pray with the ves- sels. Dance the tripudium, approach (the altar) and stop. Place the vessels at the back. Pray with the ranum, pray with the mead, pray with the wine, pray with the water. Then distribute the erus. When you wish to sacrifice
18 atre : pufatu: for atrepufatu. 20 pesni
nates sTnum < *suit-sno-, situla < *suit- ela. — 15 utur: Bu. and all after him translate ' aquam. ' t stands for d and the noun is an -r/n-stem cognate with vdoiQ; the n-stem is represented by abl. xme in 20 below. — pistu niru: writ- ten on the bronze with a space after pis- tu wide enough for one letter, but without the word-divider. The fertu which fol- lows is not separated from niru by either a space or a word-divider, pistu niru is read by some editors as one word, by others as two, and there is no sure inter- pretation, but a fairly satisfactory sense can be reached through the etymologi- cal translation ' pis turn nigrum' (for pal- atahzation of g see 47 c). The pistu niru would then be black or dark meal crushed with a pestle instead of ground with a millstone. Cf. Ha 18, where salu meiletu 'ground salt' appears among the materials to be taken to the place where the dog is sacrificed to Hondus. We are not told precisely how either the dark meal or the salt is subsequently used. — vepesutra: the meaning is very uncertain but see on Ila 30. — 16 mantrsiklu: see on VIb 4, from which it is clear that the word means a cloth to be worn over
the right hand. — pvme fertu: this second mention of ptine is surprising af- ter sviseve fertu pune in 14, and it is not unlikely that we have simply a dit- tographic repetition of fertu pune, the correct reading consisting of the fertu governing mantraklu and the conjunc- tion pune which begins the new sentence. On the other hand it is curious that ranu mentioned in 19 does not appear in the list of articles to be carried to the place of sacrifice, and Vetter, p. 206, may be correct in his suggestion that pune fertu was erroneously written in place of ranu fertu. But in our translation we proceed as if the text were sound. — 16-17 piuie fesnafe benus / kcibru purtuvetu, etc.: the porrectio or second part of the sacrifice now begins inside the temple. — 17 prepesnimu: this is the opening of a series of prayers, and hence the verb is compounded with pre-; not so, however, in Ha 29-31, where persnihmu occurs 4 times uncompounded. — 18 veskles : these have not been mentioned before, but they can scarcely be the vessels de- signated by sviseve in 14-15 as the con- tainers of the mead, wine, and water, since the prayers said in conjinction
lib 14
COMMENTARY
197
with these liquids are covered by a sepa- rate instruction just below in 20. Dev., pp. 361, 364, takes pistuniru to be a part of the viscera of the victim and the veskla to be the vessels containing them, but the idea of a connection between the two rests purely on a comparison of the passages in which each occurs; see, how- ever, on 19 below. — 18-19 atrepuf- atu / arpeltu statitatu : the same series of imperatives, in the same order, occurs in Ila 31-2 among the instructions immed- iately following the porrectio in the dog- sacrifice, but vestikatu, which occurs there between the prayer with the ves- sels and the dancing of the tripudium, is not included in the present passage, — 19 vesklu pustru pestu: in Ila 32 it is the supa or under-parts of the victim which are placed at the back (of the al- tar), and similarly in VIb 5, Vila 8. Apparently here the vessels containing the supa are spoken of in place of the supa themselves. — ranu: not else- where mentioned and completely obscure, beyond the fact that it must be the name of one of the materials used in the ritual; see also above on 16 pune fertu. Bii. and PI. suggested etymological connection with Qaivco, taking ranu to be a liquid of some sort. Blum., p. 73, takes as equi- valent to L. grano; he supports the va- riation r-/gr- by reference to Hesych. 'PaiHOQ • "EXXrjv. 'Pco/j,aloi de ro y' jtQoadev- xsQ rqaiKov yacrt and other similar glos- ses, and rules out Grabovius as contrary evidence on the ground that it is a foreign name. Goldmann, St. etr., XII (1938), p. 407, cites his interpretation with ap- proval. Several scholars equate with Etr. ranem, ranvis, the sense of which, how- ever, is itself uncertain: a triple vase like the Duenos Vase according to Dev., p. 363 with bibliography; a kind of cake (libum) according to Ribezzo, II, p. 192. — 21 erus: see on VIb 16. In Ila 32 the priest is instructed to convey the ems, which is the liver if iepru is correctly interpreted, in his hand.
lib 21 - 29. Sacrifice of a bull-calf to
Jupiter Sancius. Since this offering is made on behalf of the gens Lucia of the Atiedian Brothers, it is generally regarded as having a private character in contrast to the sacrifice of the pig and goat in 1-21, which concerns all the decuviae into which the citizens are divided. One noteworthy feature of the present ritual is the disk or wheel (urfeta) which the priest holds in his hand when he dedicates the victim. As in the augural ceremony at the begin- ning of the lustration (lb 11 = VIb 49), he wears a stole over his shoulder, but in the present ceremony he removes the stole after the slaying of the victim and places it upon the cake (mefa), to replace it on his shoulder after the porrectio. The instruction arviu ustetu is included, as in sacrifices during the purification of the Fisian Mount and the lustration of the people, and the use of mead is pres- cribed, but not of wine or water. — 21 vufru: Bii. and nearly all after him translate 'votivum', so also Vetter in his glossary, but he proposes ' liberum ' in his commentary and adopts it in his translation. From a phonological stand- point this is extremely tempting, since the two forms can be fully equated on the basis of derivation from *leadherom; cf. iXevdsQog, Pael. loufir ' liber, ' Fal. loifirtato ' libertatis'; U. phonology as in 23, 43, 55 a. There are semantic dif- ficulties, however. There would be no point in calling the calf ' free ' in the sense of 'unyoked, never used for plowing,' since this would be true of all calves. Vetter, p. 206, takes the sense to be ' non sollemnis' ; in this case vufru naratu in 25 may be contrasted with sevakne naratu in 8, somewhat as sakre and perakne are contrasted in Ila 5, and the calf may be free in the sense of 'not temple-property, not fenced,' or perhaps even ' unowned. ' The argument would be less difficult if we had a really clear idea of what sevakne means, but in any case it is rather strange that the priest should have to declare the victim 'free' three times (25 just below). On
198
III. TEXT AND TRANSLATION
pune heries : / (22) fagu : eruhu : tiglu : sestu : iuvepatre pune
seste : / (23) urfeta : manuve : habetu : estu : iuku : habetu : / (24) iupater : sage : tefe : estu vitlu : vufru : sestu : / (25) purtifele : tri : iuper teitu : triiuper : vufru : naratu / (26) feiu : iuvepatre : vugiiaper : natine : fratru atiieriu / (27) pune : anpenes : krikatru : testre : e uze : habetu : ape apel/us : mefe : atentu : ape purtuvies : testre : e uze habetu / (29) krikatru : arviu : ustetu : puni fetu : /
a votive bull-calf, dedicate it to Jupiter on the same day. When you dedicate it, hold a disk in your hand. Use the following words: "Jupiter Sancius, to thee I dedicate this votive bull-calf." Three times declare it fit for presenta- tion, three times pronounce it a votive offering. Sacrifice it to Jupiter for the gens Lucia among the Atiedian Brothers. While you are slaying it, wear a stole on your right shoulder. When you have slain it, place (the stole) upon the mefa cake. While you are presenting it, wear the stole on your right shoulder. Pre- sent grain-offerings and sacrifice with mead.
25 tri : iuper: for triiuper. e : uze.
26 feiu: for fetu.
27, 28 testre : euze: for testre:
the other hand It would be quite natural to make formal declaration that the vic- tim was about to be offered in fulfil- ment of a vow. We therefore reluctant- ly abandon the equation with liberum and derive from *wog'^h-(e)ro-m, from the same root as L. voveo. Other examples of stems in -ro- with predominantly pas- sive sense in relation to the verbs from which they are derived include L. sacer to sancio, integer to tango, daJgov to Sidcojui. — 21-2 pune heries / fagu; similar formulas occur in Ila 16 heriiei fagiu arfertur ... fagia tigit, where the punctuation and construction are in doubt, and in VIb 48 pone poplo afero heries. — 22 eruhu tiglu: see on Ila 15, where the interpretation ' day, ' is advocated in preference to ' (de)dicatio, declaratio.' In the present passage also we get a better sense by translating 'on the same day,' since the formula of dedication in 24 below is not the same as in the preceding sacrifice. — sestu: cf. the ritual use of L. sisto in connection with the formal dedication of the victim in Verg., Aen., VIII, 84-5 quam (sc. suem) pius Aeneas
tibi enim, tibi, maxima luno, mactat sacra f evens et cum grege sis tit ad aram; Ov., Met., XV, 130-32 Victima labe carens ... vittis praesignis et auro sistitur ante aras auditque ignara precantem. — seste: this form has been taken as 2 sg. pres., 2 sg. fut., and 3 sg. pres. pass. Although the fut. is frequent after pone pune, it is unlikely that we have an example here, for the final s in the fut. 2 sg. is etymologically double (see 123), and all the Umbrian examples, ampenes, heries, menes, purtuvies retain a single -s. The pres. pass. 3 sg. is possible, since the final r is sometimes omitted (see 115 e). Our form would then be from *ses- ter < *sist{e)ter. The use of the passive here, however, seems unnecessary and slightly unnatural. It may therefore be best to take seste with Bii., Co., Bk., Blum., and Bott. as pres. 2 sg. The pre- sent with pone pune is less frequent than the fut., but is found in lb 1 = VIb 43, and Vllb 2. — 23 urfeta: in a formal sense fully equivalent to L. orbitam, un- less the unsyncopated e was long. Se- mantically it is closer to orb is, since it
lib 21
COMMENTARY
199
evidently indicates some circular or wheel- shaped object having a symboUc value. Livy, VIII, 20, 8, in connection with the property of the condemned rebel Vitru- vius Vaccus says: ... bona Semoni Sango censuerunt consecranda; quodque aeris ex eis redactum est, ex eo aenei orbes facti positi in sacello Sangus adversus aedem Quirini. The rock-drawings in the Val Camonica include a representation of a man holding up a solar ring in his hand and another of a warrior with an eight- spoked wheel by his side (cf. F. Altheim, Geschichte der lateinischen Sprache [Frank- furt-am-Main, 1951], pp. 103-4). The association of the solar symbol with Ju- piter Sancius and Semo Sancius may be connected with their function as protec- tors of oaths and pacts; compare also the Roman practice mentioned above on 11 of swearing oaths under the open sky. — estu iuku habetu: in III 18, not be- fore the slaughter as here, but after it, we find iuka mersuva uvikum habetu ... For the plural, which is shown by the accompanying estu in preference to es- te, cf. L. ioca, and for the use with ha- betu cf. the Latin idiom orationem habere. There is no objection on semantic grounds to the connection with iocus, -a, since the Latin sense 'jest,' like that of Lith. judkas, can easily be explained as an in- stance of specialization. The serious sense, and also the e-vowel grade, are preserved in the OHG verb jehan ' declare. ' — 25 purtifele : related to *purdi-, the weak stem-form of purdouitu, as *porricibilis would be to L. porricio. The priest three times declares the bull-calf to be qualified for presentation, and three times gives notice that it is a votive offering (?). Whichever of the two possible interpre- tations we adopt for vufru in 21, it must
refer to the status of the calf before the ceremony. It appears therefore that in the present passage teitu is used of that which becomes effective as a result of the utterance and naratu of that which was already true and is merely made known, and there is no objection to ap- plying this distinction to the other in- stances of the two verbs in Tables Ila, b. III, although nowhere else are they used in such close proximity as in the present passage. — 26 Vugiiaper: the name is probably equivalent to L. Lucia, though it has also been rendered ' Vucia ' (Br.), 'Vocia' (Bu.), 'Vovicia' (Dev.). The phraseology is similar to Petrvinia- per natine fratru Atiieriu in Ila 21 (on which see note) and 35. Vetter favors the view that Vugiia here does not de- signate an actual gens Lucia but has rather the value ' gens So-and-so ' and is therefore comparable to the Roman le- gal expression TITIAE UXORI MEAE (Gains, Inst., I, 152); an ingenious and plausible explanation, but if we accept it we must discard Devoto's view cited in note on 2 pumperias that the gentes Petronia and Lucia (Vovicia) represented the two pentads of the Brotherhood. — - 27 krikatru: mentioned also in lb 11 and in the equivalent VIb 49, on which see note. In these last two passages the stole is worn during the fire-ceremony between the first and second taking of the auspices in the lustration. — 28 mefe atentu: it is not clear what is symbolized by the placing of the stole upon the surface of the broad flat cake. According to Dev., p. 368, the krikatru indicated the three principal events in the sacrifice and imparted a certain sacred character (aliquid sacri) to the exta to which the mefa was added.
200
III. TEXT AND TRANSLATION
(III 1) esunu : fuia : herter : sume : / (2) ustite : sestentasiaru : / (3) urnasiaru : huntak : vuke : prumu : pehatu / (4) inuk : uhturu : urtes : puntis : / (5) frater :
Sacrifice is required to be performed in the final period of the regular bi- monthly festival. First purify the jar in the grove. Then the Brothers, rising in groups of five, shall elect an audor
III - rv. Sacrifice in honor of Pomo- nus and Vesona. These two tables clear- ly contain a single body of text. The reverse of III is left blank and the last sentence of III is continued on IV, and the obverse of IV is almost completely filled, with slightly less than half a line of blank space at the bottom and with the reverse, like that of III, left blank. Yet the two tables seem not to have been en- graved by the same hand. Ill contains no absolutely certain instances of error: ferime 16 must probably be emended to ferine, but vatra 31 may quite easily be correct, and aruvia 31 is probably a unique variant of arvia. IV on the other hand contains at least eight sure cases of misspelling and several more probable cases. The two tables are as difficult to interpret as any among the seven, es- pecially because of the large number of names of articles used in the ritual, many of them occurring only once and giving no good clue as to their sense eith- er through comparative etymology or through study of their context. The fol- lowing is a brief outline of the content of the two tables, with the chief emphasis on those portions where the meaning is most surely established. First the time for the sacrifice is given : the final period of the regular bi-monthly festival. A certain jar or vat in the Sacred Grove is to be purified. The Atiedian Brothers then appoint an auctor, who designates a young pig and a sheep as victims. The Brothers examine and accept the vic- tims, and all proceed to the sacred field called the Arva. The victims are taken to the field on a kletra, a sort of litter whose construction is explained in detail (III 14-20) but is striking for the large number of obscurities. First the pig is
slaughtered in honor of Jupiter and on behalf of the Atiedian Brothers and the state and tribe of Iguvium (22-5). The sheep is then slaughtered in honor of the god Pomonus Publicus on behalf of the Atiedian Brothers and the state and tribe of Iguvium (26-30), yet the two sacrifices are not similar in all respects. The sheep is to be slaughtered on the ground, for ultimate burial (31-2). Two pieces are to be cut off as burnt-offerings on one side of the altar, two at another side for presentation to Pomonus, and three at a third side for presentation to Vesona, a goddess closely associated with Po- monus, probably as his consort, the two being predominantly deities of fertility and growth. In connection with this ritual (31-IV 6), use is made of cakes called *struh5las but having a special form sometimes believed to involve sex-sym- bolism. 6-14 contain instructions to make offerings of the uncut and cut portions of meat, the sin-offerings (vempesuntres) and the underparts,to pray, approach the altar and stop; to pray to Pomonus and Vesona with the wet and dry vessels and with the spoons or sticks used for ap- plying unguent, and finally to anoint the statue. 14-22 contain instructions to offer a libation and cake (mefa) from a cup, kneeling, at the right of the altar, and in the same manner to present libations and sin-offerings below and above the statue to Hula and Tursa, goddesses whose cult in the present instance seems to be merely subsidiary to that of Po- monus and Vesona. The last lines of this passage contain instructions to fetch a cover and place it over the sin-offering and to offer certain cakes of a circular form. The final portion of the sacrifice (23-32) includes, among certain obscure
Ill 1
COMMENTARY
201
details, prayers to Pomonus and Vesona, a second distribution of the eras (the first being in 14), the grinding of cakes, with prayers over the ground cakes, and the playing of music. The final sentence of IV is not directly connected with tlie fore- going instructions for the sacrifice but is an injunction against the (private or secular) use of the mead in the jar or vat the purification of which was prescribed in III 3.
nil -2 sume / ustite: cf. Ila 15 and note. — 2-3 sestentasiaru ur- nasiaru: cf. plenasier urnasier Va 2. urnasiaru was generally rendered by earlier editors ' urnarianim' on the sup- position that the festival was named from the urns used in its celebration, but Blum, was no doubt right in equating it v/ith ordinariorum, after Linde, GL, III (1912), pp. 170-1. sestentasiaru = L. sexten- tariarum 'belonging to the sixth part (of the year) ' ; hence ' bi-monthly. ' The festival then was one which occurred at intervals of two months, or possibly at the end of the second month of the year, which agrees well with the behef that this festival was somehow connected with the Roman Ambarvalia, celebrated in May. — huntak: taken by Br. as an adv. from the same root as the second part of eri- hont, with ending as in L. istac, iliac; similarly Blum. But an object is needed for pehatu, and also for prupehast in IV 32, so that huntak is best taken as a noun, as in fact the majority of editors take it. In this case it is apparently a neut. A:-stem analogous to tuplak and derived from the root *hom < *ghom- ' earth'; cf. kondra, hondomu,hondu. Since we should expect -n(n)- from -nd-, PL, II, p. 69, suggests development from a stem in -nditak-. Several editors take the huntak to be a well or similar source of underground water: ' fontem' Bii. (but observe fondlire with /, not h), 'cisternam ' PI., 'puteum' Bk., Vet. (?), Pis. But it is hard to escape Devoto's objection, p. 369, based on the fact that the huntak is evidently intended to contain the sub-
stance poni. This much is evident from IV 32-3 by any reasonable interpretation, and it is altogether unlikely that punes would be applied to water, despite its apparent derivation from *p6i- ' drink. ' Dev. translates 'ollam,' assuming the jar to be designated, perhaps euphemisti- cally, by the earthenware from which it is made. Could it be possible that the huntak was a vat located in the grove and used for making the mead, and that IV 32-3 has the effect of a prohibition against using the mead for private or secular purposes, once the vat has been purified? Such a view would be easy to support on etymological grounds, but cannot be affirmed with any certainty.
— vuke: see on VIb 43. — pehatu: since the ceremony takes place on a set day, it is probably a routine purification or consecration rather than an expiation for sin. — 4 uhturu: recognized as the etymological equivalent of L. auctorem by nearly all editors. The abstract uh- tretie '*auctorata' accompanied by the names of the eponymous magistrates is used in Va 2, 15 for dating the regulations contained in those passages, and the ab- breviation oht accompanied by the names of two officials is similarly used in an Umbrian inscription from Assisi (no. 355 Co.). It is incredible that the Atiedian Brothers should elect a civil magistrate, yet 'elect' or 'appoint' appears to be the sense of ustentuta in 5. Vetter translates uhtretie Va 2, 15, and oht in the Assisi inscription by ' auctoritate'; thus we no longer have an eponymous magistracy to deal with, and the uhtur in III becomes simply a master of cere- monies to preside over the sacrifice. This last view is doubtless correct, but if C. Vestinius and Nero Babrius in the Assisi inscription were not uhtur, it is not clear what their position was; perhaps they were merely the sponsors of the motion.
— urtes puntis: cf. puntes 9, urtas puntes 10, which are subjects of their verbs; the present form then is abl. abs. puntis is normally translated ' quinioni-
202
III. TEXT AND TRANSLATION
ustentuta : pure : / (6) fratru : mersus : fust : / (7) kumnakle : iinuk : uhtur : vapefe : / (8) kumnakle : sistu : sakre : uvem : uhtur / (9) teitu : puntes : terkantur : inumek : sakre / (10) uvem : urtas : puntes : fratrum : upetuta : / (11) inumek : via : mersuva : arvamen : etuta : / (12) erak : pir : persklu : uf etu : sakre : uvem : / (13) kletra : fertuta : aituta : arven :
who shall be in accord VN^ith the customs of the Brothers in assembly. Then the audor shall sit on the stone seat in the meeting-place. The auctor shall desig- nate a young pig and a sheep, the groups of five shall inspect them, then the groups of five rising shall accept the young pig and the sheep. Then they shall go by the accustomed way to the field. On the way load the fire (with incense) with a prayer. They shall lift and carry
7 iinuk: for inuk.
bus ' ; the brothers then are divided into groups of five for voting purposes, and the total number of twelve brothers is made up, according to Dev., p. 305, of two pentads together with the adfertor and the fratricus. urtes is generally ren- dered 'ortis,' the sense being that of surgentibus used of voting by rising. For a similar early use of orior Bii. cites Velius Longus (Gramm. Lat., VII, Keil), p. 74, 18 oriri apud antiquos surgere frequenter significabat ut apparet ex eo quod dicilur: oriens consul magistrum populi dicat, quod est surgens. — 5 ustentuta: this verb is not elsewhere used with the title of a person as object, but the semantic development ' show > proclaim > elect ' offers no special difficulty. — 5-7 pure fratru mersus fust / kumnakle : most editors give translations having the gen- eral sense 'which shall be in accordance with the laws (or customs) of the Brothers ' (quite otherwise Devoto, who translates ' ab oriente, ' with reference to the location of the huntak in the place of meeting); but the future fust in place of a present is somewhat surprising, and the whole ex- pression is slightly otiose. We may im- prove the sense of the passage by trans- lating 'who shall be in conformity with the customs of the Brothers in convention. ' The use of pure as masc. nom. sg. re- ceives adequate support from porse perca
arsmatia habiest in VI b 63, VII a 46, 51. There is no change in the construction of mersus, which is still an ablative of accordance. The spelling with -rs- in- stead of the expected -fs- is evidence that there was only a slight difference in pronunciation between rs = rs(s) and original rs. kumnakle is used also in Va 15 to designate the place of adoption of the second portion of the decree re- garding the duties and rights of the ad- fertor. — vapere : all other instances of tliis word are accompanied by forms of the adj. auiecla, and Blum., pp. 34-5, relying on Livy's description, I, 18, 6-7, of the inauguration of Numa, suggests that in the present passage also the uhtur occupies the lapis auguralis in order to take the auspices. There is, however, no direct mention of such a purpose, and we must be satisfied to assume that the seat was simply one of stone. The Acta Fratrum Arvalium also contain references to the priests or the promagister sitting in their ceremonial chairs (cf., for exam- ple, pp. 11, 12, 14 Henzen). — 8 sakre uvem: there is no possibility of regarding sakre as a mere epithet of uvem, for the two are treated separately in 22-3 and 26-7, the former being offered to Jupiter, the latter to Pomonus Publicus. It is uncertain for what victim sakre stands, but see on I la 5, where it is
Ill 4
COMMENTARY
203
taken to be a young pig. — 9 teitu: the first of three actions connected with the two victims. Normally deitu teitu is either used with oratio recta or else its object is accompanied by a predication, as in lib 25, III 25. The present example falls into neither of these categories, and the question is precisely how teitu and upetuta differ. The sense is either that the auctor designates a young pig and a sheep and the pentads of the Brothers, after examining them (terkantur), signify their acceptance of them by rising up, or that the auctor names (the species) pig and sheep as victims and the Brothers after an examination select one of each as vic- tims. The former interpretation seems slightly preferable as conforming more closely to the normal meanings of the verbs in question. — terkantur: there are no other instances, and the initial t is etymologically ambiguous, but most connect the word with deQxo/uai 'look.' In Va 7 revestu, etymologically equi- valent to L. revisito, is similarly used of examining victims. — 11 mersuva: in all probability related to mers 'law, cus- tom ' somewhat as L. annuus to annus and therefore meaning 'accustomed.' This also gives satisfactory sense in 28, where mersuva is used with iuka. In lb 14-6 = VIb 52 the adfertor after the taking of the auspices proceeds by the uia auiecla or 'Augural Way' to Acedo- nia to perform the three sacrifices of the lustration. In the present passage, how- ever, the route is not identified by a proper name. — arvamen : an a-stem, perhaps with originally collective value, in contrast to the Latin o-stem arvum. Nonius, p. 192, however, cites an ace. pi. arvas from Naevius and Pacuvius. The exact location of the Arva of the Atiedian Brothers is of course uncertain, but Ro- senzweig, pp. 18-9, plausibly suggests the level plain near the Roman theatre where the Tables were discovered. — 12 erak: an abl. sg. fem. form used adverbially and found nowhere else. It must refer not to the Arva, in which case we might
expect ife, but to via ; the act designated by the words immediately following is therefore to be performed along the way. — pir persklu uretu: the verb, which recurs in IV 30 without pir, is generally regarded as equivalent to L. ad-oleto ' burn, load (with gifts), worship. ' Dev. assumes that it means 'load (the fire) with incense, ' depending in part for sup- port on etymological connection with L. odor, olere < *odere, where comparison with Gk. odfZTJ, ddcoda shows that d and not I was the original consonant. Since f in the phonetic environment in which it stands can have arisen from either d or /, it is quite possible that ufetu re- presents a contamination of the two roots. Vetter and Pisani in both passages take uretu as a supine forming an expres- sion of purpose which they translate 'ad adolendum (-am). ' This is free from ob- jection on formal grounds (unless ufetu made a shorter supine without e, like L. visum to videre), but it is hazardous to assume that the Umbrian supine was not largely restricted, like the Latin, to use with verbs of motion. — persklu in the present passage must be taken in its narrow sense ' prayer. ' — 13 kle- tra : the etymological equation often made with Goth, hleipra 'aKrjvri' is phonolo- gically perfect though semantically slight- ly difficult. L. clitellae 'pack-saddle' probably represents a diminutive exten- sion of *kleitro-. In any case it is clear from the context that the kletra is a litter for carrying the victims to the Field. For a recent discussion of the nature of the kletra and its component parts see G. B. Pighi, Studi in onore di Gino Fu- naioli (Rome, 1955), pp. 373-77. — fertuta aituta: Gk. and L. parallels to this pair are frequent : E 483-4 drag ovri fiol ivOdds rolov, olov x' fje, qjegoisv 'Axaioi, f] KEv dyoisv. Livy, XXII, 3, 7 res sociorum ante oculos prope suos ferri agique vidit; cf. Ill, 37, 7, XXXVIH, 15, 10. In the present passage, however, the circumstances are quite different, since it is not a question of carrying ob-
204
III. TEXT AND TRANSLATION
kletram : / (14) amparitu : ervOs. : esunu : futu : kletre : tuplak : / (15) prumum : antentu : inuk : gih^efa : ententu : / (16) inuk : kazi : ferime : antentu : isunt : fefehtru : / (17) antentu : isunt : sufefaklu : antentu : seples / (18) ahesnes : tris : kazi : astintu : fefehtru : etres : tris / (19) ahesnes : astintu : sufefaklu : tuves : ahesnes / (20) anstintu : inenek :
jects and driving beasts. Here fertuta is evidently used for the lifting of the vic- tims on to the litter, aituta for the carry- ing of the litter to the designated place ('La vittima "scara," la pecora sollevino, portino via in una lettiga, ' Devoto, T.G., p. 65). — 14 amparitu: etymology and meaning were discussed on Ila 42, where the m.edio-passive was used of the priest rising. Here the active is used of the set- ting up of the litter. — eruk : like erak in 12 this form is unique. Derivatives of tlie stem *ero- are anaphorics like L. is, ea, id, etc. while eso, etc. are used to in- troduce matter which is to follow. More- over the technical instructions in 14 kletre tuplak ... 20 anstintu do not real- ly constitute the esunu but are inserted almost parenthetically between eruk esu- nu futu and 20 inenek vixkuraen esu- numen etu. eruk then is best taken not as 'in the following manner,' but as 'then, next,' the ablative form having developed an adverbial value. — tup- lak: a double object of some kind (ct. L, duplex); according to most editors a fiirca to be somehow attached to the litter; according to Vetter a sort of double floor of clay to be placed over the wooden floor of the kletra, on which glowing coals were then to be placed. But the additional three objects which are to be placed on top, the kazi, the fefelitru, and the sufefaklu are already quite enough. If the litter was of such a kind as to be easily carried but not to stand in
the young pig and the sheep on a litter. Set up the litter in the field. Then the sacrifice shall take place. First attach a fork to the litter. Then insert the lat- tice frame. Then place the kettle upon the carrier ; on the same put the chop- ping-board ; on the same put the under- board. Fasten the kettle with three bronze nails ; fasten the chopping-board with three additional bronze (nails) ; fas- ten the under-board with two bronze
place without a support, perhaps because of a round bottom, then the tuplak may have been a forked piece of wood designed to hold it in position, and in this case it is quite natural that the instruction re- garding the tuplak should be the first one given. — 15 giligefa: quite obscure, but generally brought into connection with xiyy.Ai'dsg, L. cingo, and translated ' cancellos.' It must then be a sort of lattice frame to be added to the litter. — ententu: important because the usual verb in this series of instructions is an- tentu, which governs tuplak, kazi, fe- feMru, and sufefaklu. If the kletra had holes into which the separate rods making up the gihgef a were to be inserted, the use of ententn in place of antentu would be adequately explained. — 16 kazi: many etymologies and interpreta- tions, but there is no sure solution. The z probably represents ts, and the form may be a stem in -si- or more probably in -sio- (cf. 32 a). Vetter proposed to connect with L. catinus and to assign to it a similar sense 'Kcssel' or 'Tiegel. ' His view of certain other details in the passage is not adopted here (see on tup- lak 14 above), but the notion of kazi as a kettle or pan of some sort is not necessarily in conflict with the sense adopted for the passage as a whole, if we consider that the kletra served as a sort of movable altar. - - ferime: found also in lb 25, where emendation to fe- rine is unavoidable because of the fre-
Ill 13
COMMENTARY
205
quency of the formula vatua ferine fetu in which it occurs. In the present pas- sage emendation is favored by Kent, T.C.I. , p. 36, Bk., Dev.; retention of ferime by Co., PI., Blum., Vet., Pis.; Bott. is doubtful. For a nomen inslrumenti made by adding -i- or -(i)ijo- to a verbal root, which would have to be assumed to account for ferim-e, it would be virtual- ly impossible to find a Latin parallel. Moreover andendu antentu is normally construed with a dative or locative or with an adverb (cf. Kent, loc. cit); YII a 25 superne adro trahuorfi andendu is the only instance with a preposition. It is therefore quite probable that we must assume a misspelling of ferine here as in lb 25. The sense, hovvcver, is not greatly affected, and ferira-e or ferine may be taken to indicate the flat surface on which the kazi is placed, just as fe- rine in the frequent formula uatuo ferine fetu appears to be a flat tray or similar object. — fefehtru: the suffix -tru sug- gests a nomen inslrumenti, and the r may be from d or from / (42 b, 55 b). "We should proceed, however, on the theory that fefehtru and sufefaklu in 17 are probably related, and f before a cannot be from I, unless through some analogical influence. Conway found a reasonably satisfactory solution by tak- ing as equivalent to a L. *fidetrum from the same root as findo and signifying a chopping-board. The e < i is paralleled by tefa < *diddt. — 17 sufefaklu: no less obscure than fefehtru, with which it evidently shares the same root, though vowel-extension and suffix are different. If the interpretation 'chopping-board' proposed above for fefehtru was correct, then the sufefaklu may possibly be a receptable beside and below the fefehtru designed to receive the pieces of meat after they have been chopped. — seples generally taken as equivalent to L. sim- pulis 'ladles' until it was shown that simpulum in Fest., p. 455 Li., on which the etymology depended, was a false reading for simpuvium (cf. Brinkmann,
A.L.L., XV [1908], pp. 139-43). Devote has an altogether different interpreta- tion according to which the seples are nails. This gives the most satisfactory sense when taken in connection with a(n)stintu, and the thrice repeated in- struction that the seples must be of bronze is very easily explained by the widespread tabu against the use of iron in holy places (cf. Dev. p. 381, Vv^ho cites among other references Acta Fratrum Arvalium, p. 132 Henzen, on the propi- tiatory sacrifice to atone for the bringing of iron tools into the sacred grove). De- voto's interpretation unfortunately lacks adequate etymological support, but sep- les might conceivably be from the same root as L. saepes, saepio, the semantic connection lying either in the fact that nails are used for closing wooden struc- tures or that a nail or peg has a certain resemblance to one of the poles that make up a fence. — 18 ahesnes : equivalence with L. ahenis, -eis is universally recog- nized and is beyond reasonable doubt. — astintu: used three times, having as object kazi, fefehtru, and sufefaklu. The first two actions are performed with three seples each, the last with two. Some render ' exstinguito' or ' restinguito' understanding that the fire burning on the litter is extinguished with the seples which they take to be ladles; others 'dis- tinguito' in the sense 'mark off, adorn' with the ladles or other vessels designated by seples. However, if it were merely a question of ladles being placed at various points, we might expect some expression such as *seplu antentu instead of an- stintu. The sense ' pierce ' (with nails) has a satisfactory etymological basis if we connect with ariZco, aTiy/nij, L. in- stigare, distingere; cf. Velius Longus (Gramm. lat, VII Keil), p. 67, 20 inventi sunt qui distinguere quoque sine u littera ef scribere et dicere maluerunt, adicientes et illam rationem quod distingere est inter- posito puncto dividere atque diducere. — 20 inenek: found nowhere else and hard to explain morphologically, but the va-
206
III. TEXT AND TRANSLATION
vukumen : esunumen : etu : ap /
(21) vuku : kukehes : iepi : persklumaf : karitu : vuke : pir /
(22) ase : antentu : sakre : sevakne : upetu : iuvepatre / (23) prumu : ampentu : testru : sese asa : fratrusper : / (24) atiieries : ahtisper : eikvasatis : tutape : iiuvina / (25) trefiper : iiuvina : tiglu : sevakni : teitu : / (26) inumek
(nails). Then go into the grove for the sacrifice. When you reach the grove, call (them) there to the ceremony. In the grove place fire upon the altar. The young pig without blemish, which has been selected, slay first in honor of Ju- piter, at the right of the altar, for the Atiedian Brothers, for the fire-carriers (the symbols) of the college, for the state of Iguvium, for the tribe of Iguvium. Pronounce the day solemn. Then slay
riety of pronominal and adverbial forms observed in the scanty Italic dialect re- mains is so great that we are scarcely justified in emending to inemek, as some have proposed. — vukum-en : the in- terpretation ' grove ' is to be preferred to 'temple'; see on VI b 43. — esunum- en: Bli. favored taking as an adj. 'divi- nam ' qualifying vukum, and as support for the repetition of the preposition he cited lb 14 vapefara avieklufe and Homeric ovde dojuovds. His suggestion is all the more tempting in view of persklu- mar in 21, which would lead us to expect -ar, not -en, if esunumen means 'to the sacred rites. ' Yet there are more in- stances in which the adjective following the noun is unaccompanied by the post- position, and moreover lb 14 shows an unambiguous example of esunume etu in the sense ' go to the sacrifice. ' It is therefore safer to follow tlie majority in taking esunumen as a separate phrase. — ap: a syncopated form of ape; the only other example is in IV 31. — 21 kukehes: a unique and exceptionally troublesome form. Connection with ce- hefi VI a 20 is tempting but not necessary, since there is no especially close connec- tion between the contexts in which the two forms occur. EarUer editors in ge- neral translated by forms of in- suc- cendo, understanding that the grove was to be lighted up with a fire. Etymologi- cal connections were set up with tcaico
or with L. candeo or in-cendo, but both involve difficulties of phonology because in the combination of vowel -|- h + vowel in the native alphabet the h reg- ularly has etymological value instead of being merely an indication of vowel- length. Blum., p. 74, found an easier etymology by connecting with O. kahad 'capiat' (and L. in-coharel), and is fol- lowed in substance by most subsequent editors. The sense then is not that of ' hghting up ' but simply of ' reaching ' or ' entering ' the grove. The ablaut-va- riation in kukehes: O. kahad is the same as in L. cepi : capio, but the ending is dif- ficult, ape ape is used 29 times with the fut. pf., only once with the fut. (purtu- vies lib 28). The vowel-grade in kukehes points to the pf. stem, yet a fut. pf . 2 sg. should end in -us. R. G. Kent, Lang., IX (1933), p. 217, suggested analyzing as a 3 pi. pf. subj. from *ko(m)-kehens, with the Brothers as subject, but there are no other examples of a subjunctive after ape, and besides ap kukehes seems to contain the fulfilment of the instruc- tion vukumen etu immediately preceding, as Blum., loc. cit., suggested, so that we should be inclined to expect a 2 or 3 sg. form. A hybrid form made from the long- vowel pf. stem but inflected like the fut. instead of the fut. pf. would be quite without parallel, but the difficulties by any otlier interpretation are no less se- rious. The prefix ku-, like con- in L. con-
Ill 20
COMMENTARY
207
sequor, imparts perfective aspect (in the Slavic sense) to the verb, and presents no problem. For discussion of kukehes cf., beside the editions, Kent, C.P., XV (1920), p. 366; Lang., IX (1933), p. 217; Ribezzo, II, p. 98. See also on cehefi VI a 20. — iepi : an adverb found only here and of uncertain origin. Blum., p. 74, took it to be temporal and to correspond to the preceding ap as L. lum to cum; but if ape is followed by a temporal adverb, it is normally enuk, enom, eno (cf., for example, la 30, 33, VIb 53, 63, Vila 5, 8). Most render iepi ' ibi, illic, ' etc., but there is almost no chance that it represents ife by the same type of orthographic va- riation as kutef: kutep (cf. 2 k). The first part is probably from *iyei, loc. sg. of the thematic pronominal stem seen in L. is, ea, id. The second part may be related to O. ip, Pael. ip, Mars, ceip 'hie, ' and possibly also to U. neip. The use of iepi is like the rare and late Latin use of ibi 'thither' with verbs of motion. — persklum: used in a comprehensive sense, as in I a 1 = VI a 1. — karitu : see on carsitu Vila 43. — pir / ase an- tentu: the same fire which was mentioned in 12. For the instruction to place the fire on the altar cf. I la 19-20. — 22 upetu: on the reasons for taking as pcpl. rather than imv. here and in lib 8 see on lib 1. If this view is adopted, the sentence will not end with upetu, as it does in most of the standard editions. — 23 prumu: in contrast to inumek, which introduces the sacrifice of the sheep to Pomonus at 26. — ampentu: see on Ila 20. — testru sese asa: the same expression occurs in IV 15-6, while supru sese eregluma occurs in IV 3. Although the general sense is clear, sese is obscure. There have been several attempts to de- rive it from the root *sed-, with reference to place: Br. refers it to a noun *sessis; Bii. (without exact analysis) regards it as equivalent to L, sessu, or to sessione; Bk., § 307, as equivalent to L. *sesse from sessus. Others more recently have derived it from the adverb of separation found in
sei-podruhpei, L. sed-itio, etc. (so Blum., pp. 40-1, Dev., Pis.), sese will then be taken as a postposition in close connection with testru, not as a preposition with asa, for IV 3 has supru sese alone with- out a following asa, and the combina- tion testru sese in turn will govern the abl. asa, giving the direction from which the motion proceeds, but the whole phrase may be rendered 'at the right of the altar. ' — 24 ahtisper eikvasatis : it would be satisfactory in some respects to equate ahtis with L. actis, but from an o-stem we should expect the abl. pi, to end in -es; -is is found only in vaputis lib 13 (zo-stems are not taken into con- sideration here), ahtis must therefore be abl. pi. of an z'-stem, perhaps the same noun as aMim-em in lb 12, where it was taken to designate a fire-carrier. If this interpretation is adopted, the fire- carrier must be regarded as the symbol of the Atiedian Brotherhood, or of that group affiliated with it or included in it under the name eikvases-e Atiiefie (V a 4, 16), and these symbols are then among the objects of the prayers which accompany the sacrifice. Dev., p. 386, cites, among other evidence for the sacred symbols of various priesthoods, tlie fol- lowing passages relating to the Fratres Arvales: Plin., N.H., XVIII, 6 spicea coro- na quae vitla alba colligaretur sacerdotio ei pro religiosissimo insigni data; Gellius, VII, 7, 8 cuius sacerdotii insigne est spicea corona et albae infulae. For eikvasatis see on eikvasese Va 4; whatever the etymology and sense of the latter, eik- vasatis can be easily explained as ? stem in -ati- standing toward it in the same relation as L. Arpinas to Arpinum. — tutape: see on tote Via 5. The omis- sion of the final r in -per occurs again in 28 below, and in la 12. — 23 tre- fiper: on VIb 54. Apart from III 30 it is elsewhere used only in the sen- tences of banishment and commina- tion proclaimed against the alien state of Tadinum. — ti9lu sevakni teitu: see on Ila 15, b 22 on the reasons for
208
III. TEXT AND TRANSLATION
: uvem : sevakni : upetu : puemune : / (27) puprike : apentu : tiglu : sevakni : naratu : / (28) iuka : mersuva : uvikum : habetu : fratruspe / (29) atiiefie : ahtisper : eikvasatis : tutaper / (30) iiuvina : trefiper : iiuvina : sakre : / (31) vatra : ferine : feitu : eruku : aruvia : feitu : uvem / (32) peraem : pelsanu : feitu : ererek : tuva : tefra / (33) spantimaf : prusekatu : erek : pefume : purtuvitu / (34) strugla : afveitu : inumek : etrama : spanti : tuva tefra / (35) prusekatu : efek : ere^luma :
adopting ' day ' as the translation of tiglu. The present passage becomes a virtual oratio obhqua with sevakni ' so- lemn' as predicate. — 26 upetu: as in 22, the sentence does not end with upetu, if we are correct in taking it as a pass. pcpl. — Puemune / Pup- rike: these two divine names are found only in III and IV, the former being reg- ularly accompanied by the latter, as an epithet, except in IV 5. That the names are masc. and not fem. is shown by the gen. sg. in -es IV 3-4, since the only stem-classes that can be considered are the o-stems and the consonant-stems of the Cicero-type. Puemune is in all prob- ability the male counterpart of L. Po- mona, and the Vestinian or Sabine form Poimuni-en (Go. 248) appears to be an adj. from the name of the same god as Puemune; the problem of the vocalism is briefly discussed in 4. Puprike is more difficult. Not only is it uncertain whether the r is from d or I, but it is un- natural for it to occur at all after a con- sonant; for a discussion of the phonology see Lang., XXV (1949), pp. 397, 399. If L. populus is of Etruscan origin, as some believe, Puprike may be related both to it and to Fufluns, the name of the Etrus-
the unblemished sheep, which has been chosen, in honor of Pomonus Poplicus. Announce that the day is solemn. Use the accustomed words in connection with the sheep, for the Atiedian Bro- thers, for the fire-carriers (the symbols) of the college, for the state of Iguvium, for the tribe of Iguvium. Place the young pig's ribs on a tray. With it make grain-offerings. Sacrifice the sheep upon the ground, for burial. Cut off two (pieces) of it at the side (of the altar) as burnt-offerings ; present them upon a turf-altar. Add a strusla cake. Then on the other side (of the altar) cut off two (pieces) as burnt-offerings ; pre-
can counterpart of Dionysus. It then be- comes possible to set up certain corres- pondences which may provide a satis- factory explanation. Puprike and Etr. Fufluns are related in sense to populus somewhat as L. Liber to the noun for 'people' represented by OHG Hut, Germ. leute, OCS Ijudie. The root *leudh- seems originally to have signified growth, and not only Liber and Dionysus but also Fufluns and *Pomonus Poplikos were probably, at least in part, gods of ferti- lity and growth. According to Aug., Civ. Dei, IV, 11 ; VII, 3, there was a Libera corresponding to Liber; the goddess Vesona who is worshipped in conjunc- tion with Pomonus in Table IV, her name generally followed by Puemunes Pup- fikes, may stand in much the same rela- tion to him as Libera to Liber. — 27 tiglu sevakni naratu: cf. 25, and see on II b 25 for the distinction between teitu and naratu. — 28 iuka: see on lib 23. — mersuva: it is not stated what the 'customary' words are, but the ob- jects of the prayers are the same as in the sacrifice to Jupiter above. — 30 sakre : probably not ace. but gen., because it is most natural for the object of fetu to be vatra, on which sakre will then de-
in 26
COMMENTARY
209
pend. There are no gen. sg. forms of i-stems with omission of the final -s in the older tables, but the total number of examples is small, and there is some sup- port in the o-stem forms Qerfe, Kastru- giie, etc., since the origin of the case- ending in the two classes is the same. — 31 vatra : the context so closely resembles the formula uatuo ferine fetu (see on VI a 57) that some regard vatra as merely an instance of erroneous spelling. Table III, however, unlike IV, has few errors, and it is safer to regard vatra and uatuo as stem-variants. With this variation Bii., p. 63, compares vexveg : vexqoL The exact analysis of vatra is not clear, but it is scarcely an s-stem related to uatuo as L. pecora to pecus, for the syn- cope should have occurred so early as to prevent rhotacism from taking place, if we may judge from ose, osatu (57 d). It must rather be a stem in -ro/ra-., and is better taken as a noun than as an adj., for in the latter case it w^ould probably follow instead of preceding ferine, and the sense 'place the pig on the tray in- tended for ribs ' w^ould be most unnatural. uatuo in the regular formula is never pre- ceded by another noun in the same clause but here sakre marks a change of sub- ject-matter, since 26-30 was concerned with the sheep. — aruvia: see on la 3, Ila 18. erii-ku must refer to the pig (sakre). — 32 pefaem: see on Via 58. — pelsanu: on VIb 22. The in- structions to slaughter the victim on the ground and to slaughter it for burial occur nowhere else in immediate juxtapo- sition, but both are given in connection with the dog-sacrifice, the former in II a 22, the latter in 44. — ererek: the addition of the enclitic to the gen. sg. is unusual but the form may be explained by the proportion ererek : erer = ef ek : *id, L. id. — tefra: see on II a 27. — 33 spantimaf : found only in the present passage, where spantimaf ... 34 etrama spanti ... IV 2 tertiama spanti make up a series. An adj. spantea occurs in Ila 30, but the exact sense of the latter
passage is not clear. Editors are nearly unanimous in taking spanti as ' side, ' and the sides are the front, right, and left of the altar, although it is not clear in what order ; the back of the altar is norm- ally designated by forms of postra. The etymology of spanti is uncertain; since there are no forms in the Latin alphabet, t may represent / or d, and we may have an i-stem from a root ending in a dental stop or a ti-stem from a root ending in n, but not a stem in dental stop -|- ti, which would yield /. See Lexicon. — pruse- katu: in Ila 28 two pieces of the dog were cut off as burnt-offerings and a third as an erus; here two, two, and three pieces of the sheep are cut off as burnt- offerings at three different sides of the altar. — efek: neut. sg. pronoun but referring to tuva tefra; as Latin parallels Bij. cites Plant., Trin., 152 ff. nummoram Philippeum ad tria milia. / id solus so- lum per amicitiam et per fidem j flens me opsecrauit suo ne gnato crederem; 403 ff. Le. minus quadraginta accepisti a Cal- licle. I estne hoc quod dico, Stasime ? St. quom considero, / meminisse uideor fieri. Le. quid factumst eo ? — pefume : see on VIb 24. — 34 strugla: on Via 59. — 35 eregluma: this word occurs eight times from the present passage through IV 19, under circumstances which show that, like asa 'altar', the ereglum is intended for the presentation of of- ferings. The etymology is uncertain, but connection with erus is as easy as any. We may derive from *aise-kelo- with PL, II, p. 26, or with Dev. we may assume that the suffix originally had instrumental value (-klo- < -tlo-'i) and that the for- mation came to be understood as a di- minutive, so that -glo- replaced -klo-. The ereglum has generally been taken to be a sort of side-altar or gift-table; Bii., PL, Bk., Blum., Pis. translate ' sacra- rium.' But Vetter gains an improved sense by taking it to be a cult-statue 'simulacrum (?). ' On etymological grounds this is as easy as the other view, since we can still maintain connection
210
III. TEXT AND TRANSLATION
puexnune : pupfike / (IV 1) purtuvitu : erarunt : struhglas : eskamitu : aveitu / (2) inumek : tertiama : spanti : triia tefra : prusekatu / (3) efek : supru : sese : eregluma : vesune : puemunes / (4) pupfiges : purtuvitu : struhgla : petenata : isek : / (5) afveitu : erererunt : kapifus : puemune : / (6) vesune : purtuvitu : asamaf : ereglamaf / (7) asegetes : karnus : isegeles : et • vempesuntres /
sent them at the statue to Pomonus Poplicus. Add the phallic portion of the same strusla cake. Then at the third side (of the altar) cut off three (pieces) as burnt-offerings ; present them from the top (of the altar) toward the statue, to Vesona of Pomonus Poplicus. Like- wise add a comb-cake. From the same bowls make the presentation to Pomo- nus and Vesona. At the altar and at the statue, with the uncut portions, the chopped meat, the vempersondra, and the
1 aveitu: for afveitu.
5 erererunt: for ererunt.
with eras and with Pael. aisis 'dis,' Marrue, aisos 'di,' Mars, esos ' di,' aiaoi • 6soi vnd Tvqqtjv&v Hesych. Moreover the new interpretation gains support from IV 13 ereglu timtu (for the practice of anointing cult-images see on Ila 38.) — rv 1 eskamitu: found only here and of uncertain sense, but it must .■iurely indi- cate some part of the cake called strugla. For the form of the strugla see on Via 59, especially the Festus passage. It is pos- sible that the eskamitu may be the panicula by which the cake is surmounted and also that it was a phallic symbol, as Bii. suggested; see on petenata in 4 be- low. Pisani, who follows substantially the view of Bii., finds a fairly satisfactory etymology by analyzing as a pass. pcpl. of a stem e-skami-, cf. Germ. Scham ' ver- gogna, parte pudenda. ' If we adopt this view, the prefix e- will be a prefix of per- fectivation, as in L. efficio, the verb stem will be a denominative of a stem belong- ing to Walde-Pokorny's (s)qombh-no- (II, p. 601), with semantic development from ' verkiirzt, verstiimmelt' through ' Scham,' and the literal sense of *e-skamito- will be 'that which has been made like a phallus. ' — aveitu : the misspelling seems partly due to crowding at the end of the line. The correct spelling is shown by af- 5 times, ars- 11 times, in addition to which ar- and ar- occur each once. —
3 supru sese : the two instances of tes- tru sese in III 23 and IV 15 are followed by asa, and it is not unMkely that asa is to be understood here also. The direc- tion in which the presentation to Vesona takes place is therefore from the top of the altar toward the statue. — Ve- sune Puemunes / Pupfiges: the name of the same goddess occurs in the same dative form Vesune on two dedicatory inscriptions: Co. 253 (from Antinum ; Vol- scian or Marsian) and 264 = C.I.L., P, 392 (from Ortona; Marsian-Latin). Of greater interest is the appearance of Ve- suna on an Etruscan mirror from Volsi- nu (Gerhard-Korte, Etruskische Spiegel, V, 35) wearing the chiton, nebris, and diadem and carrying the thyrsus. Her hand rests on the shoulder of the god Fuf- luns, recalling not only the association of Ariadne and Dionysus, but also of Libera and Liber and of Vesona and Po- monus in the passage now under discus- sion; see above on III 26. That Vesona is closely related to Pomonus, probably as consort, is beyond doubt, and her name in every instance in the Tables, with the exception of IV 6, is followed by Pue- munes Pupfikes (-ges). The formation of the name is similar to that of L. Bel- lona, Pomona, but the u is not graphic for o; rather it shows the change o > u, since the inscriptions Co. 253, 264 cited
IV 1
COMMENTARY
211
above, which are in the Latin alphabet, also have u. The root cannot be deter- mined with certainty. For attempts to connect the word with L. Vesta, with Vesuvius, and with Gaulish Vesontio (Be- san^on), Vesunna Petrocoriorum, etc., see Bii., p. 162, Rosenzweig, pp. 98-9, Dev., p. 391. — 4 petenata: practically all * editors translate ' pectinatam, ' the phon- ological correspondence to L. pecten being like that of Speture to L. spec to. The type of strugla offered in the present instance must therefore be a cake whose surface has been furrowed with an in- strument shaped like a comb or rake. The evidence for pecten as a symbol of female sex was assembled by Bii., p. 161, where the notion of eskamitu as a sym- bolic offering to the god Pomonus is balanced against the struhgla petenata offered to the goddess Vesona ; but it must be admitted that the evidence is not very strong: pecten is, among other meanings, the name of a type of shellfish or concha : Hor., S., II, 4, 34 pectinibus patulis iactat se molle Tarentum; cf. Pliny, N.H., IX, 101, 160, XI, 139, 267, XXXII, 150. concha is employed in Plant., Rud., 704, in an obscene pun: te (sc. Venus) ex con- cha natam esse autumant, cave iu harum conchas spernas. Cf. also Clem. Alex., Protr., p. 23 Dind. xrslg yvvaixelog, o iOTiv, Evq^rjficbg xal /uvarixcoi; einelv, fio- Qiov yvvaiKElov. — isek : the equivalent isec is found in VIb 25, and the form must be derived by addition of the enclitic -k to *esse, from the same pronominal stem as esu, etc. (109 e), with i- for e- perhaps through the influence of the stem i- represented in Latin by is, id. The adv. isek is probably one of manner, and a suitable translation is 'likewise, in like manner.' — erererunt: an error by dittography for ererunt. Not 'the same bowls (as previously), ' for none have been previously mentioned in III - IV, but the same bowls are to be used in the offerings to both deities. Dev,, p. 391, cites Roman evidence of attempts to discourage too close association of gods in
cult practice: Livy, XXVII, 25, 8-9 [pon- tifices] negabant unam cellam duobus diis recte dedicari ... neque duobus, nisi certis, deis rite una hostia fieri; Val. Max., I, 1, 8 cum M. Marcellus ... templum Honori et Virtuti ... consecrare vellet a collegia pontificum impeditus est negante unam cellam duobus diis recte dicari: futurum enim, si quid prodigii in ea accidisset, ne dinosceretur utri rem divinam fieri opor- teret. — 6 ereglumaf : the actual reading is ereglamar, but there are 5 instances of the ace. sg. in -um -u, and there is prac- tically no possibility of ereglam being a fem. adj. in agreement with asam. We must therefore follow the majority of editors in emending to ereglumaf. — 7 asegetes : this and the following line contain several difficulties, which may be partly solved by comparison with the instructions following the porrecfio of the dog (Ila 29-32) and of the goat (lib 17-21). In Ila 29-30 asegeta karne was taken as ' the uncut portion, ' and similarly in the present passage; but the use of the pi. asegetes karnus here may show that the instruction has reference to both victims, the sheep and the pig (sakre), as Dev., pp. 392-3, suggests. The contrast between vempesiintres here and sg. venperstintra Ila 30 streng- thens this belief. — isegeles: several editors including Br., Bk., and also Kent, T.C.I. , p. 37, take this to be an error for isegetes = L. insectis, non sectis, since we seem to need a pcpl. like ase- getes; but despite the bad orthography of IV the reading of the bronze here can be defended. We have either a stem in -li- like L. facilis or in -lo- like the verbal adjectives which seem to be the basis of fut. pf. forms like entelust, apelust, and which in some of the Slavic languages became the normal preterit. Since the latter formations are both predominant- ly active and not passive, the analysis of isegeles as a -/i-stem is to be preferred. The passive sense may be seen in L. fra- gilis, docilis, and although -is -is is the usual dat.-abl. pi. ending of i-stems in
212
III. TEXT AND TRANSLATION
(8) supes : sanes : pertentu : persnimu : afpeltu / (9) statitatu : veskles : snates : asnates : sevakne / (10) ere^luma : persnimu : puemune : puprike : vesune / (11) puemunes : pupfikes : klavles : persnihmu / (12) puemune : pupfikes : et : vesune : puemunes / (13) pupfikes : pustin : ereglu : inuk : ereglu : umtu / (14) putrespe : erus : inuk : vestigia : mefa : purtupite / (15) skalgeta : kunikaz : apehtre : esuf : testru • sese / (16) asa : asaxna : purtuvitu * sevakne :
uncooked under-parts stretch forth (the hands), pray, approach (the altar), and stop. With the ceremonial vessels, wet and dry, pray at the statue, to Pomo- nus Poplicus and Vesona of Pomonus Poplicus. With the smearing-sticks pray to Pomonus Poplicus and Vesona of Po- monus Poplicus at their respective sta- tues. Then anoint the statues (and dis- tribute) the erus of each (deity). Then kneeUng apart at the right side from the altar you yourself shall present at the altar from a cup a libation and a mefa cake, to Purdovif, and pronounce
12 pupfikes: for pupfiko.
Umbrian, there is some support for -e- in aves la 1, sevakne IV 9. A Latin form i(n)siciiim, -a, signifying some sort of chopped meat, is known from several literary passages: Varro, L.L., V, 110 insicia ab eo quod insecta caro, ut in car- mine Saliorum est quod in extis dicitur nunc porrectum. Macr., Sat., VII, 8, 1 isicium, quod ab insectione insicium dic- tum: amissione enim litterae postea quod nunc habet nomen obtinuit. Donatus, ad Ter. Eun. 257 fartores qui insicia et farci- mina faciunt. — vempesuntres : see on VIb 24, Ila 30. — 8 supes sanes: for supes see on VIb 5. sanes is doubt- less the etymological equivalent of L. sa- nis, but its precise sense is not clear. ' Whole, not cut up ' is probably too close to the sense of asegetes just above. It is safest therefore to talie it with Dev. as equivalent in sense to 'crudis,' that is, 'uncooked' {'quae inollis non reponuniur'). — pertentu: see on Ila 30. It is ap- parently transitive there, but intransitive here, being followed by three other verbs persnimu afpeltu statitatu, which would not govern an ace. *supa sana as object. — persnimu: in Ila 29-31 the arrangement is different, and persnihmu
is not one verb in a series quite as here. The instructions called for prayers in connection with the part of the victim called ampefia, with the uncut portion, with the venpersuntra; then the plac- ing of the under-parts at the side of the altar, then the prayer with the vessels, then the series vestikatu ahtrepufatu afpeltu statitatu, the first two of which are not found in III - IV. — afpeltu / statitatu: on Ila 32. — 9 veskles snates asnates: see on II a 19. The pas- sage most closely parallel, however, is Ila 37, where the priest prays (to Hondus) at the Obelisk just as he prays here to Pomonus and Vesona at the statue. — sevakne: the omission of -s in the dat.- abl. pi. ending of i-stems is unusual, but the space at the end of the line here is insufficient for the word. Cf. also the dat. pi. forms Klaverniie, Satane, Imes- kane, Peraznanie in lib 3-7. — 11 klavles: see on Ila 33, where the sense is closely similar to that of the present passage. — 13 pustin: this preposition has distributive force in Ila 25 pustin angif, as has also pusti posti in several passages of Va, b. In the present passage, however, its force must be local, the more
IV 7
COMMENTARY
213
so when we consider the resemblance be- tween 'pray with the spoons (or sticks) behind the statue' and 'pray with the ceremonial spits beyond (that is, on the other side of) the obelisk; likewise with the spoons (or sticks) ' Ila 35-6. — tuntu: although there are certain dif- ferences of detail, the resemblance to Ila 29-38 continues: in 38 the Obehsk was to be anointed. — 14 putrespe: taken by nearly all editors as gen. sg. equivalent to L. utriusque; that O.-U. pronouns form the masc.-neut. gen. sg. in -eis, unlike L. ems, cuius, etc. is known from O. eizeis, U. erer. There is no ad- vantage, therefore, in taking putrespe as a plural dativus sympatheticus with Havers, G/., V (1914), p. 2. The interpre- tation of the passage suffers both from the uncertainty of the meaning of erus and from the lack of a verb. For the former see on VIb 16, where the view adopted is that the erus is a part of the victim distributed at the end of the cere- mony to certain persons present. Since putrespe erus cannot be in apposition with the action indicated by the preceding ereglu uixitu, we must assume an ellipsis, probably of teftu, but it will be no more startling than the omission of the verb in I a 18-9 kapif purtitaf sakref etraf purtitaf etraf / sakref, II a 25 tiu puni tiu vinu, etc. — vestigia mafa : the of- fering of the libation and cake to the god whose name is written Purtupite close- ly resembles that in honor of Fisovius Sancius in VIb 5 immediately after the sacrifice behind the Tesenacan Gate; cf. also Vila 37. The resemblance is increased by the inclusion of skalgeta kunikaz = scalsie conegos 'from a cup, kneeling.' — Purtupite: hardly a corruption of purtuvitu, which would stand after, not before, ktinikaz, as in 18, 20 below, and would be largely superfluous in view of purtuvitu in 16. Besides, a divine name in the dative case is needed here, and by almost any reasonable view we have the title of a deity in some way associated with purtuvitu, the verb used to desig-
nate the porrectio or second stage in the sacrifice, just as Speture Ila 5 is asso- ciated with augury (spectio), Ahtu Ila 10, 11, with oracular utterance, etc. It is natural to wish to keep the reading of the bronze, but there is difficulty both in the proposal of Bii., who compares L. Seis- piiem, Sospifam, with the root found in potestas, for the vowel-weakening which would have to be assumed is extremely rare in Umbrian, and in that of Pisani, who compares Skt. pra-dapifa-, pass. pcpl. of the causative of da- ' give, ' for there is scarcely any evidence of a causative in -pey/pi- outside of Indo-Iranian. We must therefore follow the majority of editors in emending to Purtuvite, an easy correction, since the only difference between v and p is in the presence or ab- sence of the lower cross-bar. The most satisfactory solution will then be to derive purtuvite from a pres. pcpl. *por-dowin- tei with unthematic inflection of the tense- stem, as if Latin had *audinti instead of audientl, and with loss of n before t (== d from earlier 0; cf. 52 d, 60 a. *Purtuvif, the god who assists in the porrectio, may be identical with Jupiter; so much at least seems possible from a comparison with III 22-3. — 15 apeh- tre: similar to dne^, L. ah + extra, but, like O. ehtrad, formed from *ek- without s. The priest in making the present of- fering kneels at a slight distance from the altar, at the right (testru sese asa), per- haps just outside the rectangular space which immediately encloses the altar, and makes his presentation toward the altar (asama). — esuf: see on Ila 40. — 16 sevakne: the form can be ace. sg. or ace. pi., and it is not clear just what the adfertor declares to be sevakne. For a list of things to which the word is ap- plied, see on Ha 21. It is not hkely that we should assume an ellipsis of tiglu, comparing the expression tiglu sevakni teitu (naratu) III 25, 27. It is also un- likely that at this late stage in the cere- mony the victims would be declared se- vakne(f). The word is never used with
214
III. TEXT AND TRANSLATION
sukatu : / (17) inumek : vesve9a : persuntru : supu • ere9le : hule / (18) sevgikne : skalgeta : kunikaz : purtuvitu : inuntek / (19) vestigia : persuntru : turse : super : eregle : sevakne / (20) skalgeta : kunikaz : purtuvi6u : inumek : tehterim / (21) etu : veltu : efek : persuntre : antentu : inumek / (22) arglataf : vasus : ufestne : sevaknef : purtuvitu / (23) inumk : pruzufe : kebu : sevakne : persnihmu / (24) puemune : puprige : inumek :
(the offering) perfect. Then kneehng of- fer at the bottom of the statue a libation and persondro from a cup to Hula. Then kneeling offer at the top of the statue a solemn libation and persondro from a cup to Tursa. Then go select a cover ; place it upon the persondro. Then offer ceremonial circlet-cakes from sealed ves- sels. Then with the remaining ceremo- nial food pray to Pomonus Poplicus. Then at the litter pray with ceremo-
17 vesvega: forvestega. inumek.
18 iimntek: for inumek ( MM for \Y\).
23 inumk: for
uestisia or me fa, but is used with vinu in Ila 39 and with the cakes known as arglataf below in 22. It therefore seems not unreasonable to take it in the present passage as ace. pi. applied to both the libation and the cake. — stikatu: a dna^ Xeyoi^evov clearly equivalent in sense to deitu or naratu. It is most satisfactor- ily explained, with Bii., PI., I, p. 276, and others, as cognate with pru-sikurent, L. in-sece, Gk. evvsne, evi-ane, all from root *sek'^- 'say.' The vowel-gradation and stem-type are the same as in Lith. sakaii, sakyti 'say,' and the develop- ment of the labio velar to k rather than to p may have resulted from the analogy of forms in which k'^ stood before t and so lost its labial element. — 17 vestega: the bronze reads vesvega, which is un- doubtedly an error. — supu: in obvious contrast with super in 19. Since the exact nature of the ereglu is uncertain, there is some doubt as to whether the libation and persondro are offered at the foot of it or actually under it. If it is a table, the latter is possible. In any case the presentation under or at the foot of the ereglu seems to symbolize the chthonic or infernal character of Hula, in contrast to Tursa, to whom in 19 a
similar offering is made above the ereglu. — Hule : known only from this passage, but the similarity between 17-18 and 19-20, where the offering is made to Tur- sa, makes it plain that Hule is the name of a deity. The name has been compared with holtu, a verb of uncertain sense in the curse in VIb 60 = Vila 49; with Horse = Hufie in VIb 43 = lb 2; and with the Roman god Helernus, cf. Ov., F., II, 67-8 Tunc quoque vicini lucus celebratur Helerni, qua petit aequoreas advena Thybris aquas, with Frazer's note. Just as Prestota Serfia and Tursa Ser- fia are associated in the ceremonies of the lustration in VIb 57 ff., so Hula and Tursa are associated here, and just as the chthonic character of Prestota seems to be symbolized by the offering of dark- colored pigs (red or black) in Vila 6 = lb 27, so the chthonic character of Hula is symbolized by the fact that the of- fering is made to her below, to Tursa above, the ereglu. On purely linguistic grounds it cannot be determined whether Hule is masc. or fem., but the great majority of editors take the deity to be female. — 19 Turse: see on VIb 57-58. — super eregle: the chiasmus supu eregle Hule ... Turse super
IV 16
COMMENTARY
215
eregle places the last phrase in emphatic position. — 20 purtuviOu: the spel- ling with the character o occurs also in furfae lb 1 = furfant VIb 43. — tehtefim: almost certainly a cover to be placed over the personaro, as appears from the next line. Vetter takes it to be a cover used for baking the persondro, comparing Cato, R.R., 76, 4 testo caldo operito, pruna insuper et circum operito. The root is that of L. tego, tectum, etc.; the stem-formation is uncertain, but is probably in -dio- or -lio-, if the conditions for the change I > r rs in 55 h are cor- rectly formulated. — 21 veltu: usually taken as related to ehueltu, ehvelklu, from the root of L. volo, velle, and trans- lated 'deligito.' Since veltu is preceded by etu 'go, ' we may assume that there is no cover at the actual scene of the sacri- fice but that the adfertor or some other official has to go a short distance and select one from the place where they are kept. The form veltu can as easily be supine as imv., and it makes little dif- ference for the sense. Its position, how- ever, is the decisive factor, for the su- pine a(n)seriato in Via 1, 6, b 48 = lb 10 stands before, not after, the verb of going. — 22 arglatai : nearly all edi- tors explain with the help of Festus, 110 Li. arculata dicebantur circuit qui ex farina in sacrificiis fiebant. Arculum appella- bant circulum quern capiti imponebant ad sustinenda convnodius vasa quae ad sacra publica capite portabantur. We might be tempted to regard these cakes as crescent-shaped, but since the arculum in the second Festus passage was plainly a full circle, it is safest to assume the same for the arculata and the arglataf of our passage. They may then have had some resemblance to doughnuts. — ufestne: quite obscure and without a clearly marked case-ending; but while sevaknef plainly goes with arglataf, it is most reasonable to take ufestne with vasus. Bii. explained the word as a cognate of L. offendix, offendimentum, which are known from Festus, pp. 222-3
Li. as terms signifying the knot of the strap by which the apex was bound under the chin; further cognates would be Goth, bindan, Eng. bind, Skt. badhnati all from root *bhendh-, and the pit. re- construction of our stem would be *op- festo-no-. The vessels then are to be sealed or otherwise fastened shut, and the suffix -no, the appearance of which fol- lowing the participial stem has caused some embarrassment, probably serves to designate the type of vessel, where the participial stem alone would mean that they were actually closed. Biicheler's theory is substantially followed by most scholars and adopted, at least tenta- tively, here ; Pisani, however, derives with the suffixes -sti-no from the stem of the dialectal word reflected in L. of fa. — 23 pruzure kebu sevakne: best treated together as forming a single problem. kebu from any viewpoint is perplexing because of the unpalatalized k before e. If it is equivalent to L. cibo, as practi- cally all editors assume, the two forms cannot have been independently inher- ited from a form with IE bh (which would give U. /) nor with IE g'^ (which would give L. v); IE b is possible, but was very rare. Since there are no known etyma outside Italic, it is best simply to assume that U. borrowed the word from L., or that both borrowed from a common un- known source, the Umbrian borrowing having taken place so late that the k failed to change to 9. pruzure in a form- al sense is sometimes compared with proseda used by Plant., Poen., 266 in the sense 'prostitute'; cf. Festus, p. 252 Li. prosedas meretrices Plautus appellat, quod ante stabula sedeant. Eaedem et prostibulae. This is of little help in the explanation of our passage, but it is striking that the rare character z which occurs in pruzure occurs also in zefef = serse (L. sedens). pruzure may then be derived from *pr6(d)-sod-i with the o-grade of the root *sed-. It is generally translated 'praestante' or 'prostante,' the contrast being between the food which
216
III. TEXT AND TRANSLATION
kletra : veskles : / (25) vufetes : sevaknis : persihmu : vesune : / (26) puemunes : pupf ges : inumek : svepis : heri : / (27) ezariaf : antentu : inumek : erus : ta9ez : / (28) tertu : inumek : kumaltu : afkani / (29) kanetu : kumates : persnihmu : esuku / (30) esunu : ufetu : tapistenu : habetu : pune / (31) frehtu : habetu : ap : itek : fakust : purtitu / (32) futu : huntak : pif i
nial vessels of whichever type you wish to Vesona of Pomonus Poplicus. Then if anyone wishes, he may place the food- baskets upon (the htter). Then silently distribute the erus. Then grind (the grain). (The flute-player) shall play the accompaniment. Pray with the ground (grains). In connection with this cere- mony load (the fire with incense). One shall have a caldron and have boiled mead. When he has done thus, (the ceremony) shall have been completed.
25 persihmu: for persnihmu.
26 pupfges: or pupfiges.
28 tertu: for teftu.
is still on hand and that which has been presented in the porrectio or distributed as erus. The cakes may be included as well as the remaining parts of the victims, since the mefa was offered in 14 and the ♦arglatas in 22, while the grinding does not take place until 28. — 24 kletra: mentioned here for the last time. There Is some disagreement with regard to the final disposition of the litter; it was burned according to Bii., p. 154, on the evidence of esunu uretu 30; not so ac- cording to Vetter; Dev. suggests that it was simply taken apart. There does not seem to be sufficient ground for believ- ing that it was actually burned. — 25 vufetes: see on Ila 31, where it was suggested that vufetes may mean L. ' libitis, ' that is, that the exact type of vessel to be used is left to the discretion of the adfertor. This translation in the present passage seems not to involve any more difficulty than the usual 'votis. ' — 27 ezariaf: very uncertain, but usually translated 'escas,' or by some other deri- vative of the root *ed- ' eat. ' Ribezzo, II, p. 100 compares Festus, 54 Li. es- cariae mensae quadratae vocantur,in quibus homines epulantur, and translates 26-7 inumek ... antentu 'tunc si quis vult, epularias (mensas) imponito.' Direct equivalence between ezariaf and escarias
is impossible on phonological grounds, but *ed(e)s-dria- will provide a satisfactory source for the Umbrian form. It is still not certain, hov.ever, just what is to be done with the ezariaf; according to Bii. they are to be added to the fire, just as at the funeral of Julius Caesar private individuals made offerings after the con- clusion of the public sacrifices; according to Dev. they were simply to be added to the remains of the victims; according to Vet. they were food-baskets and were to be placed on the kletra and taken back to town without being a burden to their owners, I have adopted his explanation as being the most satisfactory, especially in view of svepis heri, which shows that the action is optional, or at least that no obligation rests upon any particular in- dividual and that therefore it may not be a part of the ritual at all. — erus tagez / teftu: there was already a distri- bution of the erus in 14, where, however, the sense is slightly obscure because of the ellipsis of the verb. Twofold distri- butions of the erus occur also in VIb 16, 16 (in honor of Fisovius Sancius), 38, 38 (Tefer Jovius), Ila 28, 32 (Hondus). — 28 inumek kuxnaltu : the distribution of the cms is similarly followed by the grinding (of the cakes) in la 34 and Ila 9; compare also the complex procedure
IV 23
COMMENTARY
217
in the ceremonies of tlie lustration, at Vila 43-4 = lb 34-7, where three sacri- fices are in progress at once at stations a short distance apart. — afkani / kanetu: both are ana^ Xeyoneva, yet tlie etymology and sense are reasonably clear. The Latin equivalent to afkani would be *accinium, while actually known cog- nates include accano Varro L.L., VI, 75 and accenhis, first known from Quinti- lian, I, 5, 22 as an apparent loan-trans- lation of TiQoawSia but found in Amm. Marc, XVI, 12, 36 of the sound of bra- zen trumpets, ksjietu corresponds to L. canito but is with much better reason assigned to the second than to the third conjugation, since there is no plausible explanation of why a short vowel before the / would escape syncope. It is quite possible that kanetu represents an in- tensive type corresponding to L, cantare as upetu seems to correspond to op tare (cf. 118 b) but weakened in meaning just as Italian cantare, French chanter show weakening of the proper meaning of L. cantare. Who does the singing is not fully clear, but Bii. is probably correct in taking it to be a flute-player playing his instrument rather than the priest chanting a formula. The purpose of the music is presumably to drown out any other sounds M^hich might be ill-omened. — 29 esuku / estsnu ufetu: the conclud- ing portion of Table IV is almost hope- lessly obscure in certain details, especial- ly because of the uncertain nature of the article called tapistenu and the uncertain meaning of frehtu. The form esunu could be either ace. or abl., but the latter is preferable because otherwise esu(ku) would be left with no noun to which it could refer, and because esunu as ace. would not be a suitable object for ufetu. This verb in III 12 governs pir, and in the present passage w^e may have to as- sume an ellipsis of the same noun; Dev. in fact does assume such an ellipsis. On the etymology of ufetu and the possibi- lity of its meaning 'load (the fire) with incense' see on III 12. If this view is
correct, *ofe- in contrast to L. olere < *odere must have a causative sense which however is not outwardly shown because the two characteristic marks of the IE causative, o-grade vocalism and stem in -eye jo-, were already present. The sense 'in connection with this sacrifice load (the fire with incense) ' is more satisfac- tory than a mere instruction to kindle or heap up the fire, which would be quite pointless in view of the fact that the fire was kindled some time previously. — 30 tapistenu: there is no sure hope of a solution, but since the view of pune frehtu adopted here is that it is boiled or warmed mead and not chilled mead (see on Ila 26 frehtef), tapistenu may be the name of the utensil in which it was heated. Bk., MuUer, and Pis. translate ' caldariolam,' deriving from the same root as L. tepor, etc. The a in place of the expected e must represent a reduced grade, and the explanation of the stem offered by Bii., v^^ith slight adaptations, is as good as any: "a tepore fit tepesta ut ab honore hones ta, exinde tepestinum ut ab sale salinum." — pune / frehtu: the mead v;as to be used for putting out the fire according to Bii., who compares Verg., Aen., VI, 226-7 postquam conlapsi cineres et flamma quievit, reliquias vino et bibulam lavere favillam. He takes frehtu in the sense ' frigidam' ; of course either hot or cold mead could be used for extinguishing the fire, although there is no positive assurance that this was its purpose. The reasons for taking frehtef as 'boiled portions' were discussed in connection with Ila 26, and it is most reasonable to derive frehtu in the present passage from the same source. The use of the utensil called tapistenu and of the mead which was presumably heated in it must have served some ritual purpose which is not altogether clear. — 31-2 purtitu / futu: the same expression oc- curs in Ila 43 at the end of the dog-sacri- fice, with esunu immediately preceding, and esunu is possibly to be supplied in sense here; compare the optional use in
218
III. TEXT AND TRANSLATION
: prupehast : efek / (33) punes : neifhabas /
ures : When he has purified the jar, thereafter
they shall not use any of that mead.
Vila 45 enom purditom fust — lb 38-9 enu estmu purtitu fust. — 32-3 hun- tak piri prupehast, efek / ures punes neifhabas: this final sentence is not a part of the instructions for the sacrifice proper but a special precaution relating to the jar (or vat?) the purification of which was prescribed in III 3, on which see note. There are several difficulties of grammar and interpretation. For exam- ple, it is not fully clear whether pifi and efek are pronouns or whether pifi is a conjunction and efek an adverb, both being of course pronominal in origin ; and there is some doubt whether the in- junction is against the use of the jar or of the mead which is in the jar. It is almost certain, however, that ures punes is not dat. pi., as it was taken by Bij., PL, and Bk., but gen, sg., as it is taken by Blum, and all subsequent editors, since poni- by almost any of the various mean- ings assigned to it is a noun of material and therefore unlikely to be used in the plural. As a gen. sg. it must be partitive. It will now be most unnatural to translate 'let them not use it for (any) of that mead ' because the gen. would have to be a par- titive indirect object or substitute for some phrase equivalent to L. ad -\- ace, and also because ures ' that ' would have
no clear reference; it could scarcely refer to the hot (?) mead in 30-1, since our present passage belongs in thought rather with the opening lines of III than with the conclusion of the sacrifice. The in- junction is therefore against the use of the mead, presumably for private or sec- ular purposes, and ' that mead ' means the mead which is in the jar that has been purified. By this interpretation efek as a pronoun becomes superfluous, and pifi ... efek must be taken as correlatives in the sense 'when ... thereafter.' The two verbs in the sentence call for brief ex- planation, prupehast is fut. of pehatu compounded with pru-, but the corres- pondence with prumu pehatu in III 3 should not lead us to assume that pru- pehast was erroneously written for pru- mu pehast, as A.-K. suggested, neif- habas is from *nei afhabas (34), pres. subjunctive of the verb equivalent to L. adhibeo. It is hard to see why the form might not be 2 sg. with indefinite reference ' do not use, ' but practically all editors take it as 3 pi. (-as < -ans). The reference is still indefinite' let them not use.' The shift from the second- to the third-conjugation type should be noted; cf. L. tuor beside tueor.
IV 32
COMMENTARY
219
Va 1 - b 7. Two decrees of the Atie- dian Brothers prescribing the duties of the adfertor in procuring the materials for the sacrifice, the remuneration to which he is entitled, and the penalty im- posed on him if he has failed to perform his duties satisfactorily. The native al- phabet is still used, but rhotacism of final s is regular when the s has not re- sulted from simplification of a consonant- cluster (57 e). For the relative age of V see Introduction, 17-19.
Va 2 eitipes: only here and in 14. It is almost certainly a 3 pi. pf. form of a verb whose approximate meaning is ' de- creed, resolved, ' but the etymology is uncertain. Probably from *eit(om h)epens ' ratum habuerunt' with *eitom < *aik^- (o)tom with early syncope followed by loss of labialization as in 49 d (cf. O. aikdafed. 'decreed'); for *hepens see 124 c. Devoto, p. 401, makes the verb a denominative to *eituk'^o- from the same root as O. eitua 'money' ('that which is official') but, unlike L. manducare, which he compares, our form would have to be a denominative of the 3rd conjugation, and moreover PL, I, p. 134, on whom his suggestion is based, does not account satis- factorily for the present stem. — ple- nasier urnasier: variously interpreted. plenasier is apparently 'belonging to the full (year) ' with reference to a festival occurring when the year is completed. urnasier was formerly taken to be con- nected with L. urna and understood either as a designation of a festival or as a re- ference to the use of urns in voting. Lin- de, GL, III (1912), p. 170, proposed to equate it with L. ordinariis ('regular'; d was lost after syncope of the vowel which originally followed it), and his new inter- pretation has been adopted by almost all subsequent editors. — uhtretie : for the uhtur or auctor see Introduction, 18, and note on III 4. The phrase oht C. U. Uistinie Ner. T. Babr. on an Umbrian inscription from Assisi (no. 355 Co.), perhaps of the SuUan period, gives evi- dence of an eponymous dual magistracy
of similar name there. In Tables lb 45, I la 44, however, the date is given by means of kvestretie. — 3 T. T. Kas- trugiie: the father's praenomen stands between the son's praenomen. and nomen ; so also in some of the minor Umbrian inscriptions and in Volscian and certain other dialects, in contrast to the arrange- ment in Latin, Oscan, and Paelignian, illustrated by the formula Q. Marcius L. f. S. Postumius L. f. (C.I.L., P, 581 = Senatusconsultum de Bacchanalibus). See A.J.P., LXXII (1951), p. 118. — 4 eikvasese: only here and in 16, but cf. eikvasatis in III 24, 29. There is no clear etymology, but both words are pro- bably related to L. aequus, and may have reference to collegae united in a collegium, eikvasese is in all probability abl. pi. 4- postposition -e < en. The ei, the failure to show p < k^ it connected with L. aequus (qu of labiovelar origin: cf . Skt. ekas and 37), and the two instances of unrhotacized s all suggest that a very archaic form of the word has been pre- served through ritual use. The expression *fratruse Atiiefier might rather be ex- pected, but eikvasese may possibly de- signate a smaller group among the bro- thers, eligible for duty as adfertor. — ri esune: cf. L. re divina facta, Plant., Amph., 968, etc. For the dat. with ku- raia see 143 b. — 6 si : for the use with herte see 152 c. — pure: this relative form recurs in 25, 28, b 4 with masculine antecedent and clearly refers to per- sons. — sis: equivalent to sins, L. sint. It depends on herte, whose force carries over to the end of the sentence. — 6-7 sakreu / perakneu: the expres- sion probably includes not only victims, in which case we might expect the masc, but all materials for the sacrifice. For the spelling see 10 c. For a discussion of the etymology and interpretation of pe- rakne see on I la 5. — 7-8 revestu pure terte / eru emantur herte: for the syntax see 138 c, 142 b, 152 c. — 9 tribficu: the abl. tribirsine is used in Via 54 in connection with the propitia-
220
III. TEXT AND TRANSLATION
(Va 1) esuk . f rater : atiiefiur :
/ (2) eitipes : plenasier : urnasier : iihtretie / (3) t : t : kastrugiie : arfertur : pisi : pumpe : / (4) fust
: eikvasese : atiiefier : ere : ri :
esune : / (5) kuraia : prehabia :
pif e : uraku : ri : esuna : / (6) si : herte : et : pure : esune : sis :
sakreu : / (7) perakneu : upetu :
revestu : pure : tefte : / (8) eru :
emantur : herte : et : pihaklu : pune : / (9) tribrigu : fuiest : akrutu
: revestu : / (10) emantu : herte : arfertur : pisi : pumpe : / (11)
fust : erek : esunesku : vepurus :
felsva : / (12) afputrati : fratru : atiiefiu : prehubia : / (13) et : nufpener : prever : pusti : kastruvuf : /
The Atiedian Brothers resolved as fol- lows at the regular annual festival dur- ing the auctorship of Titus Castrucius, son of Titus. Whoever shall be adfertor among the members of the Atiedian col- lege, he shall look after the religious ceremony, and provide what is neces- sary for that ceremony, and decide what persons shall be present at the ceremony. He shall select the sacred articles and those brought from elsewhere. Whatever are offered, he shall examine whether any of them should be accepted, and when there is to be a triad of propitiatory offerings, he shall make an inspection in the field, whether they should be ac- cepted. Whoever shall be adfertor, he shall provide vegetables in connection with the unburned sacrifices, at the discretion of the Atiedian Brothers, and at the rate of one new dupondius per head.
tory sacrifice of oxen. For the formation see 60 e, 88, and Lexicon, — fuiest: a future nearly equivalent to h.fiet (120 a, 123). — akrutu: found only here, and very uncertain because the k may stand for either a voiceless or a voiced stop, and the -tu may be the same post- postion 'from' found in scalseto, pureto, etc., or else the imv. sg. ending. Earlier editors (Br., Bii., PL, Bk.) took the first part as equivalent to L. agro, but Blum, felt the need for a verb before ravestu to balance upetu revestu in 7, and pro- posed (p. 47) to derive akrutu as imv. from *akriut6, translating 'saginalo' and assuming a connection with peracri- ' sehr fett. ' Devoto, p. 404, made akrutu an imv. from *agrdt6d ;^s cognate with Aeol. dyQiiut 'catch' and rendered 'sumito,' but in T.G., pp. X, 69, he reverts to the old interpretation and translates ' dal campo (luogo) d'origine.' For an alto- gether different interpretation see Ri-
bezzo, I, p. 190 (' deinceps,' Ital. 'da capo,' with formation partly as in dy.Qodev; followed by Pis. and Vet.). It is probably safest to follow the traditional view and maintain connection with L. ager, agro. The loc. *akre might appear more na- tural, but we may compare certain uses of L. a, ab with stationary value, as in ab ea parte, ab oriente, etc., and antici- pation of emantur may also have been a contributing factor. — 11 vepurus: found only here but almost certainly cognate vith vepuratu, I la 41, and ge- nerally understood in the sense 'without fire.' Blum., p. 43, takes esunesku ve- purus in the sense ' bei unreinen Op- fertieren,' the felsva then being a fine which the adfertor must pay for his failure to procure proper victims; a satisfactory interpretation of vepurus, since its Greek analogue dnvQoc. can, like dOvrog, mean 'unfit for sacrifice,' but Blum, does not offer any etymology for felsva, nor does
Va9
COMMENTARY
221
he make it clear why a different term than the muta mutu of Vb 2, 3, 6 should have been used for a fine. It might be possible to find etymological support for his interpretation through Brugmann's connection with Goth, fra-gildan 'ver- gelten ' (Ber. kon. sacks. Ges. Wiss., 1890, p. 226), but he does not here cite Brug- mann, nor does Brugmann himself take felsva to refer to a fine. Devoto, p. 407, translates 'verbis,' connecting with enoq (< *wek'^os), but apart from the sense it is difficult to see how the medial syllable would escape syncope (as in 29 j); the traditional etymology gives us u < u, not u < 0, and in any case the lack of syncope in this instance could be ex- plained by analogy with the simplex pur-. For use of the term ' fireless ' of vegetable offerings cf. Eur., fr. 912 (Nauck) av 8e fioi Qvaiav anvQov nayxaqneiac, de^ai nXriQrj TCQoxvBelaav — felsva : found on- ly here and variously interpreted. It seems best with Bii., Bk., Mu., (PL un- certain) to maintain connection with L. holus and to derive from pit. *xeles-wa, with / of dialectal origin as in L. folus (Paul, e Fest. 74 Li.) beside normal holus. — 12 arputrati: too similar to L. ar- bilralu to be reasonably taken otherwise than as its equivalent, even though the p could stand for p as well as for b. R. G. Kent {T.A.P.A., LVII [1926], p. 56, fn. 29, following a verbal suggestion of J. Whatmough, suggested contamination with the equivalent of L. putare 'think' in order to explain the u contrasting with L. /, an explanation which would not work if we should render 'adventui' with Dev. ; moreover the type of stem- formation here {-tr-atu-) is not character- istic of abstract nouns from verbs of motion in Latin and was probably not used in making such nouns in Umbrian. Since the quantity of felsva to be fur- nished is fixed by the phrase nufpener prever pusti kastruvuf in 13, the sense of arputrati fratru Atiiefiu must be that the question whether to include felsva is left to the discretion of the
brothers. — 13 et: used here to intro- duce new matter defining more precisely what has preceded, in much the same way as Germ, und zwar. — nurpener: obscure, but a comparison with 17-18 mimeklu habia numer prever pusti kastruvuf shows clearly that it has re- ference to money. Biicheler's nullipondiis must be discarded because it does not give a satisfactory basis for deriving r from /. f therefore must have come from d before syncope of the vowel which fol- lowed. Blum., pp. 44-5, connected the first member with L. nudus in the sense ' ohne alle Abziige, ' Ital. ' netto, ' the whole compound then meaning literally 'Nackt- gewicht' and standing for as librarius. Devoto, p. 405, followed him in his inter- pretation but took the word to be a Roman borrowing, because of the diffi- culty of deriving nuf- from *nog^edho-, a difficulty perhaps not insuperable, since there is some possibihty that the medial stop varied between IE dh and d, and nuf- < *n3do- < *nobdo- < *nobedo- < nog^edo- might be admissible, although decidedly compUcated. Pauli, Altitalische Studien, V (Hannover, 1887), p. 87, with- out discussion of the sense, derived from novido- as an extension of the stem equi- valent to L. novus; there is no support for such a stem (L. albidus beside albus being rather a complement to albeo after the usual correspondence of adj. in-idus to vb. in -eo), but the suggestion of R. G. Kent, Lang., IX (1933), p. 215, deriving from *novo-du-pend-iyo- 'new dupondius, ' may be the correct answer. If the cur- rency reform assumed by such an inter- pretation was a local one, it is not possible to say more about it, but the reference may be to a change in the Roman standard. According to Pliny, N.H., XXXIII, 44, the Roman as was reduced during the First Punic War to 4- of its previous value, and again, in the dictatorship of Q. Fa- bius Maximus during the Second Punic War, to ~ of its original value. The no- tion of the spread of Roman currency favors the second rather than the first
222
III. TEXT AND TRANSLATION
(14) f rater : atiiefiur : esu : eitipes : plenasier : / (15) urna- sier uhtretie : k . t. kluviier :
kumnah/kle : atiierie : ukre : eikvasese : atiiefier : / (17) ape : apelust : muneklu : habia : numer : / (18) prever : pusti : kastruvuf : et : ape : purtitu : / (19) fust : muneklu : habia : numer : tupler : / (20) pusti : kastruvu : et : ape : subra : spafu : fust / (21) muneklu : habia : numer : tripler : pusti : / (22) kastruvu : et : ape : frater : gersnatur : furent : / (23) ehvelklu : feia : fratreks : ute : kvestur : / (24) sve : rehte
The Atiedian Brothers resolved as follows at the regular annual festival dur- ing the auctorship of C. Cluvius, son of Titus. At the Atiedian assembly on the Mount, among the members of the Atie- dian college, when (the adfertof) has slain (the victims) he shall receive a do- nation of a single nummus per head, and when the presentation has been made he shall receive a donation of two nummi per head, and when the distribution has been made he shall receive a donation of three nummi per head, and when the brothers have dined the broth- er-superior or the quaestor shall call for a vote whether (the dinner)
of these changes, and the relative age of Table Va, which is in the native alphabet but shows rhotacism of final s, and con- tains the Roman official title kvestur = quaestor (line 23), may also be taken into account. — prever: for the sense cf.' Aul. Gell., X, 20, 4 veteres priva dixe- runt quae nos singula dicimus; Lucr., V, 733 inque dies privos aborisci, of the notion of the moon being destroyed and replaced by a different moon every day. — kastruvuf: the interpretation of this word, which has varied between ' fundos ' and 'capita,' depends on a comparison of its four occurrences in Va with the recurring formula nome nerf arsmo ueiro pequo castruo fri in Via, b, Vila, and with O. castrid in line 8, castrous in line 13 of the Tabula Bantina (no. 28 Co.). The Oscan passages seem to show a contrast between criminal charges involving a fine (eituas) and those involving the death penalty; so Buck (p. 236, n. 1) interpreted the forms, but in his second edition he adopted ' in fundos ' in place of 'in capita' and is followed by Pis., Vet., and Bott. The interpretation of the two Oscan forms as ' praedio, -i' rather than ' capite, -is' is ably defended by Vet., pp. 19-20, and it is natural to
suppose that the Oscan and Umbrian forms have the same meaning; yet they may represent two different stages in the semantic evolution of the same word, whatever its precise etymology and se- X mantic history were (cf., for example, \ the difference between the two etymolo- "^gjcal equivalents U. karu 'piece' and L. earo 'meat'). In any case the sense 'capita' is more suitable in the present passage and also in the recurring formula uiro pequo castruo, where uiro (and hence also pequo) appears to be gen. pi. rather than ace. pi., so that the meaning 'fundos ' would scarcely fit; see on Via 30. There remains the problem whether the meaning is ' per person ' or ' per victim. ' The last part of the sentence shows an awkward lack of balance, with its absence of any verb for the words following prehubia, but by any interpretation the force of prehubia must carry over to the end of the sentence. nufpener, if taken in the sense suggested above, as the equi- valent of the dupondius after the second reduction of the Roman as, may indicate the amount to be paid for vegetable of- ferings of all kinds in proportion to each victim; an amount considerably less than the monetary value of the victims them-
Va 13
COMMENTARY
223
selves, if we may judge by the sum paid in recompense to the adfertor in 17-22 below.
Va 14 - b 7, A second and distinct de- cree, as appears not only from the repe- tition of the opening formula but also from the name of the new eponymous uhtur. Moreover the paragraph division is indicated on the bronze by a short horizontal line engraved under the char- acters et : n at the beginning of line 13. — 14-16 : nearly all editors end the first sentence with Kluvier, except Br. and Vet., who end with Atiierier, and Dev., who ends with ukre. The view of the majority seems preferable because it is most natural for the sentence to end, as in 4, with the name of the magistrate by whose term of office the document is dated, and also because ape apelust, with nothing to indicate the setting, would make a very abrupt opening for the second sentence. In any case there can be no reasonable doubt that eikvasese Atiie- rier belongs in the second sentence, as in 4 above. The use of this expression after kumnahkle Atiiefie might appear pleonastic, but eikvasese may indicate some sort of special board within the larg- er group, as suggested above on 4. That the banquet itself was attended, at least in theory, by all the brothers appears from ape frater gersnatur furent in 22 and from mestru karu fratru Atiie- fiu pure ulu benurent in 24-5, 27-8. — 17-22 : after each of the three principal acts in the ceremony the adfertor is to receive a donation. For apelust see on Ila 20; for purtitu fust see Lexicon. Imv. forms of subra spa- are used in VIb 17 and Vila 39, which are nearly identical in content, and in VIb 41, and are translated ' superiacito ' by Bk.,Blum., Dev., Vet. Here, however, the reference seems to be to a distribution of the pieces of meat which are to be eaten by the brothers at the dinner, a sense which conforms well to the notion of spreading which is clearly visible in many of the cognates of U. spahatu. The change from
the active apelust to the impersonal pas- sives purtitu fust and spafu fust may, as Dev. (p. 409) suggests, indicate that in early times the actual slaying was per- formed by the adfertor while the porrectio and superiectio were left to others. — 17 niTineklu: generally translated ' ma- nasculum' or ' sportulam' and roughly equivalent to L. munusculum, although the formation is not precisely the same; see 21 c, 75 b. — 23 fratreks : this offi- cial is mentioned below in the closely similar passage b 1 ff,, where he is to call upon the brothers to fix the fine in case the adfertor has not performed his duty satisfactorily, and also in Vllb 1, 4, where he himself is directed to furnish twelve heifers at the time of the lustration of the people during his term of office, and to pay a fine of 300 asses if he fails to do so. The usual derivation of frat- reks, equating it with L. *fratricus, should be maintained, and the noun L. vilicus, which BU. cites, is a reasonably close parallel so far as formation is con- cerned, despite Blum., pp. 82-3, who pre- fers to derive from *fratri-rex with hap- lologic loss of a syllable. Apart from the fact that he is still obliged to assume a stem *fratreko- as source of an analogical spread of the c into the oblique and de- rived forms (e.g. fratreci, fratreca, fratre- cate, all in the Latin alphabet, where c cannot stand for the g of L. reg-, etc.), -rex as second member of the compound is scarcely compatible with the notion of an office held for a fixed term (cf. VI lb 1 sueso fratrecate) in an organization which seems to have been essentially de- mocratic; cf. Stegmann von Pritzwald, GL, XXI (1933), pp. 134-6. — kvestur: here and below in b 2 he may act in place of the fratreks as the official who calls for the vote, while kvestretie as designation of an eponymous office ap- pears at the end of lb and of Ila. Dev., p. 304, suggests that the fratreks and kvestur shared the duties previously per- formed by the ubtur alone, and it is reasonable to regard the kvestur, whose
224
III. TEXT AND TRANSLATION
: kuratu : si : sve : mestru : karu : / (25) fratru : atiieriu : pure : ulu : benurent : / (26) prusikurent : rehte : kuratu : eru : efek : / (27) prufe : si : sve : mestru : karu : fratru : atiier/iu : pure : ulu : benurent : prusikurent / (29) kuratu : rehte : neip : eru : enuk : fratru / (Vb 1) ehvelklu : feia : fratreks : / \2) ute : kvestur : panta : muta : / (3) arferture : si : panta : muta : fratru : / (4) atiieriu : mestru : karu : pure : ulu : / (5) benurent : arferture : eru : pepurkure/nt : herifi : etantu : mutu : arferture / (7) si : /
(Vb 8) clauerniur . dirsas . herd . fratrus . atiersir . posti . acnu j
(9) farer . opeter . p . IIII . agre . tlatie . piquier . martier . et . sesna /
(10) homonus . duir . puri . far . eiscurent . ote . a . VI . clauerni j
has been provided in a satisfactory manner. If a majority of the Atiedian Brothers who have come there declare that it has been provided in a satisfac- tory manner, it shall be well. If a ma- jority of the Atiedian Brothers who have come there declare that it has not been provided in a satisfactory manner, then the brother-superior or the quaestor shall call for an expression of opinion on the part of the brothers as to how great a fine shall be (imposed) on the adfertor. Whatever fine a majority of the Atie- dian Brothers who have come there demand shall be (imposed) on the ad- fertor, so great a fine shall be (imposed) on the adfertor.
The {decuvia of the) Clavernii are re- quired to give to the Atiedian Brothers each year four pounds of choice spelt from the Ager Tlatius of Picus Martius, and dinner for the two men who come to fetch the spelt, or else (to give) six
title was apparently borrowed from the Romans, as belonging to a later period than the uhtur, yet in 2 and 15 of the present table uhtretie is used to designate the eponymous magistracy. On the func- tions of the kvestur, fratreks, and uhtur in general see G. B. Pighi, Studi in onore di Gino Funaioli (Rome, 1955), p. 374, and on the relation of the magistra- cies to the whole problem of the chrono- logy of the Tables see Introduction, 18. — 24 karu : for the meaning of this word which is probably 'part, section' in all passages in the Iguvine Tables, never 'meat,' see on lla 1. — 27 prufe si: for this use of an adverb in the predicate in place of an adjective see 139 a. It is easier to admit a usage of this kind in the present passage than to assume an /-stem adj. in contrast to L. probus, -a, -um, as Bii. did. — b 2-3 panta rauta arferture si :
for the subjunctive see 152 d, — 5-6: for the syntax of the verbs see 152 c. Vb 8 - 18. This passage, which is in the Latin alphabet, must have been written some time after the completion of the preceding portion of the table, and was inserted in the blank space on the reverse of V, although its content has no direct connection with that of Va 1 - b 7. 8-18 is a pair of contracts between the Atie- dian Brothers and two of the decuviae, whereby the latter are expected to make annual contributions of grain to the brotherhood and to provide dinner or its monetary equivalent to the men who come to fetch the grain, while the decu- viae in return receive from the brother- hood at the festival of Semo certain por- tions of meat in addition to dinner or its equivalent. The decuviae to which the present passage applies are the Cla-
Va23
COMMENTARY
225
vernii and the Casilas. Since it appears from lib 3, 6, that the Clavernii had been reorganized into two new decuviae and the Casilas into three, it is immedia- tely clear why the latter must give one and a half times as much grain and re- ceive one and a half times as much meat as the former. The extant tables contain no similar contracts between the brother- hood and the remaining decuviae; it is therefore a plausible assumption that such contracts formed the content of one of the lost tables. See Introduction, 16. — 8 posti acnu: in view of O. akenei, which almost certainly means ' year, ' it is probably best to maintain the trans- lation ' each year ' in the present passage, despite Devoto's objection (pp. 416-17 'secundum moles'). The fact that posti acnu stands before farer opeter but (in 12) after pelmner sorser may have no special significance beyond the desire to place pelmner sorser in a position of emphasis, since it contrasts with cabriner, while farer opeter is not a member of a contrasting pair. Moreover 'each year ... at the fes- tival of Semo ' is not necessarily a redun- dancy, since the former gives the fre- quency with which the contribution must be made, while the latter gives the pro- per time for making it. — 9 agre Tlatie: gen. sg., as appears from the parallel agre Casiler in 14. The field cannot be identified. Huschke, p. 473, compared Dolates cognomine Salentini in Pliny's list of Umbrian states (N.H., III, 14, 19), but the improbability of t for d in the Latin alphabet and of syncope in the initial syllable is against such an identifi- cation. Most editors see a relationship with Latium,yvhich is quite possible in an etymological sense, but, as Conway s.v. pointed out, it is unlikely that the refer- ence is to Latium in the geographical sense. In the present case the name of the decuvia and the name of the field as- signed to it do not agree, but in 13-14 be- low Casilos and agre Casiler agree except for the stem-variation. Since we have only the two contracts it is difficult to
make generalizations, but it may be ob- served that among the names of the decuviae in lib 2-7 six are of the type of L. Arpinas, -atis, while the remaining four are gentilicia like L. Cornelii, etc. Devoto, pp. 355, 413, suggests that ori- ginally all bore ethnics of the type in -5//- derived from territorial names, but that several later replaced these ethnics by gentile names of the type of Claverniur. — Piquier Martier: for the peico (L. picus) from which Piquier is derived see Via 1 and note. Evidence of the associa- tion of the picus with Mars is found in Dion. Hal., Antiqu. Rom., I, 14, 5 iv ravrrj [i.e. Tiora Matiana in the Sabine country] Aeyerai XQiqaTriQiov "Aqsoq yevea- dai ndvv aqxalov... Tcagd Se rolg 'A(ioQi- ylai deoTie/uTiTog ogvig, ov avrol /nev nlKov, "EXkrivsQ de dQvoxoMTTrrjv xaXovaiv, eni xiovoQ ^vkivov (paivo/usvog... ; Plin., N.H., X, 40 ... pici Alartio cognomine insignes et in auspicatu magni. In Table Vb not only the agre Tlatie of 9 but also the agre Ca- siler of 14 are in some manner sacred to Picus Martins. The position of sanctity occupied by this bird is one of the most striking examples of totemism in early Italic religion. — ■ 9 sesna ... 10 ote a. y/, repeated in 15: there are several dif- ficulties. The decuvia Casilos as a result of its threefold division furnishes a higher quota of grain and receives a higher al- lotment of meat than the Clavernii, yet both furnish dinner for two men and in turn receive the same from the brother- hood, as shown by the identity of the sum of money set as the equivalent of the dinner. The value itself has been a second source of difficulty, because it appears excessive when compared with certain passages in Greco-Roman literature: Po- lybius, II, 15 (with reference to Cisalpine Gaul) Tioiovvrai ydg rag xarakvaeig oi dio- devovreg rr]v ;faigaj' ev rolg jiavdoxeioig, ov av[ji,(pu)vovvreg Tcegi rcov xard fisgog inirr]- delcov, dXX EQcoTMVTEg TtoGov rov avSqa dexe- xai. 'Qg fiiv odv eni ro noXv Ttagievrai rovg KaraXvxag oi jzavdoxelg, (bg, Ixavdndvr' ex^iv TO TiQog rrjv /eeiav, '^fxiaaaaQiov... ; Mar-
226
III. TEXT AND TRANSLATION
(11) dirsans . herti . f rater . atiersiur . sehmenier . dequrier / (12) pelmner . sorser . posti . acnu . uef . X . cabriner . uef . V . pretra / (13) toco . postra . fahe . et . sesna . ote . a . VI . casilos . dirsa . herti . fratrus / (14) atiersir . posti . acnu . farer . opeter . p . VI . agre . casiler . piquier / (15) martier . et . sesna . homonus . duir . pari . far . eiscurent . ote . a . VI j (16) casilate . dirsans . herti . frateer . atiersiur . sehmenier . dequrier j (17) pelmner . sorser . posti . acnu . uef .XV. cabriner . uef . VIIS . et I (18) sesna . ote . a . VI I
tial, IV, 68 invitas centum quadrantibus et bene cenas. j Ut cenem invitor, Sexte, an ut invideam'i Devoto suggests (p. 405) that the price of the dinner here, and also the fine imposed on the fratricus in Vllb 4, belong to the period after the reduction of the as to -— - of its original value; that is, after 89 B.C. according to the usual interpretation of Pliny, N.H., XXX, 44. In any case there can be no doubt of the general sense of the passage. Among individual words eiscurent alone is obscure, and all editors translate it by arcessierint or some verb of fairly similar meaning. — 11 frater: the T and the E are separated by a space sufficient for one letter, and it is clear that the wrong letter had been originally inscribed after the T and then erased. The letter was ap- parently an R, and the mistake may have been induced by the oblique case-forms or possibly even by the Latin nom. pi. — 11-12 sehmenier dequrier / pelmner sor- ser ... cabriner: cf. lib 1 semenies teku- ries Sim kaprum upetu, with note. — 12 uef: twice here and twice in 17 below. It indicates some unit of measure by weight or volume, but its etymology and precise meaning are uncertain. Thur-
asses. The Atiedian Brothers are re- quired to give to the Clavernii at the decurial festival of Semo each year ten portions of pork and five portions of goat-meat, the former pickled, the latter roasted, and dinner or six asses. The (decuvia) Casilas is required to give to the Atiedian Brothers each year six pounds of choice spelt from the Ager Casilus of Pious Martius, and dinner for the two men who come to fetch the spelt, or else (to give) six asses. The Atiedian Brothers are required to give to the {decuvia) Casilas at the decurial festival of Semo each year fifteen portions of pork and seven and a half portions of goat-meat or six asses.
neysen (cited by PL, I, p. 288) suggested that it may be an abbreviation for *uefraf, the equivalent of L. libras, and is followed by several recent editors, but the change I- > u- V- is not surely attested before front vowels (see 55, n. 2). Blum., p. 76, equated uef in form with L. vehes, -is 'wagon-load' and in sense with ON vett, vset 'weight.' Although, as Blum, points out, there is no serious objection to admit- ting one unit of measure, p., for grain and another for meat, it seems safest to as- sume a root-stem *weif- from the root widh- of L. di-vido, after PI., I, pp. 288, 455, II, p. 182, Buck, § 136 a. While trans- lating 'portions' we may understand uef to be a technical term indicating a defi- nite quantity which we cannot identify. — 12-13 pretra ... postra: these adjectives, which must be in agreement with uef, show its gender to be fem. Certain speci- fications are given for the pork and goat-meat here but are omitted in 17 be- low, either because they applied only in the case of the Clavernii or because hav- ing once been stated they were assumed to apply to both. In the word pretra the second R was at first omitted and then inserted above the line between the T
Vb 10
COMMENTARY
227
and the A. — 13 toco: the ending, and consequently the syntax in the sentence, are not clear, but there is little doubt that the word is somehow connected with Gallo-Latin tucca, tuc{c)etum, and also taxea: Schol. Pers., XI, 42 tucceta apud Gallos Cisalpinos bubula dicitur condi- mentis quibusdam crassis oblita ac macera- ta; et ideo toto anno durat. Solet etiam porcina eodem genere conditaservari; Corp. Gloss. LaL, II, 202, 52 tucca- xaraxv^ia C(o- fiov; Isid., Orig., XX, 2 taxea lardum est gallice dictum; unde et Afranius in Rosa: Galium sagatum pingui pastum taxea. — fahe: here it is not possible to say anything with certainty. The ending -e could indicate an adverb in -e or an ace. pi. of an I'-stem adj. like seuacne in Vllb 1.
Bii. translated ' confecfas' and conjectured an J-stem *fahis related to fa^ia, L. facio, as L. vehes, -is to veho, but the medial h is an obstacle, since we should expect ^, and it is doubtful whether an uncom- pounded form of this sort would have developed the meaning 'preserved, pickled (meat). ' PL, I, pp. 281, 439-40, 461, suggests connection with (pchyoj 'roast,' Germ, backen, which would mean that the meat was to be roasted before being presented to the Glavernii. Another pos- sible connection is with L. fagus, Gk. (prjyog, the sense being ' beech-cured (ham) ' in contrast to toco as ' fat bacon ' ; cf . What- mough. Dialects of Ancient Gaul, 181; but here again the Umbrian h is a diffi- culty.
228
III. TEXT AND TRANSLATION
(Via 1) este . persclo . aueis . aseriater . enetu . parfa . curnase dersua . peiqu . peica . mersiu . poei
(The ad/erfor) shall commence this cer- emony by observing the birds, the parra and crow in the west, the woodpecker
Via 1 - b 47 (= la 1 - b 9). Purifica- tion of the Fisian Mount. For an outline of the content of the present version, as well as of that of Table I, see Introduction, 13, and for the relation between the two versions see Introduction, 18. Both des- cribe the same ceremony: the same vic- tims, in groups of three, are offered to the same gods at the same three gates and at two additional stations, but the account in the later tables is far more de- tailed; not only are a number of specific instructions found in VI which were omitted in I, especially in regard to the tal<.ing of the auspices, but the prayers accompanying the sacrifices, which are completely lacking in I, are presented in full in VI.
Via 1 - 18 (- la 1 - 2). Taking of the auspices, which must be completed be- fore the beginning of the purification proper. In la the instruction for taking the auspices was compressed into the first line and a half, witli no specifica- tions beyond the use of the terms per- naies pusnaes, while Via specifies the birds to be watched, the formulas to be repeated by the adfertor and the augur, and the demarcation of the templum. Despite the very considerable amount of study that has been devoted to this whole passage, many details remain obscure. EspeciaUy is this true in tire Ust of land- marks for defining the iemplum. A sum- mary of conclusions in regard to the auspicium is given below after the note on 14. — 1 persclo: best understood as including the augural ceremony as well as the purification itself; at least this seems likely in view of the fact that Via enters immediately into detailed instruc- tions for taking the auspices. The abla- tive absolute then means 'by observing the birds ' rather than ' after observing the birds'; for this usage cf. Tac, Ann.,
VI, 45 damnum Caesar ad gloriam vertit, exsolutis domuum et insularum pretiis; Hist., II, 85, pessimum f acinus audet, misso ceniurione ad inter ficiendum Tettium lulianum. — aueis: for the exceptional importance attached to birds in augury among the Umbrians cf. Cic, Div., I, 41, 92 Phryges autem et Pisidae et Cilices et Arabum natio avium significationibus plurimum obtemperant, quod idem facti- taium in Umbria accepimus. In Roman augural practice the birds were classified according as they gave omens by flight or by voice: Festus, p. 308 Li. Oscinum tripudium est, quod oris cantu significat quid portendi, cum cecinit corvus, comix, nociua, parra, picas ... Oscines aves Ap. Claudius esse ait, quae ore canentes faciant auspicium, ut corvus, comix, noctua: alites quae alis ac volatu; ut buteo, sanqualis, aquila, inunulsus, vu I turius: pic am aut Mar- tins Feroniusque et parra et in oscinibus, et in alitibus habentur [picam autW: pica aut X; ed. princ. : picus autem Aug. : fort, pica avis\. Comparison is often made between the hst of birds in Via 1 and Plaut., As., 259-61 impetritum, inauguratumst: quo- vis admittunt aves, / picus et comix ab laeva, coruos, parra ab dextera j consuadent, in which the similarity to Table Via 1 is striking enough to suggest the possibility that Plautus had in mind the augural practice of his native Umbria. — enetu: the logical subject is the adfertor, al- though the actual observation of the birds is the duty of the augur; both have a part in the ceremony, however. — parfa: this bird has not been identified with certainty. It agrees, at least in an ety- mological sense, and presumably also in a zoological sense, with L. parra; any other hypothesis would be highly impro- bable in view of the Plautine passage, in which all modern editors adopt parra, the reading of B^, in place of the porro
vial
COMMENTARY
229
of the majority of mss. But the identity of the parra itself is unknown, and ety- mological comparisons, as, for example VfithyxiQ 'starling,' OHG spanvari 'spar- row-hawk, Sperber, epervier, ' will not give the answer. The bird in Hor., C, III, 27, 1 Impios parrae recinentis omen / ducat has been identified as an owl (so Orelli, Ussani, Wickham, etc.) or as the lapwing (O. Keller, Die an tike Tierwelt, II [Leipzig, 1913], p. 178), but with slight probability. Devoto's identification of the bird as the green woodpecker (pic- chio verde, T.G., p. 29) has a better chance of being correct, but if Feronius in the corrupt Festus passage cited above is a second variety of picus distinguished from the picus martius, it may easily be the green woodpecker, and the parra imme- diately following must be a different bird. The translation 'jay' is here offered with no great confidence in its correctness, but partly as a result of the elimination of several theories which appear less pro- bable, and partly in view of the fact that the comparative size and striking appear- ance and shrill cry of the European jay (Garrulus glandarius) might make it suit- able for inclusion among augural birds. Thomas Chase, The Works of Horace (Philadelphia, 1884) on Hor., loc. cit., sug- gested identification of parra with the jay. For a fuller discussion see Studies Presented to David M. Robinson, Vol. II (St. Louis, 1953), pp. 469-76. — cur- nase: like L. comix, which differs only in the vowel of the second syllable, this is the crovv^ (probably Corvus cornix cor- nix). — dersua: see below after the note on peica, where dersua and merstu are treated together. — peiqu : like Plautus' picus, this is generally taken to be the large black European woodpecker Dryo- copus martius martius, which was very prominent in early Italic religion and mythology (see on Vb 9), — peica: evi- dently the magpie (Pica pica pica); its absence from the passage cited from the Asinaria above, if Plautus is thinking of Roman rather than Umbrian augury, may
be significant when observed in connection with Pliny, N.H., X, 78 nuper et adhuc tamen rara ab Appennino ad urbem versus cerni coepere picarum genera quae longa insignes cauda variae appellantur. — dersua ... merstu: the syntax presents some difficulty at this point. In place of the plural adjectives which we should ex- pect with the two pairs of nouns we have singular adjectives, and merstu agrees not with peica but with the more remote peiqu. Breal suggested that the second bird in each pair was merely an alterna- tive to the first. In favor of his suggestion is the obvious fact that the greater the number of birds which must be seen, the greater the difficulty of obtaining the proper auspices required as a condition for the ceremony which was to follow. Breal himself, hovv'ever, did not use this argument, and it is not as cogent as it might at first appear, since there is some slight possibility that birds were caught and kept in order to be released at the time of the taking of the auspices; cf. Goidanich, Hist., VIII (1934), p. 256, where hov/ever no actual evidence is given. So far as the use of the singular in place of the plural is concerned, it would be quite natural if the conjunction to be supplied in the two pairs of bird-names is or rather than and; yet, while the omis- sion of et is extremely frequent (137), there are no known instances of the omis- sion of the disjunctive ote. The explana- tion, however, may be that ote is avoided because the parfa and curnase are not mutually exclusive; either one or both, in the proper situation, will provide a satisfactory omen, and similarly with the peiqu and peica. The sense may then be similar to that in expressions where Eng- lish occasionally makes use of the combi- nation and /or. dersua and merstu them- selves have defied all attempts at inter- pretation. The former, in addition to its 12 occurrences in Table VI, is found as tesvara, with parfam, in lb 13, while merstu occurs only in Via. marsuva in III 11, 28 may be, but is not necessari-
230
III. TEXT AND TRANSLATION
angla . aseriato . / (2) eest eso . tremnu . serse . arsferture . ehueltu . stiplo asenaia . parfa . dersua . curnaco dersua / (3) peico . mersto . peica . merst . a . mersta . auuei . mersta . angla . esona . arfertur . eso anstiplatu /
ly, related to it. dersua, according to an etymology proposed by Panzerbieter, Quaestiones Umbricae (Progr. Meiningen, 1851), p. 11, followed by Bii., PI., I, pp. 406-7, Buck, is from *ded-es-wo-, adj. to an es-stem built on the reduplicated root of the verb 'give.' merstu according to Bii., PI., and Buck, following Ebel, K.Z., IV (1855), p. 200, VII, pp. 271-2, is from *med-es-to-, with the same root contained in mers mefs, O. ixied-diss, L. modus, modestus, etc. The omens then would be of two degrees of value: merstu, in Latin iusto, allowing the ceremony to be undertaken, and dersua, in Latin pros- pera, actually favoring the undertaking. — aves quae non admit Ian t tantum sed secundent (Bii., p. 43). This interpreta- tion of the words has a certain advantage when applied in 3 to mersta auuei, mersta angla esona, where there is no direct con- trast with a dersua and it is difficult to understand mersta in a local or directional sense. Yet for several reasons it seems advisable to abandon it. In the first place *de-d-es-wo- is not a probable formation nor one for which good support can be found in cognate languages. Moreover there is no good evidence of a distinction between omens which merely gave per- mission and those which gave encourage- ment, nor do the etymologies adopted as a basis for this interpretation really justi- fy giving a stronger sense to one adject- tive than to the other. The correspond- ing passage in la 1-2 uses the terms pernaies pusnaes, regularly translated anticis posticis, and it seems highly pro- bable that in Via also the terms used of the birds should refer to direction, al-
and magpie in the east. The one who goes to observe the messengers, sitting shall call out from the tent to the ad- fertor : "Demand that I may observe a parra in the west, a crow in the west, a woodpecker in the east, a magpie in the east, in the east birds, in the east divine messengers." The adfertor shall
though there is no serious objection, des- pite Breal, to taking dersua, merstu in reversed correspondence to pernaies, pusnaes, if the etymologies adopted are more satisfactory under this arrangement. How far Umbrian and Roman augural procedure agreed is uncertain, but if there is to be any hope at all of under- standing this portion of Table Via, we must take account of Latin passages bearing on the matter. The Plautine passage cited above seems to assume that a Roman audience would understand the allusion to the birds on the left and right. There is almost no chance, however, that dersua means 'right,' despite Blum., p. 77, who connects it with Goth, taihswa ' de^wg' and derives it from *deks-wo-. The meanings ' right ' and ' left ' belong to destre and nertru respectively, dersua and merstu then must refer to points of the compass or to other contrasting di- rections, as pernaies pusnaes in la 2 do. In Roman augury the best evidence is for southward facing or for eastward fac- ing. Literary passages which give sup- port to the notion that the augur faced south include: Varro, L.L., VII, 7 Eius templi partes quattuor dicuntur sinistra ab oriente, dextra ab occasu, antica ad me- ridiem, postica ad septentrionem. Festus 244 Li. Sic etiam ea caeli pars, quae sole illustratur ad meridiem, antica nominatur, quae ad septentrionem postica; rursumque dividuntur in duas partes orientem at- que occidentem. Passages which support eastward facing include: Livy, I, 18, 6-7 inde ab augure ... deductus (sc. Numa) in arcem in lapide ad meridiem versus con- sedit. augur ad laevam eius capite velato
via 1
COMMENTARY
231
sedem cepit, dextra manu baculum sine nodo aduncum tenens, quern lituum ap- pellarunt. inde ubi prospectu in urbem agrumque cap to deos precatus regiones ab oriente ad occasum determinavit, dextras ad meridiem partes, laevas ad septentrio- nem esse dixit. Isid., Etym., XV, 4, 7 locus designatus ad orientem a contempla- iione templum dicitur, cuius partes quat- tuor erant, antica ad ortum, postica ad occasum, sinistra ad septentrionem, dextra ad meridiem spectans. The fact that there were two confUcting orientations appears very plainly through the use of antica for the south by Festus but for the east by Isidore. The confusion may arise part- ly from the fact that two officials took part in the taking of the auspices, in Rome the rex or other official and the augur, in Iguvium the adfertor and the official poei angla aseriato eest. But apparently there were actually two or more systems in use in Italy, reflecting not only the varying purposes for which augury was used, but also an overlapping of different cultures. The view adopted here is based on the interpretation which Devoto pro- posed (pp. 143-4) for the two adjectives on etymological and other grounds. dersua he derives from *drotsvo-, cf. L. dorsum < *d^tso-; since there is no cer- tainty as to the etymology of dorsum itself, a safe procedure may be to main- tain the connection between dersua and dorsum but to abstain from equations outside of Italic, -er- and -or- may re- present an ablaut-variation, or -er- may be from -ro- (32 d) if either the L. -or- or the U. *-ro- resulted from metathesis. The adjective dersua then is to L. dorsum as annuus to annus, a hypothesis which receives some support from the -uo-stem indirectly attested by L. dorsualis. It is to be assumed that dersua passed through the same semantic evolution as L. pos- ticus, which became ' north ' (as in Festus) or 'west' (as in Isidore). Devoto takes merstu, with Goth, maurgins ' tempus ma- tutinum,' etc., from the root *mer-k- with its variations (cf. Pokorny, Idg.
etym. Wb., p. 734). It is then comparable to Ger. Morgenland in both origin and sense, and it is apparent that the orien- tation assumed is an eastward one; on the relation of the templum to the topo- graphy of Iguvium see below in the sum- mary of 1-18. — poei angla aseriato eest: since the augur must surely have had some official title, the circumlocution here is probably an instance of tabu, as Devoto (p. 145) suggests. This reluctance to name the augur (for fear of being over- heard by the birds?) has a rough simi- larity to the hunter's tabu in certain northern European languages against the name of the bear, which has led to such substitutes as Russ. medvedb, literally ' honey-eater. ' — 2 tremnu : possibly a local abl. with serse, but since this sense would normally call for the locative, it is better, with most editors, to take it as abl. of place whence depending on ehueltu, despite the word-order. Precise- ly what sort of structure the tremnu was cannot be determined, but it was without doubt similar to the Roman tabernaculum mentioned in Cic, N.D., II, 4, 11; Div., I, 17, 33; II, 35, 75; Livy, IV, 7, 3, used in the taking of the aus- pices. — ehueltu: cf. ehvelklu 'choice, expression of opinion. ' Both contain eh- (: L. e, ex) + wel-, the root of L. volo, velle, with semantic development 'wish > select > request, ' in the case of the verb. — stiplo: separated from ehueltu by a space equal to about one letter and containing a point, stiplo is no longer taken as an infinitive depending on ehueltu, with Hu., Bii., PI., but by most editors as pres. imv., beginning the for- mula in oratio recta. The subjunctive aseriaia must then be 1 sg., although the form is ambiguous because of the lack of a final consonant. — 3 mersta: the second instance of mersta, which is not in direct contrast with dersua, is some- what awkward when taken as 'in the east,' but it may represent a provision that any birds beside those already spe- cified may be counted as good omens if
232
III. TEXT AND TRANSLATION
(4) ef . aserio . parfa . dersua . curnaco . dersua . peico . mersto . peica . mersta . mersta . aueif . merstaf j (5) anglaf . esona . mehe . tote . iioueine esmei . stahmei . stahmeitei . sersi . pirsi . sesust . poi . angla J (6) aseriato . est . erse . neip . mugatu . nep . arsir . andersistu . nersa . courtust . porsi . angla . anseriato f (7) iust . sue . muieto . fust . ote . pisi . arsir . andersesusp disleralinsust /
(8) uerfale . pufe . arsfertur .
7 andersesusp: for andersesust.
thus demand : "There observe a parra in the west, a crow in the west, a woodpecker in the east, a magpie in the east, in the east birds, in the east divine messengers, for me, for the state of Iguvium, for this estabhshed ordi- nance. " While he who goes to ob- serve the messengers sits in the seat, during that time no one shall make a noise nor shall any other person sit in the way until he who has gone to observe the birds returns. If a noise is made or any other person sits in the way, he will make (the ceremony) in- valid.
The templum where the adfertor re-
seen in the east, while mersta angla esona further provides for other omens than birds; for example, lightning (of. Dev., p. 150). The number of admissible omens then is less restricted than at first ap- pears. — anstiplalu: the compound in contrast to stiplo above may show "ac- tion determinee " : 'conclude an agree- ment'; but the distinction is no less subtle than when we compare L. stipulor and instipulor (the latter in Plant., Pseud. 1069, Rud., 1381). — 4 e/ aserio, etc.: aserio is a pres, imv. similar in formation to stiplo in 2. The augur's response comes below in 17-18, where the verb re- quired to complete the elliptical sentence might be a form of aserio. - — 5 mehe, etc. : for these datives signifiying the objects of divine favor cf. Cato, R.R., 141 diiis boncmn saluiem mihi domo familiaequc nostrae; Livy, XXIX, 27, 2 mihi populo plebique Romanae, sociis nominique La- tino. ■ — tote: there are more than 130 examples in the Tables, mostly accom- panied by some form of Iioueine. It signifies 'state' in the sense of the total citizen-body. Cf. O. touto similarly used in the Tabula Bantina (Co. 28.) — stah-
mei is almost certainly from *sid- ' stand, ' and most translate by dat. statui, al- though it is not quite clear in what sense; stahmeitei is without doubt pcpl. of a denominative verb from the former, in agreement with it and making up a fi- gura etymologica. Devoto's ' in hac super- ficie extensa ' is very tempting, since in some respects a loc. stahmei understood as a designation of place gives a better sense than a dat.; yet there are certain objections: 1) the ending -ei is more easily taken as dat., whether of an o- or of an i-stem: cf. dat. Tefrei VIb 22, but loc. ocre Via 26, 36, b 29, and see 20, 93 b, c, 97; 2) esmei stahmei stahmeitei as dat. gives a construction fully coordi- nate with the datives which precede; 3) it is difficult to understand stahmei of the augural templum, since uerfale in 8 is used in this sense. We therefore trans- late ' for this estabhshed ordinance, ' as suggested by R. G. Kent, Lang., XIV (1938), p. 215. Dev., p. 151, distinguishes the use of the dat. in such formulas from the abl. -1- -per: "postpositionis -per ope ea indicantur quibus actiones sacrae pro- sunt; casus dativus ea indicare videtur
via 3
COMMENTARY
233
'quorum nomine' actiones peragantur." esmei stohmei sfahmeitei as dat. may then have an underlying sense 'for compliance with this established ordinance. ' — 5-6 sersi pirsi sesusf, etc.: injunction against any disturbance during the taking of the auspices. The pronominal forms pirsi ... erse are here used with conjunctional and adverbial value respectively. — 6 mu- gatu: although assumed to be related to L. mugire, it must refer to noise of any kind, and the prohibition must refer not only to the augur but the adfertor and any other individual (hence the singular) in the vicinity. — arsir: the principal source of difficulty in the passage. Bii., PI., Bk., Blum., and Vet., and also Er- nout-Meillet^ p. 40, equate it with L. alius (see 32 a). Certain other scholars, however, take it as a dat. pi. : Br. in the sense ' precationibiis' (?) as cognate with aQa, Dev., Pis. as ' dedicationibus,' cog- nate with arsie, after Brugmann, Ber. sacks. Ges. Wiss., 1911, p. 173. Admitted- ly the lack of a pisi before the first arsir, to correspond to pisi arsir in 7, is surpris- ing if arsir is 'other,' but the lack is equally strange by the other interpreta- tion 'no one shall make a noise or inter- rupt the ceremonies, ' because in any case the subject of the two imvs. must be indefinite, arsie 24, b 8, 27, in address to divinities, is not a good parallel to our passage, and in the same lines the inter- pretation of arsier (b 8 asier) is itself un- certain, although the contrast in spelling between arsier and arsir is worthy of notice. I prefer therefore, though with some hesitation, to take arsir as 'alius' and to understand it as a provision that no unauthorized third person shall ap- pear on the scene and cause a distur- bance. — 7 muieto fust: impersonal. On the form of the participle see 30 c, 47 b, 116 a. — andersesust, disleralin- sust: after the word-divider that follows arsir the bronze has andersesuspdisleral- insust. There is no doubt that p must be emended to /, making the end of a 3 sg. verb, but disleralinsust presents some
difficulty. The fact that the word-divi- sion is neglected after the p gives us the right to assume that there may be a se- paration between disler and alinsust also, as earlier scholars and more recently Dev. have done. According to this view disler is abl. pi. of tigel, tiglu (t = d), and the general sense is that the formulas become ineffective. There is a certain obscurity about disler. No other instances occur in the Latin alphabet, and the contexts where tigel, tiglu occur are not close pa- rallels to Via 7; but the real difficulty is in finding any satisfactory etymology for alinsust (< *alisensust with a-privative and root as in L. licet, Breal, pp. 35, 372; cognate with oXXv/ui and meaning '[cae- rimonia] irrita erit,' Dev.). It seems best therefore with Bii., PI., Bk., etc., to re- gard disleralinsust as a denominative verb from *dis-leisd-li-: cf. h.de-lirus, delirare, but with the notion of failure or ineffective- ness in contrast to the Latin sense of mental aberration. There is no expressed subject; we may regard the augural cer- emony as logical subject or regard the verb as impersonal.
8 - 14. Instructions for defining the templum in which the birds, or other omens, are to be observed. The demarca- tion is made vv^ith relation to the augural seats, the asa deueia, and certain land- marks in the city, none of which can be clearly identified. — 8 uerfale: the ety- mological equivalent of L. verbale, but used here to designate the templum. The semantic development is considerab- ly clarified by two passages in Varro, L.L.: VI, 53 Hinc effata dicuntur, qui augur es finem auspiciorum caelestum ex- tra urbem agri<s> sunt effati, ut esset; hinc effari templa dicuntur: ab auguribus effantur qui in his fines sunt; VII, 8 In terris dictum templum locus augurii aut auspicii causa quibusdam conceptis verbis finitus. — arsfertur: although the actual taking of the auspices is performed by the augur, the adfertor himself puri- fies the Mount. — trebeit: related to tremnu, O. triibum, tribarakkiuf, etc.
234
III. TEXT AND TRANSLATION
trebeit . ocrer . peihaner . erse stahmito eso tuderato est . angluto / (9) hondomu . porsei . nesimei . asa . deueia . est . anglome . somo . porsei . nesimei . uapersus . auiehcleir / (10) est . eine angluto . somo . uapefe auiehclu . todcome tuder . angluto . hondomu . asame . deueia . todcome / (11) tuder . eine . todceir . tuderus . seipodruhpei . seritu j
(12) tuderor . totcor . uapersusto auieclir . ebetrafe . ooserclome . presoliafe . nurpier . uasirslome /
(13) smursime . tettome . miletinar . tertiame . praco . pracatarum . uapersusto auieclir . carsome /
(14) uestisier . randeme . rufrer .
mains for the purpose of purifying the Mount, when established, is bounded thus: from the lowest angle, which is next to the Ara Divina, to the highest angle, which is next to the augural seats : thence from the highest angle to the au- gural seats to the city boundary, from the lowest angle to the Ara Divina to the city boundary. Then within the city boundaries he shall make observation in either direction.
The city boundaries are : from the au- gural seats to the Exit, to the observa- tion-post, to the fore-area of Nurpius, to the Vale, to the Temple of Smurcia, to the house of the gens Miletina, to the third tower of the rampart; and from the augural seats to the avenue of Ves- ticius, to the garden of Rufer, to the
having to do with buildings of various types. The semantic development in the Umbrian verb is from ' dwells ' to ' re- mains. ' — ocrer peihaner : for the syntax of. 19 below and 153 i. — stahmito: 'siaiivum' Bii., PI., ' statuium' Bk., Blum., Pis., Bott., 'extensum' Dev., ' stabilitum' Vet. It serves to show that the templum for the auspices precedingthe purification of the Mount was a fixed and permanent one, while tuderato serves to introduce the list of points defining it. — 8-9 angluto / hondomu porsei nesimei asa deueia est: the first of a succession of imaginary lines is drawn from the "low- est angle," a point near the asa deueia, to the "highest angle," near the augural seats. The asa deueia was in all probabi- lity situated in the forum, which may very easily have been in the same part of the city as the present Piazza dei Quaranta Martiri (formerly the Piazza Vittorio Emanuele), the low elevation of which accords well with the expression angluto hondomu. From a comparison of lb 14, Via 16-17, b 51, it appears that the
adfertor occupies the augural stone-seat, while the augur, who announces the omens to him, is a short distance away. The augural seats must have been located on the Fisian Mount or arx, for the loca- tion of which see Introduction, 4. an- gluto somo uapefe auiehclu and the parallel passage angluto hondomu asame deueia present a difficulty: do the phrases with postposition -e merely repeat in more concise language what v;as previously expressed by porsei nesimei ... est; in other words does the usage resemble that of L.ad indicating proximity without motion? So Buck, § 301. 2, takes it. Yet it seems better with Dev., p. 162, to treat uapefe auiehclu and asame deueia as phrases involving motion, like the todcome tuder which follows. The lines are then drawn from the highest and lowest angles to the seats and the altar respectively, so that these objects become in effect points of transition between the two angles and the city limits indi- cated by todcome tuder. — 12 ebetrafe: cf. the incorrectly spelled hebetafe in b
via 8
COMMENTARY
235
53. The majority analyze as from *e- bait-rd- (of. L. e, ex and baeto, -ere), with sense ' exitus.' Since it refers to one single landmark, the value of the pi. is similar to that in L. fores. This exit was appar- ently distinct from the three familiar gates of Iguvium, none of which is men- tioned in the present passage, but nothing more can be said about it. — ooserclome : generally taken as from *ob-ser-klom-en or *au(i)-ser-klom-en, in either case with syncope of / or u(e) before -Mom. De- voto, however, sees in it the name of a rivulet from * Auser-klom, comparing L. Auser and the modern Serchio. The deri- vation from *au(i)-ser-klom should be dis- carded, since we already have one place for watching birds, but there appears to be no objection to regarding the ooserclom as an observation-post for mil- itary purposes. — presoliafe: common- ly assumed to be related to L. solum, but it is uncertain whether solum would have to be understood as ' foundation ' or as 'ground' and whether presoliaf are the foremost portions of a solum or certain things in front of the solum. — Nurpier: nowhere found as the name of a god, and more probably a gentile name, but noth- ing certain can be affirmed with regard to it. — uasirslome: very uncertain; sometimes taken as connected with L. vacerra < * maker zd; by Dev., p. 165, as < *vakrikelo- : cf. Skt. vakra- 'curved'. The meaning then would be ' vale ' for which Swiss-Ital. curva ' avvallamento ' is cited as a semantic parallel. — smursime : none of the various conjectures can be regarded as at all certain. A Roman god- dess Murcia, whose temple was at the foot of the Aventine, is known from Varro, L.L., V, 154; Livy, I, 33, 5; she is general- ly identified with Venus, and is mentioned by Aug., Civ. Dei, IV, 16, as a goddess who causes sloth. If our form is ace. of a stem in -io-, as seems likely, it may possibly be a neuter name of a temple of *Smurcia. For the U. sm-: L. m- see 57 b. — tettome: found here and twice in 14 be- low : it is not easily equated with L. tectum,
because we should expect -ht-< -kt- (461), yet it is hard to believe that the two are not equivalent. Blum., p. 78, arguing from the foreign appearance of the names attached to tettome, suggests that tet- tome itself may be foreign. — Miletinar: sometimes compared with /neiXixog, fiedixioQ, O. Meeilikiieis for the root, and with L. Caprotina, Libitina for the suffix. Blum., p. 78, suggests connection with L. miles (a goddess of soldiers?). Since, however, Noniar and Salter appear not to be divine names, and since 'house' fits tettom better than 'temple,' it is better to assume that Miletinar is an adjective from a gentile name, taking its gender from the equivalent of gens or some similar word. — tertiame praco pracatarum: undoubtedly the third in a certain series of objects, praco may be ace. sg. or gen. pi. used partitively, but in either case it is fem., and so is probably a consonant-stem, since fem. o-stems are rare in those IE languages which possess them at all. pracatarum must be gen. pi. of the pcpl. of a denominative verb from praco. It is most natural to derive both from the same root as L. compesco, comperce (cf. Lexicon, and MuUer, p. 320), the basic sense being one of fencing in.
— carsome : numerous etymologies have been proposed, all uncertain. The view tentatively adopted here is that of Blum., p. 79, substantially followed by Dev., whereby carsom is equivalent to L. cursus, Ital. corso, with a through the influence of Gaulish carrus, unless the whole word is a Gaulish loan. — 14 Uestisier: pos- sibly the same as the god Vestige Sage in I la 4. — randeme: extremely uncer- tain; Dev. translates 'ad hortum,' but without adopting any of the etymological conjectures of previous editors. Pis., as etymological support for the same inter- pretation, proposes to derive from *ran-ti-, from the same root as gaivai, Qavril^m, assuming an initial variation rjsr-. The reference then would be to a garden irri- gated or watered in some special way.
— Rufrer: a place Rufrae in Campania is
236
III. TEXT AND TRANSLATION
tettome . noniar . tettome . salier . carsome . hoier \. pertome . padellar /
mentioned in Verg., Aen., VII, 739, and a gentile Rnfrius, Riibrius \s common on inscriptions from several parts of Italy, but no proper name from a stem Rufro- is known. Moreover it is not likely that our form is from *Rufrio- with loss of i by S3 d, since there is no evidence of such treatment of -ri-, while speturie, dequrier tekuries constitute evidence against it. We can probably not hope to do better than assume a name Rufer < * Rufros and decline to make any conjectures re- garding the bearer of the name. — Noniar: Blum., p. 78, suggested dialectal origin on the ground that in Umbrian we should expect *Nouim- in a derivative of the ordinal from '9'; cf. nuvime Ila 26. It is not certain, however, that this last form represents the ordinal from ' 9, ' since derivation from a superlative stem to *newo- (L. novus), after Vet., p. 197, is also possible. For the value of the name in other respects see on Miletinar in 13 above. — Salier: the failure of / to change to rs (55 b) arouses suspicion that the word is a dialect borrowing, and Blum. (p. 78) is doubtless right in af- firming connection with L. Salii. The masculine, in contrast to Miletinar, No- niar, shows that here an individual and not a gens is meant. — Hoier: possibly connected with Hule IV 17, if z is erro- neously written for I through omission of the cross-bar, as suggested by R. G. Kent, T.C.I., p. 41, or with Horse Hufis as suggested by Dev., p. 257, although the loss of rs cannot be satisfactorily^ explained. Like Uestisier it is probably a divine name, but nothing can be affirmed with certainty in regard to it. — pertome: this word may reasonably be taken as cognate with L. porta, portus, in the sense 'passageway, gate.' The e-grade of the root is attested by ON fiprdr 'firth,' nom. pi. firder, but the Germanic forms, like L. partus, are u-stems; only pertome shows a transfer to the o-declension. —
house of the gens Nonia, to the house of Salius, to the avenue of Hoius, to the
Padellar: usually taken as name of a goddess related to L. Panda, O. dat. sg. Patanai, both from *Patnd, with L. -nd- < -tn- and O. anaptyxis; for the phono- logy of Padellar see Lexicon, with refe- rences. Panda w^as sometimes identified with Ceres: Varro apiid Non., p. 44 M. hone deam Aelius putat esse Cererem; sed quod in asylum qui confugiunt panis dare- tur, esse nomen fictum a pane dando, pan- dere, quod est aperire. To the gate bear- ing the name of Padella there is a Roman parallel in the porta Pandana mentioned by Varro, L.L., V, 42; cf. Fest., p. 246 Li. Pandana porta dicta est Romae, quod semper pateret. — Devoto, p. 168, re- marks that the first set of boundaries (ebe- trafe ... pracatarum) consists chiefly of natural landmarks and the second of buildings and similar structures; the first group then lies along the mountain slope and the second along the plain, an ingen- ious and valuable suggestion, although naturally it is dependent on the interpre- tation of the individual words, many of which are very obscure.
The conclusions reached in the inter- pretation of the opening portion of Via may now be briefly summarized. Ac- cording to the view adopted here dersua and merstu do not mean ' right ' and ' left, ' nor do they mean ' prospera' and ' iusto. ' Like pusnaes and pernaies they have reference to omens appearing 'behind' and ' in front, ' but the etymology adopted for merstu (after Dev.) presupposes an orientation in v/hich merstu must be toward the east. If the Fisian Mount was on Monte Foce, and the asa deueia was toward the eastern part of the area now occupied by the Piazza dei Quaranta Mar- tiri, it is possible to conceive of the augur sitting on the mount and facing in a south- eastward direction. A facing directly toward the east cannot be reconciled with the description of the templum given in Via, because the mountains, on one of
via 14
COMMENTARY
237
the slopes of which the augur would have had to sit, lie on the northern and eastern sides of Gubbio. This southeastward fac- ing is admitted by Devoto ("aliquantum ad orientem, " p. 168; cf. T.G., p. 17), Moreover Goidanich (i/isforfa, VIII [1934], p. 245), while he favors southward facing, nevertheless remarks that the decumanas in Iguvium probably ran not due west and east but northwest and southeast, because of the obstacle which the moun- tains presented against the normal view of the sunrise. In 15-16 hondra esto tudero, etc., it is said that below the limits marked out the parra and crow should be seen dersua, and above the same limits the woodpecker and magpie should be seen mersta. Apparently then, birds seen above the limits are to be counted as higher up the slope of the mountains lying to the
east of the city, and consequently in the (south)eastern region, while those seen below are in the lovver and consequently more western portion of the city. If we insist on a close equivalence between pernaies pusnaes of la 2 and the terms of direction in Via, then pusnaes may be used of those omens which are less forward (literally 'in the back'), and so dersua or western. For the higher and lower situation of augural birds cf. Ser- vius on Aen., Ill, 361 praepetes sunt quae secundo auspicio ante eum volant qui aus- picatur ... Sed praepetes aut superiora tenent et praepetes vocanlur, aut injeriora et dicuntur inferae. There is also a possi- bility that the different distribution of the birds in Plaut., As., 260, cited earlier in the commentary on this portion of Via, may have some connection with the
Monte Foce
Fisian Mount
augural seats i highest angle
238
III. TEXT AND TRANSLATION
(15) hondra . esto . tudero . porsei . subra . screihtor . sent . parfa . dersua . curnaco . dersua . seritu . subra . esto / (16) tudero . peico . mersto . peica . mersta . seritu . sue anclar . procanurent . eso . tremnu . serse / (17) combifiatu . arsferturo . nomne . carsitu . parfa . dersua . curnaco . dersua . peico . mersto . peica . meersta / (18) mersta . aueif . mersta . ancla . eesona . tefe . tote . iiouine . esmei . stahmei . stahmitei . esisco . esoneir . seueir j (19) popler . anferener . et . ocrer . pihaner . perca . arsmatia . habitu . uasor . uerisco . treblanir . porsi . ocrer / (20) pehaner . paca . ostensendi . eo . iso . ostendu . pusi . pir . pureto . cehefi . dia . surur . uerisco . tesonocir . surur / (21) uerisco . uehieir /
gate of Padella. Below these limits which are described above, watch for a parra in the west, a crow in the west. Above these limits watch for a woodpecker in the east, a magpie in the east. If the messengers sing forth, (the augur) sit- ting shall thus make announcement from the tent and shall call the adfertor by name: "(I have seen) a parra in the west, a crow in the west, a woodpecker in the east, a magpie in the east, in the east birds, in the east divine messengers for you, for the state of Iguvium, for this established ordinance." At each of these rites for the lustration of the people and the purification of the mount (the adfertor) shall hold a ritual wand. The vessels at the Trebulan Gate which are to be exhibited for the purification of the mount he shall so exhibit as to cause fire to be kindled from fire. Simil- arly at the Tesenacan Gate, similarly at the Veian Gate.
fact that he divides them into left {picus, comix) and right (coruos, parra), while Table Via divides them into dersua (par- fa, curnase) and merstu {peiqo, peica), the latter replacing the coruos of Plautus. The preceding diagram is intended to illustrate the demarcation of the templum in Iguvium, assuming that the Fisian Mount, from which the observation was made, is the present Monte Foce. The language suggests that the "lowest" and "highest angles" are imaginary points adjacent to but not identical with the Ara Divina and the augural seats respect- ively. ri_n_rL represents the city bound- ary designated by the series of names commencing with 12 uapersusto auieclir and extending in a clockwise direction along the slope of Monte Ingino around through the lower part of the town to 14 pertome padellar (see last of notes on 14).
A partial bibliography of material deal- ing with the augural templum at Rome and Iguvium, beside the standard handbooks on Italic dialects and editions of the Iguvine Tables, would include: K. O. Miiller, Die Etrusker, 2nd ed. by Deecke (Stuttgart, 1877), II, pp. 128-64; Regell, N.Jb., CXXIII (1881), pp. 598-637, Vale- ton, Mnem., XVII (1889), pp. 275-325, 418-52, XVIII (1890) pp. 208-63; Wisso- wa, R.-E., II, pp. 2586-87; Weinstock, Riim. Mitt., XLVII (1932), pp. 95-121; Blumenthal, Klio, XXVII (1934), pp. 1-13; Goidanich, Hist., VIII (1934), pp. 237-69; Ribezzo, I, pp. 55-88, especially 66-78. — 17 combifiatu: the subject is the augur, poet cmgla aseriato east; simi- larly for carsitu. — parfa dersua, etc.: here the augur complies with the instruc- tions given him by the adfertor in 4-5. For the ellipsis of the verb see 136 c. —
Va 17
COMMENTARY
239
18 esisco esoneir: for the expression with -co cf. Va 11 estmesku vepurus. — seueir: Devoto translates ' formulis,' which suits la 5, Via 56, as well as the present passage. But it would have to be derived from a long-vowel form *sewo- in contrast to Skt. sdva- ' incitator, in- citamentum' which he compares (p. 170), for *sewo- would almost certainly show the pit. change ew > ow, as in L. novus < *newos, cf. O. Nuvellum. Since O. sii om (28 Co., line 22) requires 'entirely' or a similar meaning, it seems best to see an expression of totality in U. seueir, seuom sevum, which can be fully equated in a formal sense with O. siuom. Vetter's ' singula ' (p. 236 ; ' unum quidque ' in la 5, Via 56) preserves the notion of individuality belonging to the root *se- which appears to be contained in seueir. — 19 anferener : cf . afero VIb 48 = af erum lb 10; all are from *am(f)-fer-. As a semantic parallel L. shows circumfero : Non., p. 261 circumferre est propria lus- trare; Verg., Aen., VI, 229 idem ter socios pura circumtulit unda. The Umbrian verb like the Latin is suppletive, but makes its pf. forms from the root *d6- 'give': andersafust andirsafust atefa- fust. — perca arsmatia: generally taken as a ceremonial wand; cf. L. pertica, the measuring rod used in surveying, and also a portion of land so measured, and the abbreviated form O. perek: used of road-measurement. For the use of a staff in Roman ritual cf. Paul, e Fest., 49 Li. Commoetaculum genus virgulae qua in sacrificiis utebantur. The verb on which perca(m), perkaf depends is al- ways habiest or some other form of ha- betu, except in VIb 49. The use of anouihimu there has led Br. (p. 56), Dev., (p. 172), Bott,. (p. 267) to regard the perca as a type of toga, and indeed there is some Roman evidence for the cere- monial putting on and taking off of the toga praetexta (cf. for example, Acta fratrum arvalium, pp. 14, 21, 28 Henzen). But there is no etymological support for the interpretation ' toga, ' while the equa-
tion with L. pertica, O. perek., is easy. Moreover it would not have been diffi- cult for anouihimu to extend its meaning from ' put on (clothing) ' to ' take up, equip oneself with (articles held in the hand, etc.). ' I have therefore followed the majority in taking perca as a wand or staff. — uasor: on the case-ending see 103 a. The word is used in IV 22 of con- tainers for the cakes called ar9lataf and in VIb 40 for the persondra. — 20 cehefi dia: obscure, although the general sense seems to be that fire is to be kindled from fire, that is, not kindled anew from flint or by attrition. The practice is per- haps explained by Fab. Pict. apud Aul. Gell., X, 15, 7 Ignem a flaminia, id est flaminis Dialis domo, nisi in sacrum ef- ferri ius non est; Fest., p. 94 Li. Ignem ex domo flaminia efferri non licebat, nisi divinae rei gratia. We may therefore con- jecture that what was forbidden for sec- ular use was obligatory for sacred use. It is unlikely that cehefi and dia are both verbs, for dia would hardly fall into the same class with herter, etc. (152 c); nor is it likely that dia is an adverb placed in final position after the verb of the clause. It seems best then to take dia as pres. subj. from *dwiy-at derived, like purdouitu, pcpl. purdito, L. duint, etc., from an /-extension of the root *d6w- 'give. ' cehefi may be derived from *knd-ti-, the zero-grade of the root found in L. candeo, -cendo (e in the latter being the result of vowel-weakening). The end- ing -i may represent ace. sg. of a stem in -io- (like smursim-el, tesedil) or abl. sg. of a stem in -/- (like poni) or abl. sg. of a stem in -u- (like mani). The least awk- ward construction may be reached by making cehefi an ace. and pir the subject of dia: 'that fire may produce a blaze from fire. ' For the verb ' give ' in the sense 'cause, produce' cf. L. dare morsus, motus, stragem, ruinas. On the failure of the c to suffer palatalization before e see 46 f. On the problem of cehefi in gen- eral see, beside the standard editions, Bechtel, B.B., VII (1883), p. 2; R. G.
240
III. TEXT AND TRANSLATION
(22) pre . uereir . treblaneir . iuue . grabouei buf . treif . fetu . eso . naratu . uesteis . teio . subocau . suboco / (23)
Before the Trebulan Gate he shall sa- crifice three oxen to Jupiter Grabovius. Thus shall he speak after making the libation: "Thee I invoke as the one in-
Kent, C.P., XV (1920), p. 366; Goid^nich. A.G.I. , XXV (1931-32), pp. 52-4; Ribez- zo, I, p. 79. — 20-21 surur uerisco Teso- nocir. surur / uerisco Uehieir: for the pos- sible location of the gates, see Introduc- tion, 4. The spelling Tesonocir with ante- penultimate 0 occurs also in Vila 38, against 4 instances with e e; there is no satisfactory explanation.
Via 22 - 57 (= la 2 - 6). Sacrifice of three oxen to Jupiter Grabovius before the Trebulan Gate. The libation which precedes the sacrifice is followed bj^ a short invocation of the god on behalf of the Fisian Mount and the state of Igu- vium and the names of the Mount and of the state (22 teio subocau ... 25 Dei Gra- boue). This is followed by the prayers accompanying the sacrifice of the first ox (25 Di Grabouie ... 34 tio subocau). The prayers accompanying the sacrifices of the second and third oxen are given in full in 35-44 and 45-55 respectively, al- though in content they are identical with 25-34, except for the use of the words etru and tertiu and a reference in 54 to the triad of oxen, which serves as a sort of conclusion to the whole series of three prayers. For the resemblances and dif- ferences between the present description and the much shorter one in la see the introductory paragraph in the notes on Via and the outline in the Introduction, 13. The principal difference is of course the inclusion of the full text of the prayers in VI.
Via 22 Iuue Grabouei : the epithet oc- curs in several spellings, of which the most frequent is Grabouie in the Latin alphabet, Krapuvi in the native alpha- bet. It is used in VIb 1 = la 11 of Mars, in VIb 19 = la 21 of Vofionus, elsewhere only of Jupiter. Its resemblance to L. Gradivus, epithet of Mars, has struck
many editors, but U. b and L. d cannot be etymologically equated. Most recent authorities follow Kretschmer, Festschrift Adalbert Bezzenberger (Gottingen, 1921), pp. 89-96, in connecting Grabouie with the Illyrian proper names rgd^og, Fqaficov, Epirote Modern Greek ygafioi; ' oak, ' Russ. grab 'hornbeam,' adj. grabovyj, taking account of the intrusion of Illyrian cul- ture into Italy across the Adriatic and the close connection of the oak with the cult of Zeus and Jupiter (for details see Ro- senzweig, pp. 67-72). The Illyrian theory of the origin of Grabouie does not account for L. Gradivus. R. G. Kent, T.A.P.A., LVII (1926), pp. 56-7, who did not follow Kretschmer's view, sought to maintain a connection between the Umbrian and Latin forms by suggesting that Grabouie is in origin the same word as L. Gradivus, altered as a result of a fancied connection with the Umbrian equivalent of bos, the animal which in the Iguvine cult was sacrificed to the three gods who bore the epithet Grabouie. But it may not be too hazardous to suggest that, while the deri- vation of Grabouie from an Illyrian source is correct, at the same time the Latin form may be explained by contamination with gradior (Mars the "Strider") or with divus or with both. — fetu: cognate with L. facio, Gk. ridrjfii, although the stem-formation is not fully clear (see 46 g, j with n. 3, 122 d). The sense ' sacri- fice, offer' is very frequent, the accusa- tive being used of the victim, parts of the victim (e.g. uatuo), the persondro, and aruio, which is here taken to refer to cereal-offerings but by many editors as ' exta' (see on la 3, I la 18). The abla- tive is used of offerings which accompany the principal element of the sacrifice (uinu, poni); the case-ending of uestisia and mefa spefa in VIb 5, Vila 37, is am-
via 20
COMMENTARY
241
biguous. L. facio in the technical lan- guage of ritual is accompanied by the abl. of the victim: Cato, R.R., 139 porco pia- culo facito; Verg., EcL, III, 77 cum faciam vitula pro frugibus. — naratu: etymolog- ically equivalent to L. narrato, but in part very similar in value to persnimu; compare, for example, the present pas- sage with b 25 esoc persnimu uestis, the prayer which follows being practically identical, though the deities addressed in the two prayers are different. The three- fold repetition of the prayer introduced in Via 22 with naratu is followed in 55-6 by takes persnimu seuom. Unlike pers- nimu, however, naratu is always accom- panied by some indication of the content: either the actual words of the prayer or an expression such as tiglu sevakni III 27, or a statement on the quality of the victims (e.g., lib 8, 9), or an indication that the same words are to be used as on some other occasion (e.g.. Via 59, and through VIb and Vila), — uesteis: uestis recurs in VIb 6, 25 in a similar for- mula with persnimu. It has been taken as a pres. pcpl. and as a pf. pcpl., the lat- ter view being decidedly preferable, be- cause the pres. should probably end in -/ (cf. kutef, zeref) or in a vowel after loss of -/, and because the pouring of the libation, which would take almost no time, is not to be thought of as accompany- ing the long prayer, but rather as pre- ceding it. uestis apparently belongs to a 4th conjugation verb related to L. Vesta as punio to poena. If it is not a full de- ponent, the use of the pf. pass. pcpl. may be compared with that of L. cenatus, potus. For further etymological data see on uestisia in VIb 6. — subocau suboco: in addition to its occurrence here and twice in 24, this expression occurs in the similar prayer to Fisovius Sancius in b 6-8 and to Tefer Jovius in b 26-7. subo- cau alone occurs in Via 34, 44, 55, b 15 (twice), 36, and in Vila 20 (twice), 22, 23, 33, 34, 36 (twice), the spelhng in Vila being regularly subocauu. The instances of subocauiu) not followed by suboco are
always at or near the end of prayers. suboco without a preceding subocau does not occur. The pair present a threefold problem: 1) are the two forms both verbs, or is subocau alone a verb ? 2) what tense is subocau, and what tense is suboco, if it is a verb ? 3) if suboco is not a verb but a noun or adjective, what are its stem class and case-ending? It is difficult to take subocau as pf. because the final u in a 1 sg. verb form in the Latin alphabet could only stand for the primary ending -0, while the -uu in subocauu would pro- bably point to a y-perfect (as in L. voca- vi), for which there is no good evidence in the Italic dialects, suboco, if pf. of a denominative verb, should probably re- tain the a of the present stem, as in fut. pf. prupehast, etc., despite such forms as portust, oseto (116 a). The view adopted here is that subocau(u) is 1 sg. pres., while suboco is a noun, possibly a root-stem meaning 'the one invoked,' with pas- sive sense as in the root-nouns L, redux, coniunx, obiex, Gk. diwgv^, au^v^, etc. The fondness for the figura etymologica which characterizes the language of the Iguvine Tables has preserved the tauto- logical expression 'thee I invoke (as) the one invoked, ' and has caused suboco to stand immediately after the verb and before Dei Gr about, with which it is closely connected in syntax, while the true object teio takes its normal place be- fore the verb. For phonological details see 49 d, 54 d, and for discussion of the problem as a whole see A.J.P., LXXVI (1955), pp. 79-82. — 23 Fisiu: the sacred mount or arx bears the name of the god Fisus, to whom under the name Fisus Sansius three suckling pigs are of- fered behind the Tesenacan Gate (VIb 3 = la 14-15). The Mount was probably not Monte Ingino, which is situated to the northeast of the town, but Monte Foce, which lies more directly toward the north; see Introduction, 4. — nomneper: the mention of the name occurs in several other similar formulas of benediction, including Vila 16-17 saluo seritu popler
242
III. TEXT AND TRANSLATION
dei . graboui . ocriper . fisiu . totaper . iiouina . erer . nomneper. erar. nomneper. fos . sei . pacer . sei . ocre . fisei J (24) tote . iiouine . erer . nomne . erar . nomne . arsie . tio . subocau . suboco . dei . graboue . arsier . frite . tio . subocau / (25) suboco . dei . graboue . di . grabouie . tio . esu . bue . peracrei . pihaclu . ocreper . fisiu . totaper . iouina . irer .
voked, Jupiter Grabovius, for the Fi- sian Mount, for the state of Iguvium, for the name of the mount, for the name of the state. Be thou favorable, be thou propitious to the Fisian Mount, to the state of Iguvium, to the name of the mount, to the name of the state. In the consecration I invoke thee as the one invoked, Jupiter Grabovius; in trust of the consecration I invoke thee as the one invoked, Jupiter Grabovius. Jupi- ter Grabovius, thee (I invoke) with this perfect ox as a propitiatory offering for the Fisian Mount, for the state of Igu-
totar liouinar, totar liouinar / name nerf arsmo, etc. The magical value attached to names is shown in its positive aspect here and in its negative aspect through their deliberate avoidance as a result of tabu (see on 1 poet angla aseriato / eest, and Dev., p. 187). The usage with Tusc- Naharc- labusc- in the banishment and curse in VIb 54, 58-9, Vila 12, 47-8, lb 17 is not quite similar; it has rather a notion of comprehensiveness, as in ceivis Romanus neve nominus Latini neve so- cium quisquam, in the Senatusconsultum de Bacchanalibus. — fos sei pacer sei, etc. : the same formula with the subjunctive occurs, making allowance for orthograph- ical variations, in b 7, 26, in each in- stance near the beginning of the prayer. fiiiu fo(n)s pacer with imv. appears in a 30, b 13, 34, of the same prayers, and elsewhere. — 24 arsie tio subocau suboco ... arsier frite tio subocau: the same for- mula occurs in b 8 with Fisoui Sansi and in 27 with Tefro loui, and arsie ... arsier have generally been translated ' sancte ... sancti. ' The value of frite as a noun 'trust' related to L. fretus is reasonably sure. It recurs in VIb 15, Vila 20, 23, 33, 36 in a formula partly similar to the pre- sent one, but without arsie and with a divine name in place of arsier; moreover this second formula regularly occurs near
the end of a prayer. Our understanding of the passage is hindered by the lack of any sure etymology for arsie and the other words which are assumed to be related to it: arsmor, arsmahamu, arsma- tia; but the derivatives of the root *ars- af- in general are associated with reli- gious ritual and priestly functions, while there is no evidence of their direct ap- plication to deities. Devoto translates 'in dedicatione ... dedicationis fiducia' (with discussion pp. 188-9). In Roman ritual dedicatio and consecratio are very closely associated, but in strict usage dedicatio is the function of the magistrate or other person presenting the article, building, etc., for sacred use, while conse- cratio is the function of the priest by whose word and act it becomes sacred property (see J. Marquardt, Romische Staatsverwaltung, III [Leipzig, 1878], pp. 260-1). It may therefore be slightly bet- ter to understand arsie of the formula of consecration. — 25 tio: for the ellipsis of the verb see 136 c. — peracrei : only in Tables VI and VII, except for perakre in lb 40, which is equivalent to Vila 51. In these last two passages the epithet is applied to heifers; in VIb 52, 56 it is fol- lowed by sacris, the victims not being specified; elsewhere peracri- is preceded by bue and followed by pihaclu, and
via 23
COMMENTARY
243
peracrio must be restored in Via 54 buo peracnio pihaclo. In the note on I la 5 it was pointed out that perakne and peracri stand in a sort of supplementary distribution to one another, and also that each at times stands in apparent contrast with forms of sakra sacra. It is therefore not unnatural to understand peracri, like perakni, of victims from secular sources in contrast to those raised as temple property. Devoto in his Italian translation renders both words by 'pro- fano. ' Yet this view, despite the strong arguments in its favor, is not without its disadvantages. In the first place it seems rather strange that the god should be re- minded in the prayers in Via that the victim was ' brought from outside. ' More- over the traditional connection of pe- racri with L. acer, acris on the formal side is simple, while if we derive from *per-ag- we must assume contamination with sacri in order to account for the c in place of g. I have therefore, though hesitantly, maintained the older view whereby peracri is a compound of the same stem seen in L. acer but with a semantic development like that of dx/xaiog. The forms peracri, perakne, seuacne, and also L. sollemnis present an extraordinari- ly difficult problem, complicated prob- ably by contaminations through popular etymology resulting from an imperfect understanding of the archaic liturgical formulae in which the words often occur. — 26 orer ose : variously interpreted, orer is commonly referred to the same pronom- inal stem seen in uru, uraku, as gen. sg. by most editors, as abl. pi. by Blum. ose is taken by most as equivalent to L. opere, by Dev. as equivalent to operis, by Blum, as 2 sg. subjunctive of the verb known through imv. osatu, L. operari. This has the advantage of providing a verb on which the clause pusei neip heritu may depend, but the intrusion of the con- ditional clause persei ... sent is slightly awkward, and ose does not conform to the pres. subj. paradigm of first-conju- gation verbs as seen in aseriaia, kupi-
fiaia. ose is therefore better taken as a noun than as a verb, and both on morph- ological and syntactical grounds is more probably an ablative than a geni- tive. The sense is apparently that the sacrifice of the ox is intended to nullify the evil effect of any fire (caused by lightning) on the Mount or of any omis- sion of necessary religious rites in the city, not merely during the ceremony of purification, but perhaps during the entire time that has elapsed since the last purification. The use of the perfects orto est, subator sent in preference to the future perfect seems to bear out this latter view. — persei acre Fisie pir orto est: compare the precaution in case of fire caused by lightning in the grove of the Fratres Arvales (Act. fr. arv., p. CCXIII Henzen): ... ibi imm(olaverunt}, quod vi tempestat(is) ictu fatmin{is) ar- bor(es) sacr(i) l(uci) d(eae) attact(ae) ar- duer(int)... — 26-7 arsmor dersecor / su- bator sent : Devoto, who takes ose as abl. of time having reference only to the time of the ceremony itself, understands arsmor dersecor subator sent to refer to a premature dismissal of the priestly col- leges, and it would be desirable for arsmo(r) to have a uniform sense here and in 30 below, where the translation ' priesthoods ' is defended in the note. But if we have interpreted the context of arsmor ... subator sent correctly, it is unlikely that a ' dismissal of the priestly colleges ' would be singled out as one of two types of occurrence requiring expiation. We there- fore translate arsmor dersecor as ' due rites. ' — 27 pusei neip heritu : since there are certain objections to taking ose as a verb (see on 26 above), we must assume an ellipsis of the verb on which the pusef-clause should depend, futu, the equivalent of h.esto, may be supplied in thought, heritu is apparently abl. of the passive pcpl. of her, herti, its construc- tion being similar to the impersonal use of the ablative absolute in L. consulto, auspicato, etc. — tuer perscler: partitive genitive with no governing word expressed
244
III. TEXT AND TRANSLATION
nomneper J (26) erar . nomneper . dei . grabouie . orer . ose . persei . ocre . fisie . pir . orio . est . toteme . iouine . arsmor . dersecor / (27) subator . sent . pusei . neip . heritu . dei . crabouie . persei . tuer . perscler . uaseto . est . pesetomest . peretomest j (28) frosetomest . daetomest . tuer . perscler . uirseto . auirseto . uas . est . di . grabouie . persei . mersei . esu . bue f (29) peracrei . pihaclu . pihajei . di . grabouie . pihatu . ocre . fisei . pihatu . tola . iouina . di . grabouie . pihatu . ocrer / (30) fisier . totar . iouinar . nome . nerf . arsmo . ueiro pequo . castruo . fri . pihatu . futu . fos . pacer . pase . tua . ocre fisi j (31) tote . iiouine . erer . nomne . erar .
vium, for the name of the mount, for the name of the state. Jupiter Grabo- vius, by the effect of this (ox) (bring it to pass), if on the Fisian Mount fire hath occurred or in the state of Igu- vium the due rites have been omitted, that it be as not intended. Jupiter Grabovius, if in thy sacrifice there hath been any omission, any sin, any transgression, any damage, any dehn- quency, if in thy sacrifice there be any seen or unseen fault, Jupiter Grabovius, if it be right, with this perfect ox as a propitiatory offering may purification be made. Jupiter Grabovius, purify the Fi- sian Mount, purify the state of Iguvium. Jupiter Grabovius, purify the name of the Fisian Mount, of the state of Igu- vium, purify the magistrates, the priest- hoods, the lives of men and of beasts, the fruits. Be favorable and propitious with thy peace to the Fisian Mount, to the state of Iguvium, to the name of the mount, to the name of the state.
unless we take persei as an indefinite pronoun equivalent to L. quidquid; most assign to it conditional value as in 28 below. — 27-8 uaseio est, pesetom est, peretom est, / f rose torn est, daetom est, etc.: for the meaning of the individual participles which to go make up these perfect passive verbs see the Lexicon. The passage is the clearest example of the tendency mentioned in Introduction, 5, to make the language of the formula as full as possible, as a precaution against any omission which might make the prayer invalid. — 29-31 Di Grabouie ... erar nomne : the general sense of this por- tion of the prayer is clear, but there are certain difficulties not only in the inter- pretation of individual words but, in one place, of the syntax as well. However, the words following nome fall rather easi-
ly into three successive pairs, not alUte- rative as, for example, in the curse in VIb 60 = Vila 49, but showing nonethe- less certain contrasts of meaning which may serve as a guide to the sense of the whole. — 30 nerf: connected by nearly all scholars with dv)]Q, Skt. nar-, nf-, and several Sabine forms cited in classical authors: Suet., Tib., 1, inter cognomina autem et Neronis adsumpsit, quo signifi- caiur lingua Sabina fortis ac strenuus; Aul. Gell., XIII, 23 id autem, sive Nerio sive Nerienes est, Sabinum verbum est, eoque significatur virtus et fortitudo; Ly- dus, De ]\lens., IV, 42 vsQixt] yaq rj avdoia ioTi xai VEQOJvag Tovg dvSQsiovg ol Ea^lvoi KaXovGiv. nerf in our passage cannot, like Gk. dvdQag, be merely the semantic equi- valent of L. viros, for we have ueiro also just ahead. The key to the proper mean-
via 27
COMMENTARY
245
ing of nerf is provided by several Oscan passages: Co. 28, line 29 ... ] um nerum fust, izic post eizuc tr. pi. ni fuid; Co. 108 Minniais Kaisillieis Minateis ner. ; Co. 137f Dekis Rahiis Maralieis niir. In all these passages 'man, men' would be absurd, since the reference is clearly to public officials. In Table VIb 59, 62, etc. are found the expressions nerf sihitu an- kihitu, nenis sihitir ansihitir, followed by the contrasting iouie hostatu anhostatu, iouies hostatir anosfatir. The nerf there are contrasted with the youth but are themselves divided into two opposite class- es. Although the sense of sihitu aniihitu is uncertain, it is most probable that the nerf subdivided in this way include more than the magistrates actually in office. We may therefore translate 'elders,' in the sense ' chief citizens ' ; in other words men comparable in age and social status to those who composed the Roman se- nate. — arsmo : cf . the formula arsmor dersecor subator sent in 26-7; usually translated 'riius,' but the meaning of the other nouns in the series is concrete, and it is not really probable that the god is asked to ' purify the rites. ' The inter- pretation of the word is hindered by the lack of any really satisfactory etymology. The adj. arsmatia, of the wand carried by the adfertor, and the verb arsmahamo are related, but no equations with forms in other languages are at all convincing. arsmo is neuter, if it is correct to assume that the only nom. pi. ending for masc. o-stems is -ur and that forms like pi. arsmor are neuters v/ith extended end- ings (103 a). There is no objection to Devoto's ' sacerdotes' (p. 118, etc.), 'so- cietas sacerdolum' (p. 197), if we under- stand the neut. noun to have a collective sense like ' priesthood, ' the plural here being made to include all the priest- hoods of Iguvium. The civil (nerf) and religious authorities, as the two most im- portant elements in the state, are thus the first to be named in the prayers for purification and safekeeping. — ueiro: almost certainly equivalent to L. vir-,
though the spelling with ei here and in 32, 39 is probably evidence of long i, as in Skt. viras, Lith. vyras (see Lexicon). ueiro pequo form a pair; cf. b 10-11 du- pursus peturpursus ; Ov. Met., I, 285-6 flumina ... / ... pecudesque virosque ... rapiunt; Avestan, Videvdat, VI, 32 pa- subya viraeibya 'for cattle and men. ' The sense of ueiro, etc. in the above passages is more like that of L. homines than of viri. The ending -o, which occurs in every instance of u(e)iro pequo in the Tables, is ambiguous, since it may stand for the gen. pi. (cf. pihaclo 54, Atiersio Vllb 2) or ace. pi. (cf. ehiato Vllb 2, with note. If it is ace. pi. masc, it serves as support for the ending in ueiro ; pequo, like L. pecua, is neut., but ueiro must be masc.) At first glance it would be natural to take ueiro pequo along with the rest of the words in the series as ace, but rofu (once), toru (twice), uitlu (twice) make it appear that -u, not -o, is the normal masc. ace. pi. ending of o-stems. Moreover, if w^e take castruo as equivalent to 'capita' rather than to ' fundos, ' ueiro pequo must almost certainly be dependent genitives, and it is difficult to avoid taking castruo in this way (see on Va 13). The change of case within the series is admittedly a difficulty, but not an insuperable one, since consistency is maintained within the pair. — castruo: see on Va 13; the reasons for taking it as ' (per) head ' in the latter passage and as 'lives' in the pre- sent passage, rather than as ' estates, ' are found partly in the general sense of the passages in Va, partly in the proba- bility on grounds of spelling that ueiro pe- quo in the prayer are genitive plural, in which case the sense 'estates' would be impossible. — fri: like L. fruges it may be taken to include both tree-fruit (L. fructus) and cereals (L. frumenta). — futu fos pacer pase tua: the formula here, with b 13, 34, differs from that in 23, b 7, 26 in the use of imv. instead of subj., in the failure to repeat the verb before pacer, and in the inclusion of pase tua in asso- nance with pacer. For the seeking of
246
III. TEXT AND TRANSLATION
nomne . di . grabouie . saluo , seritu . ocre . fisi . salua . seritu . tola . iiouina . di / (32) grabouie . saluo . seritu . ocrer . fisier . totar . iiouinar . nome . nerf . arsmo . ueiro . pequo . castruo . fri . salua / (33) seritu . futu . fos . pacer . pase . tua . ocre . fisi . tote . iouine . erer . nomne . erar . nomne . di . grabouie . tio . esu . bue / (34) peracri . pihaclu . ocreper . fisiu . totaper . iouina . erer . nomneper . erar . nomneper. di . grabouie . tio . subocau I
(35) di . grabouie . tio esu . bue . peracri . pihaclu . etru . ocreper . fisiu . totaper . iouina . erer . nomneper . erar . nomneper . di / (36) grabouie . orer . ose . persei . ocre . fisie . pir . orto est . tote . iouine . arsmor . dersecor . subator . sent . pusei . neip / (37) hereitu . di . crabouie . persi . tuer . perscler . uahetom . est . pesetomest . peretomest . frosetomest . daetomest . tuer I (38) perscler . uirseto . auirseto . uas . est . di . grabouie . persi . mersi . esu . bue peracri . pihaclu . etru . pihafi . di . grabouie j (39) pihatu . ocre . fisi . pihatu . tola . iouina . di . grabouie . pihatu . ocrer . fisier . totar . iiouinar . nome . nerf . arsmo . ueiro / (40) pequo . castruo. fri . pihatu . futu . fos . pacer . pase . tua . ocre . fisie . tote . iiouine . erer . nomne . erar . nomne . di /
Jupiter Grabovius, keep safe the Fisian Mount, keep safe the state of Iguvium. Jupiter Grabovius, keep safe the name of the Fisian Mount, of the state of Igu- vium, keep safe the magistrates, the priesthoods, the hves of men and of beasts, the fruits. Be favorable and pro- pitious with thy peace to the Fisian Mount, to the state of Iguvium, to the name of the mount, to the name of the state. Jupiter Grabovius, thee with this perfect ox as a propitiatory offering for the Fisian Mount, for the state of Igu- vium, for the name of the mount, for the name of the state, Jupiter Grabovius, thee I invoke.
"Jupiter Grabovius, thee with this perfect ox as a second propitiatory offer- ing for the Fisian Mount, for the state of Iguvium, for the name of the mount, for the name of the state (I invoke). Jupiter Grabovius, by the effect of this (ox) (bring it to pass), if on the Fisian Mount fire hath occurred or in the state of Iguvium the due rites have been omit- ted, that it be as not intended. Jupiter Grabovius, if in thy sacrifice there hath been any omission, any sin, any trans- gression, any damage, any delinquency, if in thy sacrifice there be any seen or unseen fault, Jupiter Grabovius, if it be right, with this perfect ox as a second propitiatory offering may purification be made. Jupiter Grabovius, purify the Fi- sian Mount, purify the state of Iguvium. Jupiter Grabovius, purify the name of the Fisian Mount, of the state of Igu- vium, purify the magistrates, the priest- hoods, the lives of men and of beasts, the fruits. Be favorable and propitious with thy peace to the Fisian Mount, to the state of Iguvium, to the name of
via 30
COMMENTARY
247
pax deorum or security from harmful vis- itation by tlie gods cf. Lucr., V, 1229-30 non divom pacem votis adit ac prece quaesit ventorum pavidus paces ...?; Verg., Aen., Ill, 370 exorat pacem divum; Livy, VII, 2, 2 pads deum exposcendae causa ... lec~ tisternium fuit. — 31 salao seritu, etc.: cf. Cato, R.R., 141, 3 pastores pecuaque
salva servassis.
Via 35 - 55 : see above in the introduc- tory paragraph to 22-57. Comment is unnecessary, since the prayers which ac- company the sacrifices of the second and third oxen are identical with those for the first, barring minor orthographic va- riations and the use of the ordinals etru
(41) grabouie . saluo seritu . ocre fisim . salua . seritu . totam iiouina . di . grabouie . saluuom . seritu . ocrer . fisier . totar / (42) iiouinar . nome . nerf . arsmo . uiro . pequo . castruo frif . saluua . seritu . futu . fons . pacer . pase . tuua ocre . fisi . tote / (43) iiouine . erer . nomne . erar . nomne . di . grabouie . tiom . essu . bue . peracri . pihaclu . etru . ocriper . fissiu . totaper . iouina . erer / (44) nomneper . erar . nomneper . di. grabouie. tiom . subocau /
(45) di . grabouie . tiom . esu . bue . peracri . pihaclu . t . ertiu . ocriper . fisiu . totaper . iiouina . erer . nomneper . erar . nomneper . di J (46) grabouie orer . ose . pirse ocrem . fisiem . pir . ortom . est . toteme . iouinem . arsmor . dersecor . subator . sent . pusi . neip / (47) heritu . di . grabouie . perse . tuer . pescler . uasetom est . pesetom . est . peretom . est . frosetom . est . daetom est . tuer / (48) pescler . uirseto . auirseto . uas est . di . grabouie . pirsi . mersi . esu . bue .
the mount, to the name of the state. Jupiter Grabovius, keep safe the Fisian Mount, keep safe the state of Iguvium. Jupiter Grabovius, keep safe the name of the Fisian Mount, of the state of Igu- vium, keep safe the magistrates, the priesthoods, the lives of men and of beasts, the fruits. Be favorable and pro- pitious with thy peace to the Fisian Mount, to the state of Iguvium, to the name of the mount, to the name of the state. Jupiter Grabovius, thee with this perfect ox as a second propitiatory of- fering for the Fisian Mount, for the state of Iguvium, for the name of the mount, for the name of the state, Jupiter Gra- bovius, thee I invoke.
"Jupiter Grabovius, thee with this per- fect ox as a third propitiatory offering for the Fisian Mount, for the state of Iguvium, for the name of the mount, for the name of the state (I invoke). Jupiter Grabovius, by the effect of this (ox) (bring it to pass), if on the Fisian Mount fire hath occurred or in the state of Iguvium the due rites have been omitted, that it be as not intended. Ju- piter Grabovius, if in thy sacrifice there hath been any omission, any sin, any transgression, any damage, any delin- quency, if in thy sacrifice there be any seen or unseen fault, Jupiter Grabovius, if it be right, with this perfect ox as a
248
III. TEXT AND TRANSLATION
peracri . pihaclu . tertiu . pihafi . di . grabouie J (49) pihatu . ocrem . fisim . pihatu . totam . iiouinam . di . grabouie . pihatu . ocrer . fisier . totar . iiouinar . nome nerf . asmo / (50) uiro . pequo . castruo . fri . pihatu . futu . fons . pacer . pase . tua . acre . fisi tote . iiouine . erer . nomne . erar . nomne . di / (51) grabouie . saluo . seritu . ocrem . fisim . saluam . seritu . totam . iiouinam . di . grabouie . saluom . seritu . ocrer . fisier / (52) totar . iiouinar . nome . nerf . arsmo . uiro . pequo . castruo . frif . salua . seritu . futu . fons . pacer . pase . tua . ocre . fisi J (53) tote . iiouine . erer . nomne . erar . nomne . di . grabouie . Horn . esu . bue . peracri . pihaclu . tertiu . ocriper . fisiu . totaper / (54) iiouina . erer . nomneper . erar . nomneper . di . grabouie . tio . comohota . tribrisine . buo . peracnio . pihaclo / (55) ocriper . fisiu . totaper . iiouina . erer . nomneper . erar . nomneper . di . grabouie . tiom . subocau . tases . persnimu / (56) seuom . surur . purdouitu . proseseto . naratu . prosesetir . mefa . spefa . ficla . arsueitu . aruio . fetu . este / (57) esono . heri . uinu . heri . poni . fetu . uatuo . ferine . fetu /
third propitiatory offering may purifi- cation be made. Jupiter Grabovius, pu- rify the Fisian Mount, purify the state of Iguvium. Jupiter Grabovius, purify the name of the Fisian Mount, of the state of Iguvium, purify the magistrates, the priesthoods, the lives of men and of beasts, the fruits. Be favorable and propitious with thy peace to the Fisian Mount, to the state of Iguvium, to the name of the mount, to the name of the state. Jupiter Grabovius, keep safe the Fisian Mount, keep safe the state of Iguvium. Jupiter Grabovius, keep safe the name of the Fisian Mount, of the state of Iguvium, keep safe the magis- trates, the priesthoods, the lives of men and of beasts, the fruits. Be favorable and propitious with thy peace to the Fisian Mount, to the state of Iguvium, to the name of the mount, to the name of the state. Jupiter Grabovius, thee with this perfect ox as a third propi- tiatory offering for the Fisian Mount, for the state of Iguvium, for the name of the mount, for the name of the state, (I invoke). Jupiter Grabovius, bringing forward the triad of perfect oxen as pro- pitiatory offerings for the Fisian Mount, for the state of Iguvium, for the name of the mount, for the name of the state, Jupiter Grabovius, thee I invoke." He shall pray each (portion) silently. In the same manner he shall make the pre- sentation, make pronouncement upon the parts cut off, add to the parts cut off a mefa spefa cake and a ficla cake, and offer grain. He shall perform this sa- crifice either with wine or with mead. He shall place the ribs on a tray.
54 peracnio: for peracrio.
via 35
COMMENTARY
249
(35, 38, 43) and tertiu (45, 48, 53). — 54 //o comohofa tribrisine buo peracnio pihaclo: cf. Cato, R.R., 134, 2 luppiter, te hoc ferto obmovendo bonas preces pre- cor. Beside tribrisine the nom. tribfigu occurs in Va 9; for the formation see 60 e, 88, and Lexicon. peracrio must almost certainly be restored in place of peracnio, the reading of the bronze, since pe.racr(e)i is used 9 times as epithet of the oxen in 25-53, while no other instances of peracn- occur in Table VI.
Via 55 - 57. Further instructions for the sacrifice before the Trebulan Gate. The instructions in the corresponding version in la 3-6 are similar, except that kutef pesniniu appears in place of tases persnimii, arvia ustentu in place of aruio feiu, afepes arves is included but the reference to a specific type of cake is omitted, and the order in which the va- rious instructions are presented is not the same. For the general character of these instructions, which are concerned partly with the manner of performing the sacrifice, partly with the supplement- ary offerings to be included, see Introduc- tion, 15. — 55 tases persnimu: cf. Ov., Met., IX, 300 tacita quoque carmina voce dixit. — 56 purdouitu : the presentation to the god, the second of the three prin- cipal stages in the sacrifice, although the actual slaying of the victim, the first stage, has not been previously mentioned, except in so far as it is implied by the general verb fetu in 22. Cf. Cic, ad Att., V, 18, 1 ne quid inter caesa et porrecta, ut aiunt, oneris mihi addatur aut temporis; Varro, L.L., VI, 31 intercisi dies sunt per quos mane et vesperi est nefas, medio tem- pore inter hostiam caesam et exta porrecta fas. — proseseto naratu : the prosecta or parts cut off for presentation to the god are the exta, if we may judge from Cato, R.R., 134, 4 ubi exta porsecta erunt; Varro, L.L., V, 110 insicia ab eo quod insecta caro, ut in Carmine Saliorum <prosicium> est, quod in extis dicitur nunc porsectum. The exta are identified in Cic, Div., II, 12, 29 as the liver,
heart, and lungs. The adfertor pronounces upon the prosecta, presumably upon their satisfactory condition. Cf. lib 7-8 si perakne sevakne upetu eveietu, sevak- ne naratu, and similarly 10-11. — me fa spefa: this rhyming pair occurs nine times in VI and VII, while mefa alone occurs several times in the older tables and in VIb 17. Practically all editors take mefa to be a cake, and since the equation with L. mensa is easy, it may be a broad flat cake to be placed under the prosecta, recalling the famous eating of the mensae in Verg., Aen., Ill, 255-7, VII, 112-6. The most satisfactory etymological ex- planation is that of Kretschmer, G/., VIII (1916), pp. 79-82 : mefa spefa = L. mensa pensa ; spefa is cognate with L. pendo, pendeo, and the semantic development of the expression is 'what is measured and weighed > flat cake.' — ficla: this type of cake, the exact nature of which is unknown, is added to the prosecta in the sacrifices before and behind the three gates and at the other stops during the purification of the Mount, and also in the lustration of the people. If the word is from *fig-kla, with the root of L. fingo, according to the usual view, we may as- sume that the cake was kneaded or fash- ioned into a certain shape. It may have been similar to the L. fertum, since fertum and strues are associated together in Cato, R.R., 134, the former as an offer- ing to Jupiter, the latter to Janus, while ficla and struhla, the equivalent of L, strues, are associated in several Iguvine passages. — aruio feiu : see on la 3 for the relation of this expression to the arvia ustentu of the earlier tables, and on Ila 18 for the grounds on which aruio arvia throughout the present work is taken to signify vegetable offerings rather than 'exta.' — 57 heri uinu heri poni : the choice of wine or poni is permit- ted before the Trebulan and Veian Gates and at the grove of Coredius; elsewhere in the purification and lustration poni alone is prescribed, and the corresponding passages of I and VI are fully consistent.
250
III. TEXT AND TRANSLATION
(58) post . uerir . treblanir . si . gomia . trif . fetu . trebo . iouie . ocriper . fisiu . totaper . iiouina . persae . fetu . aruio . fetu I (59) pone . fetu . tases . persnimu . surur . naratu . puse . pre . uerir . treblanir . prosesetir . strusla . ficla . arsueitu /
VIb 1 pre . uerir . tesenocir . buf .
trif . fetu marte . grabouei . ocriper .
fisiu totaper . iiouina . aruio . fetu .
uatuo . ferine . fetu poni / (2) fetu .
tases . persnimu . prosesetir . farsio . ficla . arsueitu . surur . naratu . puse pre . uerir . treblanir /
Behind the Trebulan Gate he shall sa- crifice three pregnant sows to Trebus Jo- vius for the Fisian Mount, for the state of Iguvium. He shall perform (the sa- crifice) upon the ground, offer grain, sa- crifice with mead, pray silently, recite the same formulas as before the Trebu- lan Gate, and add to the parts cut off a strusla cake and a ficla cake.
Before the Tesenacan Gate he shall sacrifice three oxen to Mars Grabovius for the Fisian Mount, for the state of Iguvium. He shall offer grain, place the ribs on a tray, sacrifice with mead, pray silently, add spelt-cakes and a ficla cake to the parts cut off, and recite the same formulas as before the Trebulan Gate.
poni has been variously interpreted, but most regard it as po-ni- based on the root *p6- ' drink, ' the problem being to de- termine what the drink-offering actually was. Hu. translated 'mulso,' Br. 'lacte', Bii. ' posca, ' followed by PI. and Bk. Dev., p. 204, is probably correct in assuming it to be an alcoholic product of grain or honey established in ritual use before the knowledge of grape-wine; 'mead' is therefore a suitable conjecture. For an offering of mulsum cf. Livy, X, 42, 7 voverat lovi Victori, si legiones hostium fudisset, pocillum mulsi, priusquam te- metum biberet, sese facturum. For poni see, beside the standard editions, Tliur- neysen, GL, I (1909), p. 242; F. Edgerton, A. J.P., XXXIII (1912), p. 224 ; Goidanich, A. G. L, XXV (1931), pp. 95-112; Hof- mann, B.Jb., CCLXX (1940), pp. 89-90; Kent, Lang., XIV (1938), p. 216. — uatuo ferine fetu: this instruction occurs in connection with the sacrifices before, but not behind, the three gates, and also at the groves of Jupiter and Coredius and at the Fontuli. Many interpretations have been proposed, though none is sure. On phonological grounds it is easy to set
up a connection with L. latus, -eris, with uv < I- (55 a) and stem-variation as in L. pecu: pecus, -oris. The uatuo may then be the side-portions or ribs, which are to be placed on a tray or carrier of some kind (ferine) and reserved for the priests and other persons. For the setting aside of certain parts of the victim for profane use, see Wissowa, Religion and Kultus der Romer^, p. 419, and for detailed discus- sion of the formula see A. J. P., LXXVI (1955), pp. 77-9. The noun ferine may be related to L. ferio only in the unlikely event that uatuo indicates victims, which then are slain cultro or ictu (so Blum.). Otherwise it must be connected with fertu, L. fero, as some sort of carrier on which the uatuo are placed or presented. Via 58 - 59 (= la 7 - 10). Sacrifice of three pregnant sows to Trebus Jovius behind the Trebulan Gate. For pregnant sows as offerings cf. Cic, Div., I, 45, 101 scripium ... est, cum terrae motus factus esset, ut sue plena procuratio fieret, vocem ab aede lunonis ex arce exstitisse. — 58 Trebo Iouie: this deity is mentioned only here and in the corresponding la 8, where the 0- or a-stem form Trebe is difficult
via 57
COMMENTARY
251
to reconcile with Trebo, since both must be dat. sg. For the possibility that Trebo may show a rare a-stem ending see 25 f. If this possibility is rejected, Trebo must be an u-stem. Since a divine name is more likely to have changed from an u- to an 0- or a-stem than the reverse, the later table may here preserve an older form, as Vetter, p. 245, suggests. The sex of the deity is unknown ; the principle that female victims are offered to female deities generally holds good, but in view of the ewe-lambs offered to Tefer Jovius in VIb 22 ff. = la 24 ff., we dare not in- sist on it too rigidly. Of the two genders the masculine is more in keeping with the u-stem. If Trebo is from the same root as trebeit, tremnu, O. triibum, L. trabs, etc., as seems likely, it is probably the name of some deity of the house, in- herited from a stage of Indo-European culture when the wooden house was the rule (cf. Dev., p. 211). — persae fetu: the victims in connection with which this expression occurs are sows, young pigs, boars, heifers, she-calves, sheep, and dogs. persae is regularly followed by fetu ex- cept in I la 21-2 esunu / pefae futu, while in III 32 pelsanu intervenes. It is quite certain that the expression desig- nates the position of the victim at the moment of sacrifice, and also that per- sae pefae(m) is in some way connected with persi peri 'foot, ' and perso peruixi. The latter term, whether taken as ' fos- sam ' with most editors, or ' mound, ' as in the present work, gives a satisfactory basis for the formation of an adj. *pedaio- signifying location, like pernaiaf from perne, pustnaiaf from postne. There seems therefore to be no good reason to depart from Bucheler's view (p. 64) that the victim was stretched on the ground. The same type of sacrifice may be designa- ted by the Latin expression sacrificium humanum: Paul, e Fest., 91 Li. Humanum sacrificium dicebant, quod mortui causa fiebat; Gell., V, 12, 11-12 Simulacrum igitur dei Vediovis, quod est in aede, de qua supra dixi, sagittas tenet, quae sunt
videlicet partae ad nocendum. Quaprop- ter esse demum plerumque Apollinem esse dixerunt; immolaturque ritu humano capra eiusque animalis figmentum iuxta simula- crum stat. Cf. also VIb 5, with note on sopo. — 59 strusla-- this cake accompa- nies the sacrifices offered behind the three gates and at Rubinia, across the Sacred Way, and at Acedonia, but only in VI and VII; it is not mentioned in the corresponding portions of I. It is also used in the dog-sacrifice in Ila, and in the sacrifices in III and IV. In every one of its occurrences except in these last two tables, strusla is found in juxtaposi- tion with fie la. The nature of the strusla may be inferred from that of L. s trues, to which the Umbrian word is probably related as a diminutive: cf. Fest., 407 Li. s trues genera liborum sunt, digitorum conjunctorum non dissimilia qui superiec- ta panicula in transversum continentur.
VIb 1 - 2 (= la 11 - 13). Sacrifice of three oxen to Mars Grabovius, the second of the Grabovian trinity, before the Te- senacan Gate. The only new feature is the mention of farsio, which together with the ficla are to be added to the prosecta. The word is here used substan- tively of spelt-cakes. Cf. Fest., 78 Li. farreum genus libi ex farre factum; Plin., N.H., XVIII, 10 quin et in sacris nihil religiosius confarreationis vinculo erat, novaeque nuptae farreum praeferebant. farsio are also added to the prosecta in 44 in the sacrifice to Mars Hodius at the grove of Jupiter and in Ila 12 in the sacri- fice of a boar to Ahtus Mars. They are not included, however, in the passages of la, b corresponding to VIb.
VIb 3 - 18 (= la 14 - 19). Sacrifice of three sucking pigs to Fisus Sancius be- hind the Tesenacan Gate, followed by supplementary offerings to Fisovius San- cius. The sacrifice proper is contained in 3-5. For the victims cf. Cic, Leg., II, 12, 29 iam illud ex institutis pontificum et haruspicum non mutandum est, quibus hostiis immolandum cuique deo, cui maiori- bus, cui lactentibus, cui maribus, cui fe-
252
III. TEXT AND TRANSLATION
(3) post . uerir . tesenocir . sif . filiu . trif . fetu fiso . sansie . ocrifer . fisiu . totaper . iiouina . poni . feitu . persae . fetu . amio . fetu / (4) surur . naratu . pusi . pre . uerir . treblanir . tases . persnimu . mandraclo . difue . destr e . habitu . prosesetir . ficla / (5) strusla . arsueitu . ape . sopo . postro . peperscust . u.estisia . et . mefa spefa .
3 ocrifer: for ocriper.
minis; Am., VII, 18 ut ille tauris deus, haedis alius honorelur aut ovibus, hie lactentibus porculis, alter intonsis agnis. — 3 Fiso Sansie: the name of this god occurs nowhere else except in the equi- valent la 15, with Fise showing the same stem-variation seen in Trebe: Trebo (see on Via 58). Yet the name of the Fisian Mount is evidence of the importance of his cult in Iguvium. Fiso, Fisier are generally believed to be derived from the same root as L. fido, fides, etc. The god was then a protector of oaths and solemn pledges. This aspect of his character is further confirmed by the use of the epithet Sansie, which is found also in the equi- valent la 15 and frequently with Fisoui in VIb 6-15, Vila 37; there are also several occurrences of Sage Sagi in I la, b. The connection of Sansie with L. sancio, -ire is obvious; for the use of the Latin verb in relation to solemn oaths, cf. Livy, XXV, 8, 8 rursus fide sanxerunt liberos Tarentinos leges suas suaque omnia habi- turos. Fiso Sansie further recalls the as- sociation of the Latin names Fidius and Sancus attested by Varro, L.L., V, 66 Aelius Dium Fid<.i ^um dicebat Diovis filium ... et putabat hunc esse Sancum ab Sabina lingua [Sancum Scaliger's emen- dation of Sanctum]; Ov., Fast,Vl, 213-18 Quaerebam, Nonas Sanco, Fidione refer- rem, an tibi, Semo pater. Turn mihi
Behind the Tesenacan Gate he shall sacrifice three sucking pigs to Fisus San- cius for the Fisian Mount, for the state of Iguvium. He shall sacrifice with mead, perform (the sacrifice) upon the ground, offer grain, recite the same formulas as before the Trebulan Gate, pray silently, have a maniple folded double upon his right hand, and add to the parts cut off a ficla cake and a strusla cake. When he has put the under-parts at the back (of the altar), kneeling he shall offer a libation and a mefa spefa cake in a cup
Sancus ait: ' Cuicumque ex istis dederis, ego munus habebo. Nomina terna fero, sic voluere Cures.' Hunc igitur veteres dona- runt aede Sabini, inque Quirinali consti- tuere iugo; Fest., 276 Li. siatua eius [sc. Gaiae Caeciliae] est in aede Sancus, qui deus dius fidius vocatur. — 4 mandraclo difue destre habitu: not found in the cor- responding passage in la 14-16, but the inantra(h)klu is called for in I la 19 in the sacrifice of a dog to Hondus Jovius and in lib 16 in the sacrifice of a he-goat to Sancius Jupiter. Of particular interest are certain coins of Tuder showing a hand wrapped with a cloth two folds of which cross on the back of the hand, a possible explanation of difueinour text. SeeA.-K., I, pi. 9; R. Garrucci, Le monete dell' Italia antica (Rome, 1885), p. 29, pi. LV. The association of the cloth over the hand with the solemnization of pledges receives further support from Latin sources: Livy, I, 21, 4 Fidei sollemne instituit [sc. Numa]. ad id sacrarium flamines bigis curru arcuaio vehi iussit manuque ad digi- ios usque involuta rem divinam facere, significantes fidem tutandam sedemque eius etiam in dexteris sacratam esse; Serv., ad Aen., I, 292 canam Fidem dixit vel quod in canis nominibus invenitur; vel quod ei albo panno involuta manu sacrifi- catur, per quod ostenditur fidem debere esse secretam. The mandraclo in an etymo-
VIb 2
COMMENTARY
253
logical sense is, like L. maniele, a cloth for wiping the hands (for phonology see 18 a). It is interesting to observe that mantele is one of several medieval designa- tions of the maniple worn over the left forearm by the priest celebrating Mass; the Catholic use of the maniple, however, seems to have originated in the usage of certain Roman civil magistrates; see En- ciclopedia Cattolica under manipolo. — 5 ape sopo postro peperscust, etc.: here begins the description of special cere- monies in honor of Fisovius Sancius with a libation, a mefa spefa, and poni (9). The prayer in 6-8 agrees in almost all details with Via 22-25 as far as suboco Dei Graboue, and b 11 futu fons to its conclusion in 15 is practically a repeti- tion of Via 30-4, but the passage extend- ing from b 9 to 11 uouse auie esone is pe- culiar to the ceremony in honor of Fiso- vius Sancius. The precise relationship of Fisovius Sancius to the Fisus Sancius of 3 (= la 15) is not clear. Rosenzweig, p. 74, takes both gods to be the same, and there can be no doubt that they are at least closely connected. Moreover we cannot argue for a difference in the deities from the fact that the sacrifice to Fisus Sancius in 3-5 was a complete ceremony in itself calling for the same prayers as those used before the Trebulan Gate, while the new ceremony has differences in the prayer as well as in the use of a mefa spefa instead of the sirusla of 5; for in Vila also the sacrifice of three sows at Rubinia is followed at 8 by a new cere- mony, again introduced by ape supo postro pepescus and containing new prayers, but, like the sacrifice itself, performed in honor of Prestota Serfia. In its formation Fi- sovio- has the same relation to the u-stem Fisu- that L. Marcius has to the o-stem Marcus, but it is impossible to say with certainty whether we have here a gene- alogy of deified abstractions or merely a substitution of a patronymic for the primary name. In Vila the hbation to Prestota Serfia is followed from 37 through 40 by offerings to Fisovius Sancius, the
description of which agrees closely in many details with the present passage. — sopo: editors in general admit formal equivalence with L. suppus, but it is not certain what part of the victim is meant. With ape sopo postro peperscust compare Vila 8, where the same expres- sion occurs, and supa pustra perstu in Ila 32. The victims of which sopo, sopa, etc., is used are sows and young pigs, sheep, and dogs; in other words, some of the same victims with which the expression persae fetu is associated (see on Via 58). Moreover the supa sumtu of la 9, 16 is replaced by persae fetu in the corresponding Via 58, b 3, in environ- ments otherwise almost identical. The rare L. suppus does not give direct help in understanding sopo; the most impor- tant passages are: Fest., 370 Li. suppum antiqui vocabant quern nunc supinum di- cimus ... eius vocabuli meminit Luci<li>- us: "si vero das quod rogat, et si suggeris suppus"; Lucr., I, 1061-2 et simili ra- tione animalia suppa vagari contendunt. It is the etymological relationship and equivalence of meaning between sopo, L. suppus, and supinus, vnnoq that may provide the answer. The words of this group may originally have stood in con- trast to 'erect'; at least this is the con- trast imphed in the Lucretian passage on the manner in which animals walk. But the usual contrast is with pronus, jiQrjvrjz. sopo then seems to have been specialized to mean the parts of the victim which are on top when the victim is supine; in other words, the under-parts. Some support for this meaning of sopo may be found in a similar use of vmiog in Arist. P. A., 658a, 15-19 xai rwv ixovrcov xQixaq en' djucporega oi dvdQcojtoi /novov exovaiv. rd ycLQ TETQanoda rd)v ^cocov ev rolg vtitioiq ova exet TQLxaQ, dAA' ev toIq Ttgaveat [xaXXov • oi d'dvdQCOTtoi rovvavxLov ev rolg vnrioig [xaXXov i] ev rolg Jiqaveaiv. Devoto, p. 208, makes a distinction between proseseto, 'exta' and sopo 'viscera'; but see Kent, Lang., XIV (1938), p. 214, and Devoto, T.G., passim, where sopa is 'carni' and
254
III. TEXT AND TRANSLATION
scalsie . confgos . fetu . fisoui . sansi j
(6) ocriper . fisiu . totaper . iouina . eso . persnimu . uestisia . uestis . tio . subocau . suboco fisoui sansi . ocrider . fisiu
(7) totaper . iiouina . erer . nomneper . erar nomneper . fons . sir . pacer, sir . ocre . fisi . tote . iiouine . erer . nomne /
(8) eras nomne . arsie . tiom . subocau . suboco . fisoui . sansi . asier . frite . tiom . subocau . suboco fisoui . sansi . suront j {9)poni . pesnimu . mefa spefa . eso . persnimu . fisouie . sansie . tiom . esa . mefa . spefa . fisouina ocriper . fisiu totaper . iiouina / (10) erer . nomneper . erar . nomneper fisouie .
to Fisovius Sancius for the Fisian Mount, for the state of Iguvium. Thus shall he pray after pouring the libation: "Thee I invoke as the one invoked, Fisovius Sancius, for the Fisian Mount, for the state of Iguvium, for the name of the mount, for the name of the state. Be thou favorable, be thou propitious to the Fisian Mount, to the state of Iguvium, to the name of the mount, to the name of the state. In the consecration I in- voke thee as the one invoked, Fisovius Sancius, in trust of the consecration I invoke thee as the one invoked, Fiso- vius Sancius. " Likewise shall he pray with the mead. Thus shall he pray with the mefa spefa cake: "Fisovius Sancius, thee with this Fisovian mefa spefa cake for the Fisian Mount, for the state of Iguvium, for the name of the mount, for the name of the state (I invoke). Fiso-
5 confgos: for conegos.
6 ocrider: for ocriper.
proseseto is 'visceri. ' — uestisia: from an examination of the passages contain- ing this noun and its cognate verb uesti- catu it appears that the meaning is 'li- bation, offer a libation ' ; so, for example, in I la 24, 31, 35, where vestikatu is used in association with puni, ves- icles, kapife. Devoto, p. 185, recognizes two distinct uses of uestisia, uesticaiu, illustrated especially in Ila 7-16: the aspersio or sprinkling of the victim before sacrifice, and the libation in the strict sense, which is additional to and not a part of the sacrifice proper. The etymo- logy of the two words, with uestis, is un- certain, but a plausible explanation has been offered by Kretschmer, Gl., X (1920), pp. 154-6, in part anticipated by Biicheler, pp. 52-3: uesticatu is related to L. Vesta as ctaudicare to claudus, fabricare to faber. *wesiicd may be assumed as an intermediate stage in the formation of
the actual form uestisia. uestis may be assigned to a verb related to Vesta as punio to poena, with analogy of persnis as a contributing factor in its formation. The semantic development of the whole group shows a generalization from liba- tions poured to Vesta (or else on the hearth) to libations poured to any deity. Moreover the uestisia cannot always have been a liquid offering, since the expres- sions la 28 esmik vestigam preve fiktu, 31 esmik vestiga afiktu suggest that it was something sohd, to be imposed and fixed to the persuntru; but in no case can the precise nature of the uestisia be determined. — scalsie conegos: with minor variations this expression occurs in 16 in connection with the eras of the libation; in Vila 37, which is almost identical with the present passage, and in IV 15, 18, 20 in the offerings of hba- tions, etc., to Pomonus and Vesona, to
VIb5
COMMENTARY
255
Hula, and to Tursa. The sense of scalsie has been determined by etymological means (see Lexicon), but since it was used not only for liquids but also for the mefa spefa, which was almost certainly a type of cake, it must have been a broad, shallow bowl rather than a cup. Only in the present passage, with Vila 37, is scalsie followed by the postposition -e; all the other examples have -to -ta. If we take the forms at their face value, we must assume a difference in the manner of presentation, according as the uestisia, etc., are offered in the bowl or poured out from the bowl. — conegos (^kimikaz): the majority of scholars have assumed connection with L. coniveo, whose mean- ing 'close the eyes, wink' would then be derived from 'nod, bend over,' the sense of the root *kn.eig'^h-. But the mean- ing ' kneeling, ' which seems to be re- quired for the Umbrian forms, receives better etymological support if we connect these forms with L. genu, nitor, e-nitor, nixus^ OL gnixus, the whole group pos- sibly being further connected with gigno, etc., through the kneeling position of the mother at childbirth. Cf. R. Back, I.F., XL (1922), pp. 162-5; Kretschmer, G/., XIV (1925), p. 224; Pisani, p. 153. The kneeling position in religious ritual is illustrated by certain Oscan coins show- ing a young man kneeling during the presentation of a pig in the striking of a treaty. Cf. J. Friedlander, Die oskischen Miinzen (Leipzig, 1850), p. 81. — 9 Fisouie Sansie ... 11 uouse auie esone: this portion of the prayer is found nowhere else in the Tables, but its use is called for in Vila 37-38 through the instruction suront naratu puse post uerir Tesonocir. The succession of alliterative pairs in 11 is striking. — 10-11 dupursus peturpiir- sus: for this designation of men and beasts cf. Cic, Dom., 18, 48 hoc ministro omnium nan bipedum solum sed etiam quadripedum impurissimo; Rigveda, X, 85, 42 sam. no bhavadvipade sam catuspade 'be a blessing to our bipeds, a blessing to our quadru- peds' (Lanman). Dev., pp. 224-5, gives
additional examples from Sanskrit as well as several from Avestan and calls atten- tion to the gradation -purs- : persi ' foot, ' like jiarrjQ : evndrviQ, as evidence of the antiquity of the compounds. — 11 fato should not be equated with L. factum, as in several earlier editions, but with fatum, as implied by the more recent transla- tions 'fata' (Blum.), ' vaticinium' (Dev., Bott.), 'fatum' (Pis., Vet.). The parti- ciple of fagia is not *fato- < *fahto-, but feto- (cf. feta lib 13). fito is uncertain because the i may represent I, a if equi- valent to Skt. bhutam, or e, in which case it would be an orthographic variant of the stem seen in feta from fa§ia. It seems most natural, however, to derive it from *bhwl-tom, thus bringing it into close relationship with (piTv, O. fiiet, L. fio, OL fitum. The pair then show a contrast of sense similar to that in 'word and deed' and may be translated 'pre- diction and outcome' ('das Vorbestimm- te, das Eingetretene, ' Vetter). — perne postne: this pair presents no difficulty; cf. TiQoaau) xal onioaoj. — sepse sarsite: no solution can be regarded as certain, sar- site is generally taken as an adv. in e- as if to L. *sarcitus in place of sartus, the actual pcpl. of sarcio. The develop- ment of the notion of wholeness or entire- ty from that of mending may be illus- trated by Charis., p. 220 Keil: Sarcte pro integre. sarcire enim est integrum facere. hinc ' sarta tecta uti sint' opera publica locantur et ut Porphyrio ex Verrio et Festo ' in auguralibus ' inquit ' libris ita est' sane sarcteque ... ' sepse may be a synonym to sarsite, as 'completely' to ' wholly ' or an antonym, as ' individual- ly' to ' collectively. ' In seeking an ety- mology for it we must account for the preservation of p before s (cf. 61 d). If we derive it from the pcpl. of a verb equivalent to L. saepio, saeptus, the s in place of t would show the same analogical extension from dental stems as seen in L. labor, lapsus, and the analogy of pre- sent-stem forms might be sufficient to prevent the assimilation or loss of the p.
256
III. TEXT AND TRANSLATION
sansie . ditu . ocre . fisi . tote . iouine . ocrer . fisie . totar . iouinar . dupursus f (11) peturpursus . fato . fito . perne . postne . sepse sarsite . uou.se . auie . esone . futu . fons . pacer . pase . tua . ocre . fisi . tote . iiouine / (12) erer . nomne . erar . nomne . fisouie sansie . saluo seritu . ocrem . fisi . totam . iouinam .fisouie . sansie . saluo seritu / (13) ocrer . fisi.er . totar . iouinar nome . nerf . arsmo . uiro . pequo . castruo . frif . salua . seritu . futu . fons . pacer . pase I (14) tua . ocre . fisi . tote . iiouine . erer . nomne . erar . nomne . fisouie . sansie . tiom . esa . mefa . spefa . fisouina . ocriper . fisiu J (15) totaper . iiouina . erer nomneper . erar . nomneper . fisouie . sansie . tiom .
vius Sancius, grant to the Fisian Mount, to the state of Iguvium, to the men and beasts of the Fisian Mount, of the state of Iguvium, (success in) word and deed, before and behind, in private and in public, in vow, in augury, and in sacri- fice. Be favorable and propitious with thy peace to the Fisian Mount, to the state of Iguvium, to the name of the mount, to the name of the state. Fiso- vius Sancius, keep safe the Fisian Mount, the state of Iguvium. Fisovius Sancius, keep safe the name of the Fisian Mount, of the state of Iguvium, keep safe the magistrates, the priesthoods, the lives of men and of beasts, the fruits. Be fa- vorable and propitious with thy peace to the Fisian Mount, to the state of Iguvium, to the name of the mount, to the name of the state. Fisovius Sancius, thee with this Fisovian mefa spefa cake for the Fisian Mount, for the state of Iguvium, for the name of the mount, for the name of the state, Fisovius San-
15 nomneper (first instance): per written small above the line.
For an alternative explanation see Da- nielsson, Altitalische Studien, III (Han- nover, 1884), p. 156 {sepse < *saepicio- as uouse < *vovicio-). By either deriva- tion sepse, from the notion of fencing off which is inherent in the root, could mean 'individually' in contrast to sarsite 'col- lectively. ' — uouse auie esone : of these three forms uouse is the most difficult. The actual text is uou.se, but all editors from Biicheler to the present either read it as a single word or at least translate it on this basis. Some sort of connection with L. voveo, votum is apparent. We may derive it from *vofgyo- < *vofikyo-< *wog'"hikyo- after Kent, C.P., XV (1920), pp. 364-5, the development of the diph- thong -ou- < -of- after syncope being
parallel to that in aitu < *aget6d. auie is generally regarded as an e-extension of the stem seen in auif, L. avis, itself con- tained in the further extended forms aviekate, auiecla. auie may then be taken as an abstract noun meaning 'au- gury, taking of the auspices, ' and the three worJs will give satisfactory sense if we take them as loc. sg. (not dat.; it is the dupursus peturpursus who are ex- pected to benefit from the prayer), and understand them to refer to the three principal parts of the ceremony in the widest sense: the dedication of the vic- tim, the taking of the auspices, and the sacrifice proper. See Dev., p. 228, wiiere a diagram illustrates the division of the whole ceremony (persclo) into its three
VIb 11
COMMENTARY
257
component parts, with further subdivi- sion of the uouse into diglo and arsio and of the esono or sacrifice proper into the acts designated by the verbs ampentu, purdoiiifu, and siibra spahatu. — fulu forts ... 15 Horn subocau: cf. Via 30-4, which is closely similar. — 15-16 pes- clu I semii uesticatu atripursafu: semu is generally taken as cognate with L. semi-, rii.iA- and translated ' dimidia, media,' that is, 'when the prayer is half through, in the middle of the prayer.' Vetter translates ' unaquaque, ' from the separa- tive and individualizing force seen in se-, since there are separate prayers at the presentation of the uesiisia in 6, the pone in 9, and the m^efa spefa in 9 (cf. ibid., p. 249). But sehemu in 36 cannot be so understood unless we take it to refer to the separate parts of the prayer in 25-36, for the ceremony in honor of Tefer Jovius commencing at 24-5 does not call for separate prayers with the uestisia and with the persondro sorsom. It seems best therefore to follow the general view with regard to semu. — atripursatu: found also in the instructions for the offering of the persondro to Tefer Jovius in 24 ff., for the ceremony with black and white vessels in honor of Pres- tota Serfia in Vila, for the dog-sacrifice in I la, and for the sacrifice of a he-goat in lib. It is frequently preceded by uesticatu vestikatu. The principal Latin passages relating to the dancing of the tripudium are: Act. frat. arv., p. 26 (Hen- zen) ibi sacerdotes clusi succincti libellis acceptis carmen descindentes tripodaverunt in verba haec: enos Lases iuvate; Livy, I, 20, 4 Salios item daodecim Marti Gradivo legit [sc. Numa] ... ac per urbem ire ca- nentes carmina cum tripudiis sollemnique saltatu iussit. Cf. also E. Norden, Aus altromischen Priesterbiichern (Lund, 1939), pp. 190-1, 238-9. The prefix a- aha- a- ah- in the Umbrian forms evidently means that the adfertor in performing the dance moves away from the altar. There is no good reason to assume with Bii., p. 69, that atripursatu, hke L. tripudium,
refers to the feeding of sacred fowl for the purpose of taking auspices. — earn: the mefa spefa of 9, 14 above. — eras: the etymology and meaning are both uncertain. All instances of the word are ace. sg. with the possible exception of IV 14 (abl. pi. according to Vetter), and most depend on forms of the verb dirstu, although the governing verb is prusekatu in Ila 28, kuveitu in Ila 32, kuveitu teftu in Ila 40. A dependent genitive often precedes: hapinaru la 33, uestisiar sorsalir VIb 38, uestisiar VIb 16, Vila 38, proseseto VIb 16, 38. That the erus was a solid substance and not a liquid is apparent from Ila 32 iepru erus mani kuveitu, and Ila 27-8 katles tuva tefra / tefti erus prusekatu shows that at least in some cases the erus was a part of the victim itself. Devoto, p. 230, ar- gues in favor of the view that the erus was not offered to the gods but distri- buted to those persons who were present, pointing out that the expression erus ditu or its equivalent usually occurs near the end of the instructions for the sacrifice. It may also be significant that there is no clear instance in the Tables of dirstu used of something offered to a god; here the normal verb is purdouitu, while the instances of dirstu and its various forms are almost entirely limited to expressions with erus as the object, except for Ila 40 vinu pune tertu, VIb 10 in the prayer to Fisovius Sancius, and the contracts between the Atiedian Brothers and the two decuviae in Va, b. The commonly adopted etymology of erus (see Lexicon) is not in conflict with such a view, since the portions distributed could be regarded as ' divine ' or ' sacrificial ' by outside persons, and the rhotacism, in contrast to unrhotacized esono, might be a mark of popular as against ritual usage. It is strange that the priest should be kneel- ing while making the distribution, but Dev., p. 223, may have given the correct explanation in suggesting that the vessel was a large one and was placed on the ground. As a technical term of uncertain
258
III. TEXT AND TRANSLATION
subocau fisouie . frite . Horn . subocau .
pesclu I (16) semu . uesticatu .
atripursatu . ape . earn . purdinsust .
proseseto . ems . ditu . eno . scalseto .
uestisiar . ems . conegos / (17) dirstu .
eno . mefa . uestisia . sopa . purome .
efurfatu . subra . spahmu . eno . serse .
comoltu . comatir . persnihimu /
(18) capif . purdita . dupla . aitu . sacra .
dupla . aitu j
(19) pre . uerir . uehier . buf trif . calersu fetu uofione . grabouie . ocriper .
meaning eras is left untranslated in the present work. — 17 mefa uestisia: these two oblations are again offered to Fiso- vius Sancius in Vila 37-8, just after the sacrifice to Prestota Serfia, and to the deity whose name in written Purtupite in IV 14. — sopa: ace. sg., as may be seen from sopam in the nearly identical Vila 38. — purome: an o-stem has re- placed the consonant-stem, or the masc. ending has replaced the zero-ending of the third-declension neuter; in either case the result is the same. In a number of instances where the postposition -e is not used, the normal neut. ace. sg. pir pir appears. — efurfatu : only here and in Vila 38, where the context is identical, but furf ant iuria.Q in VIb 43 = lb 1 must be the simplex of the same verb. The sense is not certain, but Dev., pp. 232-3, has offered the most satisfactory solu- tion by taking the verb as a denominative from a noun *furfa < * forfa ,Toot*bherdh / bhordh, (cf. Olcel., OE bord ' tabula, axis '), and translating ' e lance auferto. ' For the use of lances and mensae in Roman ritual cf. Verg., G., II, 194 lancibus et pandis fumantia reddimus exia; Aen., XII, 213-5
cius, thee I invoke. In trust of Fisovius I invoke thee." In the middle of the prayer he shall pour a libation and dance the tripudium. When he has presented this (cake) he shall distribute the ems of the parts cut off. Then kneeUng, from a cup he shall distribute the ems of the libation. Then he shall remove from the platter the mefa cake and the libation and scatter them down into the fire. Then sitting he shall grind (the cakes) and shall pray with the ground (cakes). He shall move the two loaned bowls, he shall move the two conse- crated bowls.
Before the Veian Gate he shall sacri- fice three oxen with white foreheads to
sacratas / in flammam iugulant pecudes et viscera vivis / eripiunt cumulantque oneratis lancibus aras; Fest., 275 Li. mensae in aedibus sacris ararum vicem obtinent; Macr., Sat, III, 11, 5 f. in Pa- piniano enim iure evidenter relatum est arae vicem praestare posse mensam dica- tam. Ut in templo inquit lunonis Popu- loniae angusta mensa est. — subra spah- mu: similarly in Vila 39; cf. VIb 40-41 enom uaso porse pesondrisco habus / serse subra spahatu, and see on Va 20 subra spafu fust. The sense 'throw' is not too remote from ' spread ' to be acceptable in VIb and Vila, but the reason for the use of the passive in the present passage and Vila 39 against the active in VIb 41 is not clear. — serse: not found in the equivalent la 16-19, but the sitting posi- tion is prescribed in the sacrifice behind the Veian Gate (VIb 22 = la 25) and in the last portion of the supplementary offer- ings which follow (VIb 41 = la 33-4), including the ritual of the vessels, the grinding of the cereal-offerings, and the accompanying prayer. — comoltu: the instruction to grind occurs not only here but also in the description of the sacrifice
VIb 16
COMMENTARY
259
behind the Veian Gate (VIb 41 = la 34) and across the Sacred Way and at Rubi- nia (Vila 39, 44-5 = lb 37-8), and in Ila and IV. Generally no object is expressed, but Ila 41 struliglas fiklas sufafias kumaltu shows a series of partitive geni- tive objects, the first two of which are clearly types of cakes. The expression aiitakre(s) kumate(s) pe(r)sni(li)niu in lb 37-8 (but not in the corresponding Vila 44-5) and in Ila 42 seems to show a contrast between whole and ground cakes (?). Dev., p. 234, suggests that the grinding may symbolize the chewing of the cakes by the gods who receive them. All the instances of kumaltu kumultu comoltu occur shortly after the giving of the eras, and the grinding is regularly followed by a prayer, comatir may possib- ly be taken as abl. abs. 'when they have been ground, ' but it is not necessary to repeat in a new construction what has already been expressed by means of comoltu. We may therefore take comatir as an instrumental abl. indicating that with which or over which the priest prays. — 18 capif: this type of vessel is used in offering libations and the pesondro in 24, 37 - la 29, 32, and Ila 8, 34, and as a container for the drink puni in Ila 33, 41. It may have had a round bottom, since a persom, generally taken as a 'pit,' here as a 'mound,' had to be made for it (cf. la 29, 32, Ila 8, VIb 24, 37). For the distinction of two classes of these cups, purdita and sacra, cf. the cor- responding passage la 18-19. In lb 29, 32 = Vila 39-40, 45, on the other hand, only the capif sacra are mentioned. It is probably best to follow Devoto, pp. 190, 236 in making a distinction between cups which were temple property and those which were loaned or given from outside sources, similar to the distinction between victims sakre and perakne (see on Ila 5). — aitu: used with ace. pi. forms capif etc. in Vila 40 = lb 29 (sacri- fice at Rubinia) and Vila 45 =- lb 37 (sacrifice at Trans Sanctam), while aitu may be supplied in the ellipHcal sentence
la 18-19 corresponding to the present passage. The precise meaning is difficult to determine. The adverb enu in lb 37 shows that here at least the instruction is not merely to "use" kapi sakra but to perform some specific act with them at a certain point in the ceremony. The occurrence of capif ... aitu at or near the end of the particular ceremony in most cases might make it appear that the bowls were to be removed, but in this case we should probably expect to find the verb compounded with some prefix denoting separation. It is safer therefore simply to translate ' move, ' a sense close enough to the normal meanings of L. ago, Gk. ayoi. The exact nature of the motion pres- cribed is uncertain, but it may not be out of place to refer to Vila 25, where the vessels {*uesclo) are to be placed in a par- ticular position.
VIb 19 - 21 (= la 20 - 23). Sacrifice of three oxen to Vofionus Grabovius before the Veian Gate. As in the sacrifice to Jupiter Grabovius before the Trebulan Gate in Via 22, 56-7 the use of either wine or poni is prescribed, and cakes known by the names me fa spefa and fie la are to be added to the prosecta, but the instruc- tions for the two sacrifices are not given in the same order. — 19 calersu: equi- valent to L. calidus, known from Isid., Orig., XII, 1, 52 equi qui album frontem habent, calidi appellantur; Philoxenus ca- lidus Aevxo/ierconoQ. For the sacrifice of victims with a white spot on the forehead cf. further Hor., C, IV, 2, 57-60 (sc. vi tutus) f rente curvatos imitatus ignis ter- tium lunae refereniis ortum, qua no tarn duxit niveus videri cetera fulvus. For similar regulations on the marking of the victim in the Vedic horse-sacrifice see P.-E. Dumont, L' Asvamedha (Paris and Louvain, 1927), p. XII, §§ 67-72. — Uofione Grabouie: the third and most obscure of the Grabovian triad. Bii. and the majority of subsequent editors have connected the name with L. voveo, U. vufru (IE *wog'^h-), the god therefore being in some way associated with vows.
260
III. TEXT AND TRANSLATION
fisiu . totaper . iiouina . uatuo ferine . fetu . herie . uinu J (20) herie . poni . fetu . amio . fetu . tases . persnimu . proseseter . mefa . spefa . ficla . arsueitu . suront . naratu . pusi . pre . uerir / (21) treblanir / ^-^
(22) post . uerir . uehier . habina .
trif . fetu . tefrei . ioui . ocriper . fisiu .
totaper . iiouina . serse . fetu . pelsana .
fetu . aruio . feitu . poni j (23) fetu .
tasis . pesnimu . prosesetir . strusla .
ficla . arueitu . suront . naratu . puse . uerisco . treblanir . ape . habina .
Pisani, Revue des etudes indo-europeennes , I (1938), p. 11, proposed derivation from *Leudh-i6n-, from the same root found in OCS Ijudije, Germ. Leute (u v < /- as in 55a). Cf. also Pisani, Lingua dell' Ita- lia antica, p. 157. The name is therefore from the same root as L. Liber, and hlce Dionysus and Etr. Fufluns the god is one of vegetation and growth. The chief advantage of the new interpretation is that it produces a closer parallel between the Iguvine triad and the Roman triad Jupiter-Mars-Quirinus, for, even though etymological equivalence between Vo- fionus and Quirinus is of course impossib- le, we may now have in each case a third god associated especially with the farm- ing class as Jupiter and Mars were with the priestly and warrior classes under the social system sometimes assumed for Indo-European and early Italic peoples. Cf. Benveniste, Revue de I'Histoire des Religions, CXXIX (1945), pp. 6-9.
VIb 22 - 42 ( = la 24 - 34). Sacrifice of three ewe-lambs to Tefer Jovius behind the Veian Gate, followed by supplement- ary offerings to Tefer Jovius. The sacri- fice proper is contained in 22-23. poni
Vofionus Grabovius for the Fisian Mount, for the state of Iguvium. He shall place the ribs on a tray, sacrifice either with wine or with mead, offer grain, pray silently, and add to the parts cut off a mefa spefa cake and a ficla cake, and recite the same formulas as before the Trebulan Gate.
Behind the Veian Gate he shall sa- crifice three lambs to Tefer Jovius for the Fisian Mount, for the state of Igu- vium. He shall sacrifice sitting, he shall sacrifice (the victims) for burial, offer grain, sacrifice with mead, pray silently, add to the parts cut off a strusla cake and a ficla cake, and recite the same for- mulas as at the Trebulan Gate. When he has presented the lambs, the same
is used, and the cakes called strusla and ficla are added to the prosecta, as behind the Trebulan and Tesenacan Gates. — 22 habina: translated 'agnas' by nearly all editors; for phonological details see Lexicon. Vetter, p. 177, favors 'sues cas- iratas,' arguing that since the victims offered before the three gates are all of the same species (oxen), those behind the Veian Gate should, hke those behind the other two gates, be pigs; and for the ety- mology he suggests that the relation of habina to L. capus, cap(p)o is similar to that of the dialectal verb-root hab- to L. cap-. But neither the argument on grounds of ritual nor of phonology seems a sufficiently cogent reason to depart from the usual interpretation. — Tefrei Ioui: this god is mentioned only in the description of the sacrifice and supple- mentary offerings behind the Veian Gate, and is nowhere known outside of Iguvium. His name is doubtless connected with tefra, tefru-to, under which the Lexicon may be consulted for phonological details. If these forms are correctly interpreted as 'burnt-offerings, place of burnt-of- ferings,' then Tefer is evidently a deity
VIbl9
COMMENTARY
261
of the hearth, comparable in certain res- pects to the Roman Vesta. Cf. especially Rosenzweig, pp. 81-2, Dev., p. 238. — serse : the sitting position is prescribed also during the grinding of the cereal-of- ferings and the prayer which follows, in 41 = la 34, as also in VIb 17 during the grinding which forms a part of the spe- cial ceremonies in honor of Fisouius San- cius. — pelsana: this gerundive is used in the equivalent la 26, in the sacrifice of a ram to Jovius in Ila 6, in the sacrifice of a dog in Ila 43, and of a sheep in III 32, while the imperative pelsalu is used in connection with the pesondro in VIb 40. Goidanich, A.G.L, XXV (1933), pp. 70-1, connects pelsana, pelsatu with L. pulsare, offering literary and archaeological evi- dence for the practice of slaughtering the victim by a blow on the head. Ety- mologically this is easier than any other interpretation yet proposed, but it of- fers a serious difficulty in Ila 43 katel asaku pelsans ftitu and VIb 39-40 enom ... endendu, pelsatu, since in both of these passages the word occurs near the end of the series of instructions. Bott., p. 277, n. 4, in defending the interpreta- tion, suggested that the instruction in Ila 43 may be an afterthought omitted from its proper place near the beginning and added at the end. This defense, how- ever, Vkill not work for VIb 39-40. Vetter, p. 178, derives from *pel(p)sa- 'to mince meat, ' as a denominative to *pelpos, a stem-variant of pelmmr; an attractive but not altogether convincing interpreta- tion. It may be best to revert to the tra- ditional view according to which pelsatu is ' bury. ' The etymological support of- fered by L. sepelire is indeed very weak; in fact connection of the latter with Skt. saparyati 'pay honor to ' is fully as satis- factory as an analysis se-peZ-, which would be required if the Umbrian and Latin forms are to be connected. But better support for pelsatu as 'bury' may be found in Goth, ga-filhan, us-filhan 'bury (the dead). ' Two advantages in maintain- ing the sense 'bury' are that it is excel-
lently suited to the position of pelsans Ila 43, pelsatu VIb 40, at the end of the series of instructions, and that it does not depend upon the correctness of the mean- ing which we assign to persondro (pelsatu as ' strike ' or as ' mince, chop up ' would require that the persondro be an animal- offering, not a cake in the shape of an animal, unless the cake was struck to simulate slaughter). — 23 suront naratu puse uerisco Treblanir: uerisco Treblanir in place of the usual pre uerir occurs also in 44, 46, Vila 5, 42, 53 (cf. also uerisco Via 19, 20, 21). It is difficult to believe that the variation is anything more than capricious, and since we know from Via 59 that the same formulas were used be- hind the Trebulan Gate as before it, there was no possibility of error in fol- lowing out the instructions. — 23 ape habina purdinsus, etc.: the porrectio is followed by supplementary offerings to Tefer Jovius, the description of which extends through 42. The present cere- mony differs in certain respects from that which follows the sacrifice behind the Tesenacan Gate (5-18 = la 16-19). The principal distinctive feature is the per- sondro, a type of offering known also from Ila 8, b 13, IV 17, 19, 21. Because of the uncertain meaning of persondro (see on 24) as well as of uestisia and erus, and of staflare, used as an epithet of pesondro in 37, the sense of the whole passage is partly obscure. If the passage is compared with la 27-34, however, it becomes ap- parent that a persondro is offered to Tefer Jovius near the priest's right foot, then a second persondro to a deity whose name may be latinized as Stabilis Jovius, near the left foot, the same prayer being used in both instances. The prayer agrees in part with that used in connection with the offerings to Fisovius Sancius after the sacrifice behind the Tesenacan Gate; thus 25-27 = 6-8, 33-36 = 12-15. On tlie other hand the present passage con- tains nothing equivalent to VIb 10 Fi- souie Sansie, ditu ocre Fisi ... 11 uouse auie esone, while at the same time 25-36
262
III. TEXT AND TRANSLATION
purdinsus / (24) eront . poi . habina . purdinsust . destruco . persi . uestisia . et . pesondro . sorsom . fetu . capirse . perso . osatu . earn . mani f (25) nertru . tenitu . arnipo . uestisia . uesticos . capirso . subotu . isec , perstico . eras . ditu . esoc . persnimu . uestis . tiom / (26) subocau . suboco . tefro . ioui ocriper . fisiii . totaper . iiouina . erer. nomneper . erar . nomneper . fonsir . pacer . si . acre . fisi . tote / (27) iouine . erer . nomne . erar . nomne . arsie . tiom . subocau . suboco . tefro ioui arsier . frite . tiom . subocau suboco . tefro . ioui . tefre I (28) iouie . tiom . esu . sorsu persontru . tefrali . pihaclu . ocriper . fisiu . totaper . iiouina . erer . nomneper . erar . nomneper . tefre / (29) iouie orer .
ose . perse acre . fisie . pir . or to . est .
tote . iiouine . arsmor . dersecor . subator .
sent . pusi . neip . heritu . tefre iouie /
(30) perse . touer . pescler . uasetomest .
pesetomest . peretomest . frosetomest .
daetomest . touer . pescler . uirseto .
auirseto . uas . est j (31) tefre . iouie .
perse . mers . est esu . sorsu , persondru .
pihaclu . pihafi . tefre . iouie . pihatu .
acre . fisi . tola . iiouina . tefre . iouie .
pihatu I (32) ocrer . fisier . to tar .
iiouinar . nome . nerf . arsmo . uiro .
pequo . castruo . fri . pihatu . futu . fans .
one who has presented the lambs shall offer at his right foot a libation and a j)ig-persondro. He shall make a mound for the bowl; he shall hold it in his left hand until he has poured the libation; he shall set down the bowl. In like man- ner he shall distribute the erus at his foot. Thus shall he pray after making the libation: "Thee I invoke as the one invoked, Tefer Jovius, for the Fisian Mount, for the state of Iguvium, for the name of the mount, for the name of the state. Be thou favorable, be thou pro- pitious to the Fisian Mount, to the state of Iguvium, to the name of the mount, to the name of the state. In the conse- cration I invoke thee as the one invoked, Tefer Jovius. In trust of the consecra- tion I invoke thee as the one invoked, Tefer Jovius. Tefer Jovius, (I invoke) thee with this pig-persondro for Tefer as a propitiatory offering for the Fisian Mount, for the state of Iguvium, for the name of the mount, for the name of the state. Tefer Jovius, by the effect of this (bring it to pass), if on the Fisian Mount fire hath occurred or in the state of Iguvium the due rites have been omit- ted, that it be as not intended. Tefer Jovius, if in thy sacrifice there hath been any omission, any sin, any transgression, any damage, any delinquency, if in thy sacrifice there be any seen or unseen fault, Tefer Jovius, if it be right, with this j)ig-persondro as a propitiatory of- fering may purification be made. Tefer Jovius, purify the Fisian Mount, the state of Iguvium. Tefer Jovius, purify the name of the Fisian Mount, of the state of Iguvium, purify the magistrates, the priesthoods, the lives of men and of beasts, the fruits. Be favorable and pro-
VIb23
COMMENTARY
263
is fully equivalent, except for insignificant details, to the prayer used in the six sacri- fices at the three gates (cf. Via 23-34). — 24 eront poi habina purdinsust: for this use of a relative clause to designate the official cf. Via 1, 5, b 50, 53. — des- truco persi: cf. la 29 testruku peri kapire pefum feitu; in contrast the pesondro staflare is offered at the left foot in 37, 39, la 32. — pesondro : the nature of this offering cannot be determined with certainty. In addition to the pesondro sorsom here and the pesondro staflare in 37-41, with the equivalent passages in la 27, 30, a persttntru is prescribed in Ila 8, b 13, IV 17, 19, 21. Bii., Bk., and others take it to be a cake in the shape of a pig or sheep, offered in place of the animal itself, and the objection of Krause, R.-E., Suppl. V, p. 237, Blum., p. 62, that this practice did not exist in Italic ritual is hard to understand in the light of Serv. ad Aen., II, 116 sciendum in sacris simulata pro veris accipi, unde cum de animalibus quae difficile inveniuntur est sacrificandum, de pane vel cera fiunt et pro veris accipiuntur; Fest., 474 Li. Tauri verbenaeque in commentario sacrorumsigni- ficat ficta farinacea. A more serious ob- jection is that a translation such as 'ani- mal-cracker ' increases the difficulty of finding a satisfactory interpretation of venpersxmtra, which must almost cer- tainly be derived from it with the pre- fix (or combination of prefixes ?) ven- and which is used in Ila 30, IV 7 in evident association with meat. Ribezzo, II, p. 82, proposed a reasonably satisfactory ety- mology in deriving persondro from *per- sont-ro-: cf. OHG sunt, Olcel. synd 'Sun- de ' ; it was then a sin offering, but we still cannot be certain of its exact nature, and the safest procedure may be, as with erus, to leave it untranslated. The possi- bility that it was a cake shaped like an animal is not absolutely excluded, but this can hardly have been its etymolo- gical meaning. The prefix in venper- suntra then must be not negative but Intensive, or at least must strengthen
the notion of removal (of sin) which is already inherent in persondro by Ribezzo's interpretation; cf. the value of ex- in L. ex-piare and M. Niedermann, I.F., X (1899), pp. 247-56 on some similar uses of L. ve-. — perso: directed to be made for the placing of the cup in the present passage, with la 29, in VIb 37 = la 32, and in Ila 9. It is the exact etymolo- gical equivalent of jteSov and is usually translated ' fossam,' but there is good ground for beheving that it was rather a mound or a turf-altar of the type men- tioned in Verg., G., I, 337 In primis vene- rare deos atque annua magnae sacra refer Cereri, laetis operatus in herbis; Hor., C, I, 19, 13-15 Hie vivum mihi caespitem, hie verbenas, pueri, ponite turaque bimi cum patera meri; cf. Ill, 8, 3-4; Act. frat. arv., 218 (Henzen) ante aedem in cespite promag{ister) et flam(en) sacr(um) fecer(unt); cf. Goidanich, A.G.L, XXV (1933), pp. 69-70, who sees a correspon- dence between the Umbrian perso and the Roman caespes for offerings to ter- restrial gods, as the rusem and scrobiculum were used for infernal and the asa and ara for heavenly gods. — 25 uesticos: troublesome because if it is a fut. pf. 3 sg. from *westikd-us(t), we should rather expect a fut. pf. of the /-type, the usual type in verbs of the first conjugation, while if we take it as a pf. pcpl. from *westikdt(o)s, with Buck and others, we must assume an ellipsis or an erroneous omission of fust. The former of the two alternatives is, I beheve, shghtly prefer- able but the sense of the passage is not affected. — subotu : apparently the same verb as siibahtu Ila 42 and the pf. pcpl. subator Via 27, 36, 46, b 29. The 0 of subotu is unexpected and does not meet the conditions described in 7 c, but it may represent an exceptional case of o from secondary a. (same verb as hatu hahtu hatu), while the retention of a in subator may result from its occur- rence in a liturgical formula. Since the priest has just been holding the cup (eam mani nertru tenitu), subotu must be
264
III. TEXT AND TRANSLATION
pacer . pase . tua . ocre . fisi . tote /
(33) iiouine . erer . nomne . erar . nomne . tefre . iouie . saluo . seritu ocre . fisi . totam . iiouinam . tefre . iouie saluom . seritu . ocrer . fisier j
(34) totar . iouinar . nome . nerf . arsmo . uiro . pequo castruo fri . salua . seritu . futu . fons pacer . pase , tua ocre . fisi . tote . iiouine . erer / (35) nomne . erar . nomne . tefre . iiouie . tiom . esu . sorsu . persondru . tefrali . pihaclu . ocriper . fisiu . totaper . iiouina . erer . nomneper . erar / (36) nomneper . tefre . iouie . tiom . subocau . persclu . sehemu . atropusatu /
(37) pesondro . staflare . nertruco . persi . fetu . suront . capirse . perso . osatu . suror . persnimu . puse . sorsu . ape . pesondro purdinsus / (38) proseseto . erus . dirstu enom . uestisiar sorsalir . destruco persi persome . erus . dirstu . pue . sorso . purdinsus . enom / (39) uestisiam . slaflarem . nertruco . persi . sururont erus dirstu . enom . pesondro . sorsalem . persome . pue . persnis . fust . ife I (40) endendu pelsatu . enom . pesondro . staflare . persome . pue pesnis . fus . ife . endendu . pelsatu . enom . uaso . porse . pesondrisco . habus /
pitious with thy peace to the Fisian Mount, to the state of Iguvium, to the name of the mount, to the name of the state. Tefer Jovius, keep safe the Fisian Mount, the state of Iguvium. Tefer Jo- vius, keep safe the name of the Fisian Mount, of the state of Iguvium, keep safe the magistrates, the priesthoods, the lives of men and of beasts, the fruits. Be favorable and propitious with thy peace to the Fisian Mount, to the state of Iguvium, to the name of the mount, to the name of the state. Tefer Jovius, thee with this ipig-persondro for Tefer as a propitiatory offering for the Fisian Mount, for the state of Iguvium, for the name of the mount, for the name of the state, Tefer Jovius, thee I invoke." In the middle of the prayer he shall dance the tripudium.
He shall offer a persondro for Sta- bilis at his left foot: he shall make a mound for the bowl in the same manner, he shall pray in the same manner as in the case of the pig-per- sondro (for Tefer). When he has pre- sented the persondro, he shall distribute the erus of the parts cut off. Then he shall distribute the erus of the libation accompanying the pig-persondro, at his right foot, toward the mound, where he has offered the pig-persondro. Then (he shall offer) the libation for Stabilis at his left foot, and likewise distribute the erus. Then where he has prayed, there he shall place the pig-persondro (for Te- fer) on the mound and bury it. Then where he has prayed, there he shall place the persondro for Stabilis on the mound and bury it. Then, sitting, he shall throw over (his head) the vessels which he has kept with the persondra. He shall sit
VIb 25
COMMENTARY
265
the instruction to set it down (on the perso). — persico: the actual reading of tlie text is perstico; Bu. keeps this read- ing, while translating 'ad pedem,' and takes the formation to be related to persi as L. rusticus is to rus. Vetter takes it as an extension in -ic- from the pf. pcpl. of peperscust, understanding it of the parts previously placed at the back. Many editors (Bk., Kent, T.C.I., p. 43, Dev.), however, emend to persico, and despite the usual advantage of keeping the original text, it is probably necessary to follow them, in view of the close analo- gy with 38 destriico persi persome eras dirstu. — 25 -36: see the general com- ment above on 23 - 42 and for details see on Via 23 - 34, with which, mutatis mutandis, the present passage is practical- ly identical. — 37 Staflare: there is general agreement among editors in equating with L. *stabularem, but the sense is uncertain : according to Bii. and others, ' sheep-, belonging to sheep, ' as the animal kept in the stabulum in contrast to swine; according to Dev. and Vet., ' concretum, -am. ' If we admit two deities, Tefer Jovius and Stabilis Jovius, here as well as in the corresponding la 27-32, then Slaflarem has the same relation to Stafli in la 30-31 (on which see note) as Tefrali VIb 28, 35, to Tefer. There is admittedly a difficulty in the contrast between 24 destruco persi uestisia et pe- sondro sorsom fetu and 37 pesondro staflare nertruco persi fetu; 38 uestisiar sorsalir destruco persi persome erus dirstu and 39 uestisiam staflarem nertruco persi sururont erus dirstu; 39-40 pesondro sorsalem ... endendu pelsaiu and 40 pesondro staf- lare ... endendu pelsatu. It is not unnatu- ral that earlier editors saw in these con- trasting expressions a reference to two kinds of pesondro and uestisia, one si- mulating or made from a pig, the other simulating or made from a sheep (see on la 30-31 stafli and on VIb 24 for the notion of pesondro as a cake in the form of an animal). But it is apparent from la 27-8 surum pesuntru / fetu ... Tefri,
30-1 sufum pesxintrum feitu Stafli luv..., that the real contrast is between two persondro-, both sorso-, but offered in the one case to Tefer Jovius, in the other (despite the uncertainty of the text, on which see note) to Stabilis Jo- vius. This view receives some confirma- tion from VIb 28, 35 tiom esu sorsu per- sontru tefrali pihaclu ... The old interpre- tation of stafli, staflare as ' ovillum' (Buck in both editions, Kent, C.P., XV, p. 356) runs into difficulty because the whole expression sufum pestuitrum must then mean ' figmentum' with the specific sense of sufum 'pig-' lost through generalization. We may perhaps assume that the persondro- in both cases was suilli generis and that in certain instan- ces in describing the prior offering the nature of the persondro, rather than its destination for Tefer Jovius, is specified; in this way it is made to appear that the contrast is between sorso and staflare.
— 38 uestisiar sorsalir ... erus: the words uestisiar sorsalir refer to the libation which was mentioned in 24 in connection with persondro sorsom. and which may actually have been a libation of lard ; the erus, if the interpretation adopted for erus above (on 16) was correct, is probably a portion of the libation set aside for distribution. — pue sorso pur- dinsus: if this is equivalent to 'near the right foot,' as appears from 24 destruco persi uestisia et pesondro sorsom fetu, it is slightly tautological in view of destruco persi in 38 just above. — 40 uaso : there is no indication that these are identical with the bowls mentioned in 24, 25, 37. In 24-5 a mound was built for the bowl, it was held in the left hand until the libation was poured, and was then set down. In 37, which begins the instructions for the second offering of the persondro, the making of the mound for the bowl is again prescribed, but the subsequent instructions regarding it are not carried out as in the former instance.
— 41 subra spahatu: nowhere else used with uaso or any similar word as its
266
III. TEXT AND TRANSLATION
(41) serse . subra . spahatuanderuomu . sersitu . arnipo . comatir . pesnis . fust serse . pisher . comoltu . serse . comatir . persnimu f (42) purdito . fust j
(43) uocucom . iouiu . ponne . oui . furfant . uitlu . toru . trif . fetu .
marte . horse fetu . popluper . totar .
iiouinar . totaper . iiouina . uatuo .
ferine / (44) fetu poni . fetu . aruio ■
fetu . tases . persnimu . prosesetir .
fasio . ficla . arsueitu . suront . naratu . puse . uerisco . treblanir f
(45) uocucom . coredier uitlu . toru . trif . fetu . honde . serfi . fetu . popluper . totar . iiouinar . totaper . iiouinar . uatue . ferine . fetu aruio / (46) fetu heri . uinu . heri . poni . fetu . tases . persnimu . prosesetir . tesedi . ficla arsueitu . suront . naratu . puse . uerisco . treblanir eno . ocar J (47) pihos . fust . suepo . esome . esono . ander . uacose . uasetome . fust . auif . aseriatu uerofe . treblano couertu . reste . esono . feitu I
45 iiouinar (second instance): for iiouina.
object. Dev. and Pis. assume that the vessels were thrown and broken; the Acta jr. arv. mention a similar throwing away of vessels after ceremonial use (Henzen, pp. 26, 30). — andenwmu: the bronze shows spahatuanderuomu with- out word-division; it is therefore uncer- tain whether we should read anderuomu
during the interval until he has prayed with the ground (cakes). Anyone at all, sitting, shall grind (the cakes). Sitting, (the adfertor) shall pray with the ground (cakes). The sacrifice will have been completed.
At the Grove of Jupiter, while they are placing on a platter (the remains of) the sheep, he shall sacrifice three bull- calves; he shall sacrifice to Mars Hodius for the people of the state of Iguvium, for the state of Iguvium. He shall place the ribs on a tray, sacrifice with mead, offer grain, pray silently, add spelt-cakes and a ficla cake to the parts cut off, and recite the same formulas as at the Trebulan Gate.
At the Grove of Coredius he shall sa- crifice three bull-calves; he shall sacri- fice to Hondus Serfius for the people of the state of Iguvium, for the state of Iguvium. He shall place the ribs on a tray, offer grain, sacrifice either with wine or with mead, pray silently, add to the parts cut off a twisted cake and a ficla cake, and recite the same formulas as at the Trebulan Gate. Then the Mount will have been purified. If there is any interruption of these rites, they shall be invalid; he must take an obser- vation of the birds, return to the Tre- bulan Gate, and perform the rite anew.
uatue: for uatuo.
or ander uomu, and since the expression occurs nowhere else, there is no sure so- lution. The interpretation here adopted is ' in the interval, meanwhile, ' anderuomu being taken as a noun compounded from ander + uomo- (root *lem/lom-, cf. Russ. lam ' debris ' and possibly also Osc. la- matir ' caedatur'). The change I- > u-
VIb 41
COMMENTARY
267
operates within the compound as in the initial position in the simplex, and the semantic development 'break > inter- rupt(ion) > interval ' may be illustrated by comparison with Germ. Unterbrechung, Eng. without a break, etc. For further details see A.J.P., LXXI (1950), pp. 67-70. — 42 purdito fust: here, as also in Vila 45, it is evident that this expres- sion must be used not of the poirectio but of the conclusion of the whole cere- mony
VIb 43-44 ( = lb 1 - 4). Sacrifice of three bull-calves to Mars Hodius at the grove of Jupiter. The details of the cere- mony resemble those in the sacrifice to Mars Grabovius before the Tesenacan Gate in 1-2, although they are not given in precisely the same order. — ■ 43 uocucom : phonologically uocu vuku may be equated either with Gk. folxo- (in sense 'aede'; so Bu., Bk.) or v/ith L. luco (so Br., Blum., and subsequent editors). The interpreta- tion ' grove ' is preferable, since ' temple ' is approximately the sense of fesnere lib 11, fesnafe lib 16. — furfant: only here, with furfa6 in the corresponding lb 1. If the etymology and interpretation adopted for efurfatu in VIb 17 above are correct, furfant must refer to the placing of the pieces of the slaughtered sheep on a platter, presumably the lambs sacrificed in 22 ff. The platter may have been used for the distributio, at the final conclusion of which the next sacrifice could begin, assuming that the supplementary of- ferings prescribed in 24-42 had also been completed in the meantime. — Horse: only here, with Hufie in lb 2. Editors in general latinize 'Hodio,' but nothing can be affirmed as to the etymology or precise meaning of this epithet of Mars. It cannot be connected with "Odiog ex- cept on the assumption that the Umbrian word was a borrowing, since the spiritus asper in 6d6g, with OCS chodu, presup- poses an S-, which could not yield h- in Italic.
VIb 45 - 46 (= lb 4 - 7). Sacrifice of three bull-calves to Hondus Serfius at
the grove of Coredius. The instructions are the same as for the preceding sacri- fice except that l)aruio fetu is given second instead of third in order; 2) a choice of wine or poni instead of poni alone is authorized; 3) the cake called tesedi takes the place of fasio. — 45 Coredier: else- where only in the equivalent lb 4, in the form Kureties. Dev., p. 255, compares the modern place-name Goregge, the name of the decuvia Kureiate lib 3, and the name of the gens Coretia (known from Latin inscriptions from Sentinum; cf. Schulze, Lat. Eigenn., p. 355), and also suggests that Coredier is not a divine name but a gentile or ethnic. Nothing, however, can be affirmed with certainty. — Honde Serfi: it is clear that Hondus was an infernal deity, both from the ety- mology of the name and from the fact that Il9 20 calls for the sacrifice of a dog to Hondus Jovius. In the present pas- sage, in place of the epithet loui, Honde is accompanied by Serfi, which is also applied to Prestota and Tursa of Serfus Martins (see on 57 below). — 46 tesedi: the corresponding form in lb 6 is tenzi- tim, and in view of the context there is little doubt that the word designated some sort of cake, perhaps a pulled or twisted cake if the root is the same as in L, tendo, tensio (cf. Bii., p. 81, who com- pares the Greek cakes called arQSTcrovg and aneiQag). It is not possible, however, to arrive ^t a sure etymology and inter- pretation of tesedi tenzitim, and the intervocalic d unchanged to rs f may be a mark of foreign or dialectal origin. VIb 46 - 47 (= lb 7 - 9). Instruction for repeating the whole purification in case there has been any omission. lb 7 - 9 is practically identical, word for word, except for the difference of the verb forms in vagetumise, uasetome fust, and the general sense of the whole passage is clear, but there are certain difficulties of word-division and grammar. The actual reading of the bronze in the most critical passage is : lb 8 : svepu : esumek : esunu : anter : vakazevagetumisea-
268
III. TEXT AND TRANSLATION
(48) pone . poplo . afero . heries . auif . aseriato . etu . sururo . stiplatu . pusi . ocrer . pihaner . sururont .
When he wishes to perform a lustra- tion of the people, he shall go and ob- serve the birds, make the same demands as for the purification of the Mount, make
vif : aseriatu, VIb 47 .suepo . esome . esono . ander . uacose . uasetome . fust . auif . aseriatu. "With the exception of PL, II, p. 670, who was hesitant, most editors since Bii. have taken anter ander in composition with vakaze uacose de- spite tlie interpunct in both instances. ander would otherwise be difficult to construe, and anter : menzaru in I la 16 shows an instance of separation in a word Vtfhich must be understood as one. With most modern editors I follow Brugmann, Ber. kon. sdchs. Ges. Wiss., 1890, pp. 217-8, in deriving uacose va- kaze from *vakdt(i)s plus the subjunctive equivalent to L. sit. The failure to show word-division may be explained partly by the identity of final and following initial, partly by the enchtic character of the verb 'be'; cf. fonsir, mersei mersi, although in most similar word-groups the division is shown. Of course there is no need for a verb until fust, which may mark the end of the protasis, but Brug- mann's view is adopted here as providing the only satisfactory analysis of uacose vakaze; in this case the protasis ends with (s)e, uasetome fust is the apodosis, and what follows is virtually a new sentence. — uasetom vagetum may be a supine or a pf . pcpl. ; if it is a supine, the final -e, despite the word-division, must be tal^en as part of *efust, fut. pf. of etu (cf. ampr- efuus lb 20, ambr-efurent VIb 56), while ise must be from *iser, fut. pf. pass, of etu (Bii.) or else pf. subj. or fut. pf. act. of etu (PL, II, p. 389, cf. amb-issit, Plant., AmpA., 71). The construction then, at least by Biicheler's interpretation, would resemble such Latin combinations as servatum itur, factum itur, and tlie sense would be excellent: 'it shall go for naught, be invalidated. ' Yet there are several objections: 1) uasetom as pf. pass.
pcpl. is already known from Via 27, 37, 47, b 30; 2) there is no real support for ise as a perfect-system form of the verb 'go'; 3) other things being equal it is better to keep the word-division -e fust. This last argument might also be applied against Devoto's proposal (p. 259) to take ise, efust as forms of the verb ' be ' with prefixed en-, though naturally one should not attach undue weight to the word division in view of the treat- ment given by nearly all recent editors to ander . uacose just before. Dev. makes excellent sense by making uasetom the subject and translating 'vitium inerit,' but from *inse we should expect *ize or *inze (cf. Kent, Lang., XIV [1938], p. 216, Buck, § 110.1). It is probably best therefore to regard -e -i as a postposition attached to uasetom, a pcpl. used substan- tively with the following verb uncom- pounded. The literal sense then would be 'it shall be for nulUfied'; that is, 'as a thing null and void. ' This use of a prepositional phrase with e(n) in place of a predicate nominative is roughly com- parable to Latin constructions like cladi fuit, in which the dative tends to replace the nominative. — auif aseriatu, etc.: for the taking of auspices anew cf. Verg., Aen., II, 178, in Sinon's tale of the de- parture of the Greek fleet for Greece.
VIb 48 - Vila 54 (= lb 10 - 45). Lus- tration of the people. The ceremony pro- per must be preceded by the taking of the auspices, as in the purification of tlie Fisian Mount. The adfertor, accompanied by two prinuati, then proceeds to Ace- donia, where he pronounces banish- ment against any enemy aliens who may chance to be present. After he has made a circuit around the assembled people, he prays to Serfus Martins and Prestota of Serfus Martins and Tursa of Serfus
VIb 46
COMMENTARY
269
Martius with imprecations against the enemies of Iguvium and entreaties for favor toward the state and people of Iguvium. The circuit is performed and the prayers said three times. There fol- lows a sacrifice to Serfus Martius at the Fontuli, then a sacrifice to Prestota Serfia of Serfus Martius accompanied by fur- ther curses against the enemies of Igu- vium and prayers for protection of the city and its people. A libation and cake {mefa spefa) are then offered to Fisovius Sancius, after which the adfertor crosses the Sacred Way and sacrifices three calves to Tursa Serfia of Serfus Martius, After the lustration has been performed for the third time, Tursa Jovia is invoked with curses against enemy states and prayers for favor toward the state and people of Iguvium. Table Vila ends with instructions for the ceremonial cha- sing of heifers and the sacrifice to Tursa Jovia of the first three heifers which are caught.
Table lb 10 - 45 is in essential agree- ment with the longer version in VIb and Vila, but it does not contain the text of the prayers and imprecations which are prescribed. The later version on the other hand does not contain a ratification of the foregoing instructions similar to that in lb 45. Both versions present certain problems in regard to the sequence of the various parts of the lustration. The in- structions for the three sacrifices at Fon- tuli, Rubinia, and Trans Sanctam are presented commencing at Vila 3 = lb 24, after the circuit of the assembled people has been performed for the third time and the two prinuati have departed by the way they came. There are cer- tain indications that the sacrifices them- selves were partly in progress at the same time; lb 33-4, for example, shows that the distribution of the eras at Trans Sanc- tam is to follow immediately after the porrectio at Fontuli. In both versions the instructions which follow after the completion of the third sacrifice are intro- duced by the words 'after he has made
the circuit around the people for the third time. ' The instructions in Vila 46 ff. include the prayer to Tursa Jovia, then at 51 the pursuit of the heifers, while at lb 40 the clause pustertiu pane puplu atefafust is followed immediately by the pursuit of the heifers, the prayer to Tur- sa Jovia being omitted, but the omis- sion of the text of prayers is regular in the older tables, and there is no doubt that the prayer and the pursuit and sacri- fice of the heifers to Tursa Jovia (Vila 53 = lb 43) are parts of a single ceremony. The difficulty lies in the fact that the ceremony in honor of Tursa Jovia is de- scribed after the three sacrifices at Fon- tuli, etc., and yet in language which implies that it was intended to follow immediately after the third circuit of the people. Moreover two prinuati are directed to participate in the chase (Vila 52 = lb 41), although the prinuati are supposed already to have departed after the circuit was completed for the third time (VIb 65 = Vila 1 = lb 22-3). Of course there is no certainty that the prinuati were the same in the two in- stances, but even if we assume that two new prinauti appeared for the chase of the heifers, we are no closer to a solution of the question whether the chase, with the prayer to Tursa Jovia immediately preceding, followed directly after the third circuit, as the words poster tio pane pop to andirsafust seem to suggest, or whether the three circuits, the three sacrifices, and the ceremony of the heifers actually took place in the order in which they are presented. The former view is supported by Br., pp. 209-10, Dev., p. 297, the latter by Hu., p. 291, Bu., p. 113. The latter is, I believe, preferable. The pursuit of the heifers between the third circuit and the first sacrifice would have introduced an element of disorder into the lustration, and, if it had really been intended, would in all probability have been prescribed at the proper place. It is not too much to believe, as Hu. and Bii. suggest, that andirsafust designates the
270
III. TEXT AND TRANSLATION
combifiatu . eriront . tuderus . auif / (49) seritu . ape . angla . combifiansiust . perca . arsmatiam . anouihimu . cringatro hatu . destrame scapla . anouihimu . pir . endendu . pone / (50) esonome . ferar . pufe . pir . entelust . ere . fertu . poe perca . arsmatiam . habiest . erihont.aso . desire . onse. fertu . erucom . prinuatur . dur J (51) etuto . perca .
lustration in the broadest sense and in- cludes the sacrifice of the victims at the three stations round about the assembled people, which was necessary if the puri- fication was to be fully effective.
The lustration at Iguvium in its most important aspects resembles Roman ce- remonies of the same liind, although Dev., p. 261, observes certain differences. The chief literary sources for the Roman lus- tration are: Cato, R.R. 141 Agrum lus- trare sic oporlet. Impera suoviiaurilia circumagi: ''Cum divis volentibus quod- que bene eveniat, mando iibi, Mani, uti illace suovitaurilia fundum agrum ter- ramque meam quota ex parte sive circumagi sive circumferenda censeas, uti cures lus- irare." lanum lovemque vino praef amino, sic dicito: ''Mars pater, te precor quaeso- que uti sies volens propitius mihi domo familiaeque nostrae, quoius re ergo ter- rain fundumque meum suovitaurilia cir- cumagi iussi, uti tu morbos visos invisos- que, viduertatem vaslitudinemque, calami- tates intemperiasque prohibessis defendas averruncesque ; utique tu fruges, frumenta, vineta virgultaque grandire beneque eve- nire siris, pastores pecuaque salva servas- sis duisque bonam salutem valetudinem- que mihi domo familiaeque nostrae; harum- ce rerum ergo, fundi terrae agrique mei lustrandi lustrique faciendi ergo, sicuti dixi, made hisce suovitaurilibus lactenti- bus immolandis esto; Mars pater, eiusdem rei ergo made hisce suouiiaurilibus lac-
announcement in the same manner, and observe the birds within the same lim- its. When he has announced the div- ine messengers he shall hold a ritual wand, take a stole, and place it over his right shoulder. He shall place fire in (the fire-carrier). When that in which he has placed the fire is brought to the sacrifice, he who holds the ritual wand shall carry it; the same shall carry it lighted on his right shoulder. With him shall go two prinuati; they shall hold
tentibus esto," etc.; Dion. Hal., Ant. Rom. IV, 22, 1-2 tots d' ovv 6 Tv?.Aioq, enei- 6rj disrate ro jieqI raQ ri/nrjaeig, xskevaaq rove, TioXirac, anavrac, ovvekQslv eig to /ne- yiOTOv rcov jiqo Tfjg Jiokecog nsdicov exovrag rd oJiXa, xai rd^ag tovq 6' iTiJielg Hard reXr} xaljovg TTeCovQ ivcpd?.ayyi xal tovq earakfie- vovq rov xpiXixov oJiKiafxov iv rolq Idiotg exda- rovg XoxoiQ, xadaQ/uov avx&v E7ioir]aaro ravQco xai xqioj xai XQayia. Td 8' leQela ravta rgig JTEQiaxOfjvai jisql to axQaroTtedov xeXevaag edvae rep xarexovri to Jiediov "A- Q£i. Tovrov rov xaOaQ/udv ecog roJv ^^qt' i/A.e xQovcov 'Pa)/j,aloi xadaiQOvrai, fisrd rrjv avvreAeiav rwv ri/.iijaeojv vno rcov exovrcov rrjv leQcordrr]v dgxrjv Aovargov ovofjACovreg. Cf. also the description of the Ambar- valia in Tib., II, 1 ; Wissowa, Religion und Kultus der Romer^, pp. 142-3, 390-1. VIb 48 - 52 (= lb 10 - 15). Taking of the auspices. This ceremony is directed to be performed in the same manner as in the purification of the Fisian Mount, with the same formulas and within the same city-limits. The birds to be observed are not named, although in lb 10-11 we find pernaiaf pustnaiaf, the same terms of direction used in la 2. The most cur- ious feature of the present passage, however, is that, after the instructions for the ceremonial carrying of the fire, we find in 51-2 (- lb 13-4) that the tak- ing of the auspices must be repeated, the bird parfa alone being called for in this second observation. A.-K. and Br. in
VIb 48
COMMENTARY
271
explanation suggested that the repetition was necessary because the lustration was to take place outside the pomerium, while the first taking of the auspices had effect only within. Dev., p. 265, however, points out that the second observation also takes place before the adfertor has made his descent from the citadel by the uia auiecla, and as an alternative explanation suggests that the first obser- vation is intended not for the lustration but only for the preUminaries, perhaps for the fire-ceremony. — 48 stiplatu: cf. Via 2. — combifiatu: the logical subject must be the augur; cf. Via 17. The subject of anouihimu, however, is the adfertor, since it is the bearer of the perca arsmatia who pronounces the im- precations and prayers commencing at 53. — 49 ape angla combifiansiust: lb 11 instead has pune kuvurtus and omits the instruction regarding the perca ars- matiojn; for perca see on Via 19. — cringatro: similarly krenkatrum hatu in lb 11, while in lib 27, 29 the wearing of a krikatru over the right shoulder is prescribed during the slaughter (? = anpenes) and presentation of the calf. Nearly all editors translate ' cine turn' or 'cingulum,' and the notion of a girdle or encircling object receives support from the cognates OE hring, OCS krogii ' circle. ' ' Stole, ' however, seems an ap- propriate English translation for a litur- gical garment worn over the shoulder. — scapla: hardly a mere substitute in the later language for uze in lib 27, 29, since onse is used in VIb 50 just below, scapla is rather the shoulder-blade, over which the cringatro was thrown back. — pir endendu: in lb 12 aMim-em, the name of a kind of fire-carrier, depends on en- tentu. Not only is the name of the vessel omitted here, but immediately below, where it should appear as subject of ferar, it is replaced by the circumlocu- tion pufe pir entelust. The deliberate avoidance of the name of the fire-carrier is an almost sure instance of tabu. For tabu among the Scots of the western
isles against the names of certain com- mon objects see J. G. Frazer. The Gol- den Bough^, Part II = Taboo and the Perils of the Soul (London, 1911), pp. 292-3. — 50 poe perca arsmatiam habiest: another probable instance of tabu, re- peated in 53, 63, Vila 46, 51. The similar- ity of this last passage, describing the chasing of the heifers, to lb 41-2 iveka perakre tusetu / super kumne arfer- tur, prinuvatu tuf tusetutu confirms the identity of poe perca arsmatiam ha- biest with the adfertor. — aso: etymol- ogy and sense are both uncertain, but the unrhotacized s naturally suggests connection with L. assum 'roast'; what is applied in the one case to the meat is applied in the other to the fire itself; in other words there is no serious objection to following Buck's interpretation '... carry it lighted on the right shoulder. ' aso taken as a supine does not give satis- factory sense, for pir endendu in 49 shows that the fire has already been kindled, and it is not specified that it is to be car- ried for the purpose of lighting some- thing else, aso then must be taken with Buck as a pf. pcpL, probably to the verb cognate with L. ardeo. Under certain conditions we should expect *arfo (< *arsso- < *ard-to-, with development of consonants as in trahuorfi), but the dif- ficulty can be avoided if we derive from *ar- or from *as- + -so- appearing analog- ically in place of -to- as in L. pressus, pulsus, etc. Goidanich, p. 56, followed by Bott., takes aso as a noun signifying some sort of support for the ahti- in which the fire is placed in lb 12, but this is a less easy interpretation, especial- ly in view of lib 12, where taile is the name of the flat object on which the fire is carried. — prinuatur: etymology ob- scure, but the word clearly designates certain officials. They accompany the adfertor in the present augural ceremony and in the exterminatio which immediately follows, and in the actual lustration (55, 56, lb 19); they join him in the invocation to Serfus Martins, Prestota, and Tursa
272
III . TEXT AND TRANSLATION
ponisiater . habituto . ennom . stiplatu parfa desua . seso . tote . iiouine . sururont . combifiatu . uapefe . auieclu . neip / (52) amboltu prepa . desua . combifiansi . ape . desua combifiansiust . uia . auiecla . esonome . etuto . com . peracris . sacris ape . acesoniame / (53) hebetafe . benust . enom . termnuco . stahituto
the wands of a puniceatus. Then he shall demand a parra in the west for himself and for the state of Iguvium. In the same manner (the augur) shall make his announcement in the direction of the augural seats. He shall not walk about until he has announced (a parra) in the west, they shall proceed by the Augural Way to the sacrifice with the unblem- ished victims. When he has come to Acedonia, to the exits, then they shall stand at the boundary. He who holds
{persnimumo tasetur 57), and after the circuit has been completed for the third time, they depart by the same way they came (65 = Vila 1 = lb 23). They appear again as participants in the in- vocation to Tursa Jovia (Vila 46), and in the ritual pursuit of the heifers (Vila 52 = lb 41). The number of prinuati is fixed as two in the present passage and in Vila 46, while VIb 51 (= lb 15) shows that they are to be distinguished by special insignia, perca ponisiater. Not only from this fact, but from their close association with the adfertor, shown by the 3 pi. verb forms etuto 52, stahituto 53, ambretuto, ambrefurent 56, etc., we may infer that they were persons of some dignity, perhaps civil magistrates of a certain grade. — 51 ponisiater: gen. sg. and so used to describe the bearer of the perca, while arsmatia is an adj. de- scribing the perca itself. The word is found only here, with punigate in the corresponding lb 15. Nearly all editors take it as equivalent to L. *puniceatus, the relation of which to puniceus, -a, -um would be the same as that of togatus to toga. The prinuati on the present oc- casion then are to carry wands of the type used by the officials who are named for their purple-ornamented robes. The Roman toga praetexta naturally comes to mind, but just what official is here desig- nated by the wearing of the purple stripe cannot be determined with certainty.
— parfa desua: for a discussion of the possible meaning of these terms see on Via 1. — seso: the formation is not quite clear, but the word is clearly dat. sg. of the reflexive pronoun (see 107, n. 1). Table lb, where the instructions are in the second person, has tefe (13), while Via 5, with the actual words to be used by the adfertor, has mehe. — uapefe auieclu: we have learned from Via 2, 17 that the augur (poet angla aseriato eest) makes his observation from the tremnu and reports (combifiatu) to the adfertor, who apparently occupies the stone seat. Whereas Via 16 gave the source of the announcement, tremnu, the present pas- sage gives its destination, uapefe auieclu.
— neip amboltu, etc.: in Via 5-7 the prohibition was against any noise or inter- ruption before the return of the augur from the place of observation; here the augur is forbidden to move from the spot until he has announced the proper omen. That amboltu refers to the augur (so Dev.) rather than to the adfertor (so Bii.) is probable because no change of subject is indicated between combifiatu and com- bifiansi, and because if the adfertor were meant, we might expect the two prinuati to be included in the prohibition, as they are in the command etuto just below. Vetter, p. 261, denies close connection with L. ambulare on the ground that denominatives never transfer to the third conjugation, and he translates 'sich
VIb 50
COMMENTARY
273
aufmachen, ' taking am- from an- ' up ' rather than from amf-. But L. ambulare, although of rather uncertain etymology, is probably not a denominative, and its relation to amboltu may be similar to that in dicare : dicere, educare : ducere, etc. Moreover the root *bol or *pol which must be assumed as a basis for the new interpretation is without support. It seems best therefore to follow the tradi- tional view of amboltu. — 52 uia auiecla : this was the counterpart of the Via Sacra in Rome, if we may judge from Varro, L.L., V, 47 Carinae pote a caeri- <m>onia, quod hinc oritur caput Sacrae Viae ab Sfreniae sacello quae pertinet in arce<m>, qua sacra quotquot mensibus feruntur in arcem et per quam augures ex arce profecti so lent inaugurare. However, there was in Iguvium a (Uia) Sahata (= Sancta) also, on one side of which three heifers are offered to Tursa Serfia in Vila 41 (= lb 31). — peracris sacris: see on Via 25.
VIb 52-55 (= lb 15 - 18). Banish- ment of aliens. The adfertor, having arrived at Acedonia by the Via Aviecla after taking the auspices, stands at the boundary and directs all persons of the Tadinate nation and tribe and of the Tuscan, Narcan, and lapudic "name" to depart from among the assembled people of Iguvium or be taken away and punished according to the law. The banishment is proclaimed three times. The versions in VI and in I fully agree in distinguishing two categories of aliens, designated respectively by the terms to- tar Tarsinater trifor Tarsinater and Tus- cer Naharcer labuscer nomner, but there are certain differences of phraseology; thus in the present version the names of the banished nations and the instructions for apprehending loiterers are all included in the formula to be used by the adfertor, while in I the adfertor is merely instructed to banish the members of the alien nations, after which the threat svepis habe ... feitu uru pefe ixiers est is given in the direct form. For the banishment of
various classes of persons from religious ceremonies in Rome cf. Fest., 198 Li. exesto, extra esto. Sic enim lictor in qui- busdam sacris clamitabat: ' hostis, vinc- tus, mulier, virgo exesto '; Verg., Aen., VI, 258-9 'procul o procul este profani,' conclamat vafes, ' to toque absistite luco.' — 52 Acesoniame:yiIa52has Acersoniem, lb 16 Akef uniamen ; with these forms may be compared O. Akudunniad on a coin (Co. 158) from the modern town Lacedonia, situated east of Benevento and south of Foggia, the name of which may be from Macedonia through reanaly- sis of phrases containing a preposition and article. The frequent association of U. Acersoniame, O. Akudunniad with the name Aquilonia may be an error re- sulting from confusion between the two adjacent modern towns Lacedonia and Aquilonia (cf. Co., pp. 171-2). The Acersonia of the Tables is evidently a suburb located on one side of Iguvium and serving as a suitable place for as- semblies; Dev., p. 272, places it across the Camignano in the direction of Monte Ingino ; the name Crisonia by which this region was designated in the Middle Ages may be derived from Acedonia (cf. Studi etruschi, IV [1930], pp. 226-8). — 53 hebetafe: there is no parallel form in lb 16, but the word is apparently equivalent to ebetrafe in Via 12, with h possibly serv- ing as a mark of vowel-length in crasis of postposition -e 4- ebet- (although di- vided between lines), and with erroneous omission of the r. ebetrafe occurred as the first in a series of landmarks to be used in defining the augural templum in Via 12-14. If this 'exit' lay on the north- eastern side of the town, toward the slope of Monte Ingino (see note on Via 14), its location may be of some help in locating the hebetaf of b 53 and consequently Acedonia. But there is no proof that the two ' exits ' were the same, and moreover it is not likely that Acedonia, where the people were assembled, was situated on a steep incline. — termnuco stahituto: no parallel form in lb 16, but lb 19 has
274
ill. TEXT AND TRANSLATION
poi . percam . arsmatia . habiest . etursta . eso . eturstahmu . pisest . Mar I (54) tarsinater . trifor . tarsinater . iuscer nakarcer . iabuscer . nomner . eetu . ehesu . poplii . nosue . ier . ehe . esu . poplu sopir hate / (55) esme . pople . poitatu . ulo . pue .
the ritual wand shall pronounce banish- ment. Thus shall he pronounce banish- ment : " Whoever is of the Tadinate people, of the Tadinate tribe, of the Tuscan, the Narcan, the lapudic name, let him go out from this people. Unless a going out from this people takes place, if any- one is caught among this people, bring him to that place to which it is the law,
53 etursta: for eturstahmu: enough blank space for three or four letters follows.
pune prinuvatus staheren termnesku ;
perhaps the sg. denotes the boundary and the pi. the posts (two cippil) which marked the boundary and at which the two prinuati took their stand. It is quite certain that neither the "^''ebetras nor the termini were identical witli any of the three city-gates, which are called uerof, -ir, etc., accompanied by the proper name of the particular gate. The boundary at whicli the banishment was pronounced must have been located on the pomerium, and what lay beyond was not really alien territory but was within the Iguvine *trifus (see on 54 below). The aliens then were simply expelled from the area which was to be encircled by the lustral proces- sion. — 53-4 lolar / Tarsinater trifor Tarsinater: cf. the similar expression tu- tape(r) liuvina trefiper IiuvinaIII24-5, 29-30. trifor, trefi in these passages can- not refer to a division into thirds, as L. tribus does in connection with Ranw.es, Titles, and Luceres, the names of the three tribes of early Rome, and as the apparent derivation from *tri- would lead us to expect; unless Iguvium and Tadinum are to be thought of as (original- ly third) portions of some larger grouping, which is altogether unlikely. A satisfac- tory sense may be reached through com- parison with Livy, XXXI, 2, 6 ... C. Ampium ... per Umbriam, qua tribum Sapiniam vacant agrum Boiorum invadere iussit; XXXIII, 37, 1 consul per tribum
Sapiniam in Boios venit. The trifus then is a territorial unit, as the tota is a politic- al unit. — Tarsinater: the town from v/hich this ethnic is derived is commonly identified with the modern Gualdo Ta- dino, situated on the main road between Foligno and Fano, and about 20 km. southeast of Gubbio. Philipp, R.-E., 2te Reihe, VIII, pp. 1999-2000, places it near the Church of S. Maria Tadina, about one mile north of Gualdo Tadino. — Tuscer: the nearest large Etruscan city was Perusia, nearly 40 km. southwest of Iguvium. — Naharcer : the spelling Na- harkum in lb 17 shows that, originally at least, the h must have had etymological value, since the spelling with vowel -f h + vowel is not used in the earlier tables merely as a mark of length. A similar Latin spelling occurs in the expression municipi Interamnat(is) Nahartis ( C.I.L., XI, 4213: age of Augustus). On linguistic grounds there is no objection to connect- ing these forms with Nar, the ancient name of the river Nero, since the spell- ing with a simple (long) a was sufficient once the h ceased to be sounded, but the relative remoteness of the Nar from Iguvium might be an objection. Vet., p. 183, for this reason suggests that Tuscer and Naharcer both refer to a single nearby Etruscan state, perhaps in the vicinity of Perusia; yet there is no evidence of such a state to which Na- harcer could refer, and moreover it is
VIb 53
COMMENTARY
275
not inconceivable that the word Tuscer could have been meant to include all Etruscans indiscriminately in the banish- ment. It may be best therefore to assume that the Naharci were the inhabitants of a state located somewhere along the Nar. Krahe, GL, XXVI (1938), pp. 95-7, refers our form to a stem *Nah-a.r-ko- with -ar- a suffix used in such river-names as 'lodga, etc., and -ko- designating the dwellers along the river, as in Isarci. — labuscer: in a total of 8 instances (in- cluding lapuzkum lb 17, the only in- stance in the native alphabet) 5 have b and 3 have p p. The 6-forms are perhaps to be compared with L. Burrus, buxus, where b replaced the p of ITvoqoq, tiv^oq as a result of the difference of degree of tenseness in the voiceless stops in Greek and Latin. If we make allowance for the difference of suffix, the relationship of the name labuscer, lapuzkum (< *Ia- pud{i)s-ko-) to that of the lapyges in the heel of Italy is obvious. Both nations were of Illyrian origin; for the labusci cf. Strabo, IV, 207 Kai ol 'Idnodsg ds r]dr] rovTo EJtlixixTov 'IXKvQiolo, xal Kek'colc, sQvoc, Tisgl rovTovQ olxovai totiovq xal r\ "Ox^ Qa Tzkrjaiov rovxiov eaxiv. Pliny, N.H., III, 18, 127 Carnorum haec regio iunctaque lapudiim, amnis Timavus. If the proper home of the lapudes was the region be- tween the Carnic Alps and the Adriatic, the labusci banished from the Iguvine lustration were probably an offshoot far- ther south, perhaps somewhere between Ancona and Ravenna. — nomner: see on Via 23 nomneper; the nomen stands for the nation in the widest possible sense; as it is more inclusive than tola and trifus together, so is the alien char- acter of the Tusci, Naharci, and labusci more strongly emphasized than that of the Tadinates, who, we may assume, were probably an Umbrian-speaking nation. — ehesu: preposition ehe + pronoun esu are in crasis. — ier: an isolated form, with no parallel in lb 18, but best taken as a pf. subj. pass, of *i- 'go' used im- personally and thus equivalent in mean-
ing, though not in form, to L. Hum sit. — sopir habe: lb 18 reads svepis habe. habe has caused considerable difficulty. If taken as 3 sg. ind. act. equivalent to L. habet, with the majority of editors, the loss of final t (< -//) is surprising in view of the preservation of -t in tigit, trebeit (heri IV 26 may be otherwise ex- plained). Moreover it is then necessary either to understand habe as equal in sense to L. capit and supply an object signifying any alien who is caught after the sentence of banishment, or to under- stand habe in the sense of L. habitat ' dwells. ' The best solution is to follow Kent, C.P., XV (1920), pp. 359-60, in taking habe as 3 sg. ind. pass, from *haber ' (if anyone) is caught. ' Omission of the passive ending -r is otherwise unknown in Tables I-IV, but -pe for -per occurs in la 12, and in general the number of 3 sg. pass, forms in the native alphabet is not sufficient to establish norms of spelling. The objection to habe as 'capitur' on the ground that 3 sg. pass, forms in -r, as contrasted with -ter, are always imper- sonal need not be pressed too strongly, since the personal use is probably to be assumed for ferar in VIb 50. — • 55 portaiu ulo pue mersest,fetuuru pirse mers- est: the two versions agree in all essen- tials, but neither the place of punishment nor the nature of the punishment is precisely stated.
VIb 55 -Vila 2 (- lb 19 - 23). The actual lustration. The adfertor orders the Iguvine people to arrange themselves in formation and, still accompanied by the two prinuati, he makes the circuit around the people with the sacrificial victims. Having completed the circuit the ad- fertor and prinuati pray to Serfus Martins, Prestota and Tursa for afflictions to be visited on the Tadinate, Tuscan, Naharcan and lapudic nations and for divine favor toward Iguvium, after which the order ' go, Iguvines ' is given. The circuit, the prayer, and the order 'go, Iguvines' are repeated a second and a third time, after which the prinuati depart the way they
276
III. TEXT AND TRANSLATION
mersest . feiu . uru . pirse mersest . trioper . eheturstahamii . (font . termnuco . com . prinuaiir / (56) stahiiu . eno . deitu . arsmahamo . caterahamo . iouinur . eno com . prinuatir . peracris . sacris . amhretuio ape . ambrejurent / (57) iermnome . benurent . termnuco . com . prinuatir . eso . persnimumo . tasetur . serfe . martie prestota . serjia . serjer / (58) martier . tursa . serjia . serfer . martier . iotam . tarsinatem . trijo .
do with him that which is the law." Three times shall he pronounce banish- ment. He shall stand on the same spot at the boundary together with the pri- nuati. Then he shall say: "Arrange your- selves in priestly ranks and military ranks, men of Iguvium. " Then along with the prinuati they shall make the circuit with the unblemished victims. When they have made the circuit and have come to the boundary, at the boun- dary he and the prinuati shall thus pray silently: "Serf us Martius, Prestota Ser- fia of Serfus Martius, Tursa Serfia of Serf us Martius: the Tadinate state, the
came. The version in lb is considerably more condensed; it not only omits the text of the prayer but does not even name the deities to whom the prayer is address- ed. Instead of the peracris sacris of VIb 56, lb 20 has tures et pure ; and although in lb the prinuati are present, they are not included in the instructions for the procession and prayer; contrast, for example, ambretulo 56, 63, 64 and apretu lb 20, amprehtu 21, pe(r)snimumo 57, 64, 65 and pe(r)siiimu lb 21, 22. — 58 arsmahamo caterahamo Iouinur: lb 19-20 with armanu kateramu Ikuvinu is substantially the same, with arm- written for arm- and armanu corrected to read -amu by all editors except Br., who was hesitant (p. 182); the error involves omission of one oblique and one vertical stroke in the m . For tlie phonology of these pi. imv. pass, forms see 127 d. As for the sense it is clear that the Iguvini are ordered to arrange themselves in formation, and it is altogether unlikely that arsmahamo and caterahamo are mere- ly synonyms; one may refer to larger and the other to smaller military units, the former including the latter, or one may refer to infantry and the other to cavalry units (cf. Hor., Episi., II, 1, 190
dum fugiunt equitum turmae peditumque catervae, and Dion. Hal., Ant. Rom., IV, 22 on the Roman lustration, cited above on VIb 48 - Vila 54). Devoto, p. 277, however, improves on earlier interpreta- tions by seeing a distinction between priestly ranks and military ranks; the connection between arsmahamo and the cult-words, arsmor, arsmatia, arsier is most favorable to such a view whether we take arsmor as ' ritus' with the major- ity or as ' sacer datum concilia' with Dev., Bott. cateraJiamo is generally regarded as a denominative to the equivalent of L. caterva; otherwise Dev. and Bott., but there is no apparent ground for aban- doning the traditional view, once we ad- mit the loss of w after r attested by seritu in contrast to L. servare, servire (cf, 54 g). — ambretuto: for the syntax of the plural see 138 a. The sense is nearly equivalent to that of afero aferum. The latter, however, appears to be the proper ritual term for the lustration as a whole, and is used transitively, while ambretuto is used of the actual circumambulation and no- where has a direct object. Each verb has its own fut. pf,, and andirsafust, the sup- pletive to afero, governs poplo in the three passages where it occurs, yet its
VIb 55
COMMENTARY
277
use in Vila 46 = lb 40, with reference to the completion of the third circuit, re- sembles that of ambrefurent, which is used in VIb 56 of the first. — 57-8 Serfe Mar tie, Prestota Serf la Serfer / Martier, Tursa Serfia Serfer Martier: this triad is the special object of worship in the ceremonies of the lustration. None of its members is mentioned elsewhere in the Tables (although Tarse alone ap- pears in IV 19), and at the same time no other deities are honored during the lus- tration, except for Fisovius Sancius, who receives a libation and cake at Vila 37, and Tursa Jovia, who is invoked in Vila 47-51 in a prayer nearly identical with that addressed to the triad in VIb 57 ff. ^erfus Martins himself is the recipient of a sacrifice of three boars at the Fontuli in Vila 3-5, while further offerings are made to Prestota Serfia at Rubinia in Vila 6-36 and to Tursa Serfia across the Via Sancta in Vila 41-2. The closest Latin cognate to Serfus is Cerus: cf. Fast., 109 Li. ... et in carmine Saliari Cerus manus inlelligitur creator bonus; Varro, L.L., VII, 26 (Kent) ... in Carmine Saliorum sunt haec: ... laneus iam es, duonus Cerus es, duonus lanus. Serfus and Cerus then signify a male deity cor- responding to Ceres, Oscan Kerri (dat. sg. on the Agnone Dedication — Co. 175) in much the same way as the rare Tel- lurus (Mart. Cap., I, 49) to Tellus. For phonological details see 59 e. Serfe, L. Cerus, and O. Kerri are syncopated, in contrast to unsyncopated L. Ceres, Ce- reris, while the gemination is not ex- pressed in the spelling of Cerus, nor in O. Keri on the Curse of Vibia (Co. 130). L. creo, -are, cresco, creber are generally admitted into the same etymological group, ere- representing a disyllabic base *kere- and creber < *kres-ro- possibly showing the same s-extension assumed for L. Cereris < *Keres-. Serfus is then a deity associated with growth and nour- ishment, and his importance in the lus- tration is quite natural. The regular use of Martier as an epithet of Serfer is closely
connected with the debated question of the agricultural and warlike aspects of Mars. Cf. Wissowa, Religion und Kultus der Romer, p. 143, Dumezil, Les dieux des indo-europeens, pp. 28-30; Marback, R.-E., XIV2, pp. 1934-37. The impor- tance of Mars in the Roman lustration is apparent from Dion. Hal., Ant. Rom., IV, 22, 1 Td 6' leosla ravra role. Jieoia/dfjvai neol TO aTQaroTtedov xeXevaaq, edvas xq) y.a- rexovTi TO jcediov "Aqei and from the long prayer addressed to him in the lustra- tion of the field in Cato, R.R., 141. The two female associates of Serfus Martius show contrasting aspects of his character and activity. Prestota, v/ho is known only from the two versions of the lustral ceremony, has been compared with Au- terstatai on the Agnone Dedication and, more aptly, with the Sabine Praestita (C.I.L., IX, 4322), Jupiter Praestes (C.7.L. XIV, 3555), the Lares Praestites of Ov., Fast., V, 129-36 Prciestitibus Maiae Lari- bus videre Kalendae aram constitui par- vague signa deum ... Slant guogue pro nobis, et praesunt moenibus Urbis, et sunt praesenles auxiliumque ferunt; cf. also the Roman goddesses Praestana, Arn., IV, 3, Praestitia, Tert., Adv. Nat, II, 11. Praestota Serfia then represents the de- fensive power of Serfus Martius which is besought to avert evil from the people of Iguvium. Tursa on the other hand re- presents the offensive force which terri- fies and puts to flight the enemies of the city. The name is clearly related to the verb tursitu, forms of which are used as a part of the curse against the enemies of the state (60 = Vila 49) and with re- ference to the chasing of the heifers (lb 40, 41, Vllb 2), and the connection with L. terror is scarcely less evident despite the difference in vowel-grade and stem- class and the assimilation rs > rr. For the religious concept cf. the similar deification of Pavor and Pallor among the Romans: Livy, I, 27, 7 Tullus in re trepida duodecim vovit Salios fanaque Pallori ac Pavori; Serv. ad Aen., VIII, 285 {Salii) ab Hostilio vero pavorii et
278
III. TEXT AND TRANSLATION
tarsinatem . tiiscom . naharcom . iabuscom . nome j (59) totar . tarsinat . er . trifor . tarsinater . tuscer . naharcer . iabuscer . nomner nerf . sihitii . ansihitii . iouie . hostatu I (60) anhostahi . tursitu . tremitu hondii . holtu . ninctu . nepitu . sonitu . sauitu . preplotatu
Tadinate tribe, the Tuscan, the Narcan, the lapudic name, the chief citizens in office and not in office, the young men under arms and not under arms, of the Tadinate state, of the Tadinate tribe, of the Tuscan, the Narcan, the lapudic name: terrify them and cause them to tremble, cast them down to Hondus, to Hola, overwhelm them with snow, overwhelm them with water, deafen them with thunder and wound
pallorii insfituti; Lact., I., 20 Pavorem Palloremque Tullus Hostilius figuravit etcoluit. —58-60 totam Tarsinatem, etc.: the first portion of the prayer consists of a curse against the four enemy nations which were included in the banish- ment in 53-54, and the Tadinates are again placed in a separate category from the other three nations, as in the banish- ment, but this time the names of the nations are followed by a division of their populations into nerf and iouie, these classes being further divided into two sub-classes designated by affirma- tive and negative adjectives, so that the passage reads: ... Tuscer Naharcer Ia- buscer nomner nerf sihitu ansihiiu iouie hostatu anhoslatu tursitu, etc. — 59 nerf: see the discussion on Via 30, where, however, nerf w^as contrasted with arsmo. In the present passage, on the other hand, nerf is contrasted M'ith iouie and so must refer to a grouping based primarily on age. — 59 sihitu ansihitu : this pair may be connected etymologically with L. ci- tos incitos, as by Br. and Dev., or with cinctos incinctos, as by the majority ol editors, but in any case an- must be the negative and not the local preposition as in L. incinctus and (partly) incitus, for the opposition of ideas is as strong as in hostatu anhostatu below. Breal's inter- pretation, being connected with an un- tenable view of nerf {'Lares'), may be ignored here, but it is necessary to take
account of Devoto's objection to L. cincius, incinctus (p. 282). On the ground that the normal word for ' gird ' was *kringaom (inferred from cringatro kri- katru) he makes sihitu ansihitu equiva- lent to L. accitos non accitos and as sup- port for accire of summoning to office he cites Cic, Resp., II, 13, 25 ... eumque (sc. Numam Pompilium) ad regnandum ... Romam Curibus accivit; Lucan, I, 584 haec propter placuit Tuscos de more vetusto acciri vates. Yet the use of the simplex in this sense would not seem quite natural; L. cio, cire, ciius is mostly limited to compounds, while direct connection of sihitu with citus is impossible. I have therefore decided, though hesitantly, to follow the usual view whereby sihitu is equivalent to cinctos. The two roots *krengh and *keng/k may easily have coexisted in the language, the latter ap- pearing not only in sihitu but also in gihgera: indeed the two maj^ possibly even be variants of a single ultimate root altered by assimilatory or dissimilatory processes, sihitu ansihitu taken as cinc- tos incinctos may refer to men in and not in military office; so Bii., p. 99, takes it, and the use of cingitur, cincius for taking up or being under arms can be supported by Verg., Aen., II, 749 cingor fulgentibus armis; D/g^., XXXIX, 1, 38 cinctus in alia militia: etc. This gives a sufficiently comprehensive division of the population. If we were to extend the meaning of
VIb 58
COMMENTARY
279
sihitu further so as to include all in of- fices of any kind, military or civil, the support from L. cingitur, cinctus would be less good. — iouie: all editors since Bii. have recognized the equivalence of this word with L. iuvenes, only the stem being different. — 59-60 hostatu / an- hostatu: the iouie or young men of mili- tary age are divided into two groups, the hastatos and the *inhastatos. The distinc- tion is either between the heavy-armed and the light-armed or between those under arms and those not under arms. Bii., p. 100, favors the former interpre- tation, while Dev., p. 283, mentions both possibilities but uses ' alle arrai e non alle armi' in his Italian translation in T.G. The question cannot be positively decided, but Devoto's viev/ appears slightly better, since if the second class consisted of light-armed troops, they might rather be designated by some characteristic weapon than merely by their lack of spears. — 60 : here follows a series of ten verbs in- dicating the nature of the afflictions to be visited on the enemies of Iguvium. The arrangement of the verbs in five al- literative pairs is the most striking in- stance of the use of alliteration to be found anyv/here in the Tables; for other instances of its use see Introduction, 11. The verbs are all imperatives of the usual 2nd and 3rd sg. type, despite the fact that the deities addressed form a triad and that 61 below shows a consistent use of the plural, the repetition of the names of the triad being followed by fututo foner pacrer pase uestra. Dev., p. 282, notes the discrepancy and suggests that the three deities are not regarded here as the direct agents of vengeance and that the verb forms tursilu, etc., have therefore an impersonal character. — tursilu: equi- valent to L. terreto; for phonology see on Tursa in 58 above. — tremitu: equiva- lent in sense to L. Iremefacito; for the phonology and conjugational class see 118 d. — hondu: cognate with hondra, hondomu, and the name of the god Honde; for details see 60a, 119 c. — holtu:
sometimes regarded as connected with oXXv/j,i, in which case the h would have no etymological justification. Ribezzo, II, p. 103, suggested connection with the name of the goddess Hala (IV 17). If *Hola is the female counterpart of the infernal deity Hontus, the present pair of verbs would mean ' Honto Holae mit- tito, ' a tempting explanation, were it not for the fact that analysis of holtu as an actual denominative from *Hola is impossible and Ribezzo gives no really adequate explanation of the verb stem. Connection with olXvui may not be im- possible after all, for all the verbs in the curse beside hondu holtu are joined in alliterative pairs, and it would not be surprising if *oltu should here have re- ceived an unetymological initial h, at least in spelling if not in pronunciation. A possible source would be contamination with *Hola or other words of its family, or with hondu itself. The weakness of the sound of h was a source of confusion, and hebetafe VIb 53 with unetymological h, in contrast to ebetrafe Via 12, is an actual instance of this confusion. — ninctu: taken as equivalent to L. ninjuifo by practically all editors except Pis. and Vet., who object to the transitive use of ninctu in the sense 'cover with snow' and translate 'necato,' deriving from *ni-n{e)c-tdd. A stronger objection to the traditional interpretation is the failure of the form to show an initial sn-: cf. L. ninguit, but Av. snaezaiti, Goth, snai- wan, Lith. sninga, OCS subst. snegu, and L. nare but U. snata (see 57 b). Yet des- pite the loss of s- it seems best to follow the older view. Initial clusters consisting of s -f- nasal show a certain degree of instability in several IE languages, as a result of sentence-sandhi; it is difficult to find good support for a reduplicated present *ni-n{e)c-i6d; and on the other hand the use of ninctu with external accus. although without parallel in Latin, is indirectly implied for Greek by the pas- sive use in Ar., Ach., 1075 vAjtsiTa r)]Q£lv vsicpofievov raQ ecr/^oAd;. Xen., HelL, II,
280
III. TEXT AND TRANSLATION
preuilatu / (61) serfe . martie prestota . serfia serfer . martier . tursa . serfia . serfer . martier . fututo . foner . pacrer . pase . uestra . pople totar . iiouinar / (62) tote . iiouine ero nerus . sihitir . ansihiiir . iouies . hostatir . anostatir . ero . nomne . erar . nomne . ape . este . dersicurent . eno J (63) deitu . etato . iiouinur . parse . perca . arsmatia habiest. ape este . dersicust . duti . ambretuto . euront . ape . termnome / (64) couortuso . suTUTont . pesnimumo . sururont . deitu . etaians . deitu . enom . tertim . ambretuto . ape . termnome . benuso / (65) sururont . pesnimumo . sururont . deitu etaias eno . prinuatur . simo . etuto erafont . uia . pora benuso J
them, trample them under foot and bind them. Serfus Martins, Prestota Serfia of Serfus Martins, Tursa Serfia of Serfus Martius,be favoreble and propitious with your peace to the people of the state of Iguvium, to the state of Iguvium, to their chief citizens in office and not in office, to their young men under arms and not under arms, to their name, to the name of the state. " When they have said this, then he that holds the ritual wand shall say "Go, men of Iguvium." When he has said this, the same ones shall make the circuit a second time. When the boundary has been reached again, they shall pray in the same man- ner. He shall speak in the same man- ner; he shall direct them to go. Then for the third time they shall make the cir- cuit. When the boundary has been reached, in the same manner they shall pray, in the same manner he shall di- rect them to go. Then the prinuati shall go back by the same road by which they came.
4, 3 01 8s virpoixEVOi aTtfj^.dov eiz to acrrv. Cf. also Polyb., XVI, 12, 3. — nepitu: generally regarded as related to Neptu- nus and translated ' inundate. ' The in- terpretations of ninctu and nepitu are interdependent insofar as, if one is taken as a curse invoking disastrous weather, the other should preferably be taken in the same way. For curses of this type Dev. compares Gromatici, ed. Lachmann, 350 f. qui contigerit terminum moveritque damnabiiur a diis ... domus exstirpabitur, gensque ... omnia interiet ... terra a tem- pestatibus et turbinibus ... movebitur ... fructus saepe laedentur decutienturque im- bribus atque grandine caniculis inlerient, robigine occidentur. Vet. derives nepitu from *nepid-iu and translates 'ad nihilum redigito, ' thus preserving semantic simil-
arity between the two verbs in the pair (see on ninctu above). — sonifu: both verbs in the present pair are obscure, but sonilu is in all probability somehow connected with L. sonare. Devoto trans- lates 'verberato,' the sense 'beat' being derived from the notion of loud noise inherent in the root *swen- > son-, as in the derivation which he adopts for persondru. Whether this beating is to be accomplished by means of lightning and thunder is not clear, but it is alto- gether natural that a curse of this sort should include this form of divine ven- geance. In order to keep fairly close to the probable etymological sense of sonitu I have translated it 'deafen with thun- der. ' — sauilu : possibly related to L. saucius if c in the latter form is regarded
VIb 60
COMMENTARY
281
as an extension and not a part of the root. Nothing can be affirmed with cer- tainty, since the connections of saucius itself are ratlier uncertain, but the varia- tion of forms with and without -cio- could be compared with L. novus: novi- cius, or for semantic purposes with emo: empticius, suppono: suppositicius. sauitu might then correspond to a lost Latin verb whose relation to the pass. adj. saucius would be similar to that in the forms just cited, allowing for the dif- ference in their derivation through the pass. pcpl. stem. 'Wound' is as satis- factory a translation as we can find under the circumstances. — preplotatu: Vila 49 has preplohotatu, showing that the root-vowel is long. Since preuilatu — preuislatu, the other member of the pair, is clearly a verb of binding, we might expect a similar sense for preplotatu also. Ribezzo, II, p. 104, connects with L. plecto, amplecior, deriving from *prai- plokia-iod, which gives the desired sense and accounts for the length of the root- syllable (see 46 i), but the o-vowel grade is scarcely admissible in a formation of this type. It is best then to follow Bu., Co., and Dev. in regarding preplotatu as a denominative from Umbrian p lotus known from Fest., 274 Li. <plotos appel- lant^ Umbri pedibus plants natos ... et <Macci> us poeta, quia Umber Sarsinas erat, a pedum planitia, initio Plotus, postea Plautus coeptus est dici. Yet the curse is probably intended not merely to inflict lameness on the enemies of Igu- vium, as suggested by Conway's ' impe- dito, claudum facito, ' but rather to lay them flat (' sternito ' Dev.). — preuilatu : an erroneous spelling of preuislatu, which appears in Vila 49 and which nearly all editors take as equivalent to L. *praevinculato, denomma.ti\e to vinculum. For the use of pre- in composition with verbs of binding, Bii., p. 101, calls atten- tion to L. praevincio, praestringo, praeligo. — 61-2 : the second portion of the prayer is an entreaty for divine favor toward the state and people of Iguvium. The triad
of deities addressed is the same as in the curse preceding, and the population of Iguvium, like the enemy populations in the curse, is divided into nerus sihitir ansihitir, iouies hostatir anostatir. — fututo foner pacrer pase uestra, etc.: for the sense of the expression as a whole see on Via 30. For the plural forms in contrast to sg. fursitu tremitu, etc., see on 60 above.
VIb 62 -Vila 2 (- lb 21-23). In- structions for repeating the circuit, etc. — 62 dersicurent: reduplicated fut. pf. of deitu, with rs from intervocalic d. — 63 eiato liouinur: this can scarcely be an order for dismissal, despite Dev., § 155, p. 286 (' ter formula de conventu dimitten- do pronuntianda est'), for in this case we should expect some order such as arsmahamo caterahamo directing them to reassemble, and moreover it is altogether likely that they would have been expected to remain in formation until the whole lustration was completed. Possibly they were obliged to perform certain evolu- tions within the area between the bound- ary-mark and the three stations at which the sacrifices were to be offered. Cf. the description of the lustration of the Macedonian army in Livy, XL, 6, 1-2 caput mediae canis praecisae et pars ad dexter am, cum extis posterior ad laevam viae ponitur: inter hanc divisam hostiam copiae armatae traducuntur. For the form of etato see 127 c. — euront: the adfertor and the two prinuati. — termnome: to the boundary-mark at which the adfertor had pronounced the banishment in 53 ff. and at which the Iguvines had been ordered to assemble in 56. — 64 etaians deitu : subjunctive here and in 65 = Vila 1, in contrast to imv. in 63 and in lb 21, 22. — 65 eno prinuatur simo etuto erafont uia pora benuso: the two prinuati had joined the adfertor in 50-51 and had proceeded with him to Acedonia and the boundary-marks by the Via Aviecla. si- mo: commonly translated 'retro,' but the actual sense is '(back) hither' from the viewpoint of one who is in the in-
282
III. TEXT AND TRANSLATION
(Vila 1) sururont . pesnimumo . sururont . deitii . etaias . eno . prinuatur . simo . etuto . erafont . uia . pora I (2) benuso f
(3) fondlire . ahrof . trif . fetu . heriei . rofu . heriei . peiu . serfe . mar tie . feitu . popluper . totar . iioiiinar . totaper / (4) iiouina . uatiio . ferine . feitu . poni . fetu . aruio . fetu . tases . persnimu . prosesetir mefa . spefa . ficla . arsueitu j (5) suront . naratu . puse . uerisco . treblanir . ape . traha . sahata . combifiansust . enom . erus . dirstu /
(6) rubine . porca . trif . rofa . ote . peia . fetu . prestote . serfie . serfer . martier . popluper . totar . iiouinar . totaper j (7) iouina . persaia . fetu . poni . fetu . aruio . fetu . suront . naratu . pusi . pre . uerir . treblanir . tases . persnimu / (8) prosesetir . strusla . ficla . arsueitu . ape . supo .
In the same manner they shall pray, in the same manner he shall direct them to go. Then the prinuati shall go back by the same road by which they came.
At Fontuli he shall sacrifice three boars, either red or black; to Serf us Martins he shall sacrifice for the people of the state of Iguvium, for the state of Iguvium. He shall place the ribs on a tray, sacrifice with mead, offer grain, pray silently, add to the parts cut off a mefa spefa cake and a ficla cake, and recite the same formulas as at the Tre- bulan Gate. When (the porrectio) has been announced from Trans Sanctam then the erus shall be distributed.
At Rubinia he shall sacrifice three sows, red or black, to Prestota Serfia of Serfus Martius for the people of the state of Iguvium, for the state of Iguvium. He shall sacrifice (the victims) upon the ground, sacrifice with mead, offer grain, recite the same formulas as before the Trebulan Gate, pray silently, and add to the parts cut off a strusla cake and a ficla cake. When he has put the under- parts at the back (of the altar), then
terior of the town. Cf. Varro apud Non., 131, 9 Li. eo die cis Tiberim redeundum est.
Vila 1 : the exact words of VIb 65 are repeated hi order to make it clear that the new table is a continuation of the instructions for the Lustration.
Vila 3 - 5 (= lb 24 - 26). Sacrifice of three boars to Serfus Martius at Fontuli. The instructions are in general similar to those given for several of the sacrifices performed during the purification of the Fisian Mount, but no two agree in all details. — 3 Fondlire: this place is mentioned only here and in the equiva- lent lb 24. Dev., p. 288, and in Studi etruschi, IV (1930), p. 238, suggests con-
nection with the medieval Via di Fonte- vole, more recently Via delle Fonti, old names of the present Via Savelli della Porta, one of the three thoroughfares of modern Gubbio. Derivation of the word as a dimiimtive of *font- (L. fans) is more satisfactory than connection with L. fundala 'cul de sac' (Varro, L.L., V, 145), which Bii., p. 105, offers as an alternative explanation. He further com- pares Fondlire with Athenian h Aljuvaig remarking that the equivalent of ad fontes would have been fondlisco; yet the translation 'at Fontuli' renders the sense adequately in English. — abrof: a boar is offered to Ahtus Mars in Ila 11,
VIb 65
COMMENTARY
283
where the stem-variant apron- is used, possibly designating a young boar. — peiu: since rofu is 'red' peiu must also be a color-name, and 'black' (L. piceus) is the only possibility. Dark-colored vic- tims are appropriate because Serfus Mar- tius was a chthonic divinity; cf. Val. Flacc, III, 439 Tunc piceae mactantur oves (in the purification of the Argonauts from blood-guiltiness); Arn., adv. Nat., VII, 19 quia superis diis ... color laetus acceptus est ac felix hilaritate candoris at vero diis laevis sedesque habitanlibus in- feras color furvus est gratior et tristibus suffectus e fucis. — Serfie Martie: see on VIb 57. — popluper to tar liouinar, totaper liouina: similarly in 6-7, 41, since the lustration is for the people; cf. the expression ocriper Fisiu, totaper lioui- na used in Via 58, etc., during the puri- fication of the Mount. — 4 uatuo ferine feitu: see on Via 57; poni: on Via 57; aruio: on la 3; mefa spefa, ficla: on Via 56. — 5 uerisco Treblanir: on VIb 23. — traha Sahata: the prepositional phrase has become fossilized into a place-name hke that of the modern Roman Traste- vere, and Sahata then takes its case not from traha but from the relation of the whole phrase to the rest of the sentence; in the present instance the case is abl. Most take Sahata as equivalent to Sanc- tam (sc. viam), while Dev., p. 295, Bott., p. 426, prefer connection with Satanes, lib 4, the name of one of the decuviae. The question has no bearing on the syn- tax or interpretation of the passages where the phrase occurs, but it seems more probable that Sahata was a street than an extended region, if we must think of sacrifices being performed in three places at once with the adfertor moving from one to another and instructions being called aloud across the Sahata. — combifiansust: when the porrectio of the third sacrifice, in Trans Sahatam, has been announced, the distribution of the erus is to be made at FontuU; so much is clear from lb 33-4 pune purtingus, kafetu, pufe apruf / fakurent, puze
erus tefa, where the instruction is really the same as in Vila 5, but is presented in connection with the third instead of the first sacrifice.
Vila 6-8 (= lb 27 - 30). Sacrifice of three pigs to Prestota ^erfia of Serfus Martins at Rubinia. The details resemble those described in the sacrifices behind the Trebulan and Tesenacan Gates not only in the species of victims but also in the manner of slaughter (persae fetu) and in the kinds of cakes offered. More- over the sacrifice proper, like that in honor of Fisus Sancius behind the Tesena- can Gate, is followed by supplementary offerings and prayers, after which in Vila 37-40 a libation and mefa spefa are of- fered to Fisovius with precisely the same prayers, ceremony of the vessels, etc., as in VIb 5-18. — 6 Rubine: this place cannot be identified by the help of any place- or street-name in modern Gubbio, but Dev. (St. etr., IV [1930], p. 226; T.I., p. 289) suggests a possible connection with the Etruscan names Rupinas (or Srupinas'!), CLE., 4990 (from Orvieto), Rupenial, 4408 (from Perugia). — 7 persaia fetu: see on Via 58. — 8 strusla: on Via 59.
Vila 8 - 36. Supplementary offerings to Prestota Serfia. The ceremony com- mences with a libation to be poured into a trench (?). The vessels used in the ceremony are of two kinds, black and white, corresponding to the nature of the prayers in conjunction with which the vessels are used. The first prayer contains an invocation to the deity to visit evil on the Tadinate nation along with the other enemies of Iguvium (cf. VIb 58), followed by entreaties for favor toward the people of Iguvium. The sec- ond prayer, which is offered with the white vessels, is an entreaty for the avert- ing of evil from the people of Iguvium and for divine protection of men, beasts, etc. The two prayers are in most res- pects closely parallel. The same initial invocation is found, mutatis mutandis, in 10-11 and in 25-26. 11-13, with the char-
284
III. TEXT AND TRANSLATION
postro . pepescus . enom . pesclu . ruseme . uesticatu . prestote . serfie / (9) serfer . martier . popluper . Mar . iouinar . totaper . iouina . enom . uesclir . adrir . ruseme . eso . persnihimu . prestota J (10) serfia . serfer . martier . Horn . esir . uesclir . adrir . popluper . iotar . iiouinar . totaper . iiouina . erer . nomneper / (11) erar . nomneper . prestota . serfia . serfer . martier . preuendu . uia . ecla . atero . tote . tarsinate . trifo . tarsinate / (12) tursce . naharce . iabusce . nomne . iotar . tarsinater . trifor . tarsinater . tuscer . naharcer . iabuscer . nomner j (13) nerus sitir . ansihitir . iouies . hostatir . anostatir . ero . nomne . prestota . serfia . serfer . martier . futu . fons / (14) pacer . pase . tua . pople . totar . iiouinar . tote . iiouine . erom . nomne . erar . nomne . erar . nerus . sihitir . ansihitir . iouies / (15) hostatir . anostatir . prestota . serfia . serfer . martier . saluom . seritu . poplom . totar . iiouinar , salua . serituu / (16) totam . iiouinam . prestota . serfia . serfer . martier . saluo . seritu . popler . totar
with a prayer he shall pour a libation into a trench to Prestota Serfia of Ser- fus Martius for the people of the state of Iguvium, for the state of Iguvium. Then (pouring) from the black vessels into the trench he shall pray thus : " Pres- tota Serfia of Serfus Martius, thee (I invoke) with these black vessels for the people of the state of Iguvium, for the state of Iguvium, for the name of the people, for the name of the state. Pres- tota Serfia of Serfus Martius, in every way turn thou evil against the Tadinate people, the Tadinate tribe, the Tuscan, the Narcan, the lapudic name, against the chief citizens in office and not in office, against the young men under arms and not under arms, of the Tadinate state, of the Tadinate tribe, of the Tus- can, the Narcan, the lapudic name, and against their name. Prestota Serfia of Serfus Martius, be favorable and propi- tious with thy peace to the people of the state of Iguvium, to the state of Iguvium, to their name, to the name of the state, to its chief citizens in office and not in office, to its young men under arms and not under arms. Prestota Serfia of Serfus Martius, keep safe the people of the state of Iguvium, keep safe the state of Iguvium. Prestota Serfia of Serfus Martius, keep safe the name of the people of the state of Iguvium, (the
acteristic preuendu uia ecla atero tote Tarsinate, etc., corresponds to 27-28 with ahauendu uia ecla atero pople totar Iiouinar. In 13-15 the purpose of the formula futu fons pacer pase tua, etc., is to prevent the curse of the preceding lines from having adverse effect upon the people of Iguvium; it is therefore without counterpart in the second prayer. 15 Prestota Serfia Serfer Martier, saluom
seritu through23 enom uesticatu ahafripur- satu and 28 through 36 are almost com- pletely identical. — 8 ape supo postro pepescus: for supo see on VIb 5, where precisely the same clause is used to intro- duce the instructions for offerings to Fisovius Sancius. — ruseme: there is no sure etymology, but the word appa- rently means some sort of trench or pit. Devoto, p. 292, translates ' in mundum, '
Vila 8
COMMENTARY
2S5
having in mind the covered pit or mundus in the Roman comitium (cf. Fest., 255, 273 Li.), and for the etymology compares Lith. rusys (a hole for winter storage of potatoes: cf. Walde-Pokorny, II, p. 356) but the unrhotacized s is a difficulty. W.-P.'s root 2. reu- 'aufreissen, graben, aufwiihlen' has a variety of extensions; it would be tempting to connect rusem with oQviLQ ' a digging, ' but a /i-stem of *ru-g-, *ru-gh- would yield *ruh-ti-. It is possible that rusem may be from *ru-dh-ti-m with development of the dent- al cluster as in Fise, Fiso, 44 d, the ex- tended root being the same as in ON rjoda 'reuten, riiumen' and other Ger- manic forms cited by Walde-Pokorny, II, p. 354. For the ritual value of the rusem Goidanich, pp. 65-6, calls atten- tion to the character of the prayer with which it is associated, and especially to the curse contained in the prayer. Cf. also the note on persom VIb 24. — ueslicatu: see on uestisia VIb 5. — 9-10 Prestola / Serfia ... tiom, etc.: for the ellipsis of the verb see 136 c. — 11 preuendu: only here, but etymology and sense are both fairly certain; cf. ahauendu 27, and for the directive force in pre- cf. L. prae(hi)beo, praecipio, praetendo. — ecla: the sense in conjunction with uia appears to be 'in every way,' but the derivation is very uncertain; possibly from *aiko-lo- (cf. Skt. ekas 'one') with -lo- as in L. singull and semantic develop- ment 'one> individual > each individual > every'; cf. Muller, p. 11. — atero: to be directed against the enemies of Iguvium, and in 27 to be averted from the people of Iguvium; it must therefore signify some aspect of evil, but again the derivation is very doubtful, aiero has been connected with adrir by Bii., p. 108, the notion 'evil' being derived from 'black' just as the color of the vessels is symbolical; with L. alter by some on the basis of Fest., 99 Li. alter ei pro non bono ponitur, ut in auguriis altera cum appellatur avis quae utique prospera non est; sic aliier nonnumquam pro adverse
dicitur et malo; possibly from *ap-tero- with semantic development as in L. deter ior according to a suggestion of PL, I, p. 426, II, p. 203; but none of these satisfactorily accounts for the unsyncopa- ted e. Dev., p. 295, translates 'hostem,' with derivation similar to L. ex-torrem, but the root of L. terra contained an s which would not be assimilated in Um- brian (see Buck, § 115.1). The safest ex- planation, at least from a phonological viewpoint, is that of Sommer, I.F., XI (1900), pp. 14-16, who derives from *a-terom taken as an infinitive equivalent to L. -terere 'rub' but without formal distinction of voice. He then translated ' advertito omni via civitati Tarsinati *absteri [= interire, interitum]'; 'Wende auf jedem Wege dem tarsinatischen Staate den Untergang zu' bezw. 'ab.' The unsyncopated vowel is not unnatural in the radical syllable, and for the sense Vetter, I, p. 268, cites L. detrimentum as a semantic parallel. — tote Tarsinate ... 12 labuscer nomner: see on VIb 53-4.
— 13 nerus ... anostatir: see on VIb 59-60.
— Prestota Serfia Serfer Martier, futu Jons ... 15 hostatir anostatir: evidently intended to preclude any possibility of harm to the people of Iguvium from the curse contained in the lines immediately preceding, for the same precaution is observed in VIb 61-2 just after the curse containing tursitu tremitu, etc., but not in the prayer with the white vessels which commences at Vila 25 and which is in most respects closely similar to our pre- sent passage but contains no curse. A precautionary formula similar in some respects to the present passage occurs in the curse tablets of Cnidus: i/nol 6s ooia xai ekevdEQa eh] ndvrcDQ (cf. Audollent, De- fixionum Tabellae, pp. 5-19, and R. G. Kent, C.P., XX [1925], p. 252, who used the Cnidian formula as the basis for a restora- tion in line 9 of the Oscan curse of Vibia).
— 15 Prestota Serfia ... 18 erar nomne: similar to Via 31-33, except that there the prayer was addressed to Jupiter Grabovius for protection toward the Fisian Mount,
286
III. TEXT AND TRANSLATION
iiouinar . totar . iiouinar / (17) nome . nerf . arsmo . uiro . pequo . castruo . frif . salua seritu . futu . fons . pacer . pase . tua . pople . totar . iiouinar /
(18) tote . iiouine . erer . nomne . erar . nomne . prestota . serfia . serfer . martier . tiom . esir . uesclir . adrer . popluper /
(19) totar . iiouinar . totaper . iouina . erer . nomneper . erar . nomneper . prestota . serfia . serfer . martier . tiom /
(20) subocauu . prestotar . serfiar . serfer . martier . foner . frite . tiom . subocauu . ennom . persclu . eso . deitu /
(21) prestota . serfia . serfer . martier . tiom . isir . uesclir . adrir . tiom . plener . popluper . totar . iiouinar . totaper / (22) iiouina . erer . nomneper . erar . nomneper . prestota . serfiar . serfer . martier . tiom . subocauu . prestotar j (23) serfiar . serfer . martier . foner . frite . tiom . subocauu . enom > uesticatu . ahatripursatu . enom . ruseme / (24) persclu . uesticatu . prestote . serfie . serfer . martier . popluper . totar . iiouinar . totaper . iouina . ennom . uesclir / (25) alfir . persnimu . superne . adro . trahuorfi . andendu . eso . persnimu . prestota . serfia . serfer . martier . tiom /
(26) esir . uesclir . alfir . popluper . totar . iiouinar . totaper . iiouina . erer . nomneper . erar . nomneper . prestota /
(27) serfia . serfer . martier . ahauendu .
name) of the state of Iguvium, keep safe the magistrates, the priesthoods, the lives of men and of beasts, the fruits. Be favorable and propitious with thy peace to the people of the state of Iguvium, to the state of Iguvium, to the name of the people, to the name of the state. Prestota Serfia of Serfus Martins, thee (I invoke) with these black vessels for the people of the state of Iguvium, for the state of Iguvium, for the name of the people, for the name of the state. Prestota Serfia of Serfus Martins, I in- voke thee; in trust of thy favor, Pres- tota Serfia of Serfus Martins, I invoke thee. " Then in prayer he shall speak thus: "Prestota Serfia of Serfus Martius, thee with these black vessels, thee with full vessels, for the people of the state of Iguvium, for the state of Iguvium, for the name of the people, for the name of the state, Prestota Serfia of Serfus Martius, thee I invoke. In trust of thy favor, Prestota Serfia of Serfus Martius, I invoke thee. " Then he shall pour a libation and dance the tripudium. Then with a prayer he shall pour a libation into the trench to Prestota Serfia of Serfus Martius for the people of the state of Iguvium, for the state of Iguvium. Then he shall pray with the white ves- sels; he shall place them crosswise above the black vessels; thus shall he pray: "Prestota Serfia of Serfus Martius, thee (I invoke) with these white vessels for the people of the state of Iguvium, for the state of Iguvium, for the name of the people, for the name of the state. Prestota Serfia of Serfus Martius, in every way turn thou aside evil from the
22 Serfiar: for serfia.
Vila 15
COMMENTARY
287
but here to Prestota ^erfia for the people of the Iguvine state. — 17 nome nerf ... castruo frif: see on Via 20. — 20 subo- cauu : see on Via 22. — frite : see on Via 24. — ennom persclu eso deiiu, etc. : a new prayer with a new invocation to the goddess and a new set of black vessels, for these are designated in 21 as plener. The vessels used in connection with the prayer in 9-10 had apparently been emp- tied in the pouring of the libation called for in 8. — 23 ahatripursatu : see on VIb 16. • — 23 enom ruseme ... 24 totaper louina just before the prayer with the white vessels corresponds to 8 enom pesclu ruseme ueslicalu, etc., just before the prayer with the black vessels. — 25 superne adro trcdiuorfi andendu: the
symbolic significance of this act is not clear, but it appears from the language that the white vessels are placed above the black ones in a crosswise position; at least it is natural to take superne as a preposition governing adro rather than to make adro the object of andendu with superne as an adverb, for the black ves- sels, whose use is nov/ finished, would not be placed on top of the v^rhite. trahuor- fi occurs only here but is clearly equi- valent to L. transverse; the vessels must therefore have been of oblong or elliptical shape, as Dev., p. 290, suggests. — 25 Prestota Serfia ... 36: for those respects in which the prayer with the white vessels differs from that with the black see above, on 8 - 36.
uia . ecla . atero . pople . totar . iiouinar . tote . iiouine . papier . totar . iouinar / (28) totar . iiouinar . nerus . sihitir . ansihitir . iouies . hostatir . anhostatir . ero . nomne . erar . nomne . prestota . serfia / (29) serfer . martier . saluom . seritu . poplo . totar . iiouinar . salua . seritu . totam . iiouinam . prestota . serfia . serfer / (30) martier . saluom . seritu . popler . totar . iiouinar . totar . iiouinar . nome . nerf . arsmo . uiro . pequo . castruo . frif / (31) salua . seritu . futu . fons . pacer . pase . tua . pople . totar . iiouinar . tote . iiouine . erer . nomne . erar . nomne . prestota / (32) serfia . serfer . martier . tiom . esir . uesclir . alfer . popluper . totar . iiouinar . totaper . iiouina . erer , nomneper .
people of the state of Iguvium, from the state of Iguvium, from the chief citi- zens of the people of the state of Igu- vium and of the state of Iguvium, in office and not in office, from the young men under arms and not under arms, from their name, from the name of the state. Prestota Serfia of Serfus Martius, keep safe the people of the state of Iguvium, keep safe the state of Igu- vium. Prestota Serfia of Serfus Martius, keep safe the name of the people of the state of Iguvium and of the state of Iguvium, keep safe the magistrates, the priesthoods, the lives of men and of beasts, the fruits. Be favorable and pro- pitious with thy peace to the people of the state of Iguvium, to the state of Iguvium, to the name of the people, to the name of the state. Prestota Serfia of Serfus Martius, thee with these white vessels for the people of the state of Iguvium, for the state of Iguvium, for the name of the people, for the name of
288
III. TEXT AND TRANSLATION
erar J (33) nomneper . prestota . serfia . serfer . martier . Horn . subocauu . prestotar . serfiar . serfer . martier . foner . frite . tiom / (34) subocauu . ennom . persclu . eso . persnimu . prestota . serfia . serfer . martier . tiom . isir . uesclir . alfer . tiom . plener / (35) popluper . totar . iiouinar . totaper . iiouina . erer . nomneper . erar . nomneper . prestota . serfia . serfer . martier . tiom J (36) subocauu . prestotar . serfiar . serfer . martier . foner . frite . tiom . subocauu . enom . uesticatu . ahatripursatu / (37) uestisa . et . mefa . spefa . scalsie . conegos . fetu . fisoui . sansii . popluper . totar . iiouinar . totaper . iiouina . suront / (38) naratu . puse . post . uerir . tesonocir . uestisiar . erus . ditu . enno . uestisia . mefa . spefa . sopam . purome . efurfatu j (39) subra . spahamu . traf . sahatam . etu . ape . traha . sahata . couortus . ennom . comoltu . comatir . persnihimu . capif I (40) sacra . aitu j
(41) trahaf . sahate . uitla . trif . feetu . turse . serfie . serfer . martier . popluper . totar . iiouinar . totaper . iiouina . persaea . fetu . poni / (42) fetu . aruio . fetu . tases . persnimu . prosesetir . strusla . ficlam . arsueitu . suront . naratu . puse . uerisco . treblaneir . ape / (43) purdinsiust . carsitu . pufe . abrons . facurent . puse . erus . dersa . ape . erus .
the state, Prestota Serfia of Serfus Mar- tius, thee I invoke; in trust of thy fa- vor, Prestota Serfia of Serfus Martins, I invoke thee." Then in prayer shall he pray thus: "Prestota Serfia of Ser- fus Martins, thee with these white ves- sels, thee with full vessels, for the people of the state of Iguvium, for the state of Iguvium, for the name of the people, for the name of the state, Prestota Serfia of Serfus Martins, thee I invoke ; in trust of thy favor, Prestota Serfia of Serfus Martins, I invoke thee. Then he shall pour a libation and dance the tripudium. Kneeling he shall offer a libation and a mefa spefa cake in a cup to Fisovius Sancius for the people of the state of Iguvium, for the state of Iguvium. He shall recite the same formulas as behind the Tesenacan Gate. He shall distribute the erus of the libation. Then he shall remove from the platter the libation and the mefa spefa cake and scatter them down into the fire. He shall go to Trans Sanctam. When he has returned from Trans Sanctam, then he shall grind (the cakes) and pray with the ground (cakes). He shall move the consecrated bowls.
At Trans Sanctam he shall sacrifice three heifer-calves to Tursa Serfia of Serfus Martins for the people of the state of Iguvium, for the state of Iguvium. He shall sacrifice (the victims) upon the ground, sacrifice with mead, offer grain, pray silently, add to the parts cut off a strusla cake and a ficla cake, and re- cite the same formulas as at the Trebulan Gate. When the porrectio has been com- pleted, the order shall be called to the place where they have sacrificed the boars, to distribute the erus. When
Vila 37
COMMENTARY
289
Vila 37 - 40 : Fisovius Sancius is again honored with a libation and mefa spefa, as in the ceremony described in VIb 5-18. Not only do the two passages agree in most details, but Vila 38 contains a specific direction to use the same formulas as behind the Tesenacan Gate. But apart from lb 29 kapi sakra aitu there is nothing in the earlier version to corres- pond to Vila 37-40 as la 16-19 corres- ponds to VIb 5-18. Devoto, p. 295, § 162, may therefore be right in suggesting the possibility that an error led to the addi- tion of this passage after the ceremony in honor of Prestota. The actual dif- ferences between the present passage and VIb 5-18 are: 1) the clause VIb 5 ape sopo postro peperscust is lacking in Vila; 2) the ceremony in VIb is performed for the Fisian Mount and the state, that in Vila for the people and state; 3) the prayers are given in VIb in full but in Vila only by cross-reference to the sacri- fice behind the Tesenacan Gate; 4) the instruction pescZu / semu uesticatu atripur- satu VIb 15-16, the reference to the erus of the prosecta, and the use of the expres- sion scalseto ... conegos in connection with uestisiar erus are without parallel in VI la; 5) the direction to the priest to go across the Sancta and to return are naturally not found in VIb; 6) VIb 18 mentions cups both purdila and sacra, Vila 39-40 only sacra. — 37 uestisa: see on VIb 5. — mefa spefa: on Via 56. — scalsie conegos: on VIb 5. — Fisoui Sansii: on VIb 5. — 38 erus : on VIb 16. — so- pam purome efurfatu: see 139 b. — 39 subra spahamu: on VIb 17, — traf Sahatam efu. ape traha Sahata couortus: the clause beginning with ape is obvious- ly the same instruction as enem Rubiname postro couertu given in 44 below in con- nection with the sacrifice to Tursa, for both passages are followed by comoltu, comatir persnihimu (et) capif sacra aitu. The adfertor leaves Rubin ia for Trans Sanctam in order to perform the sacrifice to Tursa, or at least so much of it as re- quires his personal presence and activity,
and remains until after the porrectio, when he gives the order for the distribu- tion of the erus at Fontuli and then re- turns to Rubinia. — ennom comoltu, comatir persnihimu: see on VIb 17. — capif I sacra aitu: on VIb 18.
Vila 41-45 (=Ib 31 - 39). Sacrifice of three heifer-calves to Tursa Serfia at Trans Sanctam. The details agree closely with those of the sacrifice behind the Trebulan Gate (Via 58-9) as well as with that at Rubinia, although in each case the order in which the separate instruc- tions are given is somewhat different. When the porrectio has been completed, the adfertor calls out the order for the distribution of the erus at Fontuli, which is followed by the distribution at Rubinia, and finally at Trans Sanctam. He then goes to Rubinia and performs the cere- mony of the grinding, after which he returns to Trans Sanctam and repeats the ceremony there. — 41 traha f Sa- hate: for the grammatical character of this place-name see on 5 above. — uitla: heifer-calves are offered as victims only here and in the equivalent lb 31. — Turse: see on VIb 58. — 43 carsitu ... combifiatu ... 44 combifiatu: in lb 33-5 the verbs are similarly distributed except that the third verb is subjunctive kupi- fiaia. The difference between the two verbs is that carsitu is used of calling out an order, while combifiatu is used of giv- ing information. In the present passage notice is given that the erus has been distributed at Fontuli, and this notice is the signal for the distribution at Ru- binia (43 postro combifiatu Rubiname erus I dersa). After a similar notice from Rubi- nia the erus is distributed at Trans Sanc- tam; on the whole passage cf. Dev., p. 296. — 43 pufe abrons facurent: the same circumlocution is used in lb 33-4; Dev., pp. 146, 288, regards it as another instance of euphemism. The form abrons is peculiar, for line 3 has abrof, lb 24, 33 apruf. Since there are no known ace. pi. forms in -ns in the second declension, editors in general have been inclined to
290
III. TEXT AND TRANSLATION
dirsust . postro . combifiatu . rubiname . ems I (44) dersa . enem . traha . sahatam . combifiatu . ems . dersa . enem . mbiname . postro . couertu . comoltu . comatir . persnimu . et / (45) capif . sacra . aitu . enom . traha . sahatam . couertu . comoltu . comatir . persnihimu. enom . purditom . fust / (46) postertio . pane . poplo . andirsafust . parse . perca . arsmatia . habiest.et . prinuatur . dm . tefruto . tursar . eso . tasetur j (47) persnihimumo . tursa . iouia . totam . tarsinatem . trifo . tarsinatem . tuscom .
the erus has been distributed, the in- struction shall be passed back to Rubin- ia to distribute the erus. Then the instruction shall be passed to Trans Sanc- tam to distribute the erus. Then (the adfertor) shall return to Rubinia, grind (the cakes), pray with the ground (cakes) and move the consecrated bowls. Then he shall return to Trans Sanctam, grind (the cakes), and pray with the ground (cakes). Then the sacrifice will have been completed. "When he has made the circuit around the people for the third time, he who holds the ritual wand and the two prinuati shall silently pray thus from the (place of) burnt offerings, to Tursa: "Tursa Jovia: the Tadinate state, the Tadinate tribe, the Tuscan, the Nar-
derive abrons from the same n-stem found in abrunu Ila 11, with various explana- tions for the anomalous form (of. Daniels- son, Altitalische Studien, III, p. 146; Buck, §181 b; Vetter, p. 270; Pisani, p. 183). Yet in view of the close rela- tionship of the present passage to the three others cited it seems better to as- sume an o-stem in all four passages and an n-stem only in Ila 11, where there is good reason to believe that the difference in the stem-class corresponds to a dif- ference in the victim (see on Ila 11). PL, I, pp. 510-11, suggested that abrons was a sandhi-doublet which might have existed beside abrof under certain condi- tions, presumably under conditions re- sembling those in 58 a (original medial ns). Our form would then represent an extension of this form into a different phonetic environment. — 45 enom pur- ditom fust: here used of the completion of the whole ceremony, as in VIb 42. Vila 46 - 51. Prayer to Tursa Jovia with imprecations against the enemies of Iguvium. The language of the curse and ol the prayer for favor toward the people
of Iguvium is identical with that of the version contained in VIb 57-62, except that now Tursa Jovia alone is addressed and consequently there are certain gram- matical variations (e.g. futu for fututo, fans for foner, tua for uestra, etc.); there are also several insignificant orthograph- ical variations. — 46 postertio pane poplo andirsafust: the same formula oc- curs in lb 40, where, however, it directly introduces the instructions for the pur- suit of the heifers, the text and even the mention of the prayer being omitted. For discussion of the actual position of the pursuit of the heifers in the order of events see above on VIb 48 - Vila 54 (= lb 10-45), where the lustration as a whole is briefly outlined. — porse perca arsmatia habiest: similarly 51, VIb 50, 53, 63. — prinuatur dur: there is an apparent confUct between this passage and VIb 65, where the prinuati are sup- posed to depart from the scene; see above on VIb 48 - Vila 54. — tefruto : used to indicate the spot where the adfertor and prinuati stand while reciting the prayer, as the postposition -to plainly shows.
Vila 43
COMMENTARY
291
The similar prayer commencing in VIb 57 was recited termnuco ' at the boundary. ' The form tefra in Ila 27, III 32, 34, IV 2 is commonly understood to designate certain pieces cut from the victims, to be roasted as burnt-offerings, and Oscan tefuriim on the Agnone Dedication (Co. 175, lines a 17, b 20) may have a similar sense. The present passage, however, seems to show a transfer of meaning from the object burnt to the place of burning, the hearth (L. focus or rogus, but not ara, U. asa). — 47 Tursa louia: Tursa Serfia of Serfus Martins is one of the triad invoked in the prayer commencing at VIb 47, nearly identical with the pre- sent one, and was also the recipient of three heifer-calves in Vila 41; while Tur- sa without special epithet is honored in IV 19-20 with offerings of a vestigia and persuntru. The epithet louio- on the other hand is borne by the deities Trebus, Tefer, Ticamnus, and Hontus. Tursa Jovia, who is mentioned only here, with the accompanying account of the pursuit of the heifers (51-54 = lb 40-44), evidently signifies the personification of that aspect of Jupiter which inspires terror in the people's enemies, as Tursa Serfia signifies the same aspect of Serfus Martins. That Jupiter's power should be invoked in this prayer is quite natural in view of the fact that several of the words of the curse, e. g. ninctu, nepitu, sonitu, seem to refer to disasters arising from the weather. For these and other verbs of the series beginning with tursitu tremitu see above on VIb 60.
Vila 51 -54 (- lb 40 - 44). Ritual pursuit of the heifers. The adfertor and prinuati put to flight an unspecified number of heifers, to be caught by per- sons waiting below the Forum Seheme- niar, and the first three caught are sacri- ficed to Tursa Jovia at Acedonia. The details of the sacrifice are similar to those at Rubinia (7-8), although they are not presented in the same order. The account of the chase in lb 40-44 differs in several respects from that in VII. The adfertor
and prinuati each chase a single heifer, making three in all. The chase begins above the place of assembly, and the heifers are caught below the Forum Se- hemeniar, as in VII, but apparently by the adfertor and prinuati and not by whatever persons first succeed in catching them. The pursuit of the heifers may be a ritualistic survival from the time when victims for sacrifices actually had to be hunted. What is more important, however, is its apparent magical character. It is scarcely a coincidence that the pursuit takes place in close connection with the curse against the enemies of Iguvium, and that the victims are offered to Tursa, while the verb tursitu is used both of the chasing of the heifers and as a part of the curse. Parallels to this feature of the Iguvine ritual are by no means wanting. Plato, Critias, 119 e, in the mythical story of Atlantis, refers to a similar pur- suit of bulls : diperoiv ovvcov ravQcov iv rep rov IIoaeidwvoQ Isqw... avev aidrjQOV ^vXoiq xai ^QOXOIQ eQrjQevov, ov 8e ekoiev rcbv ravQcov TiQog rrjv arrjXrjv JiQoaayay ovrsg Kara, xoqv- (prjv avrfjg eacparrov. Diog. Laert., I, 110, says of the purification of Athens by Epimenides during a pestilence in the forty-sixth Olympiad: ... 2.afifbv nqo^ara (jLeXavdrs xai Xevxa ijyaye jiqoc, t6v"Aqsiov ndyov • xdxsldsv elaoev levai ol ^ovXolvxo, TiQoord^ag rotg dxokovQoig evda dv xaraxkivoi avTOJv sxaOTOv dvsiv rip TtQoarjxovTi, deep • xai ovrw kfj^ai ro xaxov. The Oscan gerundive eehiianasum on a Capuan inscription (Co. 117) evidently refers to the same practice; cf. ehiato in Table Vllb 2 — 51 iuenga: how many is not stated, but if a definite number was called for, it was probably twelve, since VI lb 1-2 says that the fratricus is to procure the Atie- dian Brothers' twelve victims \vhich are to be let loose and pursued after the lustration of the people. — peracrio : for the meaning see on Via 25. The form here is gen. pi. ; it must therefore be partitive with the sense ' from among the excellent ones, ' that is 'from among those most fit for sacrifice.' — 52 faro sehemeniar:
292
III. TEXT AND TRANSLATION
naharcom . iapusco . nome . totar / (48) tarsinater . trifor . iarsinater . iuscer . naharcer . iapuscer . nomner . nerf . sihitu . ansihitu . iouie . hostatu . anosiatu j (49) tursitu . tremitu . hondu . holiu . ninctu . nepitu . sunitu . sauitu. preplohotaiu . preuislatu . tursa . iouia . futu . fons I (50) pacer . pase . tua . pople . totar . iouinar . tote . iouine . erar . nerus . sihitir . ansihitir . iouies . hostatir . anhostatir . erom J (51) nomne . erar . nomne . este . trioper . deitu . enom . iuenga . peracrio . tursituto . poTse . perca . arsmatia . habiest . et J (52) prinuatur . hondra . juTO . sehemeniar . hatuto . totar . pisi . heriest . pafe . trif . promom . haburent . eaf . acersoniem / (53) fetu . turse . iouie . popluper . totar . iiouinar . totaper . iouina . suront . naratu . puse . uerisco . treblanir . aruio . fetu J (54) persaea . fetu . strusla . ficla . prosesetir . arsueitu . tases . persnimu . poni . fetu /
(Vllb 1) pisi . panupei . fratrexs .
fratrus atiersier . fust erec. sueso . fratrecate . portaia . seuacne .
fratrom j (2) atiersio . desenduf . pifi .
reper . fratreca . parsest . erom . ehiato .
ponne . iuengar . tursiandu . hertei /
(3) appei . arfertur . atiersir . pop lorn .
can, the lapudic name, the chief citizens in office and not in office, the young men under arms and not under arms, of the Tadinate state, of the Tadinate tribe, of the Tuscan, the Narcan, the lapudic name: terrify them and cause them to tremble, cast them down to Hondus, to Hola, overwhelm them with snow, overwhelm them with water, deafen them with thunder and wound them, trample them under foot and bind them. Tur- sa Jovia, be favorable and propitious with thy peace to the people of the state of Iguvium, to the state of Iguvium, to the chief citizens in office and not in office, to the young men under arms and not under arms, of the state, to their name, to the name of the state." This he shall say three times. Then he who holds the ritual wand and the prinuati shall chase the heifers (selected) from among those fit for sacrifice. Below the forum of Semonia whoever of the state wishes shall catch them. Whichever three they catch first, these (the ad- fertor) shall sacrifice at Acedonia to Tur- sa Jovia for the people of the state of Iguvium, for the state of Iguvium. He shall recite the same formulas as at the Trebulan Gate, offer grain, sacrifice (the victims) upon the ground, add to the parts cut off a strusla cake and a ficla cake, pray silently, and sacrifice with mead.
Whoever at any time shall be brother- superior among the Atiedian Brothers, during his term of office shall bring twelve victims for the Atiedian Brothers, which are required on the brotherhood's account, to be let loose when the heifers are to be chased, when the Atiedian adfertor has performed the circuit around
Vila 52
COMMENTARY
293
lb 42 has furu sehmeniar, so that the final r to all appearances is original. Von Planta, II, p. 47, and Buck, § 257.4 explain as a stem in -dri-, the latter trans- lating 'forum seminarium.' One need not object too strongly on the ground that in Latin the suffix -dri- appears only as a result of dissimilation of -dli- (e.g. militaris, popularis, but annalis, Dialis, navalis), for exfaris, coquinaris, as well as O. dekkviarim, are exceptions to the general rule. But the loss of the final i would be strange, for it is preserved in uerfale, while there are no really applic- able examples to support loss of final i in *semen.idri. Devoto translated ' con- cilii, ' deriving from *semeno-yd and com- paring Skt. samanam ' Zusammenkunft, Festversammlung ' (p, 301). The etymol- ogy is too far-fetched to be convincing, but the recognition of -ar -ar as a gen. sg. ending is an improvement over older theories. Rhotacism of final s in the ear- lier tables is decidedly unusual, but afeper in 30 is an instance. According to the view preferred here sehemeniar sehmeniar is gen. sg. of the name of a goddess identical with L. Semonia and is so rendered in the translation. The only source of our knowledge of Semonia is Macrobius, I, 16, 8 Apud veteres quoque qui nominasset Salutem Semoniam Seiam Segetiam Tulilinam, ferias observabat. On Semonia's male counterpart Semo and on the ablaut-variation in the Latin and Umbrian names see on lib 1; cf. also Pi- sani, pp. 186, 196, Pighi, Umbrica, p. 20; Bott, pp. 270, 321.
Vllb 1-4. Responsibilities of the fra- tricus in connection with the heifer-chase. The four lines of text on the reverse side of Table VII have some connection with Vila, as is apparent from the reference to the chase after the completion of the lustration. The fratricus is charged with the responsibility of obtaining the heifers, under penalty of a fine in case he fails. reper fratreca in line 2 arouses the sus- picion that the ceremony here may be the private concern of the Atiedian Brother-
hood and distinct from those performed popluper totar louinar totaper louina (cf. Dev., pp. 303, 307). But it is diffi- cult to believe that the words appei or- fertur Atiersir poplom andersafust refer to any other lustration than the one described in the last portion of VIb and Vila. Perhaps the heifer-chase, and not the lustration, may be a special one under- taken in the interests of the Brotherhood; but it is not impossible that reper fratreca merely designates the ceremony as the responsibility of the Brotherhood, without implying that it is for their special bene- fit. — 1 fratrexs: on the nature of this official and on the derivation of the word see on Va 23. — fratrus Atiersier: for the dative see 143 a. — seuacne: used here without a noun, as also in III 22, where however it stands beside the ad- jective sakre. In the present passage, since desenduf is preferably to be taken with seuacne rather than with fratrom Atiersio, seuacne can scarcely mean any- thing but the heifers which are to be pursued, even though in other passages it is used of a variety of objects connected with ritual. For a list of these objects as well as for a discussion of the meaning of seuacne see on I la 21. — 2 desenduf: Dev., p. 305, talvcs with fratrom Atiersio, which is easier so far as its position is concerned, but it is improbable that the accusative of ' two ' as a component of 'twelve' has been fossilized as an unin- flected form, and moreover it would not be natural for the number of brothers to be stated at all. It seems better there- fore to understand the whole expression as meaning twelve victims correspond- ing to the (twelve) Atiedian Brothers. reper fratreca: see above in the introduc- tory remarks to the present table. — pars est: only here, but similar to L. par est 4- inf. Bii., p. 118, sets up a dis- tinction whereby pars est is used of that which is in compliance with the law, hertei of that which has been resolved upon (edo^ev). — erom ehiato: the infi- nitive form erom and the sense of the
294
III. TEXT AND TRANSLATION
andersafust . sue . neip . portust . issoc . pusei . subra . screhto . est / (4) fratreci motar . sins . a . ccc /
whole expression are quite clear, but ehiato presents certain difficulties. The stem can be satisfactorily explained by connecting it with L. hio, -are, etc. (see Lexicon and 45 a, 48 a; for the rare transitive use of L. hio ' spew forth ' cf. Val. Fl., VI, 705-6 sabitos ex ore cruores / saucia ligris Mat). The ending is more difficult. Vetter and Bott. take the form as a supine, the -o < -um being adequate- ly supported by anseriato aseriato, but the construction of a supine with erom{ = L. esse) is so unnatural that we are almost forced to take it as a pf. pass. pcpl. The masc. instead of the expected fem. may be explained by the ambiguity of seuacne and pifi and the lack of any distinctly fem. antecedent. The alter- native is to take ehiato as neut. sg., mak- ing a pf. inf. with the participle uninflect- ed. The choice is difficult, but the former of the two alternatives is perhaps the better, erom ehiato here does not, like kuratu eru Va 26, 29, have the charac- teristic perfect meaning, but it is possible that the combination of pass. pcpl. + inf. had to serve both as pres. and as pf. inf.; at least no forms like L. amari, moneri, mitti are known in the dialects. — ponne: used with present hertei of contemporaneous events, in contrast to appei below with fut. pf. andersafust of prior event. — 3 Atiersir: sometimes taken as dat.-abl. pL, as by Buck, § 172 in additions and corrections; by others, including Vet., Pis., Bott., as nom. sg. On grounds of orthography either view has some support, but the second alter- native is preferable since no form of the stem Atiersio- is elsewhere used as a substantive; all stand with a noun, usually a form of frater, and the omission of fratrus here would be strange. Havers, Gl., V (1914), pp. 2-3, proposed taking Atiersir as an adnominal dative depending on arfertur, and indeed if it is dat. pi. it
the people. If he does not bring them just as it is written above, the fine for the brother-superior shall be 300 asses.
is hard to see how it could be construed in any other way, since the adfertor does not perform the lustration 'for' the Atiedii, and if he did we should expect abl. -|- -per. Atiersir then is best taken as nom. sg. in agreement with arfertur. Elsewhere arsfertur is not accompanied by Atiersio-, and its use here may result from a desire to emphasize the private nature of the present ceremony, as Dev., p. 308, suggests, or may imply that other adferiores sometimes performed lustrations but that the present instructions in re- gard to the heifer-chase apply to the lus- tration performed by the Atiedian ad- fertor. — 4 motar: the sense is clearly 'the fine for the fratricus shall be 300 asses, ' but the construction is difficult to classify. Br. and Bii. prior to his edi- tion (p. 119) took motar 'multae' as nom. pi., but it is in all probability gen. sg. A striking analogy is found in C.I.L., XI, 4766 (from Spoleto; about 200B.G.[ ?]) seiquis scies violasit dolo malo, lovei bovid piaclum datod et a. CCC moltai suntod; moltai though ambiguous here is shown to be gen. sg. by the sentence which follows: eius piacli moltaique dica- tore exactio estod. Dative of purpose might be conceivable for the Latin exam- ple but not for the Umbrian. It must therefore belong to the judicial class of genitives. — a. CCC: this abbreviation, stating the amount of the penalty, fol- lows sins with an intervening space nearly large enough for five letters. The size of the fine of 300 asses is, according to Bii., p. 119, an indication of the seriousness of the offense from a religious viewpoint, not of the monetary value of the cattle. But see on Vb 9, where reference is made to Devoto's suggestion that both the price of a dinner there and the fine in the present passage presuppose a date after the reduction of the Roman as to -^ of
24
its original value.
IV. LEXICON AND INDEX VERBORUM
The alphabetical order is in general the same as in Latin, but k is alphabetized with c, f is treated as equivalent to rs, s q follows s s, and u u, v are treated together.
a, abbreviation for L. asses or its U. equi- valent. Nom. pi. Vllb 4; ace. pi. Vb 10, 13, 15, 18.
aanfehtaf, 'unused' (?); according to another interpretation, ' uncooked. ' Pcpl. pass. ace. pi. fern. I la 33-4. From neg. an- -\- *feh-ta-ns. Cf. L. in-fectas. 26 d, 87 a.
abrof apruf , ' boars. ' Ace. pi. apruf lb 24, 33, abrof Vila 3, abrons Vila 43. L. aper. 58 b, 60 e, 67, 90; n. on Vila 43. abrunu, 'young boar' (?). Ace. sg. I la 11. From * apron-, cf. abrof, L. aper. 81 a; n. on loc. cit.
Acesoniam-e, Akeruniam-em, ' Acedo- nia, ' name of a locality in Iguvium. Ace. sg. (+ -em -e) Akefvuiiamem lb 16, Acesoniame VIb 52; loc. Akefunie lb 43, (+ -m) Acersoniem Vila 52. 2i, 46f, 68 e; n. on VIb 52.
acnu, 'years.' Ace. pi. Vb 8, 12, 14, 17. O. akenei, L. annus, Goth, apnam, all from *atno-. 41 f, 74 a; n. on Vb 8.
akrutu: see agre.
adro atru, ' black. ' Ace. pi. neut. atru lb 29, adro Vila 25; abl. adrir Vila 9, 10, 21, adrer Vila 18. L. ater. 3 d, 60 f , 76 a.
afero aferum, ' perform a lustration. ' Inf. aferum lb 10, afero VIb 48; gdve. gen. sg. m. anferener Via 19. For etym. see am- and fertu. 12 a n. 1, 51 e, 60 d, 119 a, 122 c, 128, 153 i; n. on Via 19.
afiktu, 'add to, add upon. ' Imv. la 31. For etym. see an-, fiktu, and n. on la 28.
agre, ' field, country. ' Gen. sg. agre Vb 9, 14; abl. (+ -tu) akrutu Va 9 (?). ager nom. sg. in minor U. inscr. no. 355 Co., L. ager, dygog, Skt. ajras, Eng. acre. 32 d, 45, 47 a, 55 a, 57 h, 76 a, 90, 142 a; n. on Va 9.
ahatripursata ahtrepufatu, ' dance the tripudium. ' Imv. ahtrepufatu I la 24, 25, 31, 38, atrepufatu lib 18; ahatripursata Vila 23, 36, atripursatu VIb 16, atropusatu VIb 36. From a- 4- * tripuddtod. Cf. L. tripudium. 10 c, 28 c, 42 b, 117 c; n. on VIb 16.
ahauendu, 'turn aside, avert. ' Imv. Vila 27. From a- -\- *wendhetdd. Cf. preuen- du. Germ, wenden. 60 n. 2, 119 a.
ahesnes, 'of bronze.' Abl. pi. Ill 18, 19, 19. L. aenus, ahenus, *ayes-no-, cf. L. aes, Germ. Eisen, Eng. iron. 33 a, 74 b.
ahtim-em, ahtis-per, of uncertain mean- ing, possibly ' fire-carrier. ' Ace. sg. (+ -em) lb 12, 12; abl. pi. (+ -per) III 24, 29. From *akti-, cf. L. ago, actio (?). 46 i, 77 b; n. on lb 12, III 24.
Ahtu, of uncertain meaning, perhaps name of a god Ahtus. Dat. sg. Ila 10, 11. Possibly connected with L. aio, ad- agium, prod-igium, etc. 25 f, 46 i, 78 b, 90; n. on Ila 10-14.
aitu, 'move, carry' (?). Imv. aitu lb 29, 37, aitu VIb 18, 18, Vila 40, 45; imv. 3 pi. aituta III 13. O. actud, inf. acum (with analogical extension of the c which comes from g before t), L. ago, ayco. 46 j, 119 a; n. on III 13, VIb 18.
aiu, '(oracular) response' (?). Nom. sg. Ila 4. Etym. uncertain; possibly from *ag-yom, cf. L. prod-igium, aio < *agy6. 68 e.
alfir, alfu, 'white.' Ace. pi. neut. alfu lb 29; abl. alfir Vila 25, 26, alfer Vila 32, 34. L. albus, dX(povQ • Xevkovq He- sych. 7 b, 40, 55 d, 67.
am-, prefix ' around, about. ' 134 b.
amboltu, 'walk about.' Imv. VIb 52. Cf. L. ambulo. 60 d, 116 a, 119 a, 152 f.
ombre tuto, amprehtu, ' make the circuit. ' Imv. amprelitu lb 21, apretu lb 20;
296
IV. LEXICON AND INDEX VERBORUM
imv. 3 pi. ambretuto VIb 56, 63, 64; fut. pf. 2 sg. amprefuus lb 20; 3 pi. ambrefurent VIb 56. From ambr- ' around, about' + etu, etc. 26 c, 60 d, 122 b, 124 d, 125, 134 c, 138 a.
amparitu, act. ' set up, ' pass. ' stand up. ' Imv. amparitu 11114; imv. pass. amparihmu I la 42. Cf. L. paro, pario (?). 11 a, 120 a, lEO d.
ampentu, 'offer up, slaughter.' Imv. ampentu Ila 20, III 23, ampetu lib 10, 11, apentu III 27; fut. 2 sg. an- penes lib 27; fut. pf. 2 sg. apelus lib 27, 3 sg. apelust Va 17. From an- + pend-, cf. L. impcndo. 3 d, 29 m, 51 d, 62 d, f, 61 e, 119 a, 123, 124 e, 151 c; n. on lla 20.
ampefia, of uncertain meaning, possibly 'part around the foot.' Abl. sg. Ila 29. From *am(J)-ped-iad, cf. L. amb- + pes (?). 68 e.
amprehtu, amprefuus: see ambretuto.
an-, prefix 'up, upon. ' 134 d.
ancla, anclar: see angla.
andendu antentu, 'put on, place upon. ' Imv. antentu Ila 20, III 15, 16, 17, 17, 22, IV 21, 27, atentu lib 28, andendu Vila 25. From an- + tend-, cf. L. iendo. 2 i, 3 d, 119 a with n. 3.
andersafust atefafust, 'make the cir- cuit. ' Fut. pf . 3 sg. atefafust lb 40, andersafust Vllb 3, andirsafust Vila 46. From am- + root of dirstu teftu, L. do. 42 b, 52 d, 117 b, 122 c, 124 d, 125.
andersistu, 'sit in the way.' Imv. ander- sistu Via 6; fut. pf. 3 sg. andersesust Via 7. From ander- + root of sersitu, L. sedeo, sldo. 29 b, 60 a, 152 f.
anderuacose, 'interruption' + enclitic vb. = L. sit. Nom. sg. contained in antervakaze lb 8, anderuacose VIb 47. Probably from * anter-wakat{i)s sit, cf. L. inter- + vacatio. 60 a, 77 b, 86 c.
anderuomu VIb 41, of uncertain meaning, possibly 'interval, ' as abl. sg. of *ander- womo-, with *wom- from *lom- 'break, ' cf. O. lamatir ' be beaten, ' Russ. lo-
matb 'break.' 55 a, 60 a, 67; n. on
loc. cit. andirsafust: see andersafust. anferener: see afero. angla, '(divine) messengers.' Nom. pi.
anclar Via 16; ace. angla Via 1, 3, 5,
6, b 49, anglaf Via 5, ancla Via 18.
From *an-kla-, cf. L. cld-mo, cld-rus.
18 a, 60 c, 65 a, 66. anglom-e, 'angle.' Ace. sg. (+ -e) an-
glome Via 9; abl. sg. (+ -to) angluto
Via 8, 10, 10. L. angulus. 60 c, 75 a.
anhostatir, 'unarmed, not under arms.' Dat. pi. anhostatir Vila 28, 50. anos- tatir VIb 62, Vila 13, 15; ace. pi. anhostatu VIb 60, anostatu Vila 48. From neg. an- + hostatir, q.v. 2 1, 17 e, 71, 87 a.
anostatir, anostatu: see anhostatir.
anouihimu, 'wear, put on, hold (clothing, etc.). ' Imv. pass. VIb 49, 49. Cf. L. ind-uo, ind-uor, Lith. aviii 'wear, put on (shoes). ' 11 a, 134 d, 150 e.
anpenes : see ampentu.
anseriato anzeriatu, 'observe.' Pres. subj. 1 sg. aseriaia Via 2; imv. azeria- tu lb 8, aseriatu VIb 47; imv. pres. 2 sg. aserio Via 4; supine anzeriatu lb 10, anseriato Via 6, aseriato Via 1, 6, b 48; pcpl. pass. abl. pi. anzeriates la 1, Ila 17, aseriater Via 1. From an- + serwid-, cf. L, servire, servare. 2 i, 14 b, 25 a, 51 e, 54 g, 57 e, 58 a, 115 a, 117 e, 126 a, 127 a, 129, 153 d.
ansihitu: see ansihitir.
anstintu, ' fasten (with nails) ' ( ?). Imv, anstintu III 20, astintu III 18, 19. Cf. L. in-stigare, di-stingere, oril,(o, ariy- 117]. 134 d; n. on III 18.
anstiplatu, 'demand.' Imv. Via 3. From an- + stiplatu. 58 a, 117 c, 134 d.
angif, 'turnings.' Ace. pi. Ila 25. From *anki-, cf. L. uncus, angulus, dyxvXoq, dyacbv. 35 d, 77 a.
ansihitir, 'not in office' (?). Dat. pi. aniihitir VIb 62, Vila 13, 14, 28, 50;
axxiparitu
297
Afmune
ace. pi. m. ankihitu VIb 59, ansihitu Vila 48. From neg. an- + Mhitir. 87 a.
antakres, 'whole, unground. ' Abl. pi. antakres Ila 42, antakre lb 36, 38. L. integer, ef. tango. 17 e, 57 h, 76 a, 86 a; n. on lb 36.
antentu: see andendu.
antermenzaru, 'between moons, inter- lunary. ' Gen. pi. Ila 16. From ander- + mens-d-, cf. L. mensis, Gk. fiijv. 58 a, 66, 86 a.
antervakaze: see anderuacose.
anzeriates, anzeriatu: see anseriato.
ape, conjunction 'when.' ape lb 34, Ila 9, b 27, 28, Va 17, 18, 20, 22, api la 27, 30, 33, ap III 20, IV 31, ape VIb 5, 16, 23, 37, 49, 52, 52, 56, 62, 63, 63, 64, Vila 5, 8, 39, 42, 43, appei Vllb 3. From *at-k'"e: L. atque. 31 b, 49 a, 61 a, 112 g, 113 1, 154 b.
apehtre, 'apart.' IV 15. From *ab- ektred, cf. O. ehtrad, L. ab + extra, ane^. 26 c, 46 i, 76 c, 112 c, 113 d.
apelus, apelust, apentu: see ampentu.
aplenia, 'sufficient' (?). Ace. pi. aple- nia Ila 23; abl. pi. aplenies Ila 23. For etym. see n. on loc. cit. 68 e.
appei : see ape.
apretu: see ambretuto.
apruf : see abrof.
arfertur: see arsfertur.
armamu: see arsmahamo.
arnipo, conjunction 'until.' VIb 25, 41. From ar- (= ax-, L. ad-) + neg. ne + pon2 (L. quom, cum) or pod (L. quod). 112 g, 154 b.
-af -a, postposition ' to, toward. ' asam- af IV 6, asam-a Ila 39, IV 16; ereglum-a III 35, IV 3, 10, ereglum- af (written ereglamaf ) IV 6 ; etram-a III 34; persklura-ar III 21 ; spantim- ar III 33; spinam-ar Ila 33, spiniam- a Ila 37. L. ad, ar- (in arbiter, etc.), cf. O. az. 4, 147 a.
afkani, ' musical accompaniment. ' Ace. sg. IV28. Probably from *ad-kaniom, cf. L. ac-cino, ac-centus. 42 c, 68 a, 86 c.
afepes, noun 'fat' (?). Abl. pi. afepes la 6, 10, 19, 23, b 4, afeper lb 30, 33, afepe lb 26, 44, Ila 7, afipes lb 7, afiper la 27, afpes la 13. L. adipeis (?). 36 b, 42 c, 68 b; n. on la 3.
arsfertur arfertur, 'adfertor,' the chief priest in the cult of the Atiedian Bro- thers. Nom. sg. arfertur lb 41, Ila 16, Va3,10, arsfertur Via 8, arfertur Via 3, Vllb 3; dat. afferture Vb 3, 5, 6, arsferture Via 2; ace. arsferturo Via 17. From *ad-bher-tor, cf. L. af-fero, arferia (of dialectal origin ?)• aqua quae infer is libabatur, Paul, e Fest., p. 10 Li. 13 b, 42 c, 55 c, 82 a, 86 d,90, 91 c.
arhabas, ' use. ' In neif habas = nei + arhabas, pres. subj. 3 pi., cf. L. adhibeant, 34, 152 b; n. on IV 33.
arsier, noun of uncertain meaning, pos- sibly 'consecration, formula of conse- cration.' Gen. sg. arsier Via 24, b 27, asier VIb 8; loc. (?) sg. arsie Via 24, b 8, 27. Probably related to arsmor, arsmahamo, arsmatia. No sure con- nections outside Umbrian. 68 e; n. on Via 24.
arsir, 'other' according to the view pre- ferred here. Nom. sg. m. Via 6, 7. From *alyos: L. alius, dXXoQ. According to another view, with almost equally good support, ' dedicationibus ' as abl. pi. of arsier. 32 a, 91 b; n. on Via 6.
arsmahamo, ' arrange in priestly ranks' ( ?). Imv. pass. 2 pi. armamu (written ar- manu) lb 19, arsmahamo VIb 56. From
* ad-ma.-, cf. arsmor, arsmatia. No sure connections outside Umbrian. 73, 117 c, 150 e.
arsmatia, adj. 'ritual. ' Ace. sg. arsmatiam VIb 49, 50, arsmatia Via 19, b 53, 63, Vila 46, 51. 68 e, 73.
arsmor, 'rites; priesthoods' (?). Nom.pl. arsmor Via 26, 36, 46, b 29; ace. pi. arsmo Via 30, 32, 39, 42, 52, b 13, 32, 34, Vila 17, 30, asmo Via 49. From
* ad-mo-, cf. arsmahamo, arsmatia. No sure connections outside Umbrian. 73, 103 a.
Af raune, an epithet of Jupiter. Dat. sg.
298
IV. LEXICON AND INDEX VERBORUM
lib 7. Probably from *ad-mdn-, cf. arsmor. 74 f; n. on loc. cit.
afpeltu, 'approach.' Imv. Ila 32, b 19, IV 8. Probably from *ad-k'"el-{e)tod, cf. niXofiai, L. colo. According to another view cognate with L. adpello, -ere. 42 c, 119 a; n. on Ila 32.
arpes: see afepes.
afputrati, 'discretion.' Abl. sg. Va 12. L. arbitratu; root and vocalism un- certain. 15 a, 39, 42 c, 78 b, 86 c.
arsueitu afveitu, 'add.' Imv. afveitu Ila 12, 29, b 13, III 34, IV 5, arveitu lb 6, aveitu IV 1, arsueitu Via 56, 59, b 2, 5, 20, 44, 46, Vila 4, 8, 42, 54, arueitu VIb 23. L. advehito, cf. kuveitu 2 i, 46 i, 119 a.
arglataf , ' circlets ' or ' crescents, ' name of a type of ritual cakes of a circular or semicircular shape. Ace. pi. IV
22. From *arkelata,- cf. L. arcus, arculum, arculata. 6, 29 e, 46 c, 71, 75 a.
arvam-en, 'field.' Ace. sg. (+ -en) arvamen III 11; loc. (-|- -en partly absorbed) arven III 13. Cf. aruio, L. aro, arvum. 69.
arueitu arveitu: see arsueitu.
aruio arvia, 'fruits of the field, grain '(?). Ace. pi. arvia la 3, 9, 26, b 3, 6, 43, Ila 18, 24, aruvia III 31, arviu la 12, 16, 23, b 25, 28, 32, Ila 6, 11, 12, b 8, 29, aruio Via 56, 58, b 1, 3, 20, 22, 44, 45, Vila 4, 7, 42, 53; abl. arves la 6, 10, 13, 19, 23, b 4, 26, 30, 33, 44, lla 7, arvis la 27, b 7. Probably related to arvam(-en), L.. arvum, as dyQiogiodyQOQ. 2 h, 54 c, 68 e; n. on la 3, Ila 18.
asa, 'altar.' Dat. (or loc?) sg. ase Ila 19, III 22; ace. (-1- -ar) asamar IV 6, (+ -a) asama Ila 39, IV 16, ( + -e) asame Via 10; abl. asa Ila 38, III
23, IV 16, asa Via 9, (+ -ku) asaku Ila 39, 43. L. ara, O. loc. sg. aasai, cf. possibly also Hitt. hassaS ' hearth. ' 7 a, 57 d, 66, 90, 145 a.
aseriaia, aseriater, aseriato aseriatu, aserio : see anseriato.
asegeta, ' uncut. ' Abl. sg. f . asegeta
Ila 29; abl. pi. f. asegetes IV 7.
Neg. a- + pass. pcpl. segeto/a-, cf.
prusekatu, pcpl. dat. sg. prusegete,
L. secare, sectus. 87 a. asiane, noun of uncertain meaning, ' rear
of altar' (?). Loc. sg. la 25. Cf. asa
(?). 74 d; n. on loc. cit. asier: see arsier. asmo : see arsmor. asnata, adj. 'dry.' Ace. pi. neut. as-
nata Ila 19, asnatu Ila 34; abl. pi.
asnates Ila 37, IV 9. Neg. a- + snata.
87 a; n. on Ila 19. aso, pcpl. 'lighted' (?). VIb 50. L.
assum, arsum, pcpls. to ardeo (?).
129; n. on loc. cit.
Agetus, name of certain deities of un- certain identity. Dat. pi. Ila 14. Per- haps identical with L. dis Ancitibus. 80 b; n. on loc. cit.
atentu : see andendu.
atero, noun 'evil' (?). Ace. sg. Vila 11, 27. Etym. uncertain; see n, on loc. cit.
atefafust: see andersafust.
Atiiefiate, name of a decuvia among the citizens of Iguvium. Dat. sg. lib 2, 2. Related to Atiersiur Atiiefiur as L. Arpinas, -atis to Arpinum. 77 c.
Atiersiur Atiiefiur, adj. ' Atiedian, ' name of the priestly brotherhood at Iguvium with whose activities the Tables are concerned. Nom. sg. (?) Atiersir Vllb 3; dat. Atiiefie Ila 1, 3; loc. Atiiefie Va 16; nom. pi. Atiiefiur Va 1, 14, Atiersiur Yh 11, 16; gen. Atiiefiu Ila 21,35, b 26, Va 12,25,27, b 4, Atiersio VI lb 2; dat. Atiersir Vb 8, 14, Atiersier Vllb 1; abl. Atiiefies III 24, Atiiefie Ila 2, III 29, Atiiefier Va 4, 16. Probably related to Attiggio, name of a village south of the modern town of Fabriano. 2 h, 13 a, 33 c, 57 e, 68 e, 90.
atrepufatu atripursatu atropusatu: see ahalripursatu.
atru : see adro.
afpeltu
299
Casiler
aveitu: see arsueitu.
auie, 'augury.' Loc. sg. VIb 11. From *awi-ye-, cf. auif, L. avis. 25 d, 79, 90, 93 e.
aviekate, pcpl, pf. pass, (or pres. act. ?) ' taking auspices. ' Dat. sg. Ila 1, 3. From denominative vb. *awie-kd-, cf. auie, auiecla, auif, L. avis. 70, 79, 117 e.
auiecla, ' augural. ' Abl. sg. f. aviekla lb 14, auiecla VIb 52; ace. pi. m. auiehclu Via 10, auieclu VIb 51, (+ -e) avieklufe lb 14; abl. auiehcleir Via 9, auieclir Via 12, 13. Probably from *awie-k(o)-lo-, cf. aviekate, auie, auif, L. avis. 2 h, 79.
auif avif, 'birds.' Ace. pi. avif lb 8, avef lb 10, auif VIb 47, 48, 48, aueif Via 4, 18, auuei Via 3; abl. avis Ila 16, aves la 1, aueis Via 1. L. avis, Vedic vi-, nom. pi. vayas, 'bird.' 6, 26 e, 29 k, 54 b, 61 d, 77 a, 90.
auirseto, 'unseen.' Nom. sg. neut. Via 28, 38, 48, b 30. Neg. a- + uirseto. 10 c, 30 c, 87 a.
auuei: see auif.
azeriatu: see anseriato.
benus, 'come.' Fut. 2 sg. menes lb 15; fut, pf. 2sg. benus lib 16; 3 sg. benust VIb 53; 3 pi. benurent Va 25, 28, b 5, benurent VIb 57; fut. pf. pass. 3 sg. benuso VIb 64, 65, Vila 2. From *g^em-, cf. O. ce-bnust 'venerit,' L. venio, ^aivco. 2 k, 49 b, 114 e, 115 g, 123, 124 b, 151 c.
berva, ' (roasting-) spits. ' Ace. pi. berva Ila 26, 33; abl. berus Ila 23, 35. From *g^eru-: L. veru, Olr. bir 'spit', Goth, qairu, gloss on hnupo' oKoXoip. 29 k, 49 b, 61 d, 78 a, 90.
bue, buf, etc., 'ox.' Ace. sg. bum Ila 5; abl. bue Via 25, 28, 33, 35, 38, 43, 45, 48, 53; gen. pi. buo Via 54; ace. buf la 3, 11, 20, buf Via 22, b 1, 19. From *g^6u-, L. (of dialectal origin) bos, ^ovq, Skt. gaus. 25 e, 49 b, 65 c, 102 e.
CCC, abbreviation for equivalent of L. trecenti 'three hundred. ' VIIb4. 104 a.
K, abbreviation for praenomen possibly equivalent to L. Gaius. Gen.sg. Va 15.
cabriner, 'of goat, goat(-meat). ' Gen. sg. Vb 12, 17. L. caprinus, cf. kabru. 11 a, 46 a, 52 a, 60 e, 74 e.
kabru, '(he-)goat. ' Gen. sg. kapres lib 12; ace. kabru lib 17, kapru lib 10, kaprum lib 1. L. caper, ON hafr. 2 b, 3 d, 6, 20 a, 46 a, 60 e, 67.
calersu kalefuf, adj. 'with white fore- head, ' as epithet of oxen. Ace. pi. m. kalefuf la 20, calersu VIb 19. L. calidus in Isid.,Orzg'.,XII, 1, 52 equi qui albam frontem habent, calidi appellantur. 42 b, 72.
kanetu, 'sing, play (music).' Imv. IV
29. Cf. L. cano. 30 b, 46 a, 118 b.
capirse kapife, 'bowl.' Dat. sg. la 29,
32, Ila 8, capirse VIb 24, 37; ace. capirso VIb 25; abl. kapife Ila 34, 41; ace. pi. capif VIb 18, Vila 39, 45, kapif la 18, kapi lb 29, 37; abl. kapifus Ila 33, IV 5. L. capis, -idis. 10 c, 31 e, 42 b, 61 b, 80 c, 90, 102 d.
kapres, kapru: see kabru.
karne, karnus: see karu.
carsitu kaf etu, ' call. ' Imv. kaf etu lb
33, kafitu III 21, carsitu Via 17, Vila 43. L. calare, xaXica. 2 a, 9 a, 55 b, 114 e, 118 a.
carsom-e, noun of uncertain meaning, pos- sibly 'street.' Ace. sg. (+ -e) Via 13, 14. Possibly related to L. cursus, Ital. corso. 67; n. on Via 13.
kartu, ' roast ' ( ?). Imv. Ha 23. No sure etym. 119 a; n. on loc. cii.
karu, 'portion, section,' (with mestru) 'majority.' Nom. sg. karu Va 24, 27, b 4; dat. karne Ha 1, 3; abl. karne Ila 30; abl. pi. karnus IV 7. L. caro, cf. Skt. kfntati 'cut,' OHG sceran 'cut.' 17 b, 58 f, 81 a, 90, 91 c; on Ila 1.
Casiler, name of a field. Gen. sg. Vb 14. Cf. Casilos, and Casilos as name of magistrate on a Lepontic gold coin
300
IV. LEXICON AND INDEX VERBORUM
from the Rhone Valley, Whatmough, Prae-Italic Dialects, II, no. 327.
Casilos, Kaselate, name of a decuvia among the citizens of Iguvium. Nom. sg. Casilos Vb 13; dat. Kaselate lib 6, 6, 6; Casilate Vb 16. Cf. Casiler. 7 h, 29 1, 77 c, 90, 91 d.
castruo kastruvu ' heads, lives ' (?); ac- cording to another view, ' estates. ' Ace. pi. kastruvuf Va 13, 18, kastruvu Va 20, 22, castnio Via 30, 32, 40, 42, 50, 52, b 13, 32, 34, Vila 17, 30. Cf. O. castrous, castrid. L. castrum, -a, and see Buck, p. 236, n. 1, which I follow for the most part despite the change of view in his Additions and Corrections. 2 h, 44 c, 54 b, 78 a, 103 a.
Kastrugiie, a gentile name ' Castrucius. ' Gen. sg. Va 3. Cf. L. Castrucius, O. Kastrikiieis (Co. no. 63, from Pom- peii). 68 e.
katel, 'puppy, dog.' Nom. sg. katel I la 43; gen. katles Ila 22, 27; dat. (?) katle Ila 15; ace. katlu Ila 18, 20, 29. L. catulus. 29 e, i, 75 a, 90, 91 b.
caterahamo kateramu, ' arrange in mil- itary ranlcs ' (?). Imv. pass. 2 pi. kateramu lb 20, caterahamo VIb 56. L. caterva. 54 g, 114 e, 117 c, 150 e.
kazi, noun of very uncertain meaning, possibly 'kettle.' Ace. sg. Ill 16, 18. Cf. L. catinus (?). See n. on III 16.
kebu, 'food' (?). Abl. sg. IV 23. L. cibum (?). 39, 46 f, 67.
cehefi, Via 20, of uncertain morphology and meaning. According to the view adopted here, ' flame, blaze, ' as ace. sg. from *kefiom < *kensiom > *kent- tiyom < *knd-tiijom, cf. L. candeo, incendium. 46 f, 124 d; n. on loc. cit.
Clauerniur, Klaverniie, name of a de- cuvia among the citizens of Iguvium. Nom. pi. Clauerniur Vb 8; dat. pi. Klaverniie lib 3, 3, Clauerni Vb 10. 13 b, 68 e.
klavlaf, according to the view adopted here ' smearing-sticks ' ; according to
others 'haunches' or 'spoons.' Ace. pi. klavlaf I la 33; abl. klavles I la 36, IV 11. L. clava, clavula (?). 75 a; n. on Ila 33.
kletra, 'litter.' Ace. sg. kletram III 13; abl. kletra III 13, IV 24; loc. Metre III 14. Goth, hleipra 'gk7]v^,' cf. L. clitellae. 20 b, 76 b.
Kluviier, a gentile name ' Cluvius. ' Gen. sg. Va 15. 68 e.
com, 'with, along with.' VIb 52, 55, 56, 57. O. com, L. cum, com-. 4, 46 a.
-com -co -kum -ku, postposition 'at, by; with.' asa-ku Ila 39, 43; testru-ku la 29, destru-co VIb 24, 38; eru-com VIb 50, eru-ku III 31; esu-ku IV 29, esis-co, Via 18; esunes-ku Va 11 nertru-ku la 32, nertru-co VIb 37, 39 persi-co VIb 25; pesondris-co VIb 40 termnu-co VIb 53, 55, 57, termnes-ku lb 19; ueris-co Via 19, 20, 21, b 23, 44, 46, Vila 42, 53; vuku-kum lb 1,4, uocu-com VIb 43,45; ura-ku Va 5; uvi-kum III 28. Etymology same as for com. 46 a. 51 e, 106 b, 147 b, 155 g.
comatir: see comoltu.
combifiatu kumpifiatu, 'announce, de- liver instructions.' Pres. subj. 2 (or 3?) sg. kupifiaia lb 35; imv. kumpi- fiatu lb 14, kupifiatu lb 35, combifiatu Via 17, b 48, 51, Vila 43, 44; fut. pf. 3 sg. combificmsiust VIb 49, combifian- sust Vila 5, combifiansiust VIb 52; pf. subj. 3 sg. combifiansi VIb 52. Probably from denominative stem *kom- bheidh-iya-, cf. neidco, L. fido. 2 b, 3 d, 43, 51 d, 60 d, 114 e, 117 e, 124 f , 125, 126 a, c.
comohota, pcpl. pf. pass. ' moved for- ward, brought forward. ' Abl. sg. f. Via 54. L. commoveo, -motus. 24 b.
comoltu kumultu, 'grind.' Imv. ku- multu la 34, kumaltu Ila 9, 41, IV 28, comoltu VIb 17, 41, Vila 39,44, 45; pcpl. pass. abl. pi. kumates la 34 (written kumats), Ila 42, IV 29, kumate lb 37, 38, Ila 10, comatir VIb 17, 41, 41, Vila 39, 44, 45. L. com-molo, cf. Skt. m^nati 'crush.
Cusilos
301
kuvertu
18 a, 28 b, 35 d, 55 d, 57 h, 90, 114 e, 119 a. conegos kunikaz, pcpl. pf. pass, 'kneel- ing.' Nom. sg. m. conegos VIb 5, 16, Vila 37, kunikaz IV 15, 18, 20. Pro- bably from *ko(n)-(g)niga-, cf. L. genu, gnixus in Paul, e Fest., p. 216 Li. gnitor et gnixus a genibus prisci dixerunt.
2 d, 3 d, 7 b, 29 i, 47 d, 114 e, 116 a, 117 b, 131 ; n. on VIb 5.
Coredier Kureties, either a divine name or a gentile name ' Coredius. ' Gen. sg. Kureties lb 4, Coredier VIb 45.
3 d, 42 d, 68 e; n. on VIb 45.
couertu kuvertu, ' return. ' Imv. kuvertu lb 9, 36, 38, Ila 39, couerlu VIb 47, Vila 44, 45; fut. pf. 2 sg. kuvurtus lb 11; 3 sg. couortus Vila 39, courtust Via 6; fut. pass. 3 sg. couorluso VIb 64. L. converto, convertito, etc. 2 a, 12 d, 17 a, 29 b, 35 d, 114 e, 115 g, 119 a, 124 b.
Crabouie, Krapuvi: see Grabouie.
krematra, 'roasting-pans' (?). Ace. pi. Ila 23. From *krema-tro-, cf. L. cremo, -are. 76 b.
krematru, 'fat-drippings' (?). Ace. pi. kreinatruf Ila 26, krematru Ila 28. Etym. same as for krematra. 76 b; n. on Ila 26.
cringatro krikatru krenkatrum, noun 'stole' (?). Ace. sg. krenkatrum lb 11, krikatru lib 27, 29, cringatro VIb 49. Cf. OE hring, OCS krogu ' circle. 2i,3d,8d, 60d,f, 76 b; n. on VIb 49.
-ku: see -co.
kukehes, vb. of uncertain meaning, pos- sibly 'arrive at, reach.' Fut. 2 sg. Ill 21. From *kon-keh-es-es, cf. O. kahad, L. in-cohare (?). 46 f; n. on loc. cit.
kvestretie, 'quaestorship.' Loc. sg. lb
45, II 44. Cf. kvestur, L. quaestor.
8 a, 68 e. kvestur, 'quaestor.' Nom. sg. Va 23,
b 2. Probably borrowed from L.
quaestor. 46 h, 82 a, 91 c. kumaltu, kumate(s): see comoltu.
kumiaf: see gomia.
kumnahkle, 'assembly, meeting-place.' Loc. sg. kumnahkle Va 15, kumnakle III 7, 8. From *kom-nd-tlo-, cf. kumne. 75 b, 86 f.
kumne, a place in Iguvium, probably 'Comitium. Loc. sg. lb 41. From *kom-nei: O. comenei. 51 c, 74 a, 86 f .
kumpifiatu kupifiatu, kupifiaia: see
combifiatu.
kumultu: see comoltu.
kunikaz: see conegos.
kuratu, 'look after, provide. ' Pres. subj. 3 sg. kuraia Va 5 ; pcpl. pf . pass. neut. nom. sg. kuratu Va 24, 26, 29. L. euro, -are, Pael. pf. 3 pi. coisatens, de- nominatives to L. cura and its Pael. equivalent. No sure cognates outside Itahc. 7a, 21 c, 53 b, 57 c, 114 e, 117 c, 126 a, 131, 149.
Kureiate, name of a decuvia among the citizens of Iguvium. Dat. sg. lib 3, 3. 77 c.
Kureties: see Coredier.
curnaco, noun ' crow. ' Ace. sg. curnaco Via 2, 4, 15, 17; abl. curnase Via 1. Cf. L. cornix, with variation of the vowel in the second syllable. 10 b, 12 b, 46 b, 80 a, 90,
kurglasiu, of uncertain meaning, pro- bably ' circular, full ' (of the moon). Abl. sg. Ila 17. From */cf- or *kor-keld- sios, cf. L. circularis (with ablaut-var- iation in the root and a different stem- suffix). 35 d, 46 c, 53 e, 57 d, 68 d, 75 c; n. on loc. cit.
kutef , pres. pcpl. ' speaking in a low tone, murmuring. ' kutef la 6, 10, 13, 19, 23, b 7, kutep lb 3. Possibly from *kautens, pres. pcpl. of *kautey6, de- nominative vb. from pf. pass. pcpl. corresponding to L. cautus. 2 k, 22, 58 d, 114 e, 118 c, 130, 139 b.
kuveitu, 'convey, carry.' Imv. Ila 32, 40. L. convehito, cf. arsueitu arveitu. 46 i, 119 a.
kuvertu kuvurtus: see couertu.
302
IV. LEXICON AND INDEX VERBORUM
daetom, pcpl. pass. nom. sg. neut. used as noun 'delinquency.' Via 28, 37, 47, b 30. From da- (O. dat, cf. O. dadid 'dediderit' and, more remotely, L. de) + *itom (or *eitom1) pcpl. pass, of eiu, cf. peretom. 30 c, 114 e, 122 b, 131, 149.
Dei: see Di.
deiiu teitu, ' say, declare, pronounce, designate. ' Imv. teitu Ila 26, b 7, 25, III 9, 25, deitu VIb 56, 63, 64, 64, 65, Vila 1, 20, 51; fut. pf. 3 sg. dersicust VIb 63; 3 pi. dersicurent VIb 62. O. deicum deikum, L. dlco. 3 d, 35 d, 46 j, 52 g, 119 a, 124 a, 125.
dequrier tekuries, ' decurial, pertain- ing to the decuviae. ' Abl. pi. tekuries lib 1, dequrier Vb 11, 16. From *dek- ur-yo-, cf. L. decuria, O. dekkviarim, L. decern. 2 a, 3 d, 45, 68 e, 90, 105 h; n. on lib 1.
dersa: see dirstu.
dersecor, adj. 'due.' Nom. pi. Via 26, 36, 46, b 29. Probably from *dedeko-, reduplicated stem from root contained in ti§it, L. decet. 67, 103 a.
dersicurent, dersicust: see deitu.
dersua tesvam, adj. of uncertain meaning, possibly 'western, in the west.' Ace. sg. f. tesvam lb 13, dersua Via 2, 2, 4, 4, 15, 15, 17, 17, desua VIb 51, 52, 52; abl. sg. f. dersua Via 1. Possibly rela- ted to L. dorsum. 3 d, 54 c, 69; n. on Via 1.
desenduf, 'twelve.' Ace. pi. Vllb 2. Cf. L. duodecim, with the components in reverse order. 17 d, 42 a, 46 b, 105 i.
desire testre, adj. 'right' (of position). Ace. sg. f. (-h -e) destrame VIb 49; abl. m. testru III 23, IV 15; (+ -ku) testruku la 29, destruco VIb 24, 38; loc. m. testre lib 27, 28, destre VIb 50; loc. f. destre VIb 4. L. dexter, dextra, de^ireQdQ, cf. O. destrst = L. dextra est. 2 a, 3 d, 8 a, 29 a, 42 a, 61 d, 76 c, 90.
deueia, ' of the god(s), divina. ' Ace. sg. deueia Via 10; abl. deueia Via 9.
Probably from *deiwiyd-, cf. O. Dei- vai, deivinais, L. d'ivus, Skt. devas. 68 a. Di, name of the god corresponding to L. Jupiter, Gk. Zevg. Ace. sg. Dei Via 23, 24, 25, 27; voc. Di Via 25, 28, 29, 29, 31, 31, 33, 34, 35, 35, 37, 38, 38, 39, 40, 41, 43, 44, 45, 45, 47, 48, 48, 49, 50, 51, 53, 54, 55, Dei Via 26. From diphthongal stem *d(i)yeu- : lu(pater), O. iiiveis, Diiivei, L. lu- (piter), lovis, OL. Diovis. More remote cognates semantically are L. dies, and morphologically L. deus, dlvus, etc. 42 a, 65 c.
dia, ' give, cause, produce ' ( ?). Pres. subj. 3 sg. Via 20. Probably from *dwiyat from *dwl- ablaut-variant of *dou- 'give' -f I, cf. purdouitu, pur- dito, L. subj. 3 pi. duint. 15 b, 54 e, 120 a, 126 a.
difue, 'folded double.' Ace. sg. VIb 4. From *dwi-fwim or *dwi-fuyom with development as in 32 a; for first com- ponent cf. di-, L. bi-, Skt. dvi-, for second cf. U. futu, L. fui, for compound cf. di-q)vrjz. 54 e, 88, 105 b.
dirstu teftu, ditu tetu, 'give, distribute. ' Pres. ind. pass. 3 sg. tefte Va 7; pres. subj. act. 3 sg. tera lb 34, 35, 36, dirsa Vb 13, dersa Vila 43, 44, 44; 3 pi. dirsans Vb 11, 16, dirsas Vb 8; imv. teftu Ila 40, 40, tertu IV 28, dirstu VIb 17, 38, 38, 39, Vila 5; imv. tetu Ila 9, b 21, titu la 33, ditu VIb 10, 16, 25, Vila 38; fut. pf. 3 sg. tefust lb 34, dirsust Vila 43. All these forms are based on reduplicated stems, cf. diScofii, Skt. daddmi, O. fut. didest; cf. also U. dia, pur-douitu, L. do, dare, etc. 2 a, 3 d, 10 c, 29 b, c, 35 d, 42 a, b, c, 52 d, 55 c, 58 g, 114 e, 116 e, 119 b, 123, 124 a, 126 a.
disleralinsust, 'become ineffective' or 'make ineffective.' Fut. pf. 3 sg. Via 7. Possibly denominative to *dis-leisdli- (gradation of root uncertain), cf. L. dis- and lira, de-llrus, de-llrdre. Germ. Geleise. 57 b, 77 h, li^ e, 120 b, 124 f; n. on loc. cit.
daetom
303
eikvases-e
ditu: see dirstu.
dupla, tupler, 'two.' Ace. pi. f. dupla VIb 18, 18; abl. pi. m. tupler Va 19. L. duplus, cf. U. tuplak, L. duplex. 3 d, 38 a, 55 c, 88, 105 b.
dupursus, 'bipeds, men.' Dat. pi. VIb 10. From *du-pod-ufs (quantity of root-vowel uncertain), cf. peturpursus, L. bipes, dinovQ. 13 c, 35 d, 42 b, 65 f, 80 c, 88, 105 b.
dur, tuf, 'two.' Nom. pi. dur VIb 50, Vila 46; dat. duir Vb 10, 15; ace. pi. tuf lb 41; ace. pi. neut. tuva Ila 27, III 32, 34; abl. tuves III 19, (+ -e) tuvere Ila 33. L. duo, dvo, Eng. two, etc. ; cf. also U. duti. 2 h, 3 d, 33 c, 42 a, 54 b, 105 b, 155 h.
duti, 'a second time.' VIb 63. From *du-tim < *du-tyom. 105 b, 112 b, 113 a.
-e -e -en -em -n -m -me -i, postposition. With ace. 'into, to, upon.' Akefuniam- em lb 16, Acesoniam-e VIb 52; ahtim- em lb 12, 12; anglom-e Via 9; arvam- en III 11 ; asam-e Via 10;av iekluf-e lb 14; carsom-e Via 13, 14; destram-e VIb 49; ebetraf-e Via 12, hebetaf-e VIb 53; esunum-en III 20, esunum-e lb 14, esonom-e VIb 50, 52; fesnaf-e lib 16; ooserclom-e Via 12; perum-e Ila 27, III 33, persom-e VIb 38, 39, 40; pertom- e Via 14; purom-e VIb 17, Vila 38; presoliaf-e Via 12; randem-e Via 14; Rupinam-e lb 35, 36, Rubinam-e VII a 43, 44; rusem-e Vila 8, 9, 23; satam-e lb 38; smursim-e Via 14; termnom-e VIb 57, 63, 64; tertiam-e Via 13; tet- tom-e Via 13, 14, 14; todcom-e Via 10, 10; uapef-e Via 10, b 51; uasirslom-e Via 12; uasetom-e VIb 47, vagetum-i lb 8; veruf-e lb 9, uerof-e VIb 47; vukum-en III 20. With loc. 'in, on, at.' arve-n III 13; eikvases-e Va 4, 16; fesner-e lib 11; Fisie-m Via 46; Funtler-e lb 24, Fondlir-e Vila 13; manuv-e lib 23; ocre-m Via 46; scalsi-e VIb 5, Vila 37; tote-me Via 26, 46. O. en, -en, L. in, Gk. iv, Eng. in. 4, 8 e, 52 h, 147 d.
eaf eaf , earn : see erec.
ebetraf-e, 'exit, exits.' Ace. pi. (+ -e) ebetrafe Via 12, hebetafe [sic] VIb 53. Probably from *e-bait-rd- or *e-baite- tra-; cf. L. e, ex and baeto 'go,' the latter possibly from root i-t- with pre- fix an ablaut-variant to eni, L. ob (see Muller, pp. 59-60). 2 1, 76 a, 86 c.
ecla, 'every' (?). Abl. sg. f. Vila 11, 27. Etym. uncertain. Possibly from *aiko- lo-, cf. Skt. ekas 'one.' 75 a; n. on Vila 11.
ekvine (written ekvi:ne), probably name of a street ' (Via) Equina. ' Loc. sg. Ila 13. L. equinus. 46 h, 74 e; n. on loc. cit.
ef: see ife.
efurfatu, 'remove from a platter' (?). Imv. VIb 17, Vila 38. From e- -\- furfa, cf. fur f ant.
ehe e, preposition with abl. 'from, away from. ' e in easa Ila 38, ehe VI b 54, (in combination ehesu) 54. O. ee- eh-, L. e, ex, eK, e|. 26 c, 147 c.
eheturstahamu etufstamu, 'banish, pro- nounce banishment against. ' Imv. pass. sg. etufstamu lb 16, eturstahmu VIb 53 (written etursta), 53, ehetursta- hamu VIb 55. From e- + *tudes-td-, cf. tuder. 26 c, 57 e, 114 e, 117 c.
ehiato, 'let loose.' Pcpl. pf. pass. ace. pi. m. (or sg. neut. ?) Vllb 2. From e- + *hiyd-, cf. O. eehiianasum, L. hio, hiare. 48 a, 114 e, 117 b.
ehvelklu, noun 'vote, expression of opi- nion.' Ace. sg. Va 23, b 1. From *e-wel-tlo-, cf. ehueltu, L. volo, velle. 26 c, 75 b, 86 c.
ehueltu, 'order, call out.' Imv. Via 2. From *e-wel-(e)tdd, cf. ehvelklu, L. volo, velle. 2S c, 119 a.
eikvasatis, ' of the college ' ( ?). Abl. pi. Ill 24, 29. Cf. eikvases-e, to which eikvasatis is related as L. Arpinas to Arpinum or Ravennas to Ravenna. 19 b.
eikvases-e 'members of a (priestly) col- lege' (?). Loc. pi. (-t- -e) Va 4, 16.
304
IV. LEXICON AND INDEX VERBORUM
Etym, and stem-formation uncertain, cf. L. aequus (?), U. eikvasatis, 19 b.
eine: see enem ene.
eiscurent, 'come to fetch.' Fut. pf. 3 pi. Vb 10, 15. Probably related to Skt. icchdti (< *is-ske-ti) 'wish,' OHG eiscon 'fragen, ' OE ascian, OCS iskati ' seek, ' all these forms representing several different vowel-grades. 114 e, 119 e, 124 b.
eitipes, 'resolved.' Pf. 3 pi. Va 2, 14. Etym. uncertain; possibly from *eit- (om h)epens ' ratum habuerunt.' 9b, 34, 51 b, 114 e, 115 d n. 1, 124 c, g, 151 b; n. on Va 2.
emantur, ' accept. ' Pres. subj. pass. 3 pi. emantur Va 8, emantu Va 10. L. emo, emantur. 8 a, 56 c, 114 e, 119 a, 126 a.
endendu ententu, 'insert, place in, on.' Imv. ententu lb 12, III 15, endendu VIb 40, 40, 49; fut. pf. 2 sg. entelus lb 12; 3 sg. entelust Ylh 50. From en- + tend-, cf. andendu, L. tendo. 3 d, 52 f, 60 a, 61 e, 114 e, 119 a with n. 3, 124 e, 125, 127 b.
enem ene, temporal adv. ' then, thence. ' ene lb 35, enem Vila 44, 44, eine Via 10, 11. Cf. eno, enumek, inumek, O. inim, inim, O. eivei/n, L. enim. 109 i, 112 d.
enetu, 'commence.' Imv. enetu la 1, enetu Via 1. From en- + etu. L. in- eo. 20 c.
eno enu enuk, enumek, etc., temporal adv. 'then.' enuk la 30, 33, Va 29, inuk lb 7, III 4, 7, 15, 16, IV 13, 14, enu lb 36, 37, 38, 38, Ila 9, b 21, eno VIb 16, 17, 17, 46, 56, 56, 62, 65, Vila 1, enom VIb 38, 38, 39, 40,40,53,64, Vila 5, 8, 9, 23, 23, 36, 45, 45, 51, enno Vila 38, ennom VIb 51, Vila 20, 24, 34, 39; enumek lb 11, 13, 16, 19, 20, 21, 22, inumek III 9, 11, 26, 34, IV 2, 17, 18 (written inuntek), 20, 21, 24, 26, 27,28, inumk IV 23, inemek III 20. Cf. enem ene and Oscan and Latin etyma listed after it. 51 e, 109 i, 112 d, 113 h.
entelust, ententu: see endendu.
erec erek ere, efek erse, etc., anaphoric pronoun ' he, it, ' etc. ; sometimes with adverbial value ' then, during that time.' Nom. sg. m. erek Va 11, ere Va 4, erec Vllb 1, ere VIb 50; neut. erek Va 26, erse Via 8; gen. sg. m. ererek III 32, erer Via 23, 24, 31, 33, 34, 35, 40, 43, 43, 45, 50, 53, 54, 55, b 7, 7, 10, 12, 14, 15, 26, 27, 28, 33, 34, 35, Vila 10, 18, 19, 22, 26, 31, 32, 35, irer Via 25; f. erar Via 23, 24, 26, 31, 33, 34, 35, 40, 43, 44, 45, 50, 53, 54, 55, b 7, 8, 10, 12, 14, 15, 26, 27, 28, 33, 35, 35, 62, Vila 11, 14, 14, 18, 19, 22, 26, 28, 31,32,35,50,51; ace. f. earn VIb 16,24; neut. efek la 30, III 33, 35, IV 3, 21, 32, erse Via 6 (adverbial value in the last two passages); abl. m. (+ -com) erucom VIb 50; f. erak III 12; neut. (+ -ku) III 31, eruk III 14 (adverbial); gen. pi. eru Va 8 (?), ero VIb 62, 62, Vila 13, 14, 28, erom Vila 50; ace. f. eaf lb 42, eaf Vila 52 ; neut. eu Ila 2, b 9, eo Via 20. O. izic, cf. L. is, ea, id. 2 j, 8 b, 20 d, 31 b, 42 b, 57 c, 106 b, d, 109, 109 a (the basic ref.), 113 c, h.
ereglu, 'statue' (?). Ace. sg. ereglu IV 13 (first instance; possibly pi.), ( + -a) eregluma III 35, IV 3, 10, ( + -af) ereglumaf (written ereglamaf) IV 6; loc. eregle IV 17, 19; ace. pi. ereglu IV 13 (second instance). Pro- bably from *aise-kelo-, cf. erus, esono. 29 e, 75 c; n. on III 35.
eretu: see her.
erietu, 'ram.' Ace. sg. Ila 6. Cf. L. aries, Lith. erytis 'lamb.' 35 d, 80 b.
erihont, eront, etc. 'the same.' Nom. sg. m. erihont VIb 50, eront VIb 24; gen. f. erarunt IV 1; abl. m. eruhu lib 22; f. erahunt lb 23; erafont VIb 65, Vila 1; nom. pi. m. euront VIb 63; abl. f. erer\ant (written erererunt) IV 5; neut. eriront VIb 48. For first element cf. erec; for second element cf. L. hie, hunc, etc. 48 a, b, 52 g, 106 c, d, 109, 109 b.
erus, noun of uncertain etym. and mean- ing, left untranslated. Possibly some-
erne
305
etato
thing distributed to persons present at the sacrifice. Ace. sg. erus la 33, b 34, 34, 35, 36, Ila 9, 28, 32, 40, b 21, IV 14, 27; erus VIb 16, 16, 25, 38, 38, 39, Vila 5, 38, 43, 43, 43, 44. Possibly from *aisus, cf. O. abl. pi. aisusis ' sacrifices ' ; cf . also U. esono, 29 n. 1, 60 n. 1,83; n. on VIb 16. eskamitu, noun of uncertain meaning, possibly 'phallic portion' (of. a. sirusla cake). Ace. sg. IV 1. Cf. Germ. Scham (?). See n. on loc. cit.
esmei, esmik, pronoun 'this, it.' Dat. sg. esmik la 28, 31, esmei Via 5, 18; loc. esme VIb 55. Skt. asmai; for the stem cf. the initial vowel of exelvoq, U. etantu. 57 b, 109, 109 c.
eso esoc esu esuk, etc., adv. of manner 'thus, as follows.' esu Ila 3, Va 14, esuk Va 1, eso Via 2, 3, 8, 16, 22, b 6, 9, 53, 57, Vila 9, 20, 25, 34, 46, esoc VIb 25, iso Via 20, issoc VI lb 3. Cf. esu, O. eksuk, exac. 109 e, 112 d, 113 c, h.
esono esunu, adj. 'of the gods, divine, religious' and neut. noun 'rite, sacri- fice.' Adj.: dat. sg. f. esune Va 4; abl. f. esuna Va 5; ace. pi. f. esona Via 3, 5, eesona Via 18. Noun neut. : nom. sg. esunu lb 38, Ila 21, 42, III 1, 14; ace. esunu lb 9, Ila 2, 20, esono VIa57, b 47, (+ -en, -e -e) esunumen III 20, esunume lb 14, esonome VIb 50, 52; abl. esunu IV 30; loc. esune Va 6, esone VIb 11; gen. pi. esunu lb 8, esono VIb 47; ace. pi, esunu Ila 2, 2; abl. esoneir Via 18, (4- -ku) esunesku Va 11. Probably from *ais(e)s-ono/d, cf. dat. pi. Marruc. aisos, Mars, esos ' dis ' ; alaoi' Bsoi vno Ti^g^Tjvcov Hesych., aesar Etrusca lingua deus, Suet., Aug., 97 ; also O. aisusis, abl. pi. ' sacrifices, ' U. erus (?). 13 a, 19 a, 25 b, 57 f, 90.
est est, vb. 'be, is,' etc. Pres. 3 sg. est lb 18, 18, Ila 15, est Via 8, 9, 10, 26, 27, 27, 27, 28, 28, 28, 36, 37, 37, 37, 37, 37,38,46,47,47,47,47, 47, 48, b 29, 30, 30, 30, 30, 30, 30, 31, 53, 55, 55, Vllb 2, 3; 3 pi, sent Via 15, 27, 36, 46, b 29;
pres. subj. 2 sg. sir VIb 7, 7, 26 (in fonsir), si VIb 26, sei Via 23, 23; 3 sg. si Va 6, 24, 27, b 3, 7, se lb 8, 8 (in antervakaze), si (in mersi) Via 38, 48, se (in anderuacose) VIb 47; 3 pi. sis Va 6, sins Vllb 4; inf. eru Va 26, 29, b 5, erom Vllb 2. O. est ist, inf. ezum, L. est, eort, Skt. asti, etc. 4, 8 a, 12 a, 31 d, 41 a, 52 d, 57 e, h, 58 g, 114 e, 115 b, 122 a, 126 b, 128. See also futu, which stands in a suppletive relation to est. este este, est^i, etc., demonstrative pron. ' this, the following. ' Ace. sg. m. estu lib 24; neut. este la 1, este Via 1, 56, b 62, 63, Vila 51; ace. pi. neut. estu Ila 2, b 23, esto Via 15, 15. L. iste (1). 106 a, 109, 109 d.
esu esu, etc., demonstrative pron. 'this.' Abl. sg. m. (+ -ku) esuku IV 29, esu Via 25, 28, 33, 35, 38, 45, 48, 53, b 28, 31, 35, 54, essu Via 43, (after eh-) ehesu VIb 54; f. esa VIb 9, 14; gen. pi. esu- mek lb 8, esome VIb 47; abl. esir Vila 10, 18, 26, 32, (+ -co) esisco Via 18, isir Vila 21, 34. Cf. eso, O. eksuk, exac. 57 d, 61 d, 109, 109 e.
esuf , intensive pron. ' oneself. ' Nom. sg. m. Ila 40, IV 15. O. essuf, L. ipse. 57 d, 58 f, 61 d, 81 n. 1, 91 c, 109, 109 f.
et et, conjunction 'and.' et lb 20, IV 7, 12, Va 6, 8, 13, 18, 20, 22, et Vb 9, 13, 15, 17, Via 19, b 5, 24, Vila 37, 44, 46, 51. L. et, sri, Skt. ati. 31 d, 41 a, 112 g.
etantu, adj. 'so great.' Nom. sg. f. Vb 6. For e- cf. the initial of esmei; for -tantu, L. tanta, cf. rov, zijv, to, etc. 2 d, 7 b, 41 a, 61 a, 71, 109 c, k.
etato etatu, 'go.' Pres. subj. 3 pi. etaians VIb 64, etaias VIb 65, Vila 1; imv. 3 pi. etatu lb 21, 22, 23, etato VIb 63. From *eitd-, related to etu as fre- quentative L. ito -are to eo ire, except that e- in the Umbrian forms probab- ly stands for the strong grade of the root. 52 d, 57 e, 58 g, 114 e, 117 d, 127 c.
306
IV. LEXICON AND INDEX VERBORUM
etru, 'other, second.' Dat. sg. f, etre lib 2, 3, 4, 5, 5, 6,; ace. f. (+ -a) etrama III 34; abl. neut. etra Via 35, 38, 43; loc. etre lib 14; dat. pi. etre lib 3, 4, 6 (first instance in each of these three lines); ace. f. etraf la 18, 18; abl. etres III 18. From e- (cf. esmei, etc.) + -tero-, with semantic develop- ment as in L. alter. 76 c, 105 b, 109 c.
etu etu eelu, 'go.' Fut. 3 sg. est Via 6, eest Via 2; imv. etu lb 10, 14, lla 33,
III 20, IV 21; etu VIb 48, Vila 39, eeiu VIb 54; imv. 3 pi. etuta III 11, etutu lb 15, 23, 23, eliito VIb 51, 52, 65, Vila 1; fut. pf. 3 sg. iust Via 7; pf. subj. pass. (?) 3 sg. ier VIb 54. L. eo, ire, elfii, Skt. emi, cf. U. etato. 13 b, 20 a, c, 122 b, 123, 124 b, 126 c, 127 b, c.
etursiahmu eturstamu: see eheturstaha- mu.
eveietu, 'consecrate' (?). Imv. lib 8,11.
Probably from *e-weig-e-iod, cf. L.
vic-tima, Germ, weihen. 47 b, 118 d;
n. on lib 8. ezariaf, 'food-(baskets)' (?). Ace. pi.
IV 27. Probably from * ed{e)s-driaf < *edes-asians, cf. L. edo ' eat, ' etc. 68 d.
facurent fakurent, fakust: see fagia.
fahe, word of uncertain form and sense in Vb 13. Possibly 'roasted.' Cf. q)c6yco (?), Germ, backen and Eng. bake (?). N. on loc. cit.
famefias 'famihes' (?). Nom. pi. lib 2. O. famel, famelo, L. familia. No sure connection outside Italic. See E-M^, under famulus, where possible Etr. ori- gin is suggested.
far, noun 'spelt. ' Gen. sg. farer Vb 9, 14; ace. far Vb 10, 15. O. far, L. far, OE bere 'barley,' cf. U. farsio. 20 a, 59 b, 83, 90.
farsio fasio fasiu, 'spelt-cakes.' Ace. pi. neut. fasiu I la 12, farsio VIb 2, fasio VIb 44. L. farrea, cf. U. far., etc. 2 i, 3 f, 8 b, 53 b, 55 b, £9 a with n. 1, 68 b.
fagefele, adj. 'fit for sacrifice.' Ace. sg. (written fagefete) lib 9. Equivalent as
if to L. *facibilem; cf. facio, U. fagia in the ritual sense ' sacrifice. ' 77 i.
fagia, fetu, etc., 'do, perform, sacrifice offer, place,' etc. Pres. subj. 3 sg. fagia Ila 17; pres. subj. 3 sg. feia Va 23, b 1 ; imv. fetu la 3, 9, 11, 12, 13, 13, 14, 16, 17, 17, 20, 22, 24, 25, 25, 26, 26, 28, 28, b, 2, 3, 3, 6, 25, 25, 27, 29, 32, 43, 44, 44, Ila 2, 4, 6, 7, 8, 9, 11, 11, 12, 13, 13, 13, 14, b 7, 10, 26 (written feiu), 29, feitu la 4, 5, 7, 29, 30, 32, 32 32, b 5, 7, 9, 18, 24, 28, 31, 32, Ila 20, III 31, 31, 32, fetu Via 22, 56, 57, 57, 58, 58, 58, 59, b 1, 1, 1, 2,3, 3, 3, 5, 19, 19, 20, 20, 22, 22, 22, 23, 24, 37, 43, 43, 44, 44, 44, 45, 45, 45, 46, 46, 55, Vila 3, 4, 4, 6, 7, 7, 1, 37, 41, 42, 42, 53, 53, 54, 54, feilu VIb 3, 22, 47, Vila 3, 4, feetu Vila 41; inf. fagiu Ila 16, fagu lib 22; pcpl. pass. abl. sg. f. feta lib 13; fut. pf. 3 sg. fakust IV 31 ; 3 pi. fakurent lb 34, facurent Vila 43. From root *dhd-k-, dhe-ik)-: O. fakiiad, L. facio, cf. rl67]/ni, Skt. dadhami, Eng. do. 16 a, 31 d, 35 d, 41 b, 43, 46 b, d, g, j, 53 d, 114 e, 115 d, 121, 122 d (the basic ref.), 124 b, 125, 126 a, 127 b, 128.
fata, '(success in) word' (?). Ace. sg. VIb 11. Pcpl. pass, from root *bhd-: L. fdtum, cf. (pr][i,i. (?). N. on loc. cit.
feetu: see fagia.
fefure, vb. of uncertain meaning, 'be disturbed, turn out for ill' (?). Pf. 3 sg, (?) Ila 4. Etym. uncertain; pos- sibly connected with q^vgco, TiecpvQxat 'mingle, confound.' 114 e, 124 a.
feia, feitu feitu: see fagia.
feliuf : see filiu.
felsva, noun of uncertain meaning, pro- bably neut. ace. pi. ' vegetables. ' Va 11. If this interpretation is correct, from *gheles-wo-, cf. L. holera, dialectic L. folus cited as "ancient" by Paul, e Fest., 74 Li.; x^or}, x^ojQog, Eng. yellow, Lith. zelvas 'greenish,' Russ. zelenyi 'yellow,' the Balto -Slavic forms giv- ing evidence of original initial palatal stop. 29 d, 48 b, 64 c, 69.
etru
307
f rater
ferar, ferest: see fertu.
ferine, ' carrier, tra,y ' ( ?). Loc. sg. ferine (in lb 25, III 16 written ferime) la 4, 13, 22, b 3, 6, 25, III 16, 31, ferine Via 57, b 1, 19, 43, 45, Vila 4. Cf. fertu, L. fero. ferine : L. *feri6ne : fero = U. natine : L. natione : na-scor. 81 b; n. on Via 57.
ferelitru, noun of uncertain meaning, possibly 'chopping-board. ' Ace. sg. Ill 16, 18. From *fide-tro-, cf. L. findo, Skt. bhindtti. (?). 76 b; n. on III 16.
fertu, 'bring, carry, lift.' Pres. subj. pass. 3 sg. ferar VIb 50; fut. 3 sg. ferest I la 26; imv. fertu I la 17, 19, 26, 27, 33, 34, b 12, 12, 13, 14, 14, 15, 15, 16, 16, fertu VIb 50, 50; imv. pi. fertuta III 13. L. fero, (psQai, Eng. bear. 8 a, 29 b, m, 40, 41 a, 42 e, 114 e, 115 e, 119 a, 123, 126 a.
f esnaf-e, ' temple. ' Ace. pi. ( + -e) fesnafe lib 16; loc. (+ e) fesnere lib 11. O. fiisnu, cf. L. fanum (-an- < -dsn-), festus, feriae. Connections out- side Italic are uncertain. 35 d, 57 b, f, 74 a, 90.
fetu fetu, feta: see fagia.
ficla fikla, a type of cake, here left un- translated. Gen. sg. fiklas Ila 41; ace. sg. fikla Ila 18, 29, ficla Via 56, 59, VIb 2, 4, 20, 23, 44, 46, Vila 4, 8, 54, fie lam Vila 42. Probably from *fig-kla, cf. L. fingo, relxog, all from root *dheigh-. 75 b, 142 b; n. on Via 56.
fiktu, ' attach, add. ' Imv. la 28. L. figo (?). 49 d, 119 a; n. on loc. cit.
filiu, adj. 'sucking.' Ace. pi. m. feliuf la 14, filiu VIb 3. L. filius, cf. felare, e-nXri, dfj?ivg. 43, 55 b, 68 e.
Fise: see Fiso.
Fisier, adj. in our texts found only with forms of ocar ukar, ' the Fisian Mount, ' and hence always sg. m. Gen. Fisier Via 30, 32, 39, 41, 49, 51, b 13, 32, 33, Fisie VIb 10; dat. Fisie Via 40, Fisi Via 30, 33, 42, 50, 52, b 7, 10, 11, 14, 26, 32, 34, Fisei Via 23; ace. Fisim
Via 41, 49, 51, Fisi Via 31, 39, b 12, 31, 33, Fisei Via 29; abl. Fisiu la 5, 8, 12, 15, 17, 21, 25, 29, 31, Fisiu Via 23, 25, 34, 35, 45, 53, 55, 58, b 1, 3, 6, 6, 9, 14, 19, 22, 26, 28, 35, Fissiu Via 43; loc. Fisie Via 26, 36, b 29, (+ -m) Via 46. Cf. Fiso. 32 a, 57 h, 68 a, 90, 93 b.
Fiso, name of a god 'Fisus. ' Dat. sg. Fise la 15; dat. sg. Fiso VIb 3. From *bh(e)idh-to-, cf. L. fido, fisus, neCdoj. (?). 25 f, 44 d, 78 a; n. on VIb 3.
Fisouie, name of a god 'Fisovius, ' gene- rally in combination with Sansi Sage. Gen. sg. Fisouie VIb 15; dat. Fisuvi la 17, Fisoui VIb 5, 6, Vila 37; ace. Fisoui VIb 8, 8; voc. Fisouie VIb 9, 10, 12, 12, 14, 15. Cf. Fiso. 68 e; n. on VIb 5.
Fisouina, adj. 'Fisovian, for Fisovius.' Abl. sg. VIb 9, 14. Cf. Fiso, Fisouie. 74 e.
Fissiu : see Fisier.
fito, '(success in) deed' (?). Ace. sg. VIb 11. Pcpl. pass, from root *bhw-l-- O. fiiet, L. fio, OL fitum. (?). N. on loc. cit.
Fondlir-e, name of a place in Iguvium, ' Fontuli, the Springlets. ' Loc. pi. (-|- -e -e) Funtlere lb 24, Fondlire Vila 3. Probably equivalent as if to L. *font-ulis. 48 b, 57 f, 60 a, 75 a.
fons, 'favorable.' Nom. sg. fons Via 42, 50, 52, b 7, 11, 13, 26 (in fonsir), 32, 34, Vila 13, 17, 31, 49, fos Via 23, 30, 33, 40; gen. foner Vila 20, 23, 33, 36; nom. pi. foner VIb 61. L. Fones, gloss. ' dei silvestres' ; cf. Faunus, faveo. No sure connections outside Italic. 22, 29 1, 52 d, 53 b, 58 e, 77 d, 90, 91 d.
frater, '(priestly) brothers.' Nom. pi. frater III 5, Va 1, 14, 22, frater Vb 11, frateer Vb 16; gen. fratrum III 10, fratru Ila 21, 35, b 26, III 6, Va 12, 25, 27, 29, b 3, fratrom Vllb 1; dat. fratrus Vb 8, 13, VI lb 1; abl. (+ -per, -pe) fratrusper Ila 2, III 23, fratrus- pe III 23. L. frater, (pQdrrjQ, Skt.
308
IV. LEXICON AND INDEX VERBORUM
bhratd, Eng. brother. 7 a, 26 h, 29 k, 40, 51 e, S9 f, 60 f, 82 b, 90, 143 a.
frafrecate, 'term of office as frairicus. ' Loc. sg. Vllb 1. Related to frairexs as L. tribunatus to tribunus, but with appar- ent transfer to the o-decl. 10 c, 70, 71.
frafrexs, adj. 'of the brotherhood' and m. noun 'frairicus, brother-superior.' Nom. sg. m. fratreks "Va 23, b 1, frairexs Vllb 1; dat. m. fratreci Vllb 4; abl. f. (as adj.) fraireca Vllb 2. Cf. frater. 10 c, 25 b, 29 i, 46 e, 70, 90, 91 b, 93 b.
frehtef , ' boiled portions ' ( ?). Ace. pi. ( ?) Ila 26. Possibly from *bhreg-ii-. For etym. discussion see n. on loc. cit.
frehtu, 'boiled, hot' (?). Ace. sg. IV 31. Possibly pass. pcpl. of vb. to which frehtef is a verbal noun. N. on loc. cit and on Ila 26.
frif, 'fruits.' Ace. pi. frif Via 42, 52, b 13, Vila 17, 30, fri Via 30, 32, 40, 50, b 32, 34. L. frUx, fruges, cf. O. fruk- tatiuf, L. fnior, frucius, frumenium, Goth, brulcjan ' gebrauchen, ' OE bru- can. 15 a, 61 b, 65 f, 80 a, 102 d.
frite, noun 'trust,' Abl. sg. Via 24, b 8, 15, 27, Vila 20, 23, 33, 36. Cf. L. freius. 9 b, 80 b.
froselom, pass. pcpl. nom. sg. neut. used as noun ' damage. ' Via 28, 37, 47, b 30. Probably pcpl. to pit. *fraussa- formed as if iterative to L. fraudo, -are, and with variation of stem-vowel as in peseiom contrasted with L. peccaium. 22, 30 c, 44 b, 131, 149.
fuia, 'be performed, be.' Pres. subj. 3 sg. fuia III 1; fut. 3 sg. fuiest Va 9. O. fiiet, L. fio, all three being from root *bha- -f L Cf. also U. fito (?), fuiu. 114 e, 120 a, 123.
Funtlere: see Fondlir-e.
furent: see fuiu.
furfant, 'place on a platter' (?). Pres. 3 pi. furfae lb 1, furfant VIb 43. From *bhordh-a-, cf. efurfalu, Germ. Bord, Eng. board. (?). 1 b, 52 g, 115 d, 117 e; n. on VIb 17, 43.
furo, 'forum.' Ace. sg. furu lb 42, furo Vila 52. L. forum, cf. fores, dvga, Eng. door, etc., the usual meaning of forum having developed from an earl- ier meaning 'enclosure around a house, ' according to E-M^, p. 444. 12 c, 43, 67.
futu futu, 'be. ' Fut. 3 sg. fust lb 7, 39, III 6, Va 4, 11, 19, 20, fust Via 7, b 39, 41, 42, 47, 47, Vila 45, b 1, fas VIb 40; 3 pi. furent Va 22; imv. futu Ila 22, 43, 43, III 14, IV 32, futu Via 30, 33, 40, 42, 50, 52, b 11, 13, 32, 34, Vila 13, 17, 31, 49; imv. pi. fuiuto VIb 61. Cf. O. fust, fufans, L. fui, (pvco, Skt. bhavati, Eng. be ; also U. fito ( ?) fuia. The absence of the change u < i (15) may point to a thematic stem *bhuwe-, unless the u is preserved by analogy of other forms in which it is not subject to change. 29 m, 40, 52 g, 57 c, 114 e, 115 c, 122 a, 123, 127 b, c. See also est, to which futu stands in a suppletive relation.
gomia kumiaf , ' pregnant. ' Ace. pi. f. kumiaf la 7, gomia Via 58. L. gumia ' glutton, ' cf . yo/noi;, ysfico. 2 a, 3 d, 35 d, 47 a, 68 a.
Grabouie Crabouie, Krapuvi, ' Grabovius, ' epithet of the gods Jupiter, Mars, and Vofionus. Dat. sg. Krapuvi la 3, 11, 21, Grabouei Via 22, b 1, Grabouie VIb 19; ace. Graboue Via 24, 25, Graboui Via 23; voc. Grabouie Via 25, 26, 28, 29, 29, 31, 32, 33, 34, 35, 36, 38, 38, 39, 41, 41, 43, 44, 45, 46, 47, 48, 48, 49, 51, 51, 53, 54, 55, Crabouie, Via 27, 37. Probably connected, through an lUyrian medium, with Epirote Mo- dern Gk. yqafioQ 'oak,' Russ. grab 'hornbeam,' etc. 2 a, b, 3 d, 25 b, 31 c, 33 c, 68 e, 90; n. on Via 22.
habe habe, babetu habitu, hahtu hatu, ' take, receive, have. ' Pres. pass. ( ?) 3 sg. habe VIb 54, babe lb 18; pres. subj. 3 sg. habia Va 17, 19, 21. Fut. 3 sg. habiest VIb 50, 53, 63, Vila 46, 51; fut. pf. 3 sg. habus VIb 40, 3 pi.
fratrecate
309
Huntia
hahurent Vila 52; imv. habetu lib 23, 23, 27, 28, III 28, IV 30, 31, habitu Via 19, b 4, imv. pi. habetutu lb 15, habituto VIb 51 ; imv. hahtu Ila 22, 22, hatu lb 11, hatu VIb 49, imv. pi. ha- tutu lb 42, hatuto Vila 52. O. hafiest, pf. subj. hipid, cf. L. habeo and capio, whose roots in O.-U. appear to have become contaminated. 2 e, 9 a, b, 13 b, 31 d, 38 b, 39, 114 e, 116 c, 118 a, 121, 121 a, b, 122 e (the basic ref.), 126 a, 127 b, c, 138 a.
habina hapinaf, 'lambs.' Gen. pi. ha- pinaru la 33; ace. hapinaf la 24, habina la 27, habina VIb 22, 23, 24. From *ag^nina- (cf. L. agniis, agninus) with dissimilatory loss of n, and un- etymological h possibly from contami- nation with the U. equivalent of L. haedinus. 2 b, 3 d, 49 b, 74 e; R. G. Kent, Lang., II (1925), pp. 189-90.
habitu, habus, hahtu hatu: see habe.
her, herti, heries, etc., vb. 'wish, ' in pass., 'be necessary, intended.' Pres. 3 sg. heri IV 26, her (in pisher) VIb 41 ; pres. pass. 3 sg. herter Ila 40, III 1, herte Va 6, 8, 10, herti Vb 8, 11, 13, 16, hertei Vllb 2; fut. 2 sg. heries lb 10, lib 21 ; 3 sg. heriest Vila 52, heries VIb 48; pf. subj. 3 sg. heriiei Ila 16; pf. subj. pass, herifi Vb 6; pcpl. pass. abl. sg. neut. heritu Via 27, 47, b 29, hereitu Via 37, eretu Ila 4. O. fut. herest, cf. L. horior, hortor, xaiQco, Skt. haryati ' delight in, ' Germ, gern, Eng. yearn. 2 h,l,4, 8 c, 9 c, 10 a, 29 m,48 a, 53 c, 57 h, 114 e, 115 c, e, 121, 121 a, b, 123, 124 b, d, 126 c, 152 d. Cf. also heri.
heri, conjunction in correlative pairs 'ei- ther ... or. ' heris la 4, b 6, 6, heri la 4, 22, 22, lib 9, 10, heri Via 57, 57, b 46, 46, herie VIb 19, 20, heriei Vila 3, 3. Fossilized 2 sg. pres. ind. form of her, etc. 2 j, 112 f, 115 b.
Hoier, proper name, probably of a god 'Hoius. ' Gen. sg. Via 14. Etym. un- certain. 68 e, n. on loc. cit.
holtu, vb. of uncertain meaning, possibly 'destroy.' Imv. VIb 60, Vila 49.
Perhaps related to oXXvixi (but not to L. ab-oleo, if E-M' are right in connect- ing the latter with alo). 119 a; n. on VIb 60, including problem of the h-.
homonus, 'men.' Dat. pi. Vb 10, 15. O. humuns, L. homo, hominibus, OL homones, cf. humus, xa/xai, ydo'iv, etc. 13 a, 45, 48 a, 57 e, 61 d, 81 a.
Honde Hunte, name of a god ' Hondus. ' Dat. sg. Hunte lb 4, Ila 20, 34, Honde VIb 45. Probably from *ghom-to-, cf. homonus, hondomu, hondra, L. humus, Xa/nai, x^(^v etc. 71.
hondomu, superlative adj. 'lowest.' Abl. sg. Via 9. 10. Cf. hondra, L. humus, etc. 45, 48 a, 60 a, 73 a.
hondra hutra, comparative adj. with value of improper preposition 'below.' hutra lb 42, hondra Via 15, Vila 52. From *ghom-t{e)rad, cf. hondomu, L. humus, etc. 2 a, 45 a, 43 a, 52 d, 60 a, 76 c, 113 Q, 14S e.
hondu, vb. of uncertain meaning, prob- ably 'cast down.' Imv. VIb 60, Vila 49. From *ghom-d^(6d (?); for first member cf. hondomu, hondra, L. humus, etc., the second member being either from root *dd- or *dhe-. 48 a, 60 a, 119 c.
Horse Huf ie, epithet of Mars ' Hodius ' ( ?). Dat. sg. Hurie lb 2, Horse VIb 43. Etym. uncertain. N. on VIb 43.
hostatir, 'armed, under arms.' Dat. pi. hostatir Ylh 62, Vila 13,15,28, 50; ace. pi. hosfatu VIb 59, Vila 48. Cf. an- hostatir, L. hastatus, hasta, Goth, gazds' xsvTQov. 35 d, 44 d, 48 a., 71, 114 e, 117 c n. 2; n. on VIb 59-60.
Hule, name of a deity 'Hula.' Dat. sg- IV 17. Etym. unknown. 66; n. on loc. cit.
huntak, 'jar' (or 'vat' ?). Ace. sg. Ill 3, IV 32. From *honditak, cf. hondomu, hondra. (?). 80 a; n. on III 3.
Hunte: see Honde.
Huntia, 'the Hondia, the festival of Hondus; things pertaining to the Hondia (?).' Nom. pi. neut. (?) Hun-
310
IV. LEXICON AND INDEX VERBORUM
tia Ila 15; ace. pi. neut. (?) Huntia Ila 17. Cf. Honde Hunte. 68 e; n. on Ila 15.
Hufie: see Horse.
hutra : see hondra.
I III, symbol for ' four. ' Vb 9. 104 a.
labuscer, lapuzknm, adj. of nationality ' lapudic. ' Gen. sg. labuscer VIb 54, 59, Vila 12, lapuscer, Vila 48; dat. labusce Vila 12; ace. labuscom VIb 58, lapusco Vila 47, lapuzkum lb 17. Cf. L. lapyges, 'Idjiodeg. 46 e, 70; n. on VIb 54.
Ikuvina: see liouinar.
iepi, adv. 'there, to that place.' Ill 21. Cf. ife; for -p- cf. O. ip, Pael. ip (?). N. on loc. cit.
iepru, 'liver' (?). Ace. sg. Ila 32. Cf.
rinaq, L. iecur, Skt. yaki't. (?). 84 n.
1; n. on loc. cit. ier: see etu. ife, adv. 'there, thither.' ife lib 12, 13,
ife VIb 39, 40, ef Via 4. L. ibi, Skt. iha,
Prakrit idha, all from *i-dhe. 112 d,
113 f.
ifont, adv. 'on the same spot.' VIb 55. Cf. ife, erafont, isnnt, suront. 113 f, i.
liouinar, Ikuvina, etc., adj. ' Iguvine, of Iguvium ' and pi. noun ' men of Iguvium. ' Gen. sg. f. liuvinas lb 2, 5, liouinar Via 32, 39, 42, 49, 52, b 32, 43, 45, 61, Vila 3, 6, 10, 14, 15, 16, 16, 17, 19, 21, 24, 26, 27, 28, 29, 30, 30, 31, 32, 35, 37, 41, 53, louinar Via 30, b 10, 13, 34, Vila 9, 27, 50; dat. f. IkuvineIbl3,7zowzneVIa 18,24,27,31, 40, 43, 50, 53, b 7, 11, 14, 33, 34, 51, 62, Vila 14, 18, 31, lioueine Via 5, louine Via 33, b 10, 27, Vila 50; ace. sg. f. liouinam Via 49, 51, b 12, 33, Vila 16, 29, liouina Via 31, 41, 54, b 31, louina Via 29, 39; abl. f. Ikuvina la 5, 8, 12, 15, 19, 21, 25, 29, 31, b 2, liuvina lb 5, III 24, 25, 30, 30, liouina Via 23, 45, 55, 58, b 1, 3, 7, 9, 15, 19, 22, 26, 28, 35, 43, 45 (vmtten iiouinar), Vila 4, 10, 22, 26, 32, 35, 37, 41, louina
Via 25, 34, 35, 43, b 6, Vila 7, 9, 19, 24, 53; loc. f. liouine VIb 29, louine Via 26, 36, (+ -777) louinem Via 46; nom. pi. m. (as. voc.) Ikuvinus lb 21, 22, Ikuvinu lb 20, liouinur VIb 63, louinur VIb 56. L. Iguvini, Iguvium. No clear etym. 2 j, 3 d, 11 a, 13 b, 47 c, 74 e, 90.
inumek : see eno enu.
loui luvi, ' Jovius, of Jupiter,' as epithet of several deities and in Ila 6 used alone as a divine name. Dat. sg. luvi la 28, luvie la 8, 24, b 43, Ila 6, 8, 20, 35, loui VIb 22, louie Via 58, Vila 53; ace. luviu lb 1, loui VIb 26, 27, 27; abl. louiu VIb 43; voc. m. louie VIb 28, 29, 29, 31, 31, 31, 33, 33, 36, liouie VIb 35; voc. f. louia Vila 47, 49. From *d(i)i)ow-(i)yo-, related to lu(pater), luve, as L. lovius to !u(piter), lovis, or Martius to Mars. 20 a, 25 b, 31 a, 33 c, 53 a, 68 a, 90.
iouies, 'young men.' Dat. pi. iouies VIb 62, Vila 13, 14, 28, 50; ace. pi. iouie VIb 59, Vila 48. From *yuwiye-, cf. L. iuvenis, Skt. yuvan-. 25 d, 35 d, 53 a, 79, 90.
isec isek, adv. 'in like manner, likewise. * isek IV 4, isec VIb 25. Cf. esu, eso, etc. 113 h.
isegeles, adj. with value of pass. pcpl. 'chopped' (?). Abl. pi. IV 7. From *en-seke-li-, cf. pro-seseto, pru-sekatu, L. seco, -are, etc. 8 e, 77 g, 86 e.
iso, issoc: see eso.
isunt, adv. 'at the same spot, on the same.' Ila 28, 36, III 16, 17. Cf. esu, isec, and for the second part ifont. 106 c, 112 d, 113 i.
itek, adv. 'thus. ' IV 31. Cf. L. ita, item, Skt. iti. 112 d, 113 h.
iuka iuku, neut. noun ' words. ' Ace. pi. iuka III 28, iuku lib 23. L. iocus, cf. Lith./iio/cas 'joke,' OHG jehan ' declare. ' 53 a, 67, 90; n. on lib 23.
iuengar, 'heifer.' Ace. sg. iveka lb 40; nom. pi. iuengar Vllb 2; ace. iveka lb 42, iuenga Vila 51. L. iuvenca, cf.
Hufie
311
mestru
iuvencus, iuvenis, U.iouies, Ski. yuvasas. 2 h, 3 d, 17 d, 32 c, 52 d, 53 f, 60 c, 70, 90.
luieskanes, name of a decuvia among the citizens of Iguvium. Dat. pi. luieskanes lib 6, luieskane lib 5. Etym. uncertain, but Dev., p. 356, sug- gests connection witli lovio- in refer- ence to the grove in lb 1 = VIb 43. 74 d.
lupater, luve, name of the god 'Jupi- ter. ' Dat. sg. luve patre I la 5, b 7, 17, 22, 26, III 22, luvip I la 10, luve la 3, 31 (word-division and interpre- tation uncertain), luue Via 22; voc. lupater lib 24. From *D(i)ijou- < *D(i)yeu- + forms of *pat(e)r- 'father. ' O. luveis, luvei, Diuvei, L. lupiter, lovis, ZsvQ naxrjQ, Skt. Dyauh pita, and more remotely L. deus, dies, etc. 82 b, 90, 98 c.
iust: see etu.
luvi: see loui.
maletu, pcpl. ' crushed, ground. ' Ace. sg. I la 18. Simplex to comoltu ku- multu, etc., q. V. 9 n. 1, 17 b.
mandraclo mantraklu, 'hand-towel, maniple. ' Ace. sg. mantraklu lib 16, mantrabklu Ila 19, mandraclo VIb 4. From *man-trdk-klo-< *man-frdg-tlo-, cf. mani, L. manus and tergeo. 2 e, 18 a, 60 a, 75 b, 89.
mani, 'hand.' Abl. mani Ila 32, mani VIb 24; loc. (+ -e) manuve lib 23; ace. pi. manf Ila 38. L. manus, cf. O. manim, OE mund, OHG munt. 15 a, 25 c, 51 a, 78 a, 90, 97.
mantraklu: see mandraclo.
manuve: see mani.
Marte, name of the god 'Mars.' Dat. Marte la 11, b 2, Marte VIb 1, 43. L. Mars, cf. Mavors, O. Mamers from Fest., p. 117 Li. Mamers Mamertis facit, id est lingua Osca Mars Martls, unde et Mamertini in Sicilia dicti, qui Messanae habitant. 51 a, 80 b.
Martier, 'Martins, of Mars.' Gen. sg.
Marties lb 28, 31, Martier Vb 9, 15, VIb 58, 58, 61, 61, Vila 6, 9, 10, 11, 13, 15, 16, 18, 19, 20, 21, 22, 23, 24, 25, 27, 29, 30, 32, 33, 33, 34, 35, 36, 41; dat. Marti lb 24, Ila 11, Margie Vila 3; voc. Martie VIb 57, 61. L. Martins. Related to Marte as loui (L. lovius) to lu(pater). 68 a.
me/a, a type of cake, here left untranslated ; generally but not always accompanied by spefa. Dat. sg. mefe lib 28; ace. msfa la 16, IV 14, mefa Via 56, b 17, 20, Vila 14, 38; abl. mefa lib 13, mefa VIb 5, 9, 9, 14, Vila 37. L. mensa, other connections very uncertain. 58 c.
mehe, first person pronoun, dat. sg. 'to
me. ' Via 5. L. mihi, cf. Skt. mahyam.
48 a, 51 a, 107 a. menes: see benus.
menzne, 'moon' (?). Abl. sg. Ila 17. :h Probably from *mens-en-i. Sab. (or ' Vest.?) mesene (no. 248 Co.), cf. L.
mensis, fiijv, Doric /usig. 29 f, 58 a,
81 a, 145 c.
mers mefs, 'law, right, custom.' Nom. sg. mers lb 18, 18, mers Via 28 (in mersei), 38, 48 (both in mersi), h 31, 55, 55; abl. pi. m.3rsus III 6. From *medos, cf. O. meddiss (similar in meaning and identical in second mem- ber with L. iu-dex), L. modus, [ledoixai. 29 j, 35 d, 42 c, 83.
mersta, adj. 'eastern, in the east' (?).
Ace. sg. m. mcrsto Via 3, 4, 16, 17; f.
mersta Via 3, 4, 16, meersta Via 17;
abl. m. merstu Via 1; ace. pi. f. merstaf
Via 4, mersta Via 3, 3, 4, 18, 18. Cf.
Goth, maurgins ( ?). 2 f, 26 h, 29 d, 71,
90; n. on Via 1. mersuva, ' accustomed. ' Abl. sg. f . mer-
suva III 11; ace. pi. neut. mersuva
III 28. From *medes-ivo~ (the expected
form being *raefsuva), cf. mers mefs.
69.
mestru, comparative adj. 'greater,' with karu 'part' making up expres- sion 'majority.' Nom. sg. f. Va 24, 27, b 4. Probably from *mais-trd <
312
IV. LEXICON AND INDEX VERBORUM
*magis-tera. L. magisfer, cf. magis, O. superlative maimas, etc. 76 c.
Miletinar, a gentile name or (less pro- bably) a divine name 'Miletina. ' Gen. sg. VIal3. Etymology unknown. 74 e; n. on loc. cit.
mo tar, muta, noun 'fine.' Nom. sg. muta Yb 2, mutu Vb 6; gen. mo tar Vllb 4; ace. rauta Vb 3. O. molto, -tas, -torn, L. miilta, all from *molk-td-. 2 d, 7 b, 12 n. 1, 17 a, 27 a, v^5 d, 71, 90, 142 c.
mugatu, 'make a noise.' Imv. Via 6; pcpl. pass. nom. sg. neut. muieto Via 7. Cf. L. mugio, /avCco, fivxdoOai. 15 c, 30 c, 47 a, b, 116 a, 117 b, 152 1
munekl-u, 'donation.' Ace. sg. Va 17, 19, 21. From *moin-e-tlo- (?), vocalism of second syllable uncertain. Cf. L. munusculum, O. muiniku, etc. 21 c, 75 b.
Museiate, name of a decuvia among the citizens of Iguvium. Dat. sg. lib 5, 5. Etym. uncertain, but. Dev., p. 356, suggests connection with the personal name Museia (C.I.L., XI, 4613, from Carsulae) and the gentile names Mussia, Museia, Musetia (Co., p. 576). 77 c.
muta mutu: see motar.
Naharcer, adj. of nationality 'Narcan. ' Gen. sg. Naharcer VIb 54, 59, Vila 12, 48; dat. Naharce Vila 12; ace. Na- barkum lb 17, Naharcom VIb 58, Vila 47. Cf. L. Nahartis, Nar. 46 e, 70; n. on VIb 54.
naraklum, ' (oracular) response. ' Nom. sg. Ila 1. From *gna-ra-tlo-, cf. naratu, L. gnarus, narro, all from an ablaut- variant of root *gn6- ' know. ' 47 d, 52 b, 75 b, 76 a, SO.
naratu, 'announce, pronounce, proclaim, speak, recite. ' Imv. naratu Ila 3, b 8, 9, 11, 25, III 27, naratu Via 22, 56, 59, VIb 2, 4, 20, 23, 44, 46, Vila 5, 7, 38, 42, 53. From *gnd-rd-, cf. narak- lum, L. gnarus, narro (with gemina- tion not easily explained), all from an
ablaut-variant of root *gn6- 'know.' 18 a, 47 d, 52 b, 76 a, 117 c, 127 b.
natine, in the social organization of Iguvium apparently equivalent to 'gens.' Abl. sg. Ila 21, 35, b 26. L. natio, cf. L. (g)nascor, gens, etc., all from root *gen-. 18 a, 31 e, 47 d, 52 b, 81 b; n. on Ila 21.
nei, prohibitive ' not. ' IV 33 (in neif ha- bas). Cf. O. nei, ni, L. ne. 112 g. Vocalism of the U. form not clear. Cf. also neip.
neip, negative and prohibitive 'not.' neip Ila 4, Va 29, neip Via 27, 36, 46, b 29, 51, Vllb 3, nep Via 6. nei- same as nei. -p < k'^(e), cf. ape, L. neque, atque. 31 b, 49 a, 52 a, 112 g, 113 1.
nepitu, vb. of uncertain meaning, possib- ly 'overwhelm with water.' Imv. VIb 60, Vila 49. Cf. L. Neptunus ( ?). 118 d.
nerf, nerus, 'elders, chief citizens.' Dat. pi. nerus VIb 62, Vila 13, 14, 28, 50; ace. nerf Via 30, 32, 39, 42, 49, 52, b 13, 32, 34, 59, Vila 17, 30, 48. O. niir, gen. pi. nerum (magistrate(s), ' dvriQ, Skt. nf- 'man, hero,' cf. Sa- bine-Latin Nero, Nerio. 29 k, 52 a, 65e,82 ; n. on Via 30.
nersa, conjunction 'until.' Via 6. Pro- bably from *ne-ddw., cf. L. ne-que and quon-dam. 112 g, 113 a, k, 154 b.
nertru, adj. 'left.' Abl. sg. (+ -ku -co) nertruku la 32, nertruco VIb 37, 39; nertru VIb 25. Gk. veqteqoq 'low^er, inferior, ' cf. Eng. north, the left being the less good or inferior side, and the north being at the left if one faces east, as conversely Skt. daksinah is 'right' and ' south. ' 29 a, 52 a, 76 c.
nerus : see nerf.
nesimei, adv. used like predicate adj. 'next to.' Via 9, 9. Apparently from *nedh-imny'd. O. nessimas, Olr. nes- sam. 9 c, 44 d, 73 a, 112 a, 113 d, 139 a.
ninctu, vb. of uncertain meaning, pos- sibly ' overwhelm with snow. ' Imv. VIb 60, Vila 49. Probably from *nink-t6d < *sning^'h-e-t6d. L. ninguit,
Miletinar
313
orer
cf. veCcpei, Eng. snow. 49 d, 52 b, 57 b, 119 d; n. on VIb 60.
niru, word-division and meaning un- certain, probably ' blaclc , dark. ' Ace. sg. niru lib 15. L. niger, nigrum (?). 47 c; n. on loc. cit.
nome numem, noun 'name.' Gen. sg. nomner VIb 54, 59, Vila 12, 48; dat. nomne Via 24, 24,31,31, 33, 33, 40, 40, 43, 43, 50, 50, 53, 53, b 7, 8, 12, 12, 14, 14, 27, 27, 33, 33, 35, 35, 62, 62, Vila 12, 13, 14, 14, 18, 18, 28, 28, 31, 31, 51, 51; ace. numem [sic] lb 17, 17, nome Via 30, 32, 39, 42, 49, 52, b 13, 32, 34, 58, Vila 17, 30, 47; abl. nomne Via 17, (+ -per) Via 23, 23, 25, 26, 34, 34, 35, 35, 44, 44, 45, 45, 54, 54, 55, 55, b 7, 7, 10, 10, 15, 15, 26, 26, 28, 28, 35, 36, Vila 10, 11, 19, 19, 22, 22, 26, 26, 32, 33, 35, 35. L. nomen, Skt. nama, cf. ovo/ua, Eng. name. 2 k, 13 a, 17 d, 29 f, 51 a, 52 h, 81 c, 90.
Noniar, proper name, possibly of a gens 'Nonia. ' Gen sg. Via 14. L. Nonius, cf. nonus ( ?). 68 e, 105 g.
nosue, conjunction 'unless.' VIb 54. From *noi-swai; for first part cf. nei; second part = sue sve. 21 a, 35 d, 112 g.
numem: see nome.
numer, name of a coin 'nummus. ' Abl. pi. Va 17, 19, 21. L. nummus, Doric vov/ijuog (Epicharmus) ; further connec- tions uncertain. 67, 145 b.
Nurpier, a personal or gentile name 'of Nurpius. ' Gen. sg. Via 12. Etym. unknown. 68 e.
nufpener, name of a coin, perhaps 'new dupondius. ' Abl. pi. Va 13. Etym. uncertain; possibly from *nowo-du- pend-iijo- with syncope of u after com- pletion of the change d > f , and iij treated as in 53 d. *nowo- : L. novus, veoQ (ew > ow is pit. ; cf. O. Nuvellum and 23). For -pend- cf. L. pendo, pondus. 35 d; n. on loc. cit.
nuviine, adv. 'last, the last time' (?). Ila 26. According to most, from or-
dinal stem *nowimo- 'ninth'; accord- ing to the view preferred here, from superlative *nowimo-, cf. L. novus, novissimus. 52 a, 73 a, 105 g, 112 b, 113 d; n. on loc. cit. nuvis, adv. 'nine times.' Ila 25. L. noviens, novies. 52 a, 105 g, 112 b.
ocar ukar, noun ' mount, ' always with reference to the Fisian Mount in Igu- vium. Nom. sg. ukar lb 7, ocar VIb 46; gen. ocrer Via 8, 19, 19, 29, 32, 39, 41, 49, 51, b 10, 13, 32, 33, 48; dat. ocre Via 23, 30, 33, 40, 42, 50, 52, b 7, 10, 11, 14,16,26,32, 34; ace. ocrem Via 49, 51, b 12, ocre Via 29, 31, 39, 41, b 31, 33; abl. ( + -per -per -pe) ukriper la 5, 8, 15, 17, 21, 25, 28, 31, ukripe la 12, ocriper Via 23, 43, 45, 53, 55, 58, b 1, 3, 6, 6, 9, 14, 19, 22, 26, 28, 35, ocreper Via 25, 34, 35; loc. ukre Va 16, ocre Via 26, 36, b 29, (+ -m) ocrem Via 46. L. ocris (Fest., p. 296 Li. ocrem antiqui ... montem confragosum vocabant), oxqiq, cf. also o^vq, L. acer. The ending of ocar ukar may reflect an alternate stem in -ari-, for otherwise we should expect -er as in pacer. 2 a, 12 a, 20 a, 35 d, 45 a, 51 e, 57 e, 77 e, 90, 91 d, 93 c, 142 d.
onse uze, ' shoulder. ' Loc. sg. uze lib 27, 28, onse VIb 50. L. umerus, cf. &no<;. 2 i, 20 a, 29 d, 52 d, 58 e, 61 a, 67.
ooserclom-e, name of a landmark used in defining the augural templum, possibly 'observation-post.' Ace. sg. (+ -e) Via 12. Probably from *ob-ser-klom, with vowel of uncertain identity syn- copated before -klom < -tlom. 2 g, 26 g, 75 b, 86 c; n. on Via 12.
opeter, upetu, 'choose.' Imv. upetu lib 1; imv. 3 pi. upetuta III 10; pcpl. pass. gen. sg. opeter Vb 9, 14; pcpl. (?) pass. ace. sg. upetu lib 8, 11, III 22, 26. O. ufteis 'voluntatis, ' L. opto, -are. 118 b; n. on lib 1.
orer ures, demonstrative and anaphoric pron., 'this, that, him, it.' Gen. sg.
314
IV. LEXICON AND INDEX VERBORUM
ures IV 33, orer Via 26, 36, 46, b 29 ; abl. sg. uru lb 18, uru VIb 55, (+ -ku) uraku Va 5. Etym. unknown. 109 h, 145 b.
orto, urtas, pcpl. pf. 'arisen, rising, oc- curred.' Nom. sg. neut. urtu I la 4, ortom Via 46, orto Via 26, 36, b 29; nom. pi. f. urtas III 10; abl. pi. urtes III 4. L. orior, ortus, Skt.rtds, cf. dgrog; or- in the Italic forms is probably from f-. 17 a, 90.
osatu, 'make.' Imv. VIb 24, 37. From *op(e)sd-. h. operor, -art, cf. U. ose (?). 57 d, 61 d, 116 a, 117 c, 124 c.
ose, noun 'effect' (?). Abl. (?) sg. Via 26, 36, 46, b 29. Probably equivalent to L. opere < *opesi; cf. U. osatu. 29 d, 83: n. on Via 26.
ostendu ustentu, 'present, exhibit, elect. ' Fut. pass. 3 pi. ostensendi Via 20; imv. ustentu la 3, 9, 12, 16, 23, 26, b 3, 6,
25, 28, Ila 6, 11, ustetu la 17, b 32, 43, Ila 12, b 9, 29, ostendu Via 20; imv. 3 pi. ustentuta III 5. From *ops- + tend-. L. ostendo, ostendito, cf. U. andendu, endendu. 2 i, 3 with d, 8 c, 12 a, 29 b, 55 c, 58 e, 60 a, 61 d, e, 114 e, 115 f, 119 a with n. 3, 123,127 c.
ote ute, conjunction 'or.' ute lb 24, 27, Va 23, b 2, 10, 13, 15, 18, ote Via 7, Vila 6. O. auti, avt, L. aut, cf. a^, avre, aiBiq. 10 b, 22, 31 e, 112 g.
out, uve, 'sheep.' Ace. sg. uvem III 8, 10, 12, 26, 31, uve Ila 10; abl. ( + -kum) uvikura III 28; ace. pi. uvef lb 1, out VIb 43. L. ovis, dig, Skt. (wis, Eng. ewe. 12 a, 54 b, 38 b, 77 a, 90.
p, abbreviation probably for L. pondo or its U. equivalent. Vb 9, 14.
paca, ablative noun with value of im- proper preposition 'for the sake of, for.' Via 20. Cf. pacer, pase, L. pax, etc. 45 a, 65, 147 f.
pacer, 'propitious.' Nom. sg. pacer Via 23, 30, 33, 40, 42, 50, 52, VIb 7, 11, 13,
26, 32, 34, Vila 14, 17, 31, 50; pi. pacrer VIb 61. From *pak-ri-. Mars. pacre, Pael. pacris, cf. U. paca, pase.
L. pax, paciscor, Skt. pdsas ' bond. ' 29 1, 32 d, 33 b, 45 a, 46 e, 53 b, 77 e, 90, 91 d.
Padellar, name of a goddess 'Padella. ' Gen. sg. Via 14. Probably from *Padenlds < * Pain Ids < *Patnolds. Cf. O. Patanai, L. Panda, pando, pateo. 32 e,41 f, 42 d, 60b, 75 a; n. on loc.cit.
pafe : see po-.
pane, conjunction 'when.' pane lb 40, pane Vila 46. L. qucunde, cf. O. pan, U. conj. pone pune. 49 a, 112 e, 113 a, k, 154 b.
panta, ' how great, whatever. ' Nom. sg. f. panta Vb 2; ace. panta Vb 3. L. quanta. 2 d, 49 a, 61 a, 71, 111 a.
panupei, adv. 'at any time.' Vllb 1. L. quandoque, cf. U. pane; for -pei cf. ape. 113 k, 1.
parfa, an augural bird of uncertain iden- tity, possibly a jay (?). Ace. sg. par- fam lb 13, parfa Via 2, 4, 15, 17, b 51; abl. parfa Via 1. L. parra, U. and L. forms both being from *par(e)sd. Con- nection with tpaQ ' starling, ' Goth. sparwa ' sparrow ' less sure. 59 e, 66,
pars, with est 'it is fitting, required.' Vllb 2. L. par (?). 59 a, 83.
pase, 'peace. ' Abl. sg. Via 30, 33, 40, 42, 50, 52, b 11, 13, 32, 34, 61, Vila 14, 17, 31, 50. L. pax, pace, cf. U. paca, pa- cer. 3 e, 65 f, 80 a.
pater, found only as a component of lupater, q.v. 38 a.
pehaner, pebatu: see pihatu.
peia, ' black. ' Ace. pi. m. peiu lb 24, peiu Vila 3; f. peia lb 27, peia Vila 6. L. piceus (?). 46 g, 68 a.
peica, 'magpie.' Ace. sg. peica Via 3, 4,
16, 17; abl. peica Via 1. L. pica, cf.
U. peicu, Skt. pika- (?) 'cuckoo,'
Germ. Specht (?). 11 c, 66. peico, 'woodpecker.' Ace. sg. peico Via
3, 4, 16, 17; abl. sg. peiqu Via 1. L.
plcus, cf. U. peica, etc. 11 c, 67. Peieriate, name of a decuvia among the
citizens of Iguvium. Dat. sg. lib 4, 4.
Etym. uncertain. 77 c.
or to
315
perne
peihaner: see pihaiu.
peiqo: see peico.
peiu peiu: see peia.
pelmner, 'meat.' Gen. sg. Vb 12, 17. From *pelp-men, with loss of p in -Ipm-; cf. L. pulmentum, pulpamentum, both with u < 0 < e before non-palatal /. 29 f, 55 d, 81 c.
pelsatu, vb. of uncertain meaning, here taken as 'bury.' Imv. pelsatu VIb 40, 40; gdve. nom. sg. m. pelsans Ila 43; ace. m. pelsanu Ila 6, III 32; ace. pi. f. pelsana la 26, pelsana VIb 22. Etym. uncertain; cf. Goth, ga- filhan ( ?), us-filhan ( ?) ' bury. ' 29 i, 90, 91 b, 153 i; n. on VIb 22.
pepersciist, pepescus: see perstu.
pepurkurent, vb. ' demand. ' Fut. pf. 3 pi. Vb 5. From *pe-p^k-usent (or *pe-pork-usent ?), cf. L. precor, posco
< pork-sko, Skt. pi'cchdti. 35 d, 52 g, 114 e, 124 a, 125, 138 a.
pequo, '(domestic) beasts.' Gen. pi. (?) Via 30, 32, 40, 42, 50, 52, b 13, 32, 34, Vila 17, 30. L. pecu, Skt. pasu, Goth. faihu. 38 a, 78 a, 90.
-per -per -pe, postposition ' for, on behalf of. ' ahtis-per III 24, 29; fratrus-per Ila 2, III 23, fratrus-pe III 28; nomne- per Via 23, 23, 25, 26, 34, 34, 35, 35, 44,
44, 45, 45, 54, 54, 55, 55, b 7, 7, 10, 10, 15, 15, 26, 26, 28, 28, 35, 36, Vila 10, 11, 19, 19, 22, 22, 26, 26, 32, 33, 35, 35; ocri-per Via 23, 43, 45, 53, 55, 58, b 1, 3, 6, 6, 9, 14, 19, 22, 26, 28, 35, ocre-per Via 25, 34, 35, ukri-per la 5, 8, 15, 17, 21,25,28,31, ukri-pe la 12; Petrunia-per IIa21 ; pop Zu-per VIb 43,
45, Vila 3, 6, 9, 10, 18, 21, 24, 26, 32, 35, 37,41, 53, puplu-per lb 2,5; re-per Vllb 2 ; tota-per Via 23, 25, 34, 35, 43, 45, 53, 55, 58, b 1, 3, 6, 7, 9, 15, 19, 22, 26, 28, 35, 43, 45, Vila 3, 6, 9, 10, 19, 21, 24, 26, 32, 35, 37, 41, 53, tuta-per la 5, 8, 12, 15, 19, 21, 25, 29, 31, b 2, 5, III 29, tuta-pe III 24; trefi-per III 25, 30; Vugiia-per lib 26. From *pf
< *pro. Cf. pro- pru- (134 o), O. pru, L. pro. 4, 32 d, 56 d, 147 g, 155 g.
perakne, adj. of uncertain meaning, generally applied to sacrificial vic- tims; possibly 'brought from away, from elsewhere' (that is, not raised as temple-property). Ace. sg. peraknem Ila 10, perakne Ila 5, 5, 12, 14, b 7, 10 ; ace. pi. neut. perakneu Va 7. peracnio Via 54 must be emended to peracrio. Possibly from *per-ag-ni- with adj. suffix -no- replaced by -ni- after i- stem sakri-; cf. L. per-ago. 10 c, 41 f, 77 d, 86 a, 90; n. on Ila 5.
peracri, adj. of uncertain meaning, pro- bably ' perfect, fit for sacrifice. ' Ace. sg. f. perakre lb 40; abl. sg. m. pe- racri Via 34, 35, 38, 43, 45, 48, 53, peracrei VI a 25, 29; gen. pi. peracrio Vila 51, Via 54 (misspelled peracnio); abl. peracris VIb 52, 56. Probably from *per-dk-ri-. L. deer, dcris, cf. U. ocar, ocri-, dxixrj, axfialoQ, with semantic development from ' sharp, peaked, ' through 'topmost.' 13 a, 77 e, 86 b, 90; n. on Via 25.
Peraznanie, name of a decuvia among the citizens of Iguvium. Dat. pi. lib 7. No clear etym. 68 e.
perca, 'wand, staff (?). Ace. sg. percam VIb 53, perca Via 19, b 49, 50, 63, Vila 46, 51; ace. pi. perkaf lb 15, perca VIb 51. L.perfica,0.perek. (for perekais). 29 g, 41 g, 70; n. on Via 19; on the whole etymological group see What- mough, Lang., XXIX (1953), pp. 297-8.
peretom, pcpl. pass. nom. sg. neut. trans- lated as noun 'transgression. ' Via 27, 37, 47, b 30. From per- (L. per, cf. O. pert) + *itom (or *eitom ?), pcpl. pass, of etu, cf. daetom. For semantic value cf. naQa-^aivoi. 30 c, 114 e, 122 b, 131, 149.
pernaiaf, adj. 'from in front.' Ace. pi. f. pernaiaf lb 10; abl. f. pernaies la 2. From *pernai-yd-, cf. perne. 19 c, 53 c, 68 c.
perne, adv. 'before.' VIb 11. Lith. pe'r- nai 'in the previous year,' cf. neqvai. For -ne cf. U. postne, superne. 112 c, 113 j.
316
IV. LEXICON AND' INDEX VERBORUM
persae pefae, adj. 'on the ground, stretched on the ground. ' Nom. sg. neut. perae I la 22; ace. sg. m. peraem I la 11, III 32; neut. pefae Ila 13, persae Via 58, b 3; ace. pi. f. pefaia lb 28, 32, 44, persala Vila 7, persaea Vila 41, 54. From *pedaiyo-, cf. perso pefum, persi peri, L. pes. 19 c, 68 c, SO, 139 b ; n. on Via 58.
persclo persklum, 'prayer, ceremony, acrifice. ' Gen. sg. perscler Via 27, 28, 37, 38, pescler Via 47, 48, b 30, 30; ace. persklum la 1, (+ -af) persklu- maf III 21, persclo Via 1; abl. pers- klu III 12, persclu VIb 36, Vila 20, 24, 34, pesclu VIb 15, Vila 8. Probably from *perk-sk-lo, with -sk- through influence of a verb equivalent to L. posco. Cf. pepurkurent, etc. 2 a, j, 3 f, 35 d, 41 e, 59 d, 75 a.
perse pefe, conjunction 'if,' sometimes 'vv'hen, while.' pefe Ila 3, perse Via 47, b 29, 30, 31, persi Via 37, 38, persei Via 26, 27, 28, 36, pirse Via 46, pifi IV 32, pirsi Via 5, 48. From '*pid-l < *k^''icl-l; for the first part cf. O. pid, L. quid. 10 c, 112 e, il3 g. See also pis.
persi pefi, 'foot.' Abl. sg. pefi la 29, 32, persi VIb 24, 37, 38, 39, (+ -co) persico VIb 25, where the actual read- ing perstico must probably be emended. L. pes, pedis, Skt. pat, padcis, (with ablaut-variation) novg, nodog. 2 a, 38 a, 42 h, 65 f, 80 c, 90.
persnimu pesnimu, deponent vb. 'pray.' Imv. pass, persnimu lb 7, 21, IV 8, 10, persnilimu Ila 27, 29, 30, 30, 31, 36, 36, 37, 38, 39, 42, IV 11, 23, 25, 29, pesnimu la 6, 10, 13, 19, 23, 26, 34, b 3, 22, 26, 30, 32, 37, 38, 44, Ila 7, 10, b 18, 18, 20, 20, 20, 20, persnimu Via 55, 59, b 2, 4, 6, 9, 20, 25, 37, 41 44, 46, Vila 4, 7, 25, 25, 34, 42, 44 54, persnihimu VIb 17, Vila 9, 39, 45 pesnimu VIb 9, 23; imv. pass, pi persnimumo VIb 57, persnihimumo Vila 47, pesnimumo VIb 64, 65, Vila 1 pcpl. pass, persnis VIb 39, pesnis VIb 40, 41. Apparently denominative from
*persk-ni- with -sk- of same origin as in persclo, q.v. Cf. persclo, pepurku- rent. 2 i, 3 with f, 11 a, 59 d, 77 d, 114 e, 120 b, 131.
perso pefum, 'mound, ground, turf- altar' (?); according to another view 'pit, trench.' Ace. sg. pefum la 29, 32, pefu Ila 9, 24, perso VIb 24, 37, (+ -e -e) pefume Ila 27, III 33, per- some VIb 38, 39, 40. From *pedom. jiedov, cf. persi pefi, L. pes. 67; n. on VIb 24.
persondru, persuntru, noun of uncertain meaning designating a type of offering, here left untranslated. Dat. sg. per- suntre IV 21; ace. persuntru IV 17, 19, persutru lib 13, pesuntru, la 27, pesuntrum la 30, pesutru Ila 8, pesondro VIb 24, 37, 37, 39, 40; abl. persondru VIb 31, 35, persontru VIb 28; abl. pi. (+ -co) pesondrisco VIb 40. per- as in L. per, etc. (?); for -sond- cf. Olcel. synd 'sin,' h. sons, in-sons. (?). 2 i, 59 d, 60 a, 76 b; n. on VIb 24.
perstu, vb. 'place, put.' Imv. perstu Ila 32, pestu lib 19; fut. pf. 3 sg. peperscustYIh 5, pepescus Vila 8. From *perk-ske-t6d cf. L. parco, com-pesco. 59 d, 114 e, 119 e. MuUer, p. 320, sug- gests ultimate identity of this root ('umgeben, einschranken') with that of persclom, pepurkurent, L. posco, pre- cor, etc. ('bedrangen, von alien Seiten her [be]fragen').
pert, preposition 'beyond, on the other side of. ' Ila 36 (but -pert in Ila 35 must be emended to -per). O. pert 'across,' cf. U. per-etom; -t < -ti as in post < *pos-ti. 147 h.
pertentu, 'stretch forth, place ... over.' Imv. Ila 31, IV 8. From per- (as in peretom) + tend-. L. pertendo, cf. U. andendu, endendu, ostendu. 3 d, 119 a with n. 3.
periom-e, 'gate' (?). Ace. sg. (+ -e) Via 14. Probably from ^perto- with transfer from u-stem; cf. L. partus, porta, ON fiprar, Eng. firth. 71; n. on loc. cit.
persae
317
po-
pescler, pesclu : see persclo.
pesetom, pass. pcpl. nom. sg. neut. trans- lated as noun 'sin.' Via 27, 37, 47, b 30. Cf. L. pecco, peccatum; further con- nections uncertain. 30 c, 34, 46 b, 116 a, 149.
pesni-: see persnimu.
pesondro, pesuntru: see persondru.
petenata, pcpl. 'shaped lilce or furrowed as if by a comb, comb-(cake) '. Ace. sg. f. IV 4. From *pekten-a-tam, pcpl. of denominative vb. to *pekten, cf. L. pec ten, xreig < *pkten-. 46 i, 71.
Petrunia-per, name of a gens (?; na- tine) ' Petronia. ' Abl. sg. f. + -per Ila 21 (misspelled -pert), 35. L. Pe- tronia (of dialectal origin), U. peturpur- sus, etc. 68 e, 105 d, n. 2.
peturpursus, 'quadruped, beast.' Dat. pi. VIb 11. From *k^etur-pod-ufs (quantity of root-vowel uncertain), cf. dupursus, L. qiiadrupes; also U. Pe- truniaper. 13 c, 42 b, 49 a, 65 f, 80 c, 88, 105 d.
pihaclu, ' propitiatory offering. ' Abl. sg. pihaclu Via 25, 29, 34, 35, 38, 43, 45, 48, 53, b 28, 31, 35; gen. pi. pihaklu Va 8, pihaclo Via 54. From *piy-a-tlo-, L. piaculum, cf. pius, O. Piihiui, U. pihatu. Further connection of this group with L. purus, etc., is possible if *piyo/a- is from *pw-iyo/a. 11 b, 13 b, 41 e, 75 b, 90.
pihatu pehatu 'purify. ' Imv. pehatu III 3, pihatu Via 29, 29, 29, 30, 39, 39, 39, 40, 49, 49, 49, 50, b 31, 31, 32; pf. subj. pass. 3 sg. pihafi Via 38, 48, b 31, pi- hafei Via 29; pcpl. pass. nom. sg. m. pibaz lb 7, pihos VIb 47; gdve. gen. sg. m, pihaner Via 19, b 48, pehaner Via 20, peihaner Via 8. From *piyd-, cf. pihaclu, etc. 2 a, d, 11 b, c, 29 i, 61 e, 114 e, 117 c, 124 d, 126 c.
Piquier, with Martier, adj. from the name of the deified Picas Martins. Gen. sg. Vb 9, 14. From *plk(u)w{i)yo-, cf. peico, L. picus. 68 e.
pir, noun 'fire.' Nom. (?) sg. pir Via
20; nom. sg. 26, 36, 46, b 29; ace. pir lb 12, 12, lla 19, b 12, III 12, 21, pir VIb 49, 50 (heteroclite form + -e) purome VIb 17, Vila 38; abl. sg. pure lb 20, (+ -to) pureto Via 20. tzvq. Germ. Feuer, Eng. fire, cf. O. purasiai. 10 b, 14 a, 15 a, 38 a, 55 c, 65 e, 82, 84 n. 1, 90. pirse, pirsi: see perse.
pis, pisi, indefinite and (Vllb 2) relative pronoun ' any, whoever. ' Nom. sg. m. pis{est) VIb 53, pisQier) VIb 41, (sve)- pis lb 18, IV 26, {so)pir VIb 54, pisi Va 3, 10, pisi Via 7, Vila 52, b 1; neut. pife Va 5, pere lb 18, pirse VIb 55; ace. pi. f. pifi Vllb 2. From */c«'/- (O. pis, pid, L. quis, quid, nq, ri) partly extended by -i. 4, 10 a, c, 42 b, 49 a, 57 d, 62 g, i, 105 a, d, 110 with a, b. See also perse pere, poe poi.
pistu, word-division and meaning un- certain, probably ' meal ' (crushed with a pestle). Ace. sg. lib 15. L. pinso, pistus, Skt. pinasti, pistas. (?). N. on loc. cit.
plenasier, 'annual festival' (occurring when the year is full). (?). Abl. pi. Va 2, 14. Equivalent as if to L. *plendrius, cf. U. plener, L. plenus. 53 e, 68 d; n. on Va 2.
plener, 'full.' Abl. pi. Vila 21, 34. L. plenus, cf. nXrjQ-t]Q, Skt. pUrnas, Eng. full. 9 a, 52 a, 55 c, 74 b.
po-, pa-, stem of indefinite (lb 8, VIb 47) and relative pron. ' any, who, which, ' Nom. sg. m. poe VIb 50, poi Via 5, b 24, 53, poei Via 1, porse VIb 63, Vila 46, 51, porsi Via 6, porsei Via 9, 9; f. (sve)pu lb 8, (sue)po VIb 47; neut. pure Ila 26, III 5, Va 7; dat. sg. pusme Ila 40; abl. f. pora VIb 65, Vila 1; nom. pi. m. pure Va 6, 25, 28, b 4, pari Vb 10, 15; neut. porsi Via 19, porsei 15; ace. f. pafe Vila 52; neut. porse VIb 40. From *k'^o/a- (O. pui, pai, pud, L. qui, quae, quod, cf. norsQog, nu>g, nfi, etc.) partly extended by i-. 12 a, 21 a, 49 a, 57 b, c, 106 d, 110
318
IV, LEXICON AND INDEX VERBORUM
with b, c, d, e and n. 1, 112 e, 113 g.
See also pis, pisi. pone pune, conjunction 'when, after.' pune lb 1, 10, 11, 12, 15, 19, 33, Ila 1,
7, b 16 (second instance), 21, 22, 27, Va
8, puni lb 20, pone VIb 48, 49, panne VIb 43, Vllb 2. From *k'^om-de. O. pon pun pun, cf. L. quom cum; for second part cf. also U. pane. 2 a, 61 e, 112 e, 113 a, k, 154 b.
poni puni, a substance of unknown iden- tity used in the sacrificial ceremonies of the Atiedian Brothers; according to the view preferred here, 'mead. ' Gen.sg. punes Ila 41, IV 33; ace. pune Ila 18, 33, 40, b 14, 16 (first instance), IV 30; abl. pvmi la 4, 9, 13, 16, 22, 26, 32, b 3, 7, 25, 29, 32, 44, Ila 7, 11, 13, 20, 24, 25, b 9, 20, 29, poni Via 57, b 1, 3,
9, 20, 22, 44, 46, Vila 4, 7, 41, 54, pone Via 59. From *p6-ni-, cf. L. potus, potio, Skt. pdnlya- 'drinkable.' (?). 13 a, 77 d, 90, 145 b; n. on Via 57^
ponisiater punigate, ' puniceatus' (a pur- ple-clad official). Gen. sg. punigate lb 15, ponisiater VIb 51. Related to L. puniceus as L. armatus to arma, to- gatus to toga, etc. 21 a, 46 d, 71, 117 c; n. on VIb 51.
poplom puplum, 'people.' Gen. sg. popler Via 19, Vila 16, 27, 30; dat. pople VIb 61, Vila 14, 17, 27, 31, 50; ace. puplum lb 10, puplu lb 40, poplom Vila 15, b 3, poplo VIb 48, Vila 29, 46; abl. poplu VIb 54, 54, (+ -per -per) pupluper lb 2, 5, pop- luper VIb 43, 45, Vila 3, 6, 9, 10, 18, 21, 24, 26, 32, 35, 37, 41, 53; loc. pople VIb 55. L. populus, cf. U.Pupfiges (?), Etr. Fufluns (?), Etr.-Lat. Populonia. No sure connections outside Italic. The whole group may be of Etr. ori- gin; see on III 26, and E.-M.^ under populus.
pora : see po-, pa-.
porca purka ' pig. ' Ace. pi. purka lb 27, porca Vila 6. L. porca, cf. porcus, Olr. ore, OE fearh, OHG farah. 45 a, 55 b, 58 b, 66.
porfatu purtatu, 'bring.' Pres. subj. 3 sg. portaia Vllb 1; imv. purtatu lb 18, portatu VIb 55; fut. pf. 3 sg. portust Vllb 3. L. porto, cf. porta, portus, TtoQog , neiQOi. 7 a, 12 d, 53 b, 114 e, 116 a, 117 d, 124 b, 125, 126 a, 127 b.
post -pus, preposition 'behind. ' pus la 7, 14, 24, post Via 58, b 3, 22, Vila 38. O. piist post, L. post, all from *posti. 31 d, 147 i. See also posiertio.
postertio pustertiu, 'when ... for the third time. ' postertio Vila 46, pus- tertiu lb 40. Combination of post + tertio< abl. *terti6d. 113 c.
posti pustin, preposition with distribu- tive force ' each, at each, per. ' pustin Ila 25, IV 13, pusti Va 13, 18, 20, 21, posti Vb 8, 12, 14, 17. O. pustin, both from *postyen< *posti + en (?). For the semantic value compare the distri- butive use of Germ, nach, Eng. after. 147 j.
postne, adv. 'behind.' VId 11. From *post{i) + ne, cf. perne, superne. 112 c, 113 j.
postra, pustra, adj. 'at the back; the latter.' Ace. pi. f. postra Vb 13; neut. pustra Ila 32, pustru lib 19, postro VIb 5, Vila 8. O. piistrei, L. posterns. 29 a, 76 c, 139 b. See also postro adv.
postro pustru, adv. ' back. ' pustru lb 34, 36, postro Vila 43, 44. From *poster6d (?) with same stem as pos- tra, q.v. 112 c, 113 c.
pracatarum, 'rampart' (?). Gen. pi. f. Via 13. Pcpl. pf. pass, probably of denominative vb. to praco. 27 b, 57, 71, 90, 100 a, 142 b. N. on loc. cit.
praco, a landmark used in defining the augural templum, possibly 'tower' (of a rampart). Ace. sg. (or gen. pi. ?) Via 13. Probably from root-stem ♦pf/c- or *pj'k- (gradation uncertain). Cf. L. com-pesco (<,*-perk-sk6 or *park- sko), comperce in Paul, e Fest., 52 L. ; comperce pro compesce dixerunt antiqui ; U. perstu, Eng. park. 65 f , 80 a, 142 b ; n. on loc. cit.
pone
319
prusegia
pre, preposition 'before, in front of. ' pre la 2, 11, 20, pre Via 22, 59, b 1, 2, 4,
19, 20, Vila 7. O. prai, L. prae. 19 a, 147 k.
prehabia prehubia, ' provide. ' Pres. subj. 3 sg. prehabia Va 5, prehubia Va 12. pre- + habia; see habe. 28 a, 126 a.
prepa, temporal conjunction 'before, un- til. ' VIb 52. From *prai-k'^'am cf. pre and L. quam, priusquam. 49 a, 112 e, 154 b.
prepesnimu, ' offer a preliminary prayer. ' Imv. pass. lib 17. pre- + persnimu.
preplotatu, ' trample under foot ' (?). Imv. preplotatu VIb 60, preplohotatu Vila 49. Probably denominative vb. in sense 'flatten' from same stem seen in plotus Fest., p. 274 Li. 117 e; n. on VIb 60.
presoliaf-e, a landmark used in defining the augural lemplum, possibly 'fore- area.' Ace. pi. (+ -e) Via 12. First part pre; for second part cf. L. solum (?). 55 b, 68 e.
Pres tote Prestate, name of a goddess 'Prestota. ' Gen. sg. Pres to tar Vila
20, 22, 33, 36; dat. Prestate lb 27, Prestote Vila 6, 8, 24; voc. Prestota VIb 57, 61, Vila 9, 11, 13, 15, 16, 18, 19, 21, 22, 25, 26, 28, 29, 31, 33, 34, 35. From *prai-std-td or *prai-std-ta, vowel- grade of root uncertain. Cf. stahitu, L. sto, stare, etc. 7 b, 28 c, 31 a, 71, 86 d, 90.
pretra, adj. 'the former.' Ace. pi. f. Vb 12, From *praitera, cf. pre, L. praeter. 76 c.
preve, adv. 'separately, in a separate place. ' la 28, Ila 9. From *preiwed, cf. L. privus, O. preiuatud 'reo'. 9 a, 20 a, 112 a, 113 d.
preuendu, transitive vb. 'turn ... against. ' Imv. Vila 11. pre- + verb from root *wend-, cf. Goth, bi-windan, wandjan. Germ, winden, Eng. vb. wind. 60 a n. 2, 119 a.
prever, adj. 'single, one ... per ... ' Abl.
pi. Va 13, 18. From *preiwo-. L. pri- vus, cf. U. preve, O. preiuatud. 69; n. on Va 13.
preuislatu, 'bind.' Imv. preuislatu Vila 49, preuilatu VIb 60. pre- + denomi- native vb. from noun equivalent to L. vinculum. 29 e, 46 c, 75 a, 117 c; n. on VIb 60.
prinuatur, name of certain officials, here latinized 'prinuati.' Nom. pi. prinu- vatus lb 19, 23, 23, prinuvatu lb 15, 41, prinuatur VIb 50, 65, Vila 1, 46, 52; abl. prinuatir VIb 55, 56, 57. Etym. obscure ; apparently pri- (cf. L. prior, etc., Pael. pri-stafalacirix) + root of unknown identity + dto- as in pf. pcpl. of first conjugation. 2 h, 57 h, 71; n. on VIb 50.
procanurent, 'sing forth.' Fut. pf. 3 pi. Via 16. From *pro-kan- cf. L. cano, U. kanetu. 28, 114 e, 134 o.
promom, adv. 'first.' prumum III 15, prumu III 3, 23, promom Vila 52. TZQOfiog ' chief, ' cf . -per < -pro. 105 a, 113 a.
proseseto, prusekatu, etc., 'cut, cut off.' Imv. prusekatu Ila 28, 28 (second instance written prusektu, probably by error). III 33, 35, IV 2; pcpl. pass. gen. pi. proseseto VIb 16, 38; dat. prusegete Ila 12, prosesetir VIb 44, 46, prosesetir Via 56, 59, b 2, 4, 23, Vila 4, 8, 42, 54, proseseter VIb 20; ace. pi. neut. pruse- getu lib 12, proseseto Via 56. From *pro-sek{d)-. L. seco, -are, Russ. prose- katb 'cut through. ' 8 a, 30 c, 117 b, 134 o.
prufe, adv. 'well. ' Va 27. L. probe, both from *pro-bhwed; cf. O, prufatted 'probavit. ' 54 d, 69, 112 a, 113 d, 139 a.
prumum prumu: see promom.
prupehast, ' purify (in advance). ' Fut. 3 sg. IV 32. pro- + pihatu. 29 m, 114 e, 123, 134 o.
prusekatu, prusegete: see proseseto.
prusegia, ' sacrificial pieces. ' Ace. pi. neut. Ila 23. From *pro-sek-yo-, cf. proseseto. 68 a, 86 e; n. on loc. cit.
320
IV. LEXICON AND INDEX VERBORUM
prusikurent, ' declare. ' Fut. pf . 3 pi. Va 26, 28. From *pro-sek-usent, cf. L. in-seque, in-sece, ev-vsjie, evi-ane, E-G7iere, U. sukatu. 9 b, 35 d, 49 d, 52 g, 114 e, 124 c, 125, 134 o, 138 a.
pruzure, adj. 'remaining' (?). Abl. sg. IV 23. Probably from *pr6(cl)-sod- with ablaut-variant of root *sed- 'sit.' Cf. zeref serse, L. prae-ses, pro-seda, etc. 35 d, 65 f, 80 c, 86 e; n. on loc. cit.
pue, conjunction 'where.' pue lb 18, pue VIb 38, 39, 40, 55. Probably from *k'^6-i, cf. poe poi, etc. under stem po-. 112 e, 113 c, g.
Puemune, name of a god latinized as ' Pomonus. ' Gen. sg. Puemunes IV
3, 11, 12, 26; dat. Puemune III 26, 35, IV 5, 10, 12, 24, 26. Cf. L. Pomona, Sab. or Vest. Poimunien (Co. no. 248).
4, 74 f.
pufe, conjunction 'where, in which.' puJe lb 33, pufe Via 8, b 50, Vila 43. O. puf, L. iibi, all from *k'^u-dhe. 31 e, 112 e, 113 f.
purape, with preceding pisi, ' whoever. ' Va 3, 10. With pisi pumpe cf. L. quicumque, the L. and U. forms being fully identical so far as the second and third parts are concerned. 49 a, 106 d, 110 a, b, 112 e, 113 1.
pumpefias, adj. 'quincurial (represent- ing five decuviae). ' Nom. pi. lib 2. O. pumperiais, cf. U. puntes, L. quinque, jievre, Aeol. sie/nTis. Stem-for- mation partly obscure. 8 g, 49 a, 68 e, 105 e; nn. on lib 1-21 and 2.
pune, puni: see pone and poni.
punigate : see ponisiater.
puntes, ' groups of five. ' Nom. pi. puntes III 9, 10; abl. puntis III 4. From *ponk{^)-ti-, cf. O. pontis ' quin- quiens,' Puntiis, L. Quintius. 8 g, 49 a, d, 53 b, 77 b, 90, 105 e; n. on III 4.
puplu: see pop lorn.
Pupfiges, epithet of the god Puemune ' Pomonus, ' tentatively latinized ' Pop- licus. ' Gen. sg. Pupriges IV 4, Pupri-
kes 11, 13, Pupfges IV 26; dat. Pu- prige IV 24, Puprike III 27, 35, IV 10, 12 (erroneously written Pupf ikes). Etym. uncertain; cf. poplom{1), Etr. Fufluns (?). 46 e, 70; n. on III 27.
purka: see porca.
purdouitu purtuvitu, 'offer, present, make the presentation or porrectio'; sometimes ' complete the sacrifice. ' Fut. purtuvies lib 28; imv. purtuvitu Ila 24, 29, III 33, IV 1, 4, 6, 16, 18, 22, purtuviSu IV 20, purtuvetu lib 17, purtuetu lib 11, purdouitu Via 56; fut. pf. 2 sg. purtingus lb 33, purtiius la 27, 30, 33, Ila 7, 9, (written purtiiu- sufu in connection with sufu follow- ing); 3 sg. purdinsiust Vila 43, purdin- sust Vlb 16, 24, purdinsus VIb 23, 37, 38; pcpl. pass, nom. sg. neut. purtitu lb 39, Ila 43, IV 31, Va 18, purditom Vila 45, purdito VIb 42; ace. pi. f. purtitaf la 18, 18, purdita VIb 18. From *pr-dow-i-. For prefix cf. L. por-ricio, por-rigo, v/ith par- < pf-, an ablaut-variant of per. For verb cf. L. do, duim, etc. lb, 2 a, c, ii, 11 b, 12 b, 15 b, 17 a, 29 m, 35 d, 46 d, 54 e, 114 e, 120 a, 123, 124 b, f, 127 b, 131, 149, 152 e.
pure: see pir and po-.
puri: see po-,
purome: see pir.
pufe: see porse.
purtatu: see portatu.
purtif ele, ' fit for presentation. ' Ace. sg. m. lib 25. From pit. *por-divi- flim, related to purdouitu, purdito, as L. *porricibilis to porricio. 77 i.
purtiius purtingus, purtitu, purtuvitu, etc.: see purdouitu.
Purtupite, probably corrupt reading for Purtuvite, name of a god *Purdouif, L. ' Porriciens.' Dat. sg. IV 14. From *por-dowint-, pres. pcpl. stem of pur- douitu purtuvitu (?). N. on loc. cit.
pus: see post.
puse puze, conjunction 'that, as,' equi- valent in use to L. ul. -pw^e lb 34, Ila
prusikurent
321
sahatam
4, b 9, pusi Via 20, 46, b 4, 20, 29, 48, Vila 7, pusei Via 27, 36, Vllb 3, puse Via 59, b 2, 23, 37,44,46, Vila 5, 38, 42, 43, 53. From *k'«ut(i)-s-i (?), cf. O. puz 'ut,' L. ut, uti. 2 a, 14 a, 49 a, e, 112 e, 113 m, 154 a.
pusme : see po-.
puste, noun 'rear' (?). la 25. o-, a-, or z-stem derived from post (?). N. on loc, cit.
pustertiu: see postertio.
pusti: see posti.
pustnaiaf, adj. 'from behind.' Ace. pi. f. pustnaiaf lb 11; abl. pi. f. pus- naies la 2. From *post{i)nai-ya-, re- lated to postne as pernaiaf to perne. 2 i, 53 c, 68 c.
pustru: see postra and postro.
putrespe, pron. ' either, each of two. ' Gen. sg. IV 14. From *k'^o-tereis-k^e. L. utriusque, tioteqqv re, cf. U. sei- podruhpei. 49 a, e, 76 c. 111 b, 113 1.
puze: see puse.
randem-e, a landmark used in defining the augural i'emp/u/rj, possibly 'garden.' (+ -e) Via 14. No plausible etym. N. on loc. cit.
ranu, an object or substance of un- known identity, here left untranslated. Abl. sg. lib 19. No plausible etym. N. on loc. cit.
rehte, adv. ' right, in a satisfactory man- ner.' Va 24, 26, 29. L. recte, both from *rekted, cf. L. rego, rectus, etc. 9 a, 46 i, 56 a, 71, 112 a, 113 d.
re-per, ri, ' thing, ceremony (with esuna, -e), account.' Dat. sg. ri Va 4; abl. ri Va 5, (+ -per) reper Vllb 2. L. res, Skt. ras 'wealth. ' 25 d, 65 c, 79, 90, 93 e, 143 b.
restatu, reste, 'offer additional sacrifice, ' (in combination with feitu) 'perform sacrifice anew.' Imv. restatu Ila 5; pres. pcpl. nom. sg. m. restef lb 9, reste VIb 47. From re- + vb. forms equivalent to L. sto, stare, and sisto.
-ere. ( ?). 29 h, 58 d, 114 e, 117 a, 119 b, 130 a, 134 q.
revestu, ' examine, make an inspection. ' Imv. Va 7, 9. L. revisito, fut. imv. to reviso; cf. U. uirseto, L. video, etc. 119 f.
ri : see re-per.
rofu, adj. 'red. ' Ace. pi. m. rofu Vila 3; f. rofa Vila 6. L. (of dialectal origin) Rufus, cf. ruber, U. rufru. 24 c, 26 e, 35 d, 67.
Rubinam-e Rupinam-e, name of a place in or adjacent to Iguvium, ' Rubinia. ' Ace. sg. (+ -e -e) Rupiname lb 35, 36, Rubiname Vila 43, 44; loc. Ru- pinie lb 27, Rubine Vila 6. Etym. uncertain; Bii., p. 106, suggests con- nection with rusem-e, L. rubus and ruscum, Dev., p. 289, with Etr. ru- pinas, rupenial. 53 d, 68 e; n. on Vila 6.
Rufrer, masc. proper name ' Rufer. ' Gen. sg. Via 14. Same stem as rufru (?). 14 a.
rufru, adj. 'red.' Ace. pi. m. rufru lb 24 ; f. rufra lb 27. L. ruber, igvOgog, cf. U. rofu. 35 d, 43, 56 a, 76 a, 90.
Rupiname, Rupinie: see Rubinam-e.
rusem-e, 'trench' (?). Ace. sg. (+ -e) Vila 8, 9, 23. Etym. uncertain. N. on Vila 8.
S, abbreviation for semis or its equivalent in numerical symbol VI IS. Vb 17.
sacra sakra, adj. 'sacred, consecrated, belonging to the temple ' ; as noun 'sacred article, victim, young pig' (?). Gen. sg. (?) sakre III 30; ace. sg. sakre Ila 5, 6, 21, III 8, 9, 12, 22; ace. pi. f. sakra lb 29, 37, sacra Vila 40, 45, sakref la 18, 19; ace. pi. neut. sakreu Va 6; abl. sacris VIb 52, 56. L. sacer, -ra, -rum and -ris, -re, U. andOL both showing the same variation between o/a-stem and I'-stem. Cf. saha- tam. No sure connections outside Ita- Uc. 10 c, 57 a, 76 a, 77 e, 90; n. on Ila 5.
sahatam, except in lb 38 always preceded
322
IV. LEXICON AND INDEX VERBORUM
by traf iraha trahaf tra, the phrase forming an inflected hypostasis used as name of a place in or adjacent to Iguvium, latinized 'Trans Sanctam. ' Ace. sg. sahta lb 35, (+ -e) satame lb 38, sahatam Vila 39, 44, 45; abl. sahala Vila 5, 39; loc. sate lb 31, sahate Vila 41. L, sancta, cf. sacer, U. sacra, Sansi. 2 e, 7 c, 26 a, 46 i, 52 e; n. on Vila 5.
Salter, personal name ' Salius. ' Gen. sg. Via 14. L. Salius (?). 68 e; n. on loc. cit.
salu, 'salt.' Ace, sg. Ila 18. L. sal, salem, aXg, aXa. 65 d.
saluo, adj. 'safe.' Ace. sg. m. saluo Via 31, 41, 51, b 12 (first instance), 33, saluom Vila 15, 29; f. salua Via 31, 41, Vila 15, 29, saluom Via 51; neut. saluo Via 32, b 12 (second instance), Vila 16, saluom Via 51, b 33, Vila 30, saluuom Via 41; pi. f. salua Via 32, 52, b 13, 34, Vila 17, 31, saluua Via 42. O. salavs, O. aalaFz, L. salvus, oXoq, Ionic ovAog, Skt. sarvas. The Italic forms in contrast to the Greek may be based on development of ] as in 18 a. 2 h, 54 c, 69.
sanes, adj. 'uncooked' (?). Abl. pi. IV 8. L. scuius (?). 74 b; n. on loc. cit.
Sanst Sage, epithet of the gods Fisus, Fisovius, Jupiter, and Vesticius, and once (lib 10) used alone. Dat. sg. Sage Ila 4, Sagi, la 15, lib 10, 17, Sansie VIb 3, Sansii Vila 37, Sansi VIb 5, 6; ace. Sansi VIb 8, 8; voc. Sage lib 24, Sansie VIb 9, 10, 12, 12, 14, 15. Cf. L. Sancus (see on Ila 4, VIb 3), sancire, which according to Kretsch- mer, GZ.,X (1919), pp. 155-6, is denom- inative to Sanco-; cf. also sacra, L. sacer, the connection of which with L. sancio etc., Kretschmer denied. 2 c, i, 31 c, 52 d, 68 a, 90, 93 b; n. on VIb 3.
sarsite, adv. of uncertain meaning, pos- sibly 'in public' VIb 11. From *sarklted, adv. as if from a pcpl. *sar- klto- to vb, equivalent to L. sarcio, with semantic development 'sewn to-
gether > collectively > publicly. ' ( ?). 112 a, 113 d; n. on loc. cit.
Sage Sagi: see Sansi.
satame, sate: see sahatam.
Satanes, name of a decuvia among the citizens of Iguvium. Dat. pi. Satanes lib 4, Satane lib 4. No sure etym. 74 d.
sauitu, vb. of uncertain meaning, pos- sibly 'wound.' Imv. VIb 60, Vila 49, Cf. L. saucius (?). 118 d; n. on VIb 60.
scalse-to, a ceremonial vessel of some kind, probably 'cup,' Abl. sg. (+ -ta -to) skalgeta IV 15, 18, 20, scalseto VIb 16; loc. (+ -e) scalsie VIb 5, Vila 37. L. calix, xdKv^, xvXi^, axvXXiov, axaXig, Skt. kaldsas are all possibly or probably related. 2 a, 29 g, 80 a, 97,
scapla, 'shoulder-blade, shoulder.' Ace. sg. VIb 49. L. scapulae. 75 a; n. on loc. cit.
screhto, screihfor, 'written.' Pcpl. pass, nom. sg. neut. screhto VI lb 3; nom, pi. neut. screihtor Via 15. O. scriftas, L. scribo, scriptus. 11 b, c, 38 b, 57 a, 61, 90, 103 a, 131.
Sehemeniar, 'of Semonia' (?). Gen. sg. (?)
Sehmeniar lb 42, Sehemeniar Vila
52. Cf. sehmenier, L. semen, Semo,
Semonia. 77 t; n. on Vila 52. sehmenier, adj. 'of Semo' (?). Abl. pi.
semenies lib 1, sehmenier Vb 11, 16.
Cf. L. Semo, semen, etc. 25 b, 68 e; n.
on lib 1. sehemu : see semu. seipodruhpei, adv. ' in either direction. '
Via 11, From *se{d)-k"'otero-k'^e, cf.
putrespe, O. putereipid, L. sedutra-
que. 9 c, 49 a, 60 f, 76 c, 87 b, 111
b, 112 e, 113 c, 1. semenies: see sehmenier. semu, adj. 'half-way through, middle of.'
Abl. sg. semu VIb 16, sehemu VIb 36.
Cf. L. semi-, rj/ni-, Skt. sami-. 67, 73 a;
n. on VIb 15-16. seples, ' nails '( ?). Abl. pi. Ill 17. From
*saip-lo- or -/a-,cf, L. saepes, saepio. (?).
75 a; n. on loc. cit.
Salter
323
somo
sepse, adv. of uncertain meaning, possibly 'in private.' VIb 11. From *saip-sed, adv. as if from a pcpl. *saip-so- to vb. equivalent to L. saepio, with semantic development 'fenced off > separately > in private.' (?). 112 a, 113 d; n. on loc. cit.
Serfe, Serfer : see Serfer.
Serfia, Serfie: see Serfiar.
seritu, 'watch, watch for, observe; keep.' Imv. seritu Via 11, 15, 16, 31, 31, 32, 33, 41, 41, 41, 42, 51, 51, 51, 52, b 12, 12, 13, 33, 33, 34, 49, Vila 15, 16, 17, 29, 29, 30, 31, serituu Vila 15, seritu Ila 24. Probably from *serwi-. L. servio ( ?), cf. serous, servare, U. anseriato. 54 g, 114 e, 120 c.
serse: see sersitu.
sersi, noun 'seat.' Abl. sg. Via 5. L. sides, cf. edog, L. sedeo, Eng. sit. 57 a, 77 a, 145 c.
sersitu 'sit.' Imv. sersitu VIb 41; pcpl. pres. nom. sg. serse Via 2, 16, b 17, 22, 41, 41, 41, zeref la 25, 33, 34. L. sedeo, ci. e^o/biai, \J. sersi. 2 a, j, 42 b, 68 d, 114 e, 118 a, 130 a.
sese, probably preposition with abl. 'from, from the direction of.' Ill 23, IV 3, 15. From *sed-sed, cf. sei-po- druhpei, h. sed,sed-itio, se (Jraude). (?). N. on III 23.
seso, pron. 'for himself.' Dat. sg. VIb 51. From *s(w)oi-so, cf. ol, etc. (?). 107 c with n. 1.
sestentasiaru, adj. 'bi-monthly' (that is, occurring in cycles of one-sixth of a year). Gen. pi. f. Ill 2. From *sekst- ent-asia-som, cf. L. sextantariarum. 53 e, 61 d, 68 d, 105 f,
sestu, 'dedicate.' Pres. 1 sg. sestu lib 24, 2 sg. (?) seste lib 22; imv, sestu lib 22. L. sisto, cf. U. restatu, restef. 10 c, 22 b, 114 e, 115 a, 119 b, g.
sesust, 'sit.' Imv, sistu III 8; fut. pf. 3 sg. sesust Via 5. L. sido, imv. sidito < *sizd{e)t6d, Skt. sldati< *sizdeti, cf. U, andersistu, andersesust, sersitu. For- mation of fut. pf. sesust not quite clear,
but perhaps derived from pf. pcpl. equivalent to L. sessus. 61 c, 119 b.
seuacne sevakne, adj. 'perfect, free from fault, unblemished (?); solemn, ceremonial. ' Ace. sg. sevakni III 25, 26, 27, sevakne Ila 21, b 8, 8, 9, 10,
III 22, IV 18, 19; abl. sevakni Ila 38, 39, sevakne IV 23; ace. pi. sevaknef
IV 22, sevakne IV 16 ( ?), seuacne Vllb 1; abl. sevaknis Ila 36, 37, IV 25, sevakne IV 9. From *sew- (seuom, O. siuom 'omnino') + akni- (?); accord- ing to another view, from se- privative (cf. seipodruhpei, L. se jraude) -{- wak-ni (cf. ander-uacose, uasetom). 41 f, 77 d, 89, 90, 101, n. 1; n. on Ila 21.
seuom sevum, 'each, every.' Ace. sg neut. sevura la 5, seuom Via 56; abl. pi. seueir Via 18. From *se-wom, O. siuom, cf. L. so-lus, se-d, the under- lying sense being that of separate in- dividuality.
sihitir, sihitu: see sihitir.
sim, 'sow, pig.' Ace. sg. sim lib 1, si lib 7; ace. pi. sif la 7, 14, sif VIb 3, si Via 58. L. sUs, suem, ig, iv. 2 j, 15 a, 57 a, 65 b, 102 e.
sistu: see sesust.
smursim-e, a landmark used in defining the augural templum, 'temple of Smur- cia' (?). Ace. sg. ( +-e) Via 13. Etym. and interpretation uncertain; possibly from * Smorkiom (by 12 b, 32 a) as neut. place-name corresponding to L. divine name Morcia. 57 b; n. on loc. cit.
snata, pcpl. pass, 'wet' (?). Ace. pi. neut. snata Ila 19, snatu Ila 34; abl. pi. snates IV 9, snate Ila 37. Cf. L. no, nare, vrjxca, Skt. sndti, 'bathe,' Olr. sndm 'act of swimming'; cf. also U. asnata. 57 b, h, 114 e, 117 a; n. on Ila 19.
somo, 'highest; final (?)• ' Ace. sg. somo Via 9; abl. somo Via 10; loc. sume Ila 15, III 1. L. summus, cf. sub, super, O. avTi, U, sub- SU-, super. 13 a, 14 b, 61 e, 73 a, 90.
324
IV. LEXICON AND INDEX VERBORUM
sonitu, 'deafen with thunder' (?). Imv. sonilu VIb 60, sunitu Vila 49. Cf. L. sono, -are and -ere, Skt. svanaii. 8 g, 54 f, 118 d; n. on VIb 60.
sopa, ' under-parts ' (?); in VIb 17, Vila 38 adj. with value of adv. 'down' (?). Ace. sg. f. sopam Vila 38, sopa VIb 17; ace. pi. f. supaf I la 22; neut. supa la 9, 16, Ila 22, 30, 32, siipo Vila 8, sopo VIb 5; abl. pi. supes IV 8. 14 c, S4 d, 69, 139 b; n. on VIb 5.
sorsalir, adj. 'pig-, accompanying the p\g-persondro. ' Gen. sg. sorsalir VIb 38; ace. sg. sorsalem VIb 39. Gf. sor- som. 77 h, 90,
sorsom surum, adj. 'pig-'; with pelmner 'pork.' Gen. sg. sorser Vb 12, 17; ace. sufum la 27, 30, sufu Ila 8, 9, sorsom VIb 24, sorso VIb 38; abl. sorsu VIb 28, 31, 35, 37; ace. pi. m. sufuf la 33. Cf. Sim, sorsalir. 14 d, 72.
spahatu, ' throw, scatter (?) ; distribute' ( ?). Imv. spahatu VIb 41; imv. pass, spah- mu VIb 17, spahamu Vila 39; pcpl. pf. nom. sg. neut. spafu Va 20. Cf. ondco, L. pando, pansus, paieo, spatium, U. spanti, spantea, Germ, spannen; all are probably related, Vvith an under- lying notion of extension in space, but some phonological details are obscure. 44 b, £8 c, 114 e, 117 a, 131 ; nn. on Va 20, VIb 17.
spantea, adj. 'at the side. ' Ace. pi. neut. Ila 30. Cf. spanti. 53 b, 68 b, 139 b.
spanti, 'side.' Ace. sg. spanti III 34, IV 2, (+ -ar) spantimar III 33. For- mation obscure, but probably related to spahatu, q.v. 77 a.
spefa, pcpl. pass, always preceded by mefa, the two making up the name of a type of cake, here left untranslated. Ace. sg. fem. spefa Via 56, b 20, Vila 4, 38; abl. spefa VIb 5, 9, 9, 14, Vila 37. Cf. L. pendo, pensus. 44 b, 58 c, 114 e, 131; n. on Via 56.
Speture, name of a god ' Spector. ' Dat. sg. Ila 5. Cf. speturie, L. specio, specto, spectio, ay.enxofjLai, etc. 46 i, 82 a; n. on loc. cit.
speturie, adj. 'augural.' Dat. sg. Ila 1,
3. From *spekt6riyo/a-, to Speture,
L. -spector, as amatorius to amator.
68 e.
spinia, 'obelisk' (?). Ace. sg. spinia Ila
36, spina Ila 38, (+ -af -a) spinamaf Ila 33, spiniama Ila 37. From *spin- yd-, cf. L. spina. 53 d, 68 e.
stakaz, pcpl. pass, 'established, ap- pointed.' Nom. sg. m. Ila 15. From *sta-kd-tos, cf. stahitu, L. sto, stare, etc. 29 i, 117 e.
staflarem, adj. 'for Stabilis (Jovius)' (?). Ace. sg. staflarem VIb 39, staflare VIb
37, 40. Equivalent as if to L. *stabu- larem; cf. sto, stare, stabilis, U. stahitu, etc. 35 d, 43, 75 d, 77 f ; n. on VIb 37.
Stafli, in combination with luve mak- ing a divine name here latinized 'Sta- biUs Jovius' (?). la 30-1. L. stabilis, both from pit. *sta-pli-, cf. L. sto, stare, status, etc. 16 a; n. on loc. cit.
stahitu, 'stand.' Fut. 3 pi. staheren lb 19; imv. stahitu VIb 56; imv. 3 pi. stahituto VIb 53. From *sta-ye/o-, sta-i. O. stait, stahint, cf. L. sto, stare, U. stahu (minor inscr. no. 355 Co.), stakaz, Stafli, stahmei, statitatu. 57 c, 114 e, 115 d, 120 a, e, 123.
stahmei, 'ordinance' (?). Dat. sg. Via 5, 18. From *sta-mo- or *std-mi-, cf. stahitu, L. sto, stare. 35 d, 73, 143 d; n. on Via 5.
stahmito, pcpl. pass, 'established.' Nom. sg. neut. stahmito Via 8; dat. stahmitei Via 18, stahmeitei Via 5. From de- nominative vb. to stem contained in stahmei, q.v. 120 b.
statita, pcpl. pass. ' remaining. ' Ace. pi.
neut. Ila 42. From denominative vb.
to noun *stati- or *statu-, cf. stahitu, L.
sto, stare. 15 b, 57 a, 120 c, 131; n. on
loc. cit. statitatu, 'stop.' Imv. Ila 32, b 19, IV
9. Iterative (without iterative sense)
to the denominative vb. assumed for
statita, q.v. 77 b, 117 d. stiplatu, 'stipulate, bargain for, demand. '
Imv. stiplatu VIb 48, 51, steplatu lb
sonitu
325
surur
13; pres. imv. 2 sg. stiplo Via 2. L. stipulor, -ari, cf. U. anstiplatu. 10 c, 114 e, 117 c, 127 a.
strusla, a type of cake, here left untrans- lated. Gen. sg. struhglas Ila 41, IV 1; ace. strusla Via 59, VIb 5, 23, Vila 8, 42, 54, struhgla Ila 18, 28, IV 4, struQla III 34. From *struw{i)- k(e)ld-, cf. L. s trues. 15 c, 29 e, g, 46 c, 142 b; n. on Via 59.
stibahtu, subator, 'take away; omit' (?). Imv. subahtu Ila 42, subotu VIb 25; pcpl. pf. nom. pi. neut. subator Via 27, 36, 46, b 29. From sub- + hahtu, cf. habe. (?). 14 a, 38 b, 103 a; nn. on Ila 42, Via 26-7.
subocau, 'invoke.' Pres. 1 sg. subocau Via 22, 24 (twice), 34, 44, 55, VIb 6, 8, 8, 15, 15, 26, 27, 27, 36, subocauu Vila 20, 20, 22, 23, 33, 34, 36, 36. From *sub-wok-a-yd, cf. L. voco, -are, U. suboco. 49 d, 54 d, 114 e, 115 a, 117 c, f ; n. on Via 22.
suboco, noun 'the one invoked' (?). Ace. sg. (?) Via 22, 24, 25, VIb 6, 8, 8, 26, 27, 27. According to the view adopted here, root-stem noun from *sub-wok-, cf. subocau. N. on Via 22.
subotu: see subator.
subra, adv. 'above, over,' and (Via 15, second instance) preposition ' above. ' subra Va 20, subra Via 15, 15, b 17, 41, Vila 39, b 3. L. supra. 14 c, 76 a, 112 c, 113 e, 139 b, 147 1.
sukatu, 'pronounce.' Imv. IV 16. Pro- bably from *sok-a-t6d, cf. pru-siku- rent, L. in-sece, ev-vsns, evi-ane, Lith. sakau ' say. ' 35 d, 49 d, 117 b ; n. on loc. cit.
sue sve, conjunction 'if. ' sve Va 24, 24, 27, (in svepu) lb 8, (in svepis) lb 18, IV 26, sue Via 7, 16, Vllb 3, (in suepo) VIb 47. O. suae svai, cf. U. nosue, sueso; L. si < set, loc. of *so-, a stem- variant of *swo-. 19 a, 54 c, 109 j, 110 b, 112 d, 154 c.
sueso, reflexive pronominal adj. 'his.' Loc. sg. sueso Vllb 1, svesu lb 45, Ila 44. First element from *swo-, cf. O.
suveis, L. suus, U. sue sve; second element of uncertain origin. 107 n. 1, 108 c, 109 j.
sufafias, noun of uncertain meaning;
'limb' (?). Gen. sg. sufafias Ila 41;
ace. pi. sufafiaf Ila 22. Cf. L. effafi-
latus, exfilati (?). 63 e, 142 b; n. on
Ila 22. sufefaklu, noun of uncertain meaning,
possibly 'under-board. ' Ace. sg. Ill
17, 19. From *sub-fid-a-tlo-, with same
root as f ef ebtru. ( ?). 75 b; n. on III 17. sviseve, 'flask' (?). Loc. sg. (+ -e ?)
lib 14, 14, 15. Cf. L. sinum, silula. (?).
N. on lib 14. suma: see somo. sumel, adv. 'at the same time.' Ila
27. Cf. L. semel, simul. 28 b, 35 d, 55
e, 105 a, 112 g.
sumtu, 'take up.' Imv. la 9, 16. L.
imv. sumito, both from * sub -erne tod;
cf. U. emantur. 29 b, 61 a, 119 a. supa, supes: see sopa. super, preposition with loc, ' above, at
the top of. ' lb 41, IV 19. L. super, cf.
vTieQ, U. subra, superne. 14 a, 38 a,
61 e, 147 m.
superne, preposition with ace. ' above. ' Vila 25. L. superne, cf. U. perne, postne (for second element). 14 c, 112 c, 113 j, 147 n.
supru, adj. with value of noun 'the top* (?). Abl. sg. IV 3. From *superdd, cf. subra, super, O. supruis, L. superus. 60 e, 76 a.
supu, preposition 'at the bottom of (?). IV 17. L. sub, cf. vjto, Skt. upa; cf. also U. sopa. 31 f.
suront surur ont, adv. 'likewise, in the same manner.' suront VIb 8, 20, 23, 37, 44, 46, Vila 5, 7, 37, 42, 53; sururont VIb 39, 48, 51, 64, 64, 65, 65, Vila 1, 1. For first part cf. surur; for second part cf. erihont, eront, if ont, L. hie, hunc. 52 g with n. 2, 109 j, 112 d, 113 i.
surur, adv. 'similarly, in the same manner. ' surur Via 20, 20, 56, 59, VIb 2, 4, suror VIb 37, suroro VIb 48.
326
IV. LEXICON AND INDEX VERBORUM
Probably from *s6-s6-s(6), cf. 6, 1), Skt. sa, sd, etc. 52 g with n. 2, 109j, 112 d, 113 c.
suf u : see sorsom.
sutentu, 'place under.' Imv. I la 23. L. imv. subtendito, cf. U. andendu, en- dendu, ostendu, pertentu. 3 d, 38 b, 119 a with n. 3.
Serfer, always foUow^ed by forms of Mar- tier, name of a god ' Serfus Martius. ' Gen. sg. gerfe lb 28, 31, Serfer VIb 57, 58, 61, 61, Vila 6, 9, 10, 11, 13, 15, 18, 19, 20, 21, 22, 24, 25, 27, 29, 29, 32, 33, 33, 34, 35, 36, Serfer Vila 16, 23, 41; dat. gerfe lb 24, Serfe Vila 3; voc. Serfe VIb 57, 61. L. Cerus, O. dat. sg. Kerri, cf. L. Ceres, creo, cresco. 2 a, 46 h, 59 e, 67, 90; n. on VIb 57-8.
Serfiar, epithet of the deities Prestota, Tursa, and Hondus, ' Serfius, Serfia. ' Gen. sg. f. Serfiar Vila 20, 23, 33, 36; dat. gerfie lb 28, 31, gefi lb 4, Serfie Vila 6, 8, 24, Serfi VIb 45, Serfie Vila 41; voc. f. Serfia VIb 57, 58, 61, Vila 10, 11, 13, 15, 18, 19, 21, 22 (erroneous- ly written Serfiar), 25, 27, 28, 29, 32, 33, 34, 35, Serfia VIb 61, Vila 16. Adj. in -yo- to Serfer, q.v. 2 c, 31 a, 68 e.
gersiaru, 'festival of Ceres, Cereaha' (?). Gen. pi. Ila 16. From *ker(e)s-siijd-, cf. Serfer, L. Ceres. (?). 68 e; n. on loc. cit.
gersnatur, pcpl. pf. ' dined. ' Nom. pi.
m. Va 22. L. cenati, cf. U. sesna. 56 b,
59 d, 114 e, 117 c, 131. sesna, 'dinner.' Ace. sg. Vb 9, 13, 15, 18.
L. cena, O. kersnu, cf. kerssnasias,
U. gersnatur, Skt. krntdti 'cut.' 52
c, 55 b, 59 d, 74 c.
sihitir, 'in office' (?). Pcpl. pass. dat. pi. sihitir Vila 14, 28, 50, sitir Vila 13, sihitir VIb 62; ace. sihitu VIb 59 sihitu Vila 48. L. cinctis, -os (?). 26 a, 46 b, 52 e; n. on VIb 59.
gihgefa, noun of uncertain meaning, pos- sibly 'lattice frame.' Ace. sg. Ill 15. Cf. >ciyK?udei;, L. cingo, cinctus, U. sihitir. (?). 26 b; n. on loc. cit.
^imo gimu, adv. 'hither, back' (to place where the speaker is), gimu lb 23, 23, simo VIb 65, Vila 1. From *ki-mdd, cf. give, L. cis, citra, citimus. 2 a, 10 a, 46 b, 69, 73 a, 112 c, 113 c; n. on VIb 65.
sitir: see sihitir.
give, adv. ' on the hither side, outside. ' lib 11. From *ki-wei, loc. of *ki-wo-, cf. simo gimu, L. cis, etc.
t: see Titis. -ta: see -to.
tafle, noun 'board' (?). Loc. sg. lib 12. L. tabula, other connections uncertain. 75 d, 90.
Talenate, name of a decuvia among the citizens of Iguvium. Dat. sg. lib 4, 5. No clear etym. 77 c.
tapistenu, noun of uncertain meaning, possibly 'caldron.' Ace. sg. IV 30. From * tapestenom, cf. L. tepor, Skt. tapas. (?). 16 b, 35 d; n. on loc. cit.
Tarsinater, ethnic adj. 'Tadinate. ' Gen. sg. Tarsinater VIb 54, 54, 59, 59, Vila 12, 12, 48, 48; dat. Tarsinate Vila 11, 11; ace. Tarinate lb 16, 17, Tar- sinatem VIb 58, 58, Vila 47, 47. L. Tadinates, Ital. Gualdo Tadino. 7 a, 77 c; n. on VIb 53-4.
tases tagez, pcpl. pf. 'silent(ly), in a murmur. ' Nom. sg. m. tagez la 26, b 26, 30, 32, 44, Ila 7, 39, IV 27, tases Via 55, 59, b 2, 4, 20, 44, 46, Vila 4, 7, 42, 54, tasis VIb 23; nom. pi. m. tasetur VIb 57, Vila 46. L. tacitus, cf. Goth. pahan, OHG dagen, ' be silent. ' 2 a, 8 e, 28, 29 i, 30 c, 41 a, 46 b, 114 e, 118 a, 131,139 b.
tekvias, noun latinized to 'decuviae,' groups each originally composed of one tenth of the citizen body. Nom. pi. lib 1. Cf. dequrier tekuries, O. dekk- viarim, L. decuria, decern, dexa, etc. 2 h, 3 d, 45, 46 h, 68 e, 105 h; n. on loc. cit.
tekuries: see dequrier.
tefe, 2 sg. personal pron. ' thee. ' Dat.
suru
327
tigel
tefe lb 13, lib 24, tefe Via 18; ace. tiu Ila 25, 25, tiom Via 43, 44, 45, 53, 55, b 8, 8, 9, 14, 15, 15, 25, 27, 27, 28, 35, 36, Vila 10, 18, 19, 20, 21, 21, 23, 25, 32, 33, 33, 34, 34, 35, 36, tio Via 24, 24, 25, 33, 34, 35, 54, b 6, teio Via 22. O. tfei, L. tibi, cf. Skt. tubhyam; ace. tiu etc. equivalent to L. te but with an added element -om. 40, 107 b. tefra, tefrufo, 'burnt-offering, (place of) burnt-offerings.' Abl. sg. (+ -to) tefruto Vila 46; ace. pi, tefra Ila 27,
III 32, 34, IV 2. O. tefurum. Pro- bably from *tepsro-, cf. L. tepor, Skt. tapas; connection of ricpga very doubt- ful. 57 g, 76 a; n. on Vila 46.
Tefrali, adj. 'for Tefer. ' Abl. sg. VIb 28, 35. To Tefrei as L. Saturnalia to Saturnus, etc.; cf. also U. Staflarem: Staflii (?). 77 h.
Tefrei, name of a god ' Tefer. ' Dat. sg. Tefre la 24, Tefri la 28, Tefrei VIb 22; ace. Tefro VIb 26, 27, 27; voc. Tefre VIb 27, 28, 29, 31, 31, 31, 33, 33, 35, 36, Cf. tefra, Tefrali. 25 b, 93 b; n. on VIb 22,
tehtefim, noun 'cover' (?). Ace. sg.
IV 20. Cf. L. tego, tectum; stem-for- mation uncertain. 32 a, 68 e; n. on loc. cit.
teio: see tefe.
teitu: see deitu.
ienitu, 'hold.' Imv. VIb 25. L. teneto; cf. also U. andendu, endendu, etc., from a different stem of the same root. 41 a, 118 a.
tenzitim: see tesedi.
terkantur, 'inspect.' Pres. subj. pass. 3 pi. Ill 9. deQxofiai, Skt. pf. dadarsa, Olr. condercar, ' conspicitur' ; no Italic connections except possibly personal names Dercennius, Dercina, Dorcatius, Durcatius (see PL, I, p. 397, n. 2). 3 d, 114 e, 119 a, 165 a.
termnom-e, 'boundary.' Ace. sg. (+ -e) termnome VIb 57, 63, 64; abl. (+ -co) termnuco VIb 53, 55, 57; abl. pi. ( + ku) terranesku lb 19. L, terminus, cf.
OL termen, Skt. tarati 'cross over.' 29 f, 56 b, 74 g.
tertiam-e, 'third.' Dat. sg. f. tertie lib 6; ace. f. (+ -a) tertiama IV 2, ( +-e) tertiame Via 13; ace. neut. terti Ila 28, tertim VIb 64; abl. sg. neut, tertiu Via 45, 48, 53; loc, tertie lib 14, L, tertius, Skt, tftiya-. 90, 105 c, 113 a. See also poster tio.
tertio: see postertio.
tertu: see dirstu.
tefa, tefte, teftu, tefust: see dirstu.
tesedi tenzitim, a type of cake, possibly ' twisted cake. ' Ace. sg. tenzitim lb 6, tesedi VIb 46. Etym. and stem-for- mation uncertain; possibly related to L. iendo, tensio. 2 i, 3 d, 42 d; n. on VIb 46.
Tesenocir, name of a gate in Iguvium, ' Tesenacan. ' Abl. pi. Tesenakes la 11, 14, Tesenocir VIb 1, 3, Tesonocir Via 20, Vila 38. Cf. modern Ital. Tessena- ra{-l). 7 b, 70; Intro d., 4. '
testre, testru: see desire.
tesvam: see dersua.
Teteies, gentile name latinized 'Tet- teius. ' Nom. sg. lb 45, Ila 44. L. Tetieia, Tettia. 32 b, 53 c, 68 e, 90; n. on lb 45.
tettom-e, a landmark used in defining the augural templum, 'house' (?). Ace. sg. (+ -e) Via 13, 14, 14. From *tek- tom. L. tectum, cf. reyoQ. (?). 71; n. on Via 13.
tetu: see dirstu.
Ti: see Titis.
Tikamne, name of a deity 'Dicamnus' (?). Dat. sg. Ila 8. From *dikdm(e)- nos, cf. L. dico, -are (?). 3 d, 10 a, 29 f, 35 d; n. on loc. cit.
tio tiom: see tefe.
tigel, 'day' (?). Nom. sg. tigel Ila 15; ace. tiglu III 25, 27; abl. tiglu lib 22. O. zicolom 'day,' djiikulus, cf. L. dies. (?). According to the more usual interpretation, from *dik-elos ' dicatio, dedicatio, declaratio.' 3 d, 9 b, 29 i.
328
IV. LEXICON AND INDEX VERBORUM
46 c, 55 e, 75 c, 91 b; nn. on Ila 15, b 22. tigit, impersonal vb. ' one is required. ' Pres. 3 sg, Ila 17. L. decet, cf. decus, dignus, Homeric 6ey.ro, Vedic dasti 'worship. ' 3 d, 8 e, 9 b, 46 b, 114 e, 115 c, 118 d.
Titis, praenomen ' Titus. ' Gen. sg. Titis lb 45, abbreviated Ti Ila 44, abbrev. T Va 3, 3, 15. L. Titus; cf. also Schol. Pers. I, 20 titi sunt columbae agrestes. 67, 80, 92, 155 e.
titu: see dirstu. tiu: see tefe.
Tlatie, adj., name of a field. Gen. sg. Vb 9. Etym. uncertain; possibly connected with L. Latium, or else with Idtus < *tlatos, pcpl, of fero. 41 d, 68 a; n. on loc. cit.
-to -ta -tu, postpostition 'from.' akru- tu (?) Va 9; anglu-to Via 8, 10, 10; pure-io Via 20; skalge-ta IV 15, 18, 20, scalse-to VIb 16; tefru-to Vila 46; uapersus-io Via 12, 13. Etym. uncertain; according to Co., related to suffix in L. caelitus, Skt. devatas (IE -tos); according to suggestion of PI., II., p. 453, from reanalysis of abl. sg. ending -t (if admitted as source of -<f) + « (L- a); Vetter compares Etr. -ta, with same value. 4, 147 o.
toco, word of uncertain form and sense in Vb 13. Possibly 'pickled, preserved by pickhng. ' Cf. Gallo-Latin tucca, tuc(c)etum, taxea (?). N. on loc. cit.
todceir, adj. 'city-, of the city.' Ace. sg. neut. (+ -e) todcome Via 10, 10; nom. pi. totcor Via 12; abl. todceir Via 11. O. tuvtiks, toutica, cf. U. tota. 29 g, 46 e, 70, SO, 103 a.
toTU turuf , with uitlu vitluf ' bull-calves ' ; in lb 20 'victims' (?). Ace. pi. turuf lb 1, turup lb 4, toru VIb 43, 45; abl. tures lb 20. L. taurus O. ravgo/j., Gk. ravQog, cf. Goth, stiur, Eng. steer. 2 k, 22, 26 e, 67.
toiar tutas, 'state' in sense of noXig, L. civitas. Gen. sg. tutas lb 2, 5, to tar Via 30, 32, 39, 41, 49, 52, b 10, 13, 32,
34, 43, 45, 53, 59, 61, Vila 3, 6, 9, 10, 12, 14, 15, 16, 16, 17, 19, 21, 24, 26, 27, 27, 28, 29, 30, 30, 31, 32, 35, 37, 41, 47, 50, 52, 53; dat. tute lb 13, tote Via 5, 18, 24, 31, 33, 40, 42, 50, 53, b 7, 10, 11, 14, 26, 32, 34, 51, 62, Vila 11, 14,
18, 27, 31, 50; ace. tuta lb 16, to tarn Via 41, 49, 51, b 12, 33, 58, Vila 16, 29, 47, tota Via 29, 31, 39, b 31; abl. ( + per -pe -per) tutaper la 5, 8, 12, 15,
19, 21, 25, 29, 31, b 2, 5, III 29, tu- tape III 24, totaper Via 23, 25, 34, 35, 43, 45, 53, 55, 58, b 1, 3, 6, 7, 9, 15, 19, 22, 26, 28, 35, 43, 45, Vila 3, 6, 9, 10, 19, 21, 24, 26, 32, 35, 37, 41, 53; loc. tote Via 36, b 29, (+ -me) toteme Via 26, 46. O. TO) fro touto, Olr. tuath, Goth. piuda, Lith. tauta; connection of L. totus uncertain. 2 j, 19 a, 23, 25 a, 41 a, 42 e, 57 e, 66, 90, 143 d.
touer: see tuer.
traf trahaf traha tra, preposition ' across. ' tra lb 31, 35, Ila 13, traf Vila 39, traha Vila 5, 39, 44, 45, trahaf Vila 41. L. trans, both being probably fossil- ized pcpl. pres. nom. sg. *trd-nt-s, cf. Skt. trdti 'bring across. ' 58 d, 147 p; n. on Vila 5.
trahuorfi, adv. 'crosswise.' Vila 25. L. transverse, cf. U. traf traha and vurtus, co-uertu, co-uortus. 17 a, 44 a, 59 c, 112 a, 113 d.
tre: see trif.
Trebe: see Trebo.
trebeit, vb. 'remain.' Pres. 3 sg. Via 8. Cf. tremnu, Trebo, O. triibiim, L. trabs, Goth, paurp. 27 c, 39, 61 e, 114 e, 118 d.
Treblanir, Treplanu, name of a gate in Iguvium. 'Trebulan. ' Ace. pi. Tre- planu lb 9, Treblano VIb 47; abl. pi. Treplanes la 2, 7, Treblanir Via 19, 58, 59, b 2, 4, 20, 23, 44, 46, Vila 5, 7, 53, Treblaneir Via 22, Vila 42. Cf. place-names Trebula, Trebia; trebeit, tremnu, etc. 2 b, 3 d, 74 d; Introd., 4.
Trebo, deity of uncertain sex, 'Trebus. ' Dat. sg. Trebe la 8; dat. sg. Trebo Via 58. Ci. trebeit, tremnu. 25f;n.onVIa58.
tlgit
329
Tuscer
tref treif: see trif.
trefiper: see trifo.
tremiiu, 'cause to tremble.' Imv. Vlb
60, Vila 49. Cf. L. tremo, rQifico, tqeco,
Skt. trasafi, none of which however
shows the same stem-class as tremitu.
118 d. fremnu, '(augural) tent.' Abl. sg. Via 2,
16. From *treb-no-, cf. trebeif, O.
triibum, L. frabs. 35 d, 52 c, 61 e, 74
a, 145 a. Treplanes, Treplanu: see Treblanir.
fribrisine, 'group of three, triad' (of sacrificial victims). Nom, sg. tribfigu Va 9; abl. tribrisine Via 54. Probably from *iri-pedik-ij6n/in-, cf. trif and L. pedica (so Brugmann, Ber. Kon. Sachs. Ges. Wiss. [1890J, pp. 207-10). The alternative derivation from *tri- plik-yon/in- is semantically easier, but it does not provide a satisfactory basis for a change Z > f . 35 d, 46 d, 58 f, 60 e, 81 b, 88, 90, 91 c, 105 c.
trif, 'three.' Ace. pi. tref la 7, 14, 20, 24, b 1, 4, 31, tre la 3, 11, b 27, 43, trif lb 24, trif Via 58, b 1, 3, 19, 22,43, 45, Vila 3, 6, 41, 52, treif Via 22; ace. pi. neut. triia IV 2; abl, pi. tris III 18, 18. O. tris, L. tres, Tgelg, Eng. three, etc. 2 h, j, 10 a, 26 e, 41 a, 53 c, 56 a, 105 c, 155 h. See also frioper.
trifo, 'tribe.' Gen. sg. trifor Vlb 54, 59, Vila 12, 48; dat. trifo Vila 11; ace. trifu lb 16, trifo Vlb 58, Vila 47; abl. (+ -per ) trefiper III 25, 30. L. tribus, the two pointing back to IE *tri-bhii-; cf. trif, if we may assume that the basis v^^as an early threefold division of certain nations (?). 14 b, 15 a, 25 c, 40, 78 a, 90, 92; n. on Vlb 54.
triia: see trif.
trioper, adv. ' three times. ' triiuper lb 21, 22, 22, lib 25, 25, trioper Vlb 55, Vila 51. From neut. pi. *triya 'three' H- -per < pert; for second part cf. O. petiropert, L. semper. 105 c, 112 b, 113 b.
tripler, 'three.' Abl. pi. Va 21. L. trip- lus, cf. triplex, xQmXovg. 88, 90, 105 c.
tris: see trif.
-tu: see -to.
tua: see tuer.
tuder, ' boundary, limit. ' Ace. sg. tuder Via 10, 11; nom. pi. tuderor Via 12, ace. tudero Via 15, 16; abl. tuderus Via 11, b 48. Cf. tuderato, ehe-turstahamu, Tuder (modern Todi in Umbria); no sure connections outside Italic. 42 d, 57 e, 83, 90, 103 a, 145 c.
tuderato, pcpl. pass, 'bounded.' Nom. sg. neut. Via 8. Denominative vb. from tuder, q.v. 42 d, 114 e, 117 c.
tuer, possessive pronominal adj. of the 2nd person 'thy. ' Gen. sg. tuer Via 27, 28, 37, 37, 47, 47, touer Vlb 30, 30; abl. f. tua Via 30, 33, 40, 50, 52, b 11, 14, 32, 34, Vila 14, 17, 31, 50, tuua Via 42. O. tuvai, L. tuus, reog, aog < *Tfog, the various forms showing ablaut-variation in the root syllable.
2 h, 108 a. tuf : see dur.
tuplak, 'fork' (?). Ace. sg. Ill 14. Cf. dinAa^, L. duplex. 3 d, 80 a, 88, 105 b; n. on loc. cit.
tupler: see dupla.
tures : see toru.
Tursar, name of a goddess 'Tursa. ' Gen. sg. Tursar Vila 46; dat. Turse IV 19, Tuse lb 31, 43, Turse Vila 41, 53; voc. Tursa Vlb 58, 61, Vila 47, 49. Cf. tursitu. 2 i, 31 a, 59 a, 66; n. on Vlb 57-8.
tursitu tusetu, 'terrify, chase.' Pres. subj. pass. 3 pi. tursiandu Vllb 2; imv. tusetu lb 40, tursitu Vlb 60, Vila 49;
" imv. 3 pi. tusetutu lb 41, tursitu to Vila 51. Cf. Tursar, L. terreo, terror, and more remotely tremo, U. tremitu.
3 f, 8 b, 12 b, 33 b, 59 a, 60 a, 114 e, 118 b (the basic ref.), 126 a, 127 b, c.
Turskum : see Tuscer. turuf turup: see toru. Tuscer, adj. of nationality 'Etruscan,
330
IV. LEXICON AND INDEX VERBORUM
Tuscan. ' Gen. sg. neut. Tuscer VIb 54, 59, Vila 12, 48; dat. neut. Tursce Vila 22; ace. neut. Turskum lb 17, Tuscom VIb 58, Vila 47. L. Tuscus, cf. Etruscus, TvQai]v6q, etc. 46 e, 70.
Tuse: see Tiirsar.
tusetu, tusetutu: see tursitu.
tuta, tutape(r), tutas, tute: see totar.
tuua: see tuer.
tuva, tuvere, tuves: see dur.
V, symbol for 'five.' Vb 12. 104 a.
uapef-e, vapef e, ' stone seat, seat. ' Abl. sg. vapef e III 7; ace. pi. (+ -em -e) vapef em lb 14, uapefe Via 10, b 51 ; abl. uapersus Via 9, (+ -to) uapersusto Via 12, 13. L. lapis; no sure connec- tions outside Italic. 55 a, 61 b, 80 c, 90, 102 d.
vaputu, noun of uncertain meaning, pos- sibly 'incense.' Abl. sg. vaputu lib 10, 17; abl. pi. vaputis lib 13. From *wap6to-, cf. L. vapor (?). 71; n. on lib 10.
uas, 'fault. ' Ace. sg. Via 28, 38, 48, b 30. From *wak(o)s, cf. uasetom, L. vaco, -are, vacuus. (?). 83.
uaseto uasetom: see uasetom.
uasirslom-e, a landmark used in defining the augural templum, 'the Vale' (?). Ace. sg. (+ -e) Via 12. Etym. uncertain; possibly from *wakri-kelom with s < s < k before er < f < ri, as in 32 d, 46 b ; cf. Skt. vakra- ' curved. ' ( ?). 75 c ; n. on loc. cit.
uasor, 'vessels.' Nom. pi. uasor Via 19; ace. uaso VIb 40; abl. vasur IV 22. L. vas, vasis and vasum,-i; no sure con- nections outside Italic. 57 d, 83, 90, 103 a, 138 b.
uasetom. pcpl. pass. nom. sg. neut. trans- lated 'sin.' uasetom Via 37, uasetom Via 47, b 30, uaseto Via 27; (+ -i -e) vagetumi lb 8, uasetome VIb 47. Cf. uas, L. vaco, -are (?). 30 c, 34, 129, 149.
vatra, noun of uncertain meaning, here taken as 'ribs.' Ace. pi. Ill 31. Stem-
variant of vatuva, if not merely an erronenous spelling of it. N. on loc. cit.
uatuo vatuva, noun of uncertain meaning, according to the view adopted here, 'ribs.' Ace. pi. vatuva la 4, 13, 22, b 3,5, vatuvu lb 25, uatuo Via 57, b 1, 19, 43, 45 (written uatue), Vila 4. From *latua, cf. L. latera. (?). 2 d, 36 e, 78 b, 90; n. on Via 57.
vea: see uia.
uef, noun designating definite quantities (of meat), here translated 'portions.' Ace. pi. Vb 12, 12, 17, 17. Etym. un- certain; possibly from proto-Umbrian *weif-f < *weidh- 'divide,' cf. L di- vido, Skt. vidhyati 'pierce.' 65 f, 80 c, 102 d; n. on Vb 12.
Uehier, name of a gate in Iguvium, 'Vei- an. ' Abl. pi. Vehiies la 20, 24, Uehier VIb 19, 22, Uehieir Via 21. Cf. L. yea'(?). 68 a; Introd., 4.
ueiro: see uiro.
veltu, 'choose, select.' Imv. IV 21. L.
volo, velle, cf. U. ehueltu, ehvelklu.
8 a, 119 a.
venpersuntra, vepesutra, noun or adj. of uncertain meaning, here left untrans- lated. Ace. sg. f. vepesutra lib 15; abl. f. venpersuntra I la 30, vepesutra lib 18; abl. pi. venapesuntres IV 7. For prefix cf. vepurus, vepuratu; for second part cf. persondru. 76 b, 87 c; n. on I la 30.
vepuratu, 'extinguish.' Imv. Ila 41, Cf. vepurus, to which vepuratu is a denominative vb, 87 c, 117 c.
vepurus, adj. 'unburned. ' Abl. pi. Va 11. From *we-pur-u-fs, cf. L. ve-cors, ve-sanus, and nvg, U. pir, vepuratu. 65 e, 82, 87 c; n. on loc. cit.
uerfale, adj. with value of neut. noun ' templum. ' Via 8. L. verbale, cf. also EQU), {f)QiJTQa, QijrcoQ, the sense of the root being 'make a statement'; uerfale ' that which has the nature of a desig- nation' may be a euphemistic substi- tute for the actual name of the templum. 10 b, 31 e, 43, 54 a, 55 b, 77 h, 90, 95; n, on loc. cit.
Tuse
331
uitla
uerir, ' gate. ' Ace. pi. ( + -e -e) veriof e lb 9, uerofe VIb 47; abl. veres la 2, 7, 11, 14, 20, 24, uerir Via 58, 59, b 1,2, 3, 4, 19, 20, 22, Vila 7, 38, uereir Via 22, (+ -co) Via 19, 20, 21, b 23, 44, 46, Vila 5, 42, 53. O. veru, veruis, cf. L. aperio, operio < *-wer-y6, Goth. warjan 'ward off; probably also O. vereias, vereiiai. Germ. Wehr-macht, Land-wehr. 25 b, 67, 90, 103 a.
uesclir, 'vessels.' Ace. pi. veskla Ila 19, vesklu lb 29, 37, Ila 34, b 19; abl. veskles Ila 31, 37, b 18, IV 9, 24, uesclir Vila 9, 10, 18, 21, 24, 26, 32, 34. Etym. uncertain; if equivalent to L. vasculis, the e is unexplained; possibly related to uesiicatu, uestis. 2d, 46 c, 55 n. 2, 75 c.
uesiicatu 'pour a libation.' Imv. vesti- katu Ila 24, 31, 35, 37, uesiicatu VIb 16, Vila 8, 23, 24, 36. Fut. pf. (?) uesticos VIb 25. From *westikat6d, cf. uestisiar, uestis, L. Vesta (?). 70, 117 c ; n. on VIb 6.
uestis, pf. (?) pcpl. 'after pouring a liba- tion. ' Nom. sg. uestis VIb 6, 25, ues- teis Via 22. Probably from *westitos, pcpl. to deponent vb. *westiyo-, cf. uesiicatu, uestisiar, L. Vesta (?). 120 c; n. on Via 22.
uestisiar, 'libation' (?). Gen. sg. uestiMar VIb 38, uestisiar VIb 16, Vila 38; ace. vestigia IV 14, 19, vestigam la 28, vestiga la 17, 31, vestega (erroneously written vesvega) IV 17, uestisiam VIb 39, uestisia VIb 6, 17, 24, 25, Vila 38, abl. vestigia Ila 27, vistiga lib 13, uestisia VIb 5, uestisa Vila 37. From *westikia-, cf. uesiicatu, uestis, L. Vesta. (?). 8 e, 46 d, 53 d, 68 e, 90; n. on VIb 5.
Uestisier, name of a god, 'Vesticius. ' Gen. sg. Uestisier Via 14; dat. Vestige Ila 4. Cf. uestisiar, uestis, uesiicatu. 68 e; n. on Ila 4.
uestra, possessive pronominal adj. of the 2nd person pi. 'your. ' Abl. sg. f. VIb 61. L. vestra, cf. vos, cf. Skt. vas. 76 c, 108 b.
Vesune, name of a goddess ' Vesona. ' Dat. sg. IV 3, 6, 10, 12, 25. Etym. uncertain. Possibly related to Vesu- vius, Gaulish Vesunna. 74 f; n. on IV 3.
vetu, vb. of uncertain meaning, probably 'divide.' Imv. lb 29, 37. Cf. L. di- vido, imv. di-vidito, U. uef. (?). 119 a; n. on lb 29.
ufestne, 'sealed' (of vessels) (?). Abl. pi. IV 22. From pit. *op-festo-no-, cf. L. offendix, offendimentum, Eng. bind. (?). 44 d, 74 b; n. on loc. cit.
uhtretie, ' auctorship, term of office as uhtur. ' Loc. sg. Va 2, 15. From *auktr-etia, abstract to uhtur < *auk- tor, as kvestretie to kvestur, with stem as in L. amicitia. 35 d, 68 e.
uhtur, title of an official, ' auctor. ' Nom. sg. uhtur III 7, 8; ace. uhturu III 4. L. auctor, cf. augeo, av^co. 22, 35 d, 46 i, 56 a, c, 82 a, 90; n. on III 4.
VI, symbol for 'six.' Vb 10, 13, 14, 15, 18. 104 a.
VII, symbol for 'seven.' Vb 17. 104 a. uia, 'way, road.' Abl. sg. vea lb 14,
23, via III 11, uia VIb 52, 65, Vila 1, 11, 27. O. viii, L. via. 2 h, 10 c, 66.
uinu, 'wine.' Ace. sg. vinu Ila 18, 40, b 14; abl. vinu la 4, 22, b 6, Ila 25, 25, 39, b 10, 20, uinu Via 57, b 19, 46. L. vinum, Volsc. uinu, folvog. 13 b, 67, 90, 145 b.
uiro, ' men ' (in sense of L. homines, -um, contrasted with pequo 'domestic ani- mals'). Gen. (?) pi. uiro Via 42, 50, 52, b 13, 32, 34, Vila 17, 30, ueiro Via 30, 32, 39. L. vir, Olr. fer < *wi- ros, Eng. were(wolf); Skt. viras, Lith. vyras ; quantity of the U. i ei unknown. 11 c, 54 a, 58 b, 76 a, 90; n. on Via 30.
uirseto, pcpl. pass. ' seen. ' Nom. sg. neut. Via 28, 38, 48, b 30. Equivalent as if to L. *videtum, pcpl. to video; cf. U. auirseto. 10 a, c, 30 c, 54 a, 114 e, 116 a, 118 a.
vistiga: see uestisiar.
uitla, 'heifer-calves.' Ace. pi. vitlaf
332
IV. LEXICON AND INDEX VERBORUM
lb 31, uifla Vila 41. L. vitula, cf. U. uitlii. 58 b, 75 a, 90.
uitlu, 'bull-calves.' Ace. sg. uitlu lib 21, 24; ace. pi. vitluf lb 1, vitlup lb 4, uitlu VIb 43, 45. L. vitulus, Aeolic eraAov, Doric srsXov ' yearling, ' cf . O. Viteliii, L. Italia; also Skt. vatsas ' calf, ' (f)eTog ' year. ' 2 k, 8 f , 29 e, 75 a, 90.
ulo ulu, adv. 'there, to that place.' ulu Va 25, 28, b 4, (in purtatulu) lb 18, ulo VIb 55. Cf. L. olle, olim. 13 d, 55 b, 109 g, 112 d, 113 c.
umen, 'unguent.' Ace. sg. umen II a 19, 34; abl. umne Ila 38. L. unguen, both from *ong^en, cf. U. umtu. 49 b, 52 h, 81 c, 90.
umtu, 'anoint.' Imv. Ila 38, IV 13. L. unguo, Skt. anakti 'anoint.' 49 e, 61 a, 119 a.
une: see utur.
unu, cardinal numeral 'one' (?). Ace. sg. Ila 6, 8. OL oinos, L. unus, oivrj, Goth. ains, Eng. one. (?). 21 c, 105 a; n. on Ila 6.
uocu-com vuku-kum, ' grove ' ( ?). Ace. sg. vuku III 21, (+ -en) vukumen III 20 ; abl. ( + -kum -com) vukukum lb 1 , 4, uocucom VIb 43, 45; loe. vuke III 3, 21. O. loe. sg. luvkei, L. lucus. (?). According to an alternative view, ' temple ' with folxog, Skt. ve&as ' house. ' 21 b, 24 a, 55 a, 67; n. on VIb 43.
Uofione, name of a god 'Vofionus. ' Dat. sg. Vufiune la 20, Uofione VIb 19. From * leudhy6n(o)- or *loudhy6n(p)-, cf. L. Liber, OCS Ijudij'e, Germ. Leute, the root being one associated with growth. (?). 49 c, 74 f; n. on VIb 19.
uouse, noun 'vow' (?). Loe. sg. VIb 11. From *wofsyo- < *wofikijo- < *ivog'^hi- kyo-, cf. L. voveo < *wog^hey6. (?). N. on loe. cit.
upetu, upetuta: see opeter.
uraku, ures: see orer.
urieta, 'disk.' Ace. sg. Ila 23. L. or- hiia, cf. orbis; no sure connections out- side Italic. 36 b, 71; n. on loe. cit.
urnasier, adj. 'regular' (?). Gen. pi. urnasiaru III 3; abl. urnasier Va 2, 15. Probably from *ordenasyo- with syncope as in 29 f and -rn- < -rnn- < -rdn- by 61 e. 53 e, 68 d, 90; n. on Va 2.
ufetu, 'load (fire with incense)' (?). Imv. Ill 12, IV 30. From *ole- (L. ad-olere) or from *ode- (L. olere, cf. odor, dd/iij, odcoda) or possibly from con- tamination of both. 55 b; nn. on both passages.
urtas, urtes, urtu : see orto.
usaie usage, pf. 3 sg. 'approved.' (?). usaie lb 45, usage Ila 44. From *op- sak-ed, cf. L. ob- and sancio. ( ?). 46 g, 114 e, 124 b ; n. on lb 45.
ustite, 'period (of time)' (?). Loe. sg. Ila 15, III 2. Etym. uncertain; pos- sibly from *op-sti-to-, cf. L. ob- and sto, stare. N. on Ila 15.
utur, 'water. ' Ace. sg. utur lib 15; abl. une lib 20. From ^udor, udn-. vdcog, Goth, wato, gen. watins, Eng. water. 3 d, 42 d, 61 e, 84.
vuke, vuku: see uocucom.
vufetes, pf. pass, pcpl., according to the view adopted here 'chosen, according to one's choice.' (?). Abl.pl. Ila 31, IV 25. Possibly from *lubhetois, cf. L, label, libet, libita. 49 c, 118 b; n. on Ila 31.
Vufiune : see Uofione.
vufru, 'votive' (?). Ace. sg. m. lib 21, 24, 25. From *wog^h-(e)ro-, cf. uouse, L. voveo, evxofiai. (?). 49 c, 76 a; n. on lib 21.
vurtus, 'turn out (unfavorably).' Fut. pf. 3 sg. Ila 2. L. verto vorto, Skt. var- tate, pf. vavarta. Germ, werden, cf. U. couertu, couortus. 54 a, 124 b, 125; n. on loe. cit.
Vugiia-per, name of a gens (?; natine) 'Lucia' (?). Abl. sg. f. ( -l--per) lib 26. L. Lucia (?), cf. U. Vuvgis. 68 a; n. on loe. cit.
vutu, 'wash.' Imv. Ila 39. L. lav o, -ere, cf. Aouco. 55 a, 119 a.
uitlu 333 zefef
Vuvgis, praenomen latinized 'Lucius.' X, symbol for ' ten. ' Vb 12. 104 a.
Nom. sg. lb 45, (written Vuegis) Ila xil, symbol for 'twelve.' lib 2. 104 a.
44. O. Liivkis, L. Lucius. 24 c, 32 ^^r u i c <£-£4. , -in, -.« ,«,
„„ oo on 04 1, TK ^K ^^' symbol for 'fifteen.' Vb 17. 104 a. a, 55 a, 68 a, 90, 91 b; n. on lb 45.
uze: see onse. zeref: see sersitu.
Printed by UNIVERSA Wetteren (Belgium),
la
mm^i
iefl>i!l^l:
15
■ MS' i
i^ s i^fl¥v:a^ fl^S^; v a B.>i:'v«is ^awsf pilis^^ fa^ftQ8:>tvH3'VQ3-:ii3Mr9SSfM
25
Va
:^\l)igA/^=ii:V:(3^'?iai'^ -■
15
15
■^es'ifMiiiSiW'
v&
35
40
50
m;htSO ■;. ;. ,■ .. . : _ ■ : ' \ '' ',''^::\ _■ ' ■
>iDLir<EAgP.OFTRirfErvwe<IEtROFV|lERIEIfflV-S£RFEMAr<TiefElTVPoa^ i s!NA.VATs/OfERlNef EITVfONJf ErvAliVlO-FETVTASES-PERSNIMV-rROStSEnRM^ ■: :^a>^/T;,i;,[:>,^1iVf vStVERlSCOTREBuv^lR-ArBKAHASAHA^^^ ENQmERvSOIRSTV
:V'-^i f:ai<CATRir~aofAOTEf£lA:fETvrRESTOTE''SERFlE-5ERFEtv \\ARriER-rorlvrERToL\rillOVI\-ARIOTAreR ■Vi>JAf£rs5AtAf£rvrONnF£rvARVICfErv-SVRONWARATVT'VSH''RE-V£RlR -R05E5,?TlRSTRvkAf!CtAAnSvarvArt5vrcm5SrROTEr£SCVj£jgoAArESCLvr<VSEME-V5SriCAW^ iSaFERAAAraiEIVPoni.VrfRTOtAR'lOVINARTOTAT'EftlOVlNAfNOWV^^^ \
•' J^^ !N^eREtii'M/NRTIERTlOM€SlR\f SCLI R-ADRIR-PDPLVrf R.T^^^
ti?ARAJOAAfV£rER-PftESlDrASERF/A- '^f RFERrMARTlER-mEV£NDV-VIAECL,VATEl?0-TOT£TARSIi'^/ATE'TRI POTARS.WATf; ' ryi5CE-NAHARCE-lABVSCeMdMMETpTARTARilNATERTRlFOR7ARSIMAreR-lV5CERMAHARCER-IABv5CER/-JOr^
r,\c£RrASt'-TNyAPorLE'TOTAR- llovjNAR40TE'l ldV|^/EEROM•NO^^N'E■ERAR•NERVS■^IHtTlRAN5lHITIRlovlES
■.'■:■^^ATiRA^JOS^^TIR•f'RESroTA•^ERFIA^ERFtR•M^RTlER■SALVOM•SERiTV•^O^LO^^T6^ARMlOVINAR•^
rOTAA/.'itpVIN/KMfRI:5TPTASERFIASERFER-AAARTlER-5ALVOS£Rir\/-iPprLER^
^^.:/A£^■€RFARSM0•VIR0fEav0CASTRV0f^0F■SALVAS£RITvfVrVfONSTACER.^ASE■FVA^ortET0rAI«
f ;[EHOVlNt£RER'NOA^Ne£RARNOMNErRESToT/>^ERFIA^ERFER-MARTIERTlOAAfSlRVESCLlRADRERPOPLVrEr;
: :--^RjpVlNARTOTAP£RvlOVINlAERER-NOMN£P£RERAR-NO/VUv/EfW'PRESTOrA^ERFIA-5ERFER-MARnERTiO .X-AVVfRESrOTAR-'SERFlAR'^ERFER-MARTlERfONERfRlTETIOM-Sv'BOCAVvENNOM'PERSCLvESOiXirv
r7 f 5roTA^ERFIA5ERFERMARTIERT(OMMSlRVESCLlRADRlR:TlOMf LENERroPLVPERToTAR- 1 lOVINAR-iyrAPEfi
io./l^l^ERER•N0MNEPER•ERAR■N0AA^iEPERPRESTprA^ERF|AR'S£RfER■^AARTIERTlO^^■SVBOCAVV.PRES!0TA^.
i t R riAR-SERFER-/y\ARTIERf6(</ERf RITETIOAVSVBOCAVV- ENOM-VESTlCATVAHATRirvRSAiyEN0A\Rv5EME
r£ft SC LV-VESTICATV-PR^EST0T&5ER FIE'iER FER/y\ARTI£R.rdPWrERTOTARliPVINAR^rAte
xLFliirER5f^'MV'SVPERNEADROTRAHVORFIANDElslDVES'6fERSNlMVpRESTOTA^ERFIA^ERR:R-^^ '
i:,>''<A^!^iCllRWlR-POPLVpERTOTAR|l:QV(NARTOTAPERIIG)^
i;:fi'-;A/^tRrER-MARTlERAH/WENpv.VlAECLAArtRp-noaE-KDmRllOVlNARTOT^
T^:>rAR-HoytN)AR'NER^/S-^'IHiTlRANSl:rli •
^ERFER^^ARilER■5AlVCWvS£RilV^0^L0TC^
^^AP^rl£|HALV0/VvSERIW:^<^PLERTpTAR•llOVINARTO•rAR!IQylNAR•NOME■N£RFARS^Ap•^^
SAL^?\S£RirvFMrs^1-ON5rAGERrASETVArorLETOTARrto^^^^
SERFfA'SERFER-ZviAtaiERTIOMESIR-VESCLIRALFERTpPLVpERToTAR ■ hlOMMCPERfREgoVA^ERflA^ERFER/VXAUTJER'TIOMSVEOcAWPRESrofAR
svl;oc/^v:sI:^;^lO^\V^RSCLV■ESp^PERS^Jl^^^^^^
i\^l\Vl^£I^tomR:iroVINARToTArER-IIOVIr^JAERERNOMNEPER-ERARN0MNEf^^^ : SvftoCAVVPRESTOFAR-^ERFIAR-'SERFERMARTiERfONERfRlTETIOM-SVBoCAVVEN.OA^^^ -
VESrfSAEEMEFASPEFASCALSIECOMEGOSfETVnSQVlSANSiiPoPLVPERTOTARllOViNAR-TOTAr^^^^
^/ARAT;y^vSE•PQSF■VERlR■rE';ONOG|R^'CStl51A)lfP07i-DlTV;EN . SVBRASnAHAN\VTRAr^vAHAFAMfWT\rETRAa^SAHATA^
■ SacrantV -,.:'■■.:■■;■■ ■; '- ., ..;.;-■. a " ',. ^ ■ •■'■ ".',
!RAH^F•SAHA^E:ylTLATRIEf£Etv1VRSE^EIlFlE^ERFERM^RT|ERm^lA/PER^^
!:£r^^(iVlOffW■^A5ESfERSKlM^V■PRQ5ii5EDR•STRV^tAFICLA^AAR^ , pw (^^^iVSrc ARsirvPyflEABRON SFACVRENF PvSEEKvSdER S/^PEf Ew^S Dl RSVSPPO^^^^ J^=fJ5sct^Jt/^^TRAii^SAHAWA^CP^ABlF(ATV;£f^^^ ■a^|:5ACRAAl3N/v ENOMTRAHASAjfAlAM^COVERTVCONAOETv^^ .;
..Jfe^jQl^/Hi^fdaaANDllWAFviFroRSE^PERCAARSrAA^
jy,i;A^^jfgMER.TR ifof^jy^f^^
IF|^n: :s'h.Ri;miTv) (d^lOV■HO^^ 'f^"l^'^S^^^V/^rof*i;E-iorARipviwARTWEW .: M0/v\nE6P, AR-M.Q/v\N£- ESTErRiePERpaTv|h40M-NENCAf ERACRlO.TVR-SIN:ror0RSE'N(^CAARSWTlAliAB . ;
l'l?lNvA^yl>f^^DteVR0■SEHEME^ilARl^ATvT0TOTAR^
"'■" "''*^^^''''^^&R:i&ar;iiov)N^ V
AflaA^RniE5Er|R■^RSVEfTvTA>[S^EP.S^I!^'^'''l^~^^ ■•"'■"•
Vila
(A
\fi^
UNIVERSITY OF FLORIDA
3 1262 07805 235 3
C '
mmm