Skip to main content

Full text of "Transactions and proceedings"

See other formats 









Na ses - 


















ie ie 
i 
fe a 
ms 
¢ , 
: ES a v4 iy 
hy .) cu) oh HO ah ( 
, = MA Leey 345 
reiehenpiyy. oh 9) ehh nN ; x 
pea ds AE) ali et Te viele ate 7 +s 
, Sate stash vers PPLE RAD, City venti trts f vas : 
Ble’ lea Mca sNearets Biy “1% 4 
‘ ‘ A he 
’ ? 
+) 
r - 
mal 


al? 
” ay a ) ue 2 t 
ed Th aS oy <* 
tei) 
* ri 
5, a 
. hi val . 


Dw 











ees sty RENN ae) 
Uae A i : 


Hele aah ; 
Fy 





reid 
to 
apt ae 

rh 












© ye 

ay ile! 

cane 
+ ey re 

















+ ie rr 
syecne Siena a 


‘ 
sath, tit , pte TS 
ty 

































testa Bie) a eee 
‘i oe poten Ge vate 
Seta 













ae 
pyy ry te ty tyes 
, Bes any 

a3 ae é 











af 
fhe. 








































yen ey 
SAAR 
Pty 





d 
Sine ey 
ry tea 
be ality ae 
‘ | 
‘ 


LAV Ne 
Te Ae 





: oe 









oh aioe 
Be ee 





Digitized by the Internet Archive 
in 2007 with funding from 
Microsoft Corporation 


https://archive.org/details/transactionsproc24ameruoft 


ae 


- 


= 


TRANSACTIONS 


OF THE 


AMERICAN 


ae 


PHILOLOGICAL ASSOCIATION. 


1893. 


VoLuME XXIV. 





PUBLISHED FOR THE ASSOCIATION BY 
GINN & COMPANY, 


9 TREMONT PLACE, Boston, MAss. 


EUROPEAN AGENTS. 


ENGLAND: EDWARD ARNOLD, 37 Beprorp Street, StranpD, Lonpon, W.C. 
GERMANY: OTTO HARRASSOWITZ, Letezic. 
FRANCE: H. WELTER, 59 Rue Bonaparte, Paris. 


yi 


40/97 


Norwood Press: 
J. S. Cushing & Co.— Berwick & Smith. 
Boston, Mass., U.S.A. 


P 
| 
AS 
yr24 






ASP RAR 
. Nov 25 1970 


4, <8 


/y ag® 
<Fsity of 1A 






II. 


III. 


IV. 


VI. 


VII. 


The Latin Prohibitions 


CONTENTS OF VOL. XXIV. 


TRANSACTIONS. 


. The Scientific Emendation of Classical Texts . 


E. A. SONNENSCHEIN. 


On the Canons of Etymological Investigation . 
MICHEL BREAL. 


Ein Ablautproblem der Ursprache 
WILHELM STREITBERG. 


Dunkles und helles 7 im Lateinischen 
HERMANN OSTHOFF. 


The Implicit Ethics and Psychology of Thucydides . 
PAUL SHOREY. 


English Words which hav Gaind or Lost an Initial Con- 
sonant by Attraction - he ee 


CHARLES P. G. SCOTT. 


“Extended” and “ Remote” Deliberatives in Greek 
WILLIAM GARDNER HALE. 


PROCEEDINGS. 


H. C. ELMER. 


On the Interpretation of Satura in Livy VII. 2 


GEORGE L. HENDRICKSON. 


Solution of Some Problems in the Dialogus 


ALFRED GUDEMAN. 
3 


17 


29 


5° 


xiii 


xvi 


4 Contents. 


Hunc inventum inveni (Plautus, Captivi 442) . 
W. S. SCARBOROUGH. 


Democracy and Education Pa oe : 
WILLIAM GARDNER Biatie 


The Connection between Indian and Greek Philosophy 
RICHARD GARBE. 


Some Problems in Greek Syntax . . . ; 
BASIL L. a eae 


On the Origin of the so-called Root-Determinatives 
MAURICE BLOOMFIELD. 


The Language of the Law Pee 
HERBERT L. BAKER. 


Report of the Committee on Spelling Reform . 
F. A. MARCH. 


Critical Notes on Certain Passages in Sophocles’ Philoctetes . 


M. L. EARLE. 


Some Suggestions derived from a Comparison of the Histories of 


Thucydides and Procopius 
_W. #H. Pains: 


hokas: in : Pliny, H. N., 35, 152 
HAROLD N. FOWLER. 


On some Greek inscribed Wax Tablets . 
HAROLD N. FOWLER. 


Tlepiréreva and Allied Terms in Aristotle’s Poetics 
HoRATIO M. REYNOLDs. 


Libration in the Periods of Cicero 
W. B. Orin. 


Varro and Chrysippus as Sources of the Dialogus of Tacitus . 


ALFRED GUDEMAN. 
The Indo-European root s¢@ ‘stand’ in Italic . 
CARL D. BUCK. 


Greek Nouns in -is, -idos Leer 
BENJAMIN IDE iincecen. 


XXili 


XXiv 


XXVli 


XXXV 


XXXVii 


xliv 


xlvii 


xviii 


TRANSACTIONS 


OF THE 


AMERICAN PHILOLOGICAL ASSOCIATION. 


1893. 





I. — The Scientific Emendation of Classical Texts. 


By Pror. E. A. SONNENSCHEIN, 


MASON COLLEGE, BIRMINGHAM. 


THE object of this paper may be defined as an attempt to lay 
down certain canons of textual criticism, and to apply them to 
and illustrate them from the text of a single author, — Plautus, 

If textual criticism is to become something more than the 
amusement of an idle hour; if we are to avoid the danger 
which sometimes seems to threaten us that the multiplication 
of light-hearted emendations will ultimately result in the dis- 
integration of the classical authors, our schools being saddled 
with a multitude of texts, each differing from the other and 
showing a constantly increasing tendency to diverge, — it is 
imperatively necessary to find a method which offers some- 
thing like a promise of a consensus of opinion. Such a method 
cannot be anything less than sczentific; it must be analogous 
to the methods by which the great results of chemistry or 
physics or other “exact” sciences have been achieved, and, 
like them, it must depend on evidence, and not on authority. 
Science is a great unifier. 

But salvation is to be found neither in “slashing surgery” 
nor in a childlike faith in the plenary inspiration of MSS. 
The scientific method cannot be adequately described by any 


1 The present paper is condensed from that presented at the meeting of the 
Philological Association. 


5 


- 


6 E. A. Sonnenschein. [1893. 


of the catch-words borrowed from the vocabulary of political 
parties, — ‘‘ conservative,” “radical,” etc. Its true nature, 
however difficult to define, is simple; simple, though often 
misunderstood. The first step is to examine into the rela- 
tions of the extant MSS. to one another, on the basis of an 
apparatus criticus, and to arrange them in families. Without 
this preliminary inquiry, all subsequent labor may be thrown 
away. The next step is to proceed, by way of inference, to 
the probable reading of the archetype or archetypes. Our 
extant MSS. are copies of copies; what we want to know is 
the reading of the original copy or copies from which they 
were derived. By this means we are enabled to discard the 
corruptions peculiar to the individual derived copies; and it 
often happens that this stage of our inquiry brings us face 
to face with the vera manus of the author. If so, we attain to 
what may be called restoration without emendation. If not, 
the next step is to bring all the resources of palaeography, 
logic, and observation of the usages of the language in general 
and the author in particular, to bear upon the problem of 
emending the text. Sometimes a very slight change will suf- 
fice, and the emendation may be regarded as practically cer- 
tain: sometimes, where the seat of the corruption lies deeper, 
the critic may have to take a more venturesome course, and 
put something of his own into the text ; but it must be some- 
thing that tallies precisely, even in the minutest points, with 
the usage of the author in question and the context of the 
passage. 

This I have called a simple programme; but it is obvious 
that there is no immediate danger of its becoming a merely 
mechanical operation, as easy as handling a pair of compasses, 
—a leveller of wits. On the contrary, its execution is fraught 
with the possibilities of error at every point, and gives the 
amplest scope to the individual génius of the critic. Nor can 
the personal equation ever be entirely eliminated. Above all, 
there is need of the seeing eye, —the power of going beneath 
the surface and seizing upon the really vital point. The true 
critic is a framer and verifier of hypotheses ; and hypotheses 
are not things that can be manufactured to order, by a process 
of mere industry. 


Vol. xxiv.] Sctentific Emendation of Classical Texts. 7 


Let me now apply these principles to the criticism of 
Plautus. The extant Plautine MSS. fall into two great 
families, the first being represented by a single MS., the 
Ambrosian palimpsest (A), the second by a number of MSS. 
called the “ Palatini” (BCDEV/).1 The latter, being of the 
same family, must be derived from a common archetype, 
which we may call 7.2, This lost archetype appears to have 
been a MS. of at least equal value with A, and probably of 
about the same date (fourth, or perhaps third, century of our 
era); BCDEVY, its offspring, are, roughly speaking, of the 
eleventh or twelfth centuries. Now, what is the relation 
in which these two great families (A and /) stand to one 
another? The problem is not yet entirely cleared up, but 
I think it may be said to be approaching a solution. There 
is a curious problem of likeness and unlikeness in these two 
groups, and neither of them can be regarded as derived from 
the other. But it is possible that they may both be derived 
from a common source, if we suppose, as Seyffert has sug- 
gested,® that that source contained the origina] text in various 
forms. Let us imagine, then, a very early MS., say of the 
first century of our era, which contained, in its margin or 
written above the text, a number of parallel variants; we may 
denote this MS. (the archetype of the archetype f and also 
of A) by the letter x The stemma of the chief Plautine 
MSS. would then be as follows: 


z 


BOD’ #£ <p 


1] neglect the minor MSS. By Z I mean the MS. discovered in 1879 at Milan 
(see Léwe and Goetz, Rhein. Mus., p. 53 ff., and Preface to Curcudio), not the 
worthless MS, that Ritschl called Z. 

2 I use small letters to indicate the archetypes, or hypothetical parent-MSS. 

3 In a private communication, 


8 E. A. Sonnenschein. [1893. 


But whence did these parallel variants, which are to explain 
the diversities of A and Z, arise? Partly, perhaps, in the 
shifting practice of companies of actors. We know that 
varying texts of Shakspere’s plays were current in his 
time, reflecting the preferences of individual actors or the 
modifications suggested by the experience of a company of 
actors. Similarly in Plautus there are indications that lines 
were sometimes introduced into the margin of the actors’ 
copies, intended to serve as a substitute for a passage which 
seemed too long or which contributed nothing to the devel- 
opment of the plot—and of such passages there are, alas, 
too many in the works of our Latin playwright.! 

But however the parallel variants may have arisen, it is 
clear that their existence in the archetype x would give rise 
to precisely the sort of likeness and unlikeness which we 
observe in A as compared with the descendants of ~. Take 
Pseud. 392, and let us suppose x to have presented somewhat 
of the following appearance : 


utrumque tibi nunc dilectum para 
Ex multis certust cedo 
Atque exquire ex illis paucis unum qui certus siet. 


The copyist would feel himself in a difficulty, and various 
methods of dealing with the text would suggest themselves. 
The copyist of g appears to have written in the second 
line, 


Ex multis exquire illis unum qui certus siet, 


whereas the copyist of A, in a more comprehensive spirit, 
turned it out as 


Ex multis atque exquire ex illis paucis unum qui certust cedo. 


Neither of them can be congratulated on the result. 


1 Capt. 958, 959, were perhaps intended as a substitute for the passage begin- 
ning in 957 and ending 969. Most. 816 is followed by two lines which merely 
repeat, with a variation, lines 845-847; the intention of the composer of them, 
probably, was to enable the actors to skip 817-848, a passage full of humor, and 
thoroughly Plautine, yet not necessary to the plot. 


Vol. xxiv.] Sctentific Emendation of Classical Texts. 9 


Again, in the Mostellaria, A has, after 715 and instead 
of 716, ‘= 


Tempus nunc est mihi hunc adloqui senem, 


a line which repeats 714 (Tempus nunc est senem ‘hunc 
adloqui mihi) in a slightly different form. This may be 
explained if we suppose that x had doth of these readings, 
the one in the text, and the other in the margin; the mar- 
ginal reading may have crept into the text of 4 after 715, 
displacing 716. 

To take another instance: in Mil. 155, BCD have (neglect- 
ing details) a4 : A 
hic illest lepidus quem dixi senex est. 

How are we to account for the meaningless est at the end? 
The key is supplied by A, which has, according to Stude- 
mund, senem at the end of the line. Probably, then, this 
reading had been introduced in a contracted form in + 


(above the line) — thus, Senex ; the € was meant to repre- 
sent em. The copyist of A understood it, and adopted 
the reading intended (senem); the copyist of # misunder- 
stood it to stand for esz, and introduced it, in the innocence 
of his heart, after the word senex. 

Other instances of the same or similar phenomena will be 
found in Mil. 552, Cas. 185, Pseud. 631, Stich. 342; Merc. 
757, Pseud. 85, 208, Trin. 52, 339; Mil. 1177, Stich. 202, 
Poen. 343. 

Seyffert has also called attention to another point in regard 
to one of the Palatinii— 2. This MS. appears to contain 
a sort of secondary tradition, derived from an independent 
source. It is well known that this MS. is corrected by a 
second hand, distinguished as B?; and it now appears that, in 
eight plays, these corrections were derived from a MS. which 
has now disappeared, but which contained a number of various 
readings of great antiquity, and, in some cases, whole lines 
not found in the extant copies. But why only in eight 
plays? The probable answer is that in the Middle Age 
the plays of Plautus were divided into two volumes, the 


10 E. A. Sonnenschein. [1893. 


first containing eight, the second twelve, plays. In the 
case of the MS. in question, only the first volume hap- 
pened to be preserved; so that, for the twelve plays of the 
second volume, this subsidiary source of information was 
not available! And there appear to have been differences, 
even in external form, between the first and the second 
volume, at any rate in the archetype of BCDE/ (i.e. ). 
Seyffert? has made it probable that the first volume of this 
MS. had only twenty lines on a page, whereas the second 
had twenty-one—a fact which is at first sight of small 
importance, but which may turn out to be of great signifi- 
cance to the investigator, especially in regard to lacune. 

So much, then, for the MSS. of Plautus and their rela- 
tion to one another. In spite of all their defects, their 
tradition is, on the whole, an excellent one, especially when 
we consider the length of time that separates us from 
Plautus. Probably our MSS. represent the vera manus 
of Plautus quite as successfully as the folio of 1623 repre- 
sents that of Shakspere. This may be illustrated by a 
few examples, in which their readings require only to be 
understood to be pronounced genuine, and the vera manus 
may be restored without emendation. Take Rudens 728, 
where the true reading® is staring us in the face in the 
apograph of A, for which we are indebted to the inde- 
fatigable labors of that scholar-hero Studemund. Or again, 
Rud. 528-538, which need only to be pronounced with the 
stammer of chattering teeth to be metrically above reproach. 
It is impossible to lay too much stress upon the importance 


1 Jt is well known that at the time of the Renaissance only eight of the 
Plautine comedies were known to the learned, viz. Amph., Asin., Aul., Capt., 
Cas., Cist., Curc., Epid. 

2 Reconstruction of a fasciculus of the Mostellaria (Berliner Philologische 
Wochenschrift, Feb. 13th and 20th, 1892). 

3 Dei tibi argentum? (“The gods pay thee money?”) But it required the 
eye of Seyffert to see what others looked upon without seeing. Another instance 
of his penetration is Mil. 1253, where the readings of BCD, taken together, point 
to ut guaeso amore perditast te misera as the reading of / (‘ how violently she is 
in love with you, poor thing!’): amore perditast = deperit, and so takes the 
accus. kaTd oUveow; cf, Cist. I. 2. 13. 


Vol. xxiv.] Scientific Emendation of Classical Texts. II 


of caution in dealing with MSS. A little knowledge often 
pronounces them to be corrupt where a wider knowledge 
reveals their soundness or the soundness of the archetypal 
reading. And there is no higher pleasure for the critic than 
to see, emerging from the gloomy places of the apparatus 
criticus, the light of intelligible sense. In this connexion 
I would refer to Minton Warren’s proof of the existence of 
an asseverative enclitic -zé in Plautus, —a suggestion which 
vindicates the readings of the MSS. in a number of passages. 

Side by side with the MS. tradition we have also, in Plau- 
tus, a grammarian’s tradition, represented in such sources as 
Festus, Paulus, Nonius. Recent investigations of Hermann 
Caesar and Carl Reblin prove that the readings of Nonius 
agree sometimes with A, sometimes with £, but occasionally - 
show clear traces of a third and different recension,! the 
precise relation of which to A and it is not yet possible 
to define with accuracy. On the relation of Festus to Ver- 
tius Flaccus, valuable light has been thrown by Nettleship,? 
in his Essays on Latin Literature. 

As illustrations of the third stage of criticism, I may men- 
tion the brilliant yet simple emendations of Ellis in Most. 
595, ue frit quidem, ‘not a particle,’ for nec ertt quidem of 
the MSS. (cf. 008 wacmadn and «ai adyvnv, Arist. Vesp. 
QI, 92), and Palmer in Cas. 994 Hector Ilius for ecastor ilius 
or hectore tllins of the MSS. Perhaps I may be pardoned 
for appending one or two more homely conjectures of my 
own: Rud. 321 ornatus for natus;* Pers. 392 eccillud for 
ecctllum (see Classical Review for November, 1892, where 
I maintain that o@paxos changed genders in passing into 
Latin); Most. 278 nimzs male for nt male of the MSS. 


1 Epid. 233, 559; Mil. 1180; Pers. 305, 347, 348; Poen. 312, 365, 908; 
Pseud. 184, 319, 382, 864; Rud. 533; Stich. 144, 348, 366; Trin. 251, 410. 

2 It is with deep regret that I record the death of Professor Henry Nettleship, 
which occurred, at Oxford, about the time when the Philological Congress was 
meeting at Chicago. In him England has lost one of her foremost scholars; and 
there are many who will feel their lives the poorer through the loss of his friend- 
ship and guidance. 

3 Langen supports me by referring to Pseud. 756 to illustrate the use of cum 
(ornatus cum virtutibus = ornatus virtutibus). 


12 E. A. Sonnenschein. [1893. 


On the other hand, I am inclined to withdraw my con- 
jecture ad Charontem in Most. 509 in favor of Bentley’s 
Accheruntem. Here, I fear, I was guilty of a fallacy which 
is too common at the present day, a one-sided and super- 
stitious reverence for what in itself is venerable enough, — 
the ductus litterarum (two MSS. have adcheruntem). We 
are too apt, in our scrupulous attention to the external 
appearance of the text, to neglect other and even weightier 
matters, —the sense and the usage of the author. Schoell’s 
Argentumdonida in Pers. 120 seems to illustrate this: nzhzlz 
parasitus est quit Argentumdonidast, ‘a parasite who is a 
giver of money is naught,’ does not really suit the con- 
text; all that the passage will bear is ‘a smonted parasite 
is naught,’ and however we are to read the corrupt words 
of the MSS. (cuz argentum domideste or domi idé), this is 
the sense to be brought out, as I maintained in Classical 
Review, November, 1892. 

It is at the third stage of criticism that the genius of 
the critic has its highest opportunity; he may be called 
upon to put into a lacuna of the text something which the 
author himself might have written. But this task demands 
not only an originality and power of initiative which is very 
rare, but also a profound knowledge of all the well-established 
results of special inquiry in many departments of scholar- 
ship. Who can tell what the author might have written, 
except one who is soaked in his thought and diction? And 
what this means is known to those who are acquainted with 
the vast literature that has grown up around each of the 
great classics. It is easy to ridicule the scholarship of 
the present day as “aping the methods of the physical 
sciences”; and, no doubt, minute research demands an 
amount of time which often involves a sacrifice of all-round 
culture. But the question for the critic is simply one of 
building on solid ground or spinning ropes of sand. He 
must be content to sacrifice something for the sake of his 
science. And it must be remembered that scientific research 
is itself a kind of culture, leading, both directly and indi- 
rectly, to a vital grasp of many things that escape the mere 
atlettante student. 


7 


Vol. xxiv.] Sczentific Emendation of Classical Texts. 13 


Besides MSS., the Plautine student has to take account 
of problems of metre and prosody. Of the versification of 
Plautus we know practically nothing except what we can 
learn by exploration of the facts contained in MSS.,— MSS. 
of Plautus, Terence, and the other old Latin dramatists. 
Attempts have recently been made to solve questions of 
prosody by a priort methods; but such attempts inevitably 
lead to a vicious circle in reasoning, and are, in my opinion, 
doomed to failure. For we have no independent tradition 
that we can trust as to the versification of the poets or the 
pronunciation of the educated classes at the time when 
Plautus lived. Nor can we safely infer from the metrical 
phenomena of Plautus to the prose pronunciation of his 
time. No doubt the versification of the old dramatists was 
based upon the phenomena of every-day speech; but the 
plain testimony of facts shows that they did not hesitate 
to subordinate the word-accent to the verse-accent, where 
they found it necessary or convenient to do so. Thus, for 
instance, though in the pronunciation of every-day life words 
like éésecré were uniformly accented on the first syllable, we 
find in Plautus an occasional odsécré, with last syllable short- 
ened by the ictus on the middle syllable; this is a phenome- 
non essentially similar to the prose pronunciations béné, mdlé, 
dvé, caléfdcere, for béné, mdlé, dvé, caléfacere (cf. Quintilian, 
Inst. I. 6. 21). In fact, the law of shortening is the same in 
prose and in verse; but in verse it may operate in cases 
in which it cannot operate in prose, because the poets allowed 
the ictus to fall on syllables on which the prose accent could 
not fall. All poets have allowed themselves such liberties 
to a greater or less extent; else they would hardly have got 
far in the work of composition. If any one, in his anxiety 
to vindicate the character of Plautus as an artist in words, 
declares that he cannot have written this or that because it 
would presuppose a scansion at variance with the normal 
speech of his time, I fear he is adopting an @ priori method 
of argument, —a method to which the recent work of Klotz 
(Grundziige der altrémischen Metrik, 1890) has lent some 
_ encouragement. 


14 E. A. Sonnenschein. [T893. 


But if we fix our eyes firmly on the facts as presented in 
the only source of information open to us, we are able to 
form a tolerably accurate idea as to how Plautus intended 
his verses to be read. The most important phenomenon of 
old Latin prosody is the law of iambic shortening, in its 
various developments; side by side with it we have a num- 
ber of isolated peculiarities of prosody. Luchs has shown 
that in the time of Plautus the general pronunciation was 
hiquidem instead of hicgutdem. Biicheler made the discovery 
that side by side with méguidem there was the pronunciation 
méquidem, even under the ictus; similarly, we find traces in 
‘the verse of Plautus of the parallel forms siguidem, siquidem ; 
Si quis, siqguis; né quis, néguis. Seyffert, carrying out the 
inquiries of Biicheler, has given reasons for believing in 
the existence of zsguidem side by side with zsguidem, hadqut- 
dem side by side with haecguidem, and so forth. To the 
same scholar is due the discovery that nempe never forms 
a complete foot in Plautus,—a discovery which Skutsch has 
rationalized by supposing that the Plautine pronunciation 
was always nemp, at the same time extending a similar 
treatment to the words wnde, inde, etc. (to be pronounced 
und, ind). Skutsch supports this contention by reference 
to the forms fer, fac, dic, duc (= fere, face, dice, duce), and 
to such scansions as rvedd for redde, Stich. 768,! mztt for 
mitte, Pseud. 239;? cf. Mil. 1067. We may add such phe- 
nomena as guodn’ for quodne, Mil. 614, necn’ for necne, Mil. 
1051, estn’ for estne, Epid. 614, 2? sz itis for tte si itis, Poen. 
1227, adicer for dicere, Merc. 282, nosn’ for nosne, Poen. 1238. 
As to final s, it has been held till recent times that it could 
fall away only before a consonant, as in the verse of Ennius 
and Lucretius. But Leo has adduced strong evidence in 
favor of the view that it might disappear also before vowels, 
with the result that the preceding vowel was elided; thus 
we find, Bacch. 401, comis incommodus is to be scanned com’ 
incommodus,® a pronunciation which is curiously reproduced 


1 redd’ cantionem uéteri pro uindé nouam. 
2 O Psetidole mi, sine sim nihili | Mitt’ mé sis, sine modo ego beam. 
8 Cf. my note on Rudens 1006. . 


Vol: xxiv.] Scientific Emendation of Classical Texts. 15 


by the first hand of the Codex Vetus, com in comodus. 
Cicero, in his Orator, § 153, quotes even more surprising 
instances of the loss of s (after a long vowel). And the 
doctrine of Leo offers, for the first time, an explanation of 
the familiar Plautine contractions scelestu’s (or scelest’ es) for 
scelestus es, nancta’st (or nanct est) for nancta est, re’st for 
ves est, etc. Numerous isolated words might be quoted, in 
which research has shown the necessity of rectifying the 
statements of dictionaries or commentators as to quantity ; 
e.g. Palmer has shown that dzerectus is a word of four sylla- 
bles, with the first long (Rud. 1170, etc.). 

In regard to many questions of metre, we are still only at 
the beginning of inquiry. The numeri innumeri of Plautus 
- attracted the attention of the writer of his epitaph, and ‘we 
have probably not yet got to the end of them. Inquiry is 
always leading us on the track of new metres, of which we 
are sometimes quite unable to say where Plautus got them 
from. Biicheler has proved the existence of hexameters in 
Plautus ; and we must probably recognize with Goetz and 
Schoell, in their smaller edition of the Casina (lines 959 f.), 
a metre of which the scheme is ZUZUUYU ZUU_ ZuUVU 
(trochee, dactyl, choriamb, dactyl, spondee) : 


Hic dabé protinam ét fugiam: heus, sta ilico amator 
'Occidt revocér: quasi non audiam adibo. 


Such, then, are the chief problems which the critic of 
Plautus has to face,—the problem of MSS., and the prob- 
lem of metre and prosody. Throughout the critical process 
he has to exhibit the qualities .of taste and power of esti- 
mating evidence. Neither of these is a matter for which 
rules can be given, yet neither is purely capricious. The 
only test to which the work of the textual critic can be 
brought is the judgment of those competent to judge. 

Two assumptions underlie the whole of my argument: 
(i) That the object of textual criticism is to restore what 
the author wrote, and not to improve upon his sentiments 
or diction. This apparently obvious proposition is implicitly 
denied when an emendation is praised or condemned on the 


16 E. A. Sonnenschein. [1893. 


ground of its intrinsic beauty or ugliness. (ii) That the ~ 
process of emending is some day to come to an end. 
The problem, indeed, can never be absolutely solved, but 
the day may come when men will be in a position to say 
that they have solved it so far as it can be solved. And 
then, if the world still cares for classical learning, a fair 
prospect opens up. The first Renaissance taught men to 
love and revere the classics; the second Renaissance, of 
Wolf and Altertumswissenschaft, to study them scientifi- 
cally; the third Renaissance, of which we already see the 
beginnings among us, will teach us to interpret and appre- 
ciate them. 


Vol. xxiv.] Canons of Etymological Investigation. 17 


II. — On the Canons of Etymological Investigation 


By MICHEL BREAL, 


PROFESSOR IN THE COLLEGE DE FRANCE, PARIS. 


THE time seems indeed to have come for revising the old 
etymological dictionaries, and putting them in accordance with 
the discoveries and principles of linguistics. This work has 
already been begun on different sides: I need only mention 
Murray’s Dictionary as a model of detailed and complete expo- 
sition, and, as specimens of abridgment, Kluge’s books for 
German, Scheler’s for French, Korting’s for Roman. These 
books, with their different qualities, afford good specimens of 
linguistic science. What I want to attempt here is to remind 
the reader of a few rules, to indicate a few desiderata, to point 
out a few possible and desirable improvements. 

There are etymological dictionaries which content them- 
selves with indicating the origin and formation of each word ; 
such is the case with the two last authors we have just named. 
These brief indications are no doubt of use; but the most 
important part — which is the history of words, the develop- 
ment of meanings —is unmentioned. These books might be 
compared to biographical dictionaries, giving the persons’ date 
and birthplace, but silent as to what, they were, how their lives 
were spent, the part they played in general history. They are 
repertories, rather than dictionaries, in the widest and fullest 
sense of the term. 

Altogether different are the works in which the history of 
the meanings is set forth. Here is a curious spectacle for 
the observer, showing how, and according to what laws, a 
people appropriates to its needs, its ideas, its new concep- 
tions, the ancient inheritance of its tongue. As a master- 


1 Translated by Miss Edith Williams. 


18 Michel Bréal. _[1893. 


piece of its kind, I may mention Jacob Grimm’s Dictionary, 
especially in the part due to its continuators. Yet this great 
work is not entirely above criticism ; going to another extreme, 
it may appear to carry divisions and subdivisions somewhat to 
excess, and to supply too lavish an abundance of examples, 
Littré affords a rare model of sobriety. An etymological dic- 
tionary may give the history of words, without pretending 
to note each separate shade. It should stop where literary 
criticism begins. However it be, there is no lack of models 
for imitation. Each people seems to pride itself upon drawing 
up an inventory of its riches; some small countries, such as 
Switzerland (I allude to its Idiotikon), show enenIselyes equal 
to the most forward nations. 

As regards the ancient tongues the work seems to me, in 
certain respects, less advanced. One might think that it 
would be an easier matter in the case of dead languages ; 
for the whole of their literature, since it has attained its 
entire completion, may be embraced at a glance. But here 
we meet with a difficulty of a peculiar kind, —the common 
stock, which would supply the primitive forms and the most 
ancient meanings, is wanting. We possess Latin, Greek, 
Gothic, Sanscrit.... But we can go no farther back. The 
idiom whence these languages have been derived is Jost for- 
ever. We are obliged to restore the words by conjecture, 
and with the help of comparison; a task that is always 
delicate, of which the inexperienced reader must beware of 
becoming the dupe. I may here remark upon the present 
widespread fashion of putting in the missing words, while 
contenting oneself with warning the reader, by means of an 
asterisk, that it is a purely hypothetical form. These voca- 
bles, issuing from the laboratory of the linguist, have but an 
artificial existence. Since they are our own work, they can 
teach us nothing. Moreover, there is nothing final in their 
form, and it may be supposed that they are destined to 
numerous and perpetual changes; it is curious to compare, 
in this respect, the different editions of Fick’s Dictionary : 
from one edition to another we see the words of the common 
Indo-European stock transforming their vowels and conso- 


Vol. xxiv.] Canons of Etymological Investigation. 19 


nants, according to the progress of science and the theories 
successively in favor among phonetists. I admit these res- 
torations as useful epitomes of our knowledge, as formulae 
meant to fix ideas. But abuse follows so close upon use, 
that these would-be Indo-European words cannot be too 
cautiously handled. When we see how so wary a mind as 
Kluge’s allowed itself to be led into creating stems which 
he styles wxgermanisch or urindogermanish, we learn to mis- 
trust these too easy creations. I will give but two examples. 
To explain the German word Ze/t, ‘tent,’ he supposes a Ger- 
manic stem, Ze/d, of which he gives this strange translation : 
“ Decken ausspannen.”” We may be allowed to doubt whether 
there was ever a stem with so peculiara sense. At all events, 
Zelt is simply the Italian and the Provengal ¢exda, the Span- 
ish ¢zenda, with the same change of zd into /d that we find 
in the English chz/d compared with the German siud. There 
must have existed, in the popular Latin, a substantive zenda, 
coming directly from the verb ¢endere. It is a word which, 
like so many others belonging to the military language, passed 
from the Roman legions to the Germans, probably through 
the intermedium of auxiliary troops in the service of Rome.! 
So that the supposed Germanic root has but an imaginary 
existence. On the same page we find the word Ze/ter, mean- 
ing a particular kind of horse, a hack. Kluge compares 
Anglo-Saxon ¢ealtrian, ‘to tremble, to rock.’ But, as may 
be seen by the Middle High German zé/tari, the word is of 
Latin origin: it is the Latin Zo/wtarius, meaning a hack. 

Here we have a material proof of the danger of these 
reconstructions. If we could miraculously lay hands upon 
this oft-quoted Indo-European tongue, we should see how 
little it resembled the picture we have drawn of it. 

When it comes to restoring words, linguists think they are 
right in accumulating, in the prototypes they invent, all the 
phonetic elements presented by their descendants. Hence 
many strange-looking vocables. For instance, I find in 


1 We can find several instances of this kind of borrowed locutions in the mili- 
tary language of all nations and of all times. An interesting one is the Sanscrit 
khalinas, ‘curb,’ ‘ bit,’ which is the Greek word xahuvds, 


20 Michel Bréal. [1893. 


’ 


Kuhn’s Journal the word “tv7tos,” which is said to mean 
‘fourth,’ and is intended to explain guartus and téraptos. 
The Ursprache, after having been praised for a time on 
account of the harmony and purity of its vocalic system 
(only three vowels, —a, z, #) and the simplicity of its con- 
sonantismus (fifteen consonants), has suddenly come to be 
the least sonorous and most rugged of tongues. Let us 
congratulate M. Brugmann on his not having enumerated, 
in the volume of /udex that he has just published, the Indo- 
Germanic forms with which he has besprinkled his Gruna- 
viss; this list would have given a most unfavorable and 
unprepossessing idea of this venerable ancestress. 

A Mussulman told me one day that, if his religion forbids 
the reproduction, in drawing, of the human face, it is because 
there is danger of committing with the pencil some sin against 
anatomy, whereby these ill-shaped personages would come and 
reproach you in the other world with their malformation: to 
how many reclamations are our modern linguists exposing 
themselves, if ever the hybrid words that they have created 
should come and appear before their eyes in another world! 

I now come to what is, properly speaking, the subject of 
this study, —— What rules are to be followed in etymology ? 
It hardly seems needful to enounce the first rule: the lexi- 
cographer must conform himself to the lessons taught by 
phonetics. If etymology has ceased to be an amusement 
and a game, and has become a science, we owe it to the 
principles established by phonetics — principles that the ety- 
mologist should never lose sight of. We all, in turn, invoke 
these principles : they are our common safeguard and defence 
against the ever-to-be-dreaded and unconjurable inroad of 
fancy and caprice. Therefore we should never speak lightly 
of the laws of phonetics: we cannot tell if we shall not 
require their aid to-morrow against some ignorant or too 
systematic mind. 

But every one knows there is a difference between the 
respect of the believer and the superstition of the bigot. 
Whereas the bigot blindly follows the law, and declares 
whatever does not fit into foreseen and authorized cases to 


Vol. xxiv.] Canons of Etymological Investigation. 21 


be illicit and impossible, the believer examines thoughtfully 
whatever he meets with, and asks himself whether the gen- 
eral law is not held in check by some special and as yet 
imperfectly known law. The rules of phonetics must never 
be overlooked, but the obedience that we owe them is an 
intelligent obedience. These rules on the permutation of 
vowels and consonants are the product of observation ; obser- 
vation, carried still farther, will show their bounds and explain 
the exceptions.... Thus the true philologist, before letting 
fall the word zmposszble, should look closely at each case and 
give only a well-pondered opinion. 

The too oft-repeated saying, that the phonetic laws act 
blindly, is one of those catchwords that it is well not to 
accept uncontrolled. The phonetic laws act blindly if we 
admit a set of conditions that are never realized anywhere ; 
viz. a perfectly homogeneous population coming into no con- 
tact with the outside world, learning everything by living and 
oral tradition, without any books, without any monuments of 
religion, —a population in which every one should be of the 
same social condition, in which there should be no differences 
of rank, of learning, nor even of age or sex. No sooner do 
you leave aside pure theory, to place yourself in presence of 
the reality, than you see the reasons appear which make the 
phonetic laws open to exceptions. The authors of etymo- 
logical dictionaries are well aware of it; and not from them 
will there ever come anything resembling the above-mentioned 
axiom. 

The rules of phonetics, while directing our researches, must 
not be looked upon as a code that has foreseen everything, 
and to which there is nothing to be added. There are facts 
which necessarily escape the eye of the grammarian, since he 
is always being brought back to the same forms. Thanks to 
the lexicographer, new phonetic rules are discovered little 
by little. It is not difficult to acquire the reputation of being 
a rigorous observer of the phonetic laws, but it is quite as 
important not to check the progress of science. We may 
even ask ourselves how phonetics could have been elaborated, 
if the linguists who preceded us had declared whatever they 


22 Michel Bréal. [1893. 


found no example of to be impossible. It is known that the 
exceptions of to-day contain the germs of rules for to-morrow. 
To declare that the Latin Deus has nothing in common with 
the Greek @ecs may satisfy those who care above all for formal 
regularity. The wary lexicographer will put a note of inter- 
rogation, and leave the solution to the future. 

When we see into what remote comparisons and surprising 
parallels the bigots of phonetics are led, we prefer to vote 
with the ignorant and the sinners. To reduce Oeds to a root 
chet, ‘to frighten,’ or to a verb dhueso, ‘to breathe,’ is one 
of those extremities which seem to me harder than to admit 
the identity of @etos and dzvus. 

I now pass on to a second rule, which may appear very 
elementary, but nevertheless deserves mention and is of capi- 
tal importance. It is that we should always be careful to 
clearly distinguish the suffixes. It were vain to conform 
oneself to all the laws of phonetics: the etymologies will be 
marred with errors if we are unable to separate, in each word, 
the formal from the material element. 

This second recommendation will appear to some still more 
superfluous than the first; yet it is seen to be forgotten at 
every instant. Need we remind the reader that quite lately 
an eminent scholar explained the Latin dveviter by breve iter; 
thus separating this adverb from the numerous series to which 
it belongs (suaviter, fortiter, segniter, etc.) ; that in the adjec- 
tives longinguus, propinguus, he sees the Sanscrit root anc, 
‘to turn’? Ignorance of the suffixes was the disease from 
which etymology suffered among the ancients. If, during a 
long course of centuries, it is impossible to point to any prog- 
ress in the science of etymology among the Romans, it is to 
be attributed to this cause. Just as Varro explained /rater 
by fere alter, five or six centuries later gloriabundus was ex- 
plained by gloria abundans, oratio by oris ratio, monumentum 
by guod moneat mentem. In fact, the malady existed almost 
everywhere. The Hindoos, to whom we owe the first lists 
of suffixes, forget them as soon as they come to decompose 
words ; they make no difficulty about explaining drakman by 
the root d7ih, ‘to grow,’ and man, ‘to think’ (what makes 


Vol. xxiv.] Canons of Etymological Investigation. 23 


thought grow), or agni by ang, ‘to anoint,’ and Zz, ‘to 
conduct’ (he who conducts the libation). 

Let us not, then, fear to inscribe this rule among those 
which should always be present to the mind of the linguist. 
The Linguistic Society of Paris undertook, a good many years 
ago, the publication of a Latin Dictionary in which the words, 
instead of being arranged according to the initial letters, are 
arranged according to the final letters. Circumstances too 
long to relate have delayed the publication of this dictionary. 
A book of this kind would doubtless render great services ; 
from the day when we see arranged in order all the words in 
mentum, — like segmentum, augmentum,—we shall no longer 
be tempted to explain the second part of avgumentum, as a 
linguist has recently done, by the verb meniscor—the same 
which has given reminiscor, comminiscor. 

I now come to athird rule. The concordance of meanings 
must be the-object of as minute an examination as the con- 
cordance of forms. We see philologists who carry the study 
of consonants and vowels to a great length, and yet prove 
singularly careless in the matter of sense, Provided there be 
some distant affinity between the sense of two words, that is 
enough: the two words are declared to be of the same origin. 
This is a grave oversight concerning quite half the history 
of words, an oversight which may mar many an etymology. 
I see, for instance, that Vani¢ek places under the stem £2, 
**to lie’ (Sanscrit ¢@) not only the verb xetpyaz, ‘I lie,’ and the 
substantive «xoirn, ‘a couch,’ but words like the Greek x@pos, 
‘a feast,’ c®pn, ‘a dwelling,’ c@as, ‘a fleece,’ the Latin czvzs, 
‘a citizen,’ caelebs, ‘a bachelor,’ tranguzllus, ‘tranquil,’ guze- 
scere, ‘to rest.’ Independently of all sorts of material difficul- 
ties, there is no plausible connection between the meanings 
of these words. Take but the last of these parallels, guze- 
scere and xeiyat are by no means synonymous terms, One 
may lie without resting (for instance, the wounded and the 
dead), and one may rest without lying (for instance, when 
one is seated), George Curtius rightly rejects this relation- 
ship because a Sanscrit ¢, a Greek 4, is never represented in 
Latin by gw: it has been seen that he might have rejected 


~ 


24 Michel Bréal. [1893. 


it, no less rightly, in taking his stand upon the difference of 
sense. The laws which govern the changes of meaning 
deserve to be studied with the same care as the permutation 
of letters. George Curtius connected the Latin penuria with 
the Greek wefva, ‘hunger.’ But pexuria belongs by its suffix 
to the family of esuries, parturio, nupturio; it could not, if we 
admit the etymology of Curtius, mean anything else than the 
desire to be hungry, which is inadmissible: it means, on the 
contrary, the desire to have provisions. There is conse- 
quently no doubt of its relationship with penus, penum, in 
spite of the different quantity. ; 

I am often astonished to see how carelessly the authors 
of etymological dictionaries go to work when they wish to 
determine the meaning of a word. They usually stop at 
the best-known signification, — the one which is at the head 
of the vocabularies, — without asking themselves if it is not 
an indirect and modern sense, perhaps the last in date. I 
see, for instance, that the adjective /aetus has been connected 
with the Sanscrit #72, ‘to rejoice,’ under pretext that it means 
‘joyful.’ But that is a derived sense: /aetus first meant ‘fat.’ 
It was said of things before being said of persons: /aetas 
segetes, ager laetus, convivium laetum. Thence the verb /ae- 
tare, ‘to fatten,’ ‘to fertilize,’ and the substantive /actamen, 
‘manure.’ It was only metaphorically that they came to say 
lactum augurium, sus laetus, frons laeta. What, then, becomes 
of the etymology referred to just now? 

In order to discover the primitive sense we must not stop 
on the surface, but look up the whole history of the word, 
examine its rare and obsolete uses. It is often the last 
meaning, thrust in at the end by the dictionaries, that is the 
primitive meaning. At other times, it will be found in the 
compounds and derivatives. Almost the only meaning now 
left to the word Muth is ‘courage’; but it once meant ‘intel- 
ligence,’ ‘soul,’ and that is still its meaning in Grossmuth, 
‘generosity, ‘greatness of soul,’ Demuth, ‘humility’ (Old 
High German deomuote, literally ‘soul of a servant’). We 
also find this sense in the derivatives suthmaassen, ‘to con- 
jecture,’ ecnmithig, ‘unanimously,’ Gemiith, ‘soul.’ In the 


Vol. xxiv.] Canons of Etymological Investigation. 25 


present day, Wie ist es dir zu Muthe, ‘in what state of mind 
are you,’ is sfill said. The primitive meaning has been kept 
in the English mood. A similar example is that of the Ger- 
man W7tz, which is hardly ever used now in any other sense 
than ‘ingeniousness,’ meaning a joke or a witticism. But 
this term had formerly a loftier signification: it indicated 
learning or wisdom. This meaning has remained in com- 
pounds like Wahnwitz, and in the verb witzigen, ‘to make 
wise.’ Goethe remembered the early meaning : 


Was lockst du meine Brut 
Mit Menschenwitz und Menschenlist 
Hinauf in Todesglut. 


Here again the English has remained more archaic: wit, 
‘intelligence.’ 

I have chosen these two examples from a living language, 
because there the study of meanings is easier and clearer. 
But the necessity of observing the filiation of the sense is 
none the less binding for dead languages. 

In the ancient tongues, likewise, it is well to examine com- 
mon expressions, in order to discover the primitive meaning. 
Thus, the Latin /tterae figures in the dictionary with the single 
meaning ‘letters.’ But it first signified the tablets they wrote 
upon, and that is the sense it has kept in /tteris mandare, in 
litteras mittere, litteras dare ad aliquem. Beneath this word 
litterae (that is generally derived from “twra, which is as un- 
satisfactory for the sense as for the form) is concealed the 
Greek word d:60épai, ‘skin prepared for writing upon tablets.’ 
The meaning of the word, when isolated, may have changed: 
it retains its former value in the sentences in which it was 
habitually introduced. 

In thus searching for the sense, one may be led into some- 
what unexpected comparisons. Thus, the Latin adjective 
serus, that is usually translated by ‘late,’ first signified ‘slow,’ 
and still earlier ‘heavy.’ It is synonymous with gravis. The 
poet Afranius says: 


Non ego te novi tristem, serum, serium. 


26 Michel Bréal. [1893. 


Sallust, in a passage preserved by Servius, says, in speaking 
of a war, serum bellum in angustits futurum. Setvius explains 
serum to mean ‘grave.’ Lastly, Virgil writes this line: 


Seraque terrifici cecinerunt omina vates. 


Here again Servius translates sera by gravia. Now, lin- 
guistics confirm this translation in the most satisfactory way. 
The Latin sevws corresponds with the Anglo-Saxon szwér, the 
German schwer, the Lithuanian sqwarus, ‘heavy.’ 

An etymological dictionary, giving the history of the mean- 
ings, should indicate the group of ideas, the form of social 
life, the series of occupations or conceptions to which a word 
owes its birth, and the different social layers through which 
it has passed. The most general terms— words meaning ‘to 
do,’ ‘to set,’ ‘to throw ’— are those that have the richest and 
most complicated history ; because all the little societies into 
which the great society is divided and of which a nation is 
composed, have taken hold of these words and set their stamp 
upon them. So we must not expect the different meanings 
of a word to be deduced one from another in a straight line, 
as through a single series. The general sense may be broken 
up into a number of special senses, all of the same age; the 
verb agere being employed simultaneously, at Rome, by those 
who had a suit to plead (agere causam, or simply agere) and 
by those who had a part to play upon the stage (agere partes, 
actor). The sacrificer, on his side, asking if he was to strike 
the victim, said, Agone? In this way, certain articles in the 
dictionary may present an epitome of the activity of a whole 
people. As to the general sense it is itself derived from 
some special sense which has faded away by degrees. Thus, 
ago first meant ‘to drive along.’ It must have been first used 
by the shepherds. £x ipse capellas Protinus aeger ago. 

Just think of the various meanings the word matter has 
assumed in English, used as it is in almost ‘every art, every 
trade, every kind of activity or study. This word, through 
the intermedium of the French mazzére, derived from the 
Latin materies, which signified the new wood grown after 
grafting, or after the top of the plant has been tied up. Such 


Vol. xxiv.] Canons of Etymological Investigation. 27 


is the explanation given by Columella in speaking of the cul- 
ture of the vine. We have here an example of the double 
movement ; that is to say, a special sense ending in a general 
sense, which, in its turn, is subdivided into an infinite number 
of special senses. 

Here I will cut short these reflections, which might be 
developed at great length; for all, or almost all, the chapter 
of linguistics treating of Semantics, or the science of mean- 
ings, has yet to be written. Yet, I would still call attention 
to one point. 

An idiom is never wholly isolated: it is in contact with 
other idioms, whence mutual loans. But these loans are not 
confined to taking words from another language. They are 
sometimes of a more hidden nature, when they consist in 
thrusting a new meaning upon a native word, in imitation of 
the foreign tongue. This will be made clear by an example. 
The Greek xécpos has two meanings: it signified the order 
existing in the world; and the order existing in the attire, 
the apparel. The Romans, who called the attire mundus, 
made mundus the equivalent of xoopos; adding to its first 
sense that of world and universe. What proves this second 
acceptation to be recent is that it has hardly furnished any 
derivatives ; whereas, from the first sense, we get zmmundus, 
munditia, emundare, etc. > 

Loans of this kind are to be noticed at every epoch. If 
the German /esex has two meanings, viz. ‘to collect’ and ‘to 
read,’ it is probably on account of the double meaning of the 
Latin /egere. We see that reading is called by different names 
in the various Germanic idioms. The influence that Roman 
civilization has exercised upon a number of German words. 
might supply the subject of an interesting study. Thus, 
the German barmherzig, ‘compassionate,’ formerly armherzt, 
is a copy of the Latin mzsericars. On the other hand, the 
French substantive avenir (the future) looks’as if it had been 
formed on the pattern of the German Zukunft; and to pass 
on to modern times, the word p/ateforme, which has entered 
into our political language, comes to us straight from the 
United States. 


28 Michel Bréal. : [1893. 


M. Hugo Schuchardt has written a curious study on these 
reciprocal influences exercised by idioms in contact with one 
another. 

Vainly do the purists in every nation seek to combat them: 
here we have an example of the slow and irresistible progress 
of civilization. These Uebertragungen, from one idiom to 
another, — far more numerous than is generally supposed, — 
are the cause by which all modern languages appear to be 
keeping step with one another. A metaphor found in one 
country immediately becomes the common property of all 
the other countries; a felicitous expression, a new and pic- 
turesque turn, are sure to be reproduced everywhere. The 
authors of etymological and historical dictionaries have here 
a vein, as yet unexplored, which will enable them to trace 
out what is called, somewhat vaguely, the genius of modern 
languages. . 


Vol. xxiv.] Ein Ablautproblem der Ursprache. - 29 


Ill.— Ein Ablautproblem der Ursprache. 


By WILHELM STREITBERG, 


PROFESSOR IN FREIBURG, SWITZERLAND. 


LEICHTE und schwere Vokalreihen sind im Indogerma- 
nischen scharf geschieden. Bei diesen ist die Lange, bei 
jenen die Kiirze zu Grunde zu legen. Nun erscheinen aber 
auch bei den leichten Reihen in ganz bestimmten Form- 
kategorien lange Vokale. Es ist klar, dass sie erst 
sekundaér durch Dehnung aus Kiirzen entstanden sein 
miissen, wenn diese mit Recht als das urspriingliche an- 
gesehn werden. Das Problem ist also das: wodurch sind 
die Langen der leichten Reihen entstanden? Welche 
Ursachen haben die Dehnung veranlasst? d 

Ich iibergehe die scharfsinnigen Erklarungsversuche von 
H. Moller (Paul-Braunes Beitrage VII 492 ff.) und A. 
Fick (Gottinger gelehrte Anzeigen 1881 S. 1452 ff.), die 
beide in der Dehnung die Wirkung eines musikalischen 
Akzentes sehn, und wende mich direkt zu der nach meiner 
Uberzeugung richtigen Deutung. Drei Gelehrte haben sie, 
unabhangig von einander, ausgesprochen. 

K. F. Johansson (GGA. 1890 S. 765) vermutet, “ oft 
sei mit der Reduktion eines Vokals die Verlangerung eines 
andern verbunden ;” die Entstehung von #dg- aus uego- uoge-, 
von péd- pod- aus pedo- pode- beruhe also auf demselben 
Prinzip wie die Entstehung der schwedischen Dialektformen 
Jor vet far aus fora veta fara. 

F. Bechtel (Hauptprobleme der idg. Lautlehre S. 181) 
zieht dieselben schwedischen Dialekterscheinungen wie 
Johansson heran und sieht hierin den Schliissel zu einer 
“mechanischen Erklarung der Dehnung.” Unter diesem 
Gesichtspunkt lasse sich die Dehnung aller Silben be- 


“s 


30 Wilhelm Streitberg. [1893. 


greifen, hinter denen die einstige Existenz einer zweiten 
angenommen werden diirfe. 

Friiher als die genannten Forscher hat Viktor Michels 
miindlich den gleichen Gedanken gedussert. Er bringt die 
Entstehung der Dehnstufe in Verbindung mit seinem Gesetz 
von der Entstehung des Zirkumflexes. Nach ihm wird ein 
betonter langer Vokal, hinter dem eine Silbe geschwun- 
den ist, geschleift; ein betonter kurzer Vokal, hinter 
dem eine Silbe geschwunden ist, wird dagegen gedehnt. 

Auch hier ist die Fassung des letzten, die Dehnstufe 
betreffenden Gesetzes trotz der vorgenommenen Ein- 
schrankung noch immer zu weit. Nicht um die Lange 
oder Kiirze des Vokals handelt sichs namlich, sondern 
—wie sich spater ergeben wird—lediglich um die Lange 
oder Kiirze der Silbe. 

Ich selber formuliere daher beide Gesetze, wie folgt : 

Schwindet eine akzentlose Silbe, so wird eine 
vorausgehnde betonte Silbe zirkumflektiert, 
wenn sie lang, gedehnt, wenn sie kurz ist. 

Man sieht, es handelt sich hier um ein ‘Gesetz des 
Morenersatzes,” von dem schon Fick aQO. gesprochen hat. 
Und in diesem, d.h. in der Annahme “dass die Lange zwei 
Kiirzen in sich vereinigt,” beruht, wie Bechtel aO. mit 
Recht hervorhebt, der gesunde Kern von Mollers und 
Ficks Dehnungshypothesen. 

Da, wie schon hervorgehoben worden ist, ein Versuch 
die vorgeschlagne Hypothese zu beweisen noch nicht 
gemacht worden ist, will ich, so gut es angeht, diese 
Liicke auszufiillen unternehmen. Eine Priifung samtlicher 
fiir die Dehnstufe vorhandnen Beispiele wird, wie ich hoffe, 
die Richtigkeit des Gesetzes, speziell der Fassung, die ich 
ihm gegeben habe, dartun. Fiir den Augenblick freilich 
muss ich es bei einer fliichtigen Musterung bewenden 
lassen; das vollstandige Material soll demnachst in den 
Indogermanischen Forschungen vorgelegt werden. 

Die unumgingliche Voraussetzung fiir die vorgeschlagne 
Erklarung der Dehnstufe ist die Berechtigung der Schwund- 
stufe nicht nur vor, sondern auch nach der Silbe, die den 


Vol. xxiv.] Fin Ablautproblem der Ursprache. 31 


Wortton traigt. Das ist unbedenklich; denn in Praxi ist 
von jeher mit der progressiven Akzentwirkung operiert 
worden. Und wenn auch hier und da rein theoretische 
Bedenken geadussert worden sind, so diirfen sie doch heute, 
nach Kretschmers reicher Sammlung KZ. XXXI 325-366, 
als beseitigt gelten. Die Bahn ist also frei. 


Ich wende mich den einzelnen Belegen zu. 


A. Nomen. 


Eine ganze Reihe von Kasus weisen Dehnstufe auf. a 
sind die folgenden : 


I. Nominativ Sing. 1. Wurzelnomina. Idg. diéus 
qous. Es scheint mir in jeder Beziehung gesichert zu 
sein, dass beide Nomina von leichten Wurzelstimmen 
(diéu- gou-) kommen. Johannes Schmidt KZ. XXV_ 54 
setzt allerdings gdu- als Wurzel an, steht jedoch hiermit 
ganz isoliert. Die Griinde, die gegen die Aufstellung eines 
langen Wurzelvokals sprechen, findet man in des Verfassers 
Schrift: Zur germanischen Sprachgeschichte (S. 51 ff.). 
Was fiir géu- gilt, trifft auch bei dzéu- zu. 

Wenn nun die Wurzel urspriinglich kurzen Vokal hat, 
auf welcher Ursache beruht dann die Lange des Nomi- 
nativs? Die Antwort ist schon gegeben, sie lautet: auf 
Silbenverlust. In beiden Wortern hat hinter dem 4 
urspriinglich noch ein kurzer Vokal gestanden. Die Urform 
ist demnach *diéuos * duos. Durch den Schwund des unbe- 
tonten Endungs-o wird der vorausgehnde kurze Tonvokal 
gedehnt. Die Dehnung des Wurzelvokals erfolgt also beim 
Ubergang des urspriinglichen ¢/o-Stamms in die sogen. 
konsonantische Flexion. 

Ein solcher Ubergang hat nichts befremdliches. We- 
nigstens nicht fiir den, der gleich mir der Ansicht ist, dass 
im Nominativ Sing. -zo- zu -2-, -yo- zu -u-, -mo- Zu -v- geworden 
ist. Man vergleiche nur lit. mzdis, Genitiv mé@dio, aind. 
tdéku- neben takvd-, griech. péyas aus idg. mégus, neben lat. 


=a 


32 Wilhelm Streitberg. [1893. 


magnus aus idg. magnés. Der Vorgang ist hier derselbe 
wie dort. Das unbetonte Endungs-o schwindet. Geht ihm 
ein Laut voraus, der selber silbebildend auftreten kann, so 
muss dieser silbisch d.h. Trager des Silbenakzents werden. 
Das ist bei dem z w m der angefiihrten Worter der Fall. 
Unter solchen Umstanden ist also die Silbenzahl des Wortes 
unvermindert bewahrt. Deshalb bleibt auch die vor der 
Schwundsilbe befindliche betonte kurze Silbe véllig unver- 
andert. 

Anders verlauft die gleiche Entwicklung, wenn der dem 
ausfallenden Endungs-o vorausgehnde Laut nicht silbisch 
werden kann, sei es, dass ihm dies seine Natur verbietet, 
sei es, dass ihn ein vorhergehnder Vokal daran_hindert. 
Der Prozess des Vokalverlustes ist hier zwar derselbe, aber 
er zieht eine Verminderung der Silbenzahl des Wortes nach 
sich. Damit aber ist die Bedingung fiir den Eintritt der 
Dehnung in der vorausgehnden kurzen Tonsilbe gegeben. 

Wer also —ich wiederhol es—an meiner Erklarung von 
médts, tdku-, wéyas keinen Anstoss genommen hat, der kann 
auch gegen die von idg. adzéus qous nichts stichhaltiges ein- 
wenden. Denn die Worter der ersten Gruppe bilden nur 
eine Unterabteilung in jener grossen Gemeinschaft, der alle 
Nominative mit urspriinglich nachtonigem und daher dem 
Schwund ausgesetzten o angehoren. Wer A sagt, muss 
auch B sagen. 

Dass wir aber ein Recht haben, bei den dehnstufigen 
Nominativen konsonantischer Stamme von alten ¢/o-Bil- 
dungen auszugehn, das beweisen aufs klarste die ungemein 
zahlreichen ¢/o-Bildungen, die ihnen zur Seite stehn. Es 
ist das grosse Verdienst Wheelers in seinem Buch iiber 
den griechischen Nominalakzent zuerst nachdriicklich auf 
diese Doppelheit hingewiesen zu haben. Man vergegen- 
wartige sich die folgenden Faille. 

Idg. dizus lat. divos; idg. pods, aind. mit betonter Endung 
paddm,; idg. uogs hat den es-Stamm idg. wegos zur Seite; 
zu lat. /ér gehdrt das e/o-Verb idg. /éghd, zu lat. rer rego. 
Ferner geh6ren zusammen griech. -8rew und Aréra, crow 
und «rérTw, AoW und Aéww (vgl. auch Aozds ‘Rinde’ und 


Vol. xxiv.] Ein Ablautproblem der Ursprache. 33 


den es-Stamm ézos), ceo und oxérropat, fee und 
Feéra, pop und dépw (dazu dopds ‘tragend, fordernd,’ vgl. 
tereo-popos fiir *redXec-popds). Zu aind. dvar- lat. foris 
(= idg. Nominativ Plur. dhudres) tritt got. daur usw., lat. 
forum, abg. dvorii; zu griech. @yp lat. ferus; zu dornp 
homerisch dotpa; zu ahd. sxuor griech. vedpov und veupd, 
zu aind. Nominativ -a@ ‘tétend’ ved. ghand- griech. dv8po- 
dovos; zu d@ (aus *dom nach Michels’ Gesetz) aind. damd-, 
griech. dduos sowie das ¢/o-Verb déuw; zu avest. zyaJ aind. 
himd- und die neutral-femininen Kollektiva russ. zzmd, lit. 
éma; zu griech. yynv aind. hkgsd-; zu lat. ros aind. rdsa- 
und rasa@- lit. rasa. 

Ich begniige mich fiir jetzt mit dieser fliichtigen Aufzah- 
lung. Die Beispiele sind simtlich ganz durchsichtig. 

Bei allen steht erstlich fest, dass die auftretenden 
Langen durch Dehnung entstanden, also sekundar sind. 
‘Das beweist einmal das Erscheinen der Kiirze auch in 
den starken Kasus. Man vergleiche aind. xdévam = griech. 
avépa, griech. woda = lat. pedem, aind. Lokativ. Sing. dydvi 
= lat. Jéve. usw. Ferner findet man in den schwachen 
Kasus zahlreiche Belege von Schwundstufenformen, die 
kurzvokalische Vollstufen voraussetzen. Z. B. aind. divds 
= griech. Avds, aind. dy#bhi3; aind. Akkusativ Plur. déras 
durds; aind. nrbhyas = avest. nar’byd, nrsu = griech. 
avipdot usw.. Diese Schwundstufen tragen das Geprage 
hoher Altertiimlichkeit, da man Schritt fiir Schritt beo- 
bachten kann, wie sie dem Drang nach Uniformierung 
erliegen, bis schliesslich die Lange im ganzen Paradigma 
herrscht, wie bei aind. vdac- lat. vdr, lat. er rér usw. 

Zweitens steht fest, dass die engsten Beziehungen 
zwischen den konsonantisch auslautenden Dehnformen und 
dehnungslosen e/o-Stammen vorhanden sind. Welcher Art 
sind diese Beziehungen? A priori lassen sich zwei ver- 
schiedne Méglichkeiten denken: 

a) Bei den konsonantischen Dehnstammen ist unbetontes 
e/o im Auslaut geschwunden, der konsonantische Stamm 
beruht also auf einem vokalischen. 

.*b) An einen urspriinglich konsonantischen Stamm _ ist 


34 Wilhelm Strettberg. [1893. 


“das Suffix e/o angetreten,” der vokalische Stamm ist also 
jiinger als der konsonantische. 

Fiir jeden, der die Theorie von den absteigenden Ablaut- 
reihen, die in den siebenziger Jahren an die Stelle der alten 
Gunatheorie gesetzt worden ist, nicht bloss als iiberlieferten 
Lehrsatz hinnimmt, sondern die Konsequenzen der neuen 
Lehre zu ziehn bestrebt ist, kann die Entscheidung nicht 
zweifelhaft sein. 

Denn das unmotivierte “Antreten”’ des “Suffixes” e/o an 
die “ Wurzel” ist um nichts leichter begreiflich als das 
“Einspringen” des steigernden a (oder e¢/o) in die ‘ Wurzel.” 

Der Ablaut ist verstandlich geworden, weil man, anstatt 
mit unbekannten Gréssen zu rechnen, diejenigen Krafte zu 
Hilfe gerufen hat, deren Wirksamkeit in der Sprachent- 
wicklung wir noch heute mit unsern eignen Augen beo- 
bachten kénnen. 

Dieselben Mittel, deren wir uns bedienen, um das Ver- 
haltnis von Aelw@ und édu7ov, von oda und iduev unserm 
Verstandnis naher zu riicken, genau dieselben befahigen 
uns auch das Verhaltnis von Zevs und divos, von Onp und 
férus, von dap und dopds, dépw zu verstehn, wie sie uns 
schon vorher die Ursachen der Doppelheit ¢déku- takva- 
zu erkennen gelehrt haben. 

Hier liegt also eine Kette vor uns, wo Glied um Glied 
ineinandergreift, bis der Ring geschlossen ist. Dort nichts 
als distecta membra: Ein Suffix e/o, das—man weiss nicht, 
wozu —antritt; eine Ausnahme des Ablautgesetzes, die — 
man weiss nicht, weshalb—eintritt ; eine Vokaldehnung, die 
—man weiss nicht, warum —auftritt. 

Doch man wird sich schwerlich an diesen R§atseln ge- 
niigen lassen, die ein Ausgehn von den konsonantisch aus- 
lautenden Formen unvermeidlich mit sich bringt, sondern 
das Erklarungsmittel zu Hilfe rufen, das schon in einem 
analogen Fall Aufschluss gegeben hat. Genau wie wir idg. 
smés aus einer vollern Urform *esmés herleiten, wenn diese 
auch nicht unmittelbar bezeugt ist, genau ebenso haben 
wir zur Erklarung eines dehnstufigen Nominativs eine vol- 
lere Urform vorauszusetzen, von deren einstiger Existenz 


Vol. xxiv.] Lin Ablautproblem der Ursprache. 35 


die den Dehnbildungen parallelen ¢/o-Staimme unzweideutig 
Zeugnis ablegen. 

Ein wurzelbetonter zweisilbiger ¢/o-Stamm muss iiberall, 
wo keine Ausgleichung vorliegt, durch die Wirksamkeit der 
allgemein herrschenden Ablautgesetze seinen unbetonten End- 
ungsvokal verlieren. Hierdurch wiirde die Silben- und die 
Morenzahl des Wortes vermindert, wenn nicht ein teilweiser 
Ersatz eintrate: die Quantitat der schwindenden Silbe iiber- 
tragt sich auf die vorausgehnde betonte Silbe. Hierdurch 
bleibt, bei verringerter Silbenzahl die Morenzahl des Wortes 
trotz des Verlustes unverandert. 

Dass diese Quantitatsausgleichung nur beim Schwund 
solcher Silben stattfindet, die der Tonsilbe folgen, nicht 
aber beim Verlust derjenigen, die ihr vorausgehn, ist fiir 
den nicht befremdlich, der sich den ganz verschiednen 
Charakter progressiver und regressiver Akzentwirkung ins 
Gedachtnis ruft. Beruht doch der Schwund einer nachto- 
nigen Silbe im wesentlichen darauf, dass sie bei der Bildung 
der Tonsilbe vorweggenommen wird. Das ist namentlich 
durch Axel Kocks Untersuchungen iiber den germanischen 
Umlaut dargetan worden. 

Ein Einwand liegt nah: Es ist unméglich—so wird 
man sagen—dzur Erklarung der dehnstufigen Nominative 
iiberall ¢/o-Stamme vorauszusetzen. Denn es finden sich 
auch Feminina darunter wie z. B. idg. wdgs. Der Ein- 
wurf halt nicht Stich. Seit Brugmanns und Wheelers Un- 
tersuchungen iiber die Entstehung des Nominalgeschlechts 
in der idg. Ursprache, darf unter allen Umstanden soviel 
als feststehend betrachtet werden, dass weder dem “ Suffix” 
@ noch dem “ Suffix” e/o von Haus aus bestimmtes Genus 
eigen war. Ja, wie odds und Genossen lehren, hat es selbst 
_ noch in historischer Zeit neben den Maskulinen auch Femi- 
nina in der e/o-Deklination gegeben, ohne dass ein ausseres 
Unterscheidungsmerkmal bestanden hatte. Da die gleiche 
Erscheinung bei allen sogen. konsonantischen Stammen 
sowie bei den ez- und ex-Stammen wiederkehrt, so haben 
wir ein wolbegriindetes Recht darauf, die aussere Trennung 
der Genera als etwas jiingeres, sekundares zu betrachten. 


“s 


36 Wilhelm Streitberg. =~ [1893. 


Damit ist der vorgebrachte Einwurf erledigt. 

Der Gegner mag das zugeben, aber er wird sich noch. 
nicht fiir besiegt erklaren. Das alles — wird er fortfahren 
—pbeweist, dass wir es mit Vorgaéngen zu tun haben, die 
sich in weitentlegnen Zeiten abgespielt haben. Glottogo- 
nische Probleme dieser Art lasst man aber am besten auf 
sich beruhn. 

Ich leugne, dass es sich bei der vorliegenden Frage um 

ein “glottogonisches” Problem handelt. Aus dem einfachen 
Grunde nicht, weil wir es mit fertigen Wortern- und 
deren Weiterentwicklung zu tun haben, nicht aber mit der 
Entstehung dessen, was man in der idg. Ursprache als 
Worter bezeichnet. Wer das nicht zugeben will, der muss 
auch die Herleitung von idg. smés aus der Urform *esmés 
fiir ein glottogonisches Problem erklaren; denn iiberliefert 
ist hier der Ausgangspunkt so wenig wie dort. Damit war 
iiber die ganze Ablautforschung der Stab gebrochen. 
- In Wirklichkeit steht es nicht so verzweifelt. Denn, wie 
schon der Ausdruck sagt, befasst sich nicht der mit glotto- 
gonischen Problemen, der vom fertigen idg. Wort ausgeht, 
sondern vielmehr der, dessen Bestreben es ist, die fertigen 
Worter in lauter Atome zu zerlegen, indem er auf Schritt 
und Tritt “ Wurzeldeterminative” wittert, bis schliesslich 
vom Worte kaum ein Laut mehr iibrig bleibt. 

Dass die vorgeschlagne Erklarung der Dehnstufe richtig 
ist, wird auch abgesehn von den vorhergegangnen Erwa- 
gungen durch den auffallenden Parallelismus wahrscheinlich, 
worin die langstammigen Wurzelworter zu den kurzstammi- 
gen stehn. Wahrend bei den letzten, wie gezeigt, die Ton- 
silbe gedehnt wird, bekommt sie bei den ersten nach 
Michels’ Gesetz schleifenden Ton. 

So steht neben idg. ditus aus *diéuos ein ae Nomi- 
nativ nds aus *nduos. Der gleiche Unterschied besteht 
zwischen dor. was (so ist statt mas bekanntlich zu 
schreiben) von der leichten ‘Wurzel’ jpedo- pode- und 
vedisch das (Oldenberg Hymnen des Rigveda I 173) von 
einer schweren Wurzel. Zweisilbig wird im Veda gemessen, 
-das in sudas sudasam. Auch hier liegt die schwere Wurzel 


Vol. xxiv.] Ein Ablautproblem der Ursprache. 37 


do- zu Grunde. ' Sehr instruktiv ist endlich der Zirkumflex 
in griech. yAadé neben yAaveds. Die Urform ist *g/dukos 
gewesen. Der Schwund des unbetonten Endungs-o hat 
keine Verlangrung der vorausgehnden Tonsilbe verursachen 
kénnen, da diese als geschlossne Silbe schon vorher lang 
gewesen ist. Es hat daher nach Michels’ Gesetz Akzent- 
wechsel stattgefunden. 
* Bartholomae BB. XVII 105 ff. hat das Verhialtnis idg. 
ditus : idg. ndiis geahnt, wenn er den durch Dehnung 
entstandnen langen Vokalen- der leichten Ablautreihen 
iiberlange gedehnte Vokale bei den schweren Reihen 
gegeniiberstellt und die Proportion bildet : 


idg. uogs : wegos =idg. *urdks (hom. Nom. Pl. payes) : pyyvupe. 


2. Neutrale Nominative auf -d. Eine zweite, nur kleine 
Gruppe dehnstufiger Nominative bilden die Neutra auf -d, 
deren Flexion zuerst Johannes Schmidt erkannt hat. Mit 
Sicherheit sind nur idg. s@/d ‘Salz’ und idg. kérd ‘Herz’ 
hierherzurechnen. Es sind Nominative urspriinglicher 
e/o-Stamme, die mit dem pronominalen “Neutralsuffix” 
-d gebildet sind, anstatt mit dem nominalen -m. Ein 
Nominativ idg. sa@/d kérd steht also auf einer Stufe mit 
altlat. a/id aus idg. aliod. Als Urformen sind’ daher *sélod 
*kérod anzusetzen. Der Zirkumflex in griech. «jp stammt 
aus den obliquen Kasus. ; 


‘3.  « wrouny — Hryepov. 
TaTnp — pyTop - 
evyevrns — Has cides maior 


Anté Ante (aind. sékhda) 
Bacwrevs (apers. bazau5). 


- Die obliquen Kasus mit starker Suffixstufe beweisen, 
dass die Normalform des Suffixes kurzen Vokal besitzt, 
dass die Linge des Nominativs erst einer Dehnung ihre 
Entstehung verdankt. © 

- Von Baouwrevs (bazau5) abgesehn, sind die angefiihrten 
Nominative samtlich ohne das Kasussuffix -s gebildet. 
Worauf das beruht ist unklar. Nur bei den o7-Stammen 


38 Wilhelm Strettberg. [1893. 


lasst sich die s-Losigkeit durch einen Hinweis auf die 
Analogie der @Stamme begreiflich machen. “Bei den 
andern fehlt jeder Anhaltspunkt. Man konnte die alte 
Regel, dass die einsilbigen konsonantischen Stamme den 
Nominativ Sing. mit, die mehrsilbigen ihn ohne -s bilden, 
dahin umformen, dass man von zwei- und mehrsilbigen 
e/o-Stammen spricht —hiermit ist jedoch nichts erklart. 

- Man muss sich also mit der Konstatierung der blossen 
Tatsache begniigen, wenn man nicht folgenden Deutungs- 
versuch, der mir persdnlich sehr verlockend scheint, gut- 
heissen will. : 

Wenn -ss im Auslaut schon in der Urzeit vereinfacht 
worden ist, wie Brugmann Grundriss II § 370 S. 7o1 
Anmerkung annimmt, so Jasst sich der Nominativausgang 
2s -os anstandlos auf Altres -Zss -dss, entstanden aus 
urspriinglichem *-ésos *.Jsos, zuriickfiihren. Ferner ist es 
erlaubt -ér -dr aus ilterm *-évs *-drs herzuleiten. Den 
lautgesetzlich entstandnen Nominativausgingen -és -és 
und -éy -ér kann dann -é -dn nachgebildet sein. Dass 
namlich auch bei den exz-Stammen ein Nominativ auf -s 
bestanden habe, scheint mir durch die isolierten Einsilbler 
avest. sa und aind. kas, avest. zy@J, sowie namentlich 
durch die Partizipia auf -uevo-, die formell wie begrifflich 
in engster Beziehung zu den ez-Stammen stehn, bewiesen 
zu werden. 

Natiirlich hat dieser Deutungsversuch nichts mit der 
Erklarung der Nominativdehnung zu schaffen.. Diese 
bleibt auch von seiner Ablehnung unberiihrt. Nur dass 
alsdann die Doppelheit aind. #3as und griech. syeuav so 
dunkel ist wie zuvor. 

Dass wir auch bei den Nominativbildungen dieser Gruppe 
von urspriinglichen ¢/o-Stammen auszugehn haben, beweist 
aufs klarste die schon angefiihrte Partizipialform auf -seno-. 
Dass formell ein *-¢évo- als Urform fiir -t2y vorausgesetzt 
werden darf, lehren die begrifflich allerdings weitablie- 
genden Komparative auf -tevo-.. Neben den es- und ex- 
Bildungen haben sich keine ¢/o-Formen erhalten. Aind. 
sakha jedoch hat lat. soctus neben sich, wodurch die postu- 


Vol. xxiv.] Ein Ablautproblem der Ursprache. 39 


lierte Urform bestatigt wird. Denn es besteht die Pro- 
portion : 
sdkha : socius = *sog(h)éto : *sog(h)té-s. 


4. Es bleibt noch eine Nominativform iibrig, die eine 
scheinbare Ausnahme bildet: namlich der Nominativ Sing. 
der Partizipia auf -z¢-. Sie haben in der Urzeit unzweifel- 
haft kurzen Vokal besessen. Vgl. aind. bhdvan adan 
sowie das gleichflektierende Substantiv dém ‘Zahn.’ 

Dass auch hier ¢/o-Stamme zu Grunde liegen, beweisen 
isolierte Formen wie aind. hémantd- = griech. d-yeipavrtos, 
aind. vasantd-, namentlich auch aind. va@ta- = lat. ventus 
‘Wind, wehend.’ Auch die Bildungen wie lat. cognomentum 
= dvopata, stramentum = griech. otpopara, ahd. hliumunt = 
aind. $rématam sind hierherzuziehn. 

Trotz des o-Schwundes fehlt die Dehnung des voraus- 
gehnden kurzen Tonvokals. Mit Recht, da die Tonsilbe 
geschlossen, also lang ist. Unter diesen Umstanden ware 
Zirkumflektierung der betonten Suffixalsilbe zu erwarten. 
Wie stimmt dazu ddovs ddovs ? 

Bei den griechischen Formen liegt aueniings eine Unre- 
gelmassigkeit vor, die sich aber leicht erklaren lasst. Das 
Urspriingliche ist daneben jedoch auch noch erhalten: Der 
Nominativ Sing. des aktiven Partizipiums ist im 
Litauischen schleifend betont. Vgl. sukas N. suk@, 
sukés. Hier ist der gesetzmassige Schleifton unverandert 
bewahrt, wahrend im Griechischen der Nominativ die 
Akzentqualitat der obliquen Kasus angenommen hat. 

' Zu beachten ist, dass die arischen van¢-Stamme schon 
in urarischer Zeit den Nominativausgang -v@s von den 
vas-Stammen entlehnt haben, der erst spaiter einen Nasal 
von den obliquen Kasus bezogen hat, vgl. Brugmann 
Grundriss II § 198 S. 536. Ich verwerfe daher mit Brug- 
mann schon aus diesem Grund den von Bartholomae 
KZ. XXIX 449 ff. konstruierten Nominativausgang *-xénts. 


II. Nominativ Dualis. Die vollste Form des Nom. 
Du. endet auf -du, das, wie ich entgegen meiner friihern 


40 Wilhelm Streitbherg. [1893. 


Ansicht jetzt annehme, den Wortton getragen hat. Darauf 
weist, wie Hirt mich belehrt, die Schwundstufigkeit der 
Wurzelsilbe bei alten isolierten Formen wie idg. duéu 
usw. Wir verdanken dem Scharfsinn Meringers die Deu- 
tung dieser Bildung: es ist der Nominativ Sing. eines 
ou-Stammes, der die Parigkeit ausdriickt. Die Erklarung 
ist also dieselbe wie bei den oz-Stammen Anto saékha. 


III. Nominativ Plur. Neutr. Wir haben hier Dop- 
pelformen : : 

a) Ohne Kasussuffix: avest. daman ved. dhama, vielleicht 
lat. guattuor. 

b) Mit Suffix: aind. dhaman-i catvar-i. 

Die Formen der ersten Art sind kollektive Singulare, also 
wie die friiher behandelten Bildungen zu beurteilen. 

Die zweite Kategorie hat Joh. Schmidt ebenso gedeutet, 
indem er aind. z als idg. z gefasst und dem 2z in vart usw. 
gleichgesetzt hat. Ich glaube nicht, dass diese Auffassung 
- haltbar ist, vgl. Brugmann, M.U. V 52 ff. Vielmehr muss 
man ar. 2 = idg. a setzen. Dabei lage die Versuchung 
nahe, die Dehnung des Suffixalvokals daraus zu erklaren, 
dass -2 aus dem feminin-neutralén Suffix -@ gekiirzt sei, das 
Wort also eine More verloren habe. Diese Auffassung ware 
bedenklich. Erstlich ist damit die Vokallange der suffixlosen 
Nom. Plur. N. nicht erklart, die doch kaum von den a-Formen 
losgelést werden konnen. Zweitens fehlt die sonst stets beo- 
bachte Verschiebung der Silbengrenze. Daher muss 
man die suffixlosen Formen als kollektive Singulare fassen. 
Die z-Formen sind durch Anfiigung des Suffixes -a aus ihnen 
weitergebildet. 


IV. Instrumental Sing. Hirt, IF. I 13 ff., hat -m 
als Suffix erwiesen. Vor diesem erscheint gedehnter Vokal : 
aind, prataram usw. War das Suffix urspriinglich -mo, wie 
Hirt ebenfalls schon aus andern Griinden vermutet hat, so 
ist die iiberlieferte Lange des dem m vorausgehnden Vokals 
erklart. 


Vol. xxiv.] Ein Ablautproblem der Ursprache. 4! 


V. Lokativ Sing. Es interessieren hier zwei Bil- 
dungsweisen. . 

a) Mit Dehnung: idg. ogné(i); aind. agna, got. anstai 
= ahd. emsti, abg. pati, abg. Inf. dati —lit. diiti. 


idg. sunéu : aind. siindu, got. sunau = ahd. suniu. 
idg. domén : kret. Sounv. 


b) Ohne Dehnung: idg. domen : ved. kdérman, griech. 
Sowev abg. hamen-e. 

Warum hier Normalstufe, dort Dehnung ? 

Die Frage war um vieles leichter zu beantworten, wenn 
zuvor eine andere gelést ware, die R. Meringer in 
seiner gehaltvollen Rezension von Bloomfields Schrift tiber 
Suffixangleichung (IF. Anz. II 23) folgendermassen formu- 
liert: ‘Man achte darauf, dass bei vielen mehrsilbigen 
t- u- r- n-Stammen der Lokativ und Nominativ ganz gleich 
gebildet gewesen sein diirften. Was war der Grund der 
gleichen Form des Lokativs und des Subjektkasus ?” 

Gleich Meringer konstatier ich die Tatsache der Gleich- 
heit, ohne sie erklaren zu kénnen. Die Tatsache allein 
hilft schon weiter. 

Der suffixlose Lokativ ist allerdings formell nichts anders 
als der Nominativ und —wie ich hinzufiige—der Vokativ. 
Wenn Séunv = rourv, so ist Sdiuev = aind. Suda griech. 
xvov, ahar = whtep. Hierher gehort auch lit. ¢i/te, das ich 
fiir einen regelrechten suffixlosen Lokativ der ¢/o-Stamme 
halte. 

Vergleicht man die Nominative ratip pijteap Laxpdrns 
Saiwwv mit den Vokativen mdtep pirop Lwxpates Saipor, 
so fallt zweierlei ins Auge: 

a) Ein Unterschied im Akzent. 

b) Ein Unterschied in der Quantitat des Suffixvokals. 

Ich bin nun der Ansicht, dass ein Kausalnexus 
zwischen beiden Erscheinungen besteht. Und 
zwar glaub ich, dass die Zuriickziehung des 
Akzents auf die Anfangssilbe die Ursache 
der Kiirze des Suffixvokals im Vokativ ist. 


42 Wilhelm Strettberg. [1893. 


Da beide Bildungen als sogen. Lokative wiederkehren, 
ist auch die im Lokativ bestehnde Doppelheit erklart. 

Die Zuriickziehung des Akzents im Vokativ beruht, wie 
H. Hirt gefunden hat, auf der Enklise. Derselbe Gelehrte 
ist auch in der Lage, enklitische Lokative nach Art des 
lateinischen 7//ico fiirs Indogermanische nachzuweisen. 
Somit ist eine fiir die kurzvokalischen Formen des Voka- 
tivs, der nichts anders als der enklitische Nominativ ist, 
und des kurzvokalischen Lokativs in gleicherweise zutref- 
fende lautgesetzliche Rechtfertigung gegeben. _ 


VI. Akkusativ Sing. a) Mit Dehnung: idg. dizm 
(aind. dyam und griech. Zhv), idg. gom (aind. gam und 
griech. Bov). Die Urformen *adiéuom *qiuom werden 
lautgesetzlich zu den Dehnformen *diéum *qoum. Vor 
m kann aber nach langem Vokal kein unsilbisches zw stehn. 
Es schwindet also nach Meringers Gesetz. Infolgedessen 
tritt nach Michels’ Gesetz Akzentwechsel ein. Damit ist 
die letzte Stufe, der iiberlieferte Formenstand, erreicht. 

Die Probe auf die Rechnung ermdéglicht der Akkusativ 
von idg. za@a#s. Die Urform ist *xayom. Da die Wurzel- 
silbe von Haus aus lang ist, muss durch den Schwund 
des o Akzentwechsel in der vorausgehnden Silbe hervor- 
gerufen werden. Wir erhalten demnach *xd@m. Eine 
solche. Form widerspricht den idg. Lautgesetzen. Auf der 
einen Seite kann: nach Meringers Gesetz # nach langem 
Vokal vor labialem Nasal nicht geduldet werden. Auf 
der andern Seite verliert nach Bezzenberger-Hirt ein 
geschleifter Langdiphthong niemals seinen zweiten Kom- 
ponenten. 

Hieraus folgt, dass m# silbisch werden muss. Wir ge- 
langen also zur iiberlieferten Form idg. ~@-um =. aind. 
navam, griech. vifa, lat. ndvem. Die Probe stimmt dem- 
nach, 

b) Akkusative ohne Dehnung: Sie erscheinen in allen 
iibrigen Fallen. Im Griechischen heisst es eda ora. 
Kurzer Vokal erscheint im Akkusativ ferner bei allen 
en-Stammen, die iiberhaupt noch die Abstufung gewahrt 


Vol. xxiv.] Ein Ablautproblem der Ursprache. 43 


haben. Bei den erv-Stammen haben die Nomina agentis 
mit Endbetonung in allen Kasus, also auch im Akkusa- 
tiv, langen Suffixvokal; die Nomina agentis mit Anfangs- 
betonung haben kurzen Vokal: darwp, aber Sropa. 
Besonders wichtig sind die Verwantschaftsnamen, da sie 
mehr als alle andern die urspriinglichen Abstufungs- 
verhdltnisse bewahrt haben. Sie haben im Akkusativ 
ausnahmslos kurzen Vollstufenvokal, mag dieser nun e 
oder o sein. Die es-Stamme wie der os-Stamm ’Has 
haben stets kurzen Vokal im Akkusativ. In der Ilias 
t 240 ist nach Ausweis des Metrums noch die unkontra- 
hierte Form ’Héa erhalten. — 

Auf indischem Boden erscheint eine Dehnung im 
Akkusativ Sing. nur bei a, niemals bei z und ‘zw. 
Ferner: kurzes a steht im Indischen iiberall da im 
Akkusativ, wo es griechischem e entspricht. Daher ist 
bei den geschlechtigen es-Stammen und den Verwant- 
schaftsnamen auf -¢ev- @ im Akkusativ Sing. unerhGrt. 
Wie die ¢ev-Stamme flektiert auch xar-: aind. ndram = 
griech. avépa. ‘ 

Langes @ haben von den Verwantschaftsnamen nur 
zwei: svdsar- und ndptar-; ausserdem schwankt wSds- 
zwischen wsasam und usdsam. 

Warum heisst es svdsaram (ndptaram) aber pitdéram, 
warum wsadsam aber digirisam ? 

Die Antwort drangt sich mit zwingender Gewalt auf: 
weil svdsar- und ndptar- die einzigen Verwantschafts- 
namen sind, die nicht den Suffixvokal e, sondern den 
Suffixvokal o haben, wahrend wds- der einzige os-Stamm 
ist. Die Nebenform w3ésam vérdankt ihr kurzes a dem 
Einfluss der es-Stamme. 

Folglich ist Brugmanns bekanntes und viel- 
bestrittnes Gesetz von der Vertretung des mit 
e ablautenden idg. o durch arisch @ in offner 
Silbe erwiesen. 

Ich selbst gestehe, dass mich dies Ergebnis iiberrascht 
hat, denn ich habe nicht zu den Anhangern der Brug- 
mannschen Theorie gehért. Die eben dargelegten That- 


44 Wilhelm Streitberg. [1893. 


sachen haben mich jedoch eines bessern belehrt. . Wer 
nicht die vollendete Ubereinstimmung zwischen Griechisch 
und Indisch fir blossen Zufall erklaren und dadurch 
planloser Willkiir Tiir und Tor 6ffnen will, der muss 
sich, davon bin ich iiberzeugt, daran gewohnen mit Brug- 
manns Gesetz zu rechnen. 

Das Ergebnis der vorausgegangnen Erérterungen ist 
dies: weder das Griechische noch das Arische kennen 
Dehnung des Suffixvokals im Akkusativ Sing. Woher 
kommt das? Warum heisst es idg. patérm usw. aber 
diém gom ? 7 

Die Antwort ist die. 

Bei den Urformen *diéuom und *qiuom muss durch den 
Schwund des Endsilben-o regelrechter Weise eine Ver- 
minderung der Silbenzahl eintreten. Das  urspriinglich 
zweisilbige Wort wird einsilbig. 

Anders bei den andern. : 

Urformen wie *pidom —*pédom *udgom *patérom *poimé- 
nom *dusésom miissen gesetzmissig ebenfalls ihr Endungs- 
o verlieren, genau wie *diéuom *qduom. Aber durch diesen 
Verlust werden sie nicht wie jene einsilbig. Denn der 
Nasal muss nach Lauten die schallarmer sind als er sil- 
bisch werden. Folglich bleibt die Silben- und Morenzahl 
der angefiihrten Worte auch nach dem Schwund des 0 
unverdndert. Es heisst also idg. pédm— pédm, udgm, 
patérm poiménm, ausdsm, wie iberliefert ist. Folglich 
kann eine Dehnung der betonten kurzen Suffixvokalé 
nicht eintreten, da die Grundbedingung dafiir nicht 
vorhanden ist. 


VII. Zwei scheinbare Ausnahmen sind die Genitive 
der ez- und eu-Stamme,. idg. ognois und sénoiis. Hirt 
sieht in ihnen bekanntlich ebenso wie in efuas das Geni- 
tivsuffix -es -os, setzt also Urformen wie *ognotes *siinoues 
an. Ware diese Annahme richtig, so hatten wir eine 
Ausnahme des Dehnungsgesetzes anzuerkennen. Aber 
der Zirkumflex erklart sich einfacher. 

_Neben dem Genitivsuffix -szo steht -so, vgl. abg. ce-so 


Vol. xxiv.] Ein Ablautproblem der Ursprache. 45 


usw. Dieses Suffix -so hat urspriinglich nicht nur bei 
den ¢/o-Stammen existiert, sondern bei allen vokalischen 
Stammen, wahrend -es -os das ‘Suffix’ des Genitivs konso- 
nantischer Stamme ist. 

Unter dieser Voraussetzung kann man ekuas ognois 
sinots auf die Urformen *ekud-so *ognéi-so *siindu-so 
zuriickfiihren. Der Schwund des auslautenden o rief nach 
Michels’ Gesetz Akzentwechsel hervor. 


VIII. Zum Schluss dieses Abschnitts noch ein Wort 
iiber die Dehnung in der sekundaren Nominalbil- 
dung. Am haufigsten erscheint sie auf indischem Sprach- 
gebiet, doch fehlt es auch auf europdischem Boden nicht 
an Belegen. 

Von vornherein ist klar, dass in Fallen wie saptam: 
saptd, sahasram: sahdsram u.a. die Dehnung nicht auf 
Silbenverlust beruhn kann. Dies sowie der Umstand, dass 
mit dem Auftreten der Dehnstufe zugleich eine Bedeu- 
tungsveranderung verkniipft ist, scheidet die Falle dieser 
Art scharf von den Belegen fiir rein mechanische Dehnung, 
wie sie bisher behandelt worden sind. 

Dennoch ist es wahrscheinlich, dass ein Zusammenhang 
besteht. Joh. Schmidt Pluralbildungen S. 145 Fussnote 
hat auf das Nebeneinander von vac- und vacas-, nabh- und 
nabhas- aufmerksam gemacht. Ich bin der Ansicht, dass 
diese feminin-neutralen Einsilbler mit ihrer lautgesetzlichen 
Dehnstufe den Ausgangspunkt fiir die Entstehung der 
Sekundardehnung abgegeben haben. Denn es war ihnen 
von Haus aus Kollektivbedeutung eigen. Demnach 
ist nabh- eher ‘Gewolk,’ zabhas- aber die einzelne ‘ Wolke,’ 
vac- die ‘Rede,’ vdcas- dagegen das einzelne ‘ Wort.’ 
Gleicherweise entsprechen sich muor N. und mart, gruose 
F. : gras, buost : bast, snuor : veupa, var: ver F., géns 
‘Frauenzimmer, im Sinn des Altern Nhd.’ : gind ‘Weib’ 
u. dgl. m. 

Wenn Dehnung und Kollektivbedeutung zusammenfielen 
und daher als zusammengehdérig betrachtet wurden, so war 
damit die Méglichkeit gegeben, neue Kollektivbildungen 


46 Wilhelm Strettberg. [1893. 


durch Vokaldehnung zu schaffen. So denk ich mir saptam 
und alle jene Bildungen entstanden, wo lautgesetzliche 
Erklarung der Dehnung ausgeschlossen ist. Sie sind 
nichts anders als Nachbildungen der alten lautgesetzlichen 
Muster. 


B. Das VERBUM. 


Minder reich als das Nomen ist das Verbum an Dehn- 
stufenbildungen. Zwar fiihrt Bechtel eine Reihe_ von 
Kategorien an, wo Dehnung stattgefunden haben soll, aber 
die Mehrzahl halt der Priifung nicht Stich. 

Zuerst das Kausativ. Hier ist Brugmann unbestreitbar 
im Recht, wenn es die Existenz langer Wurzelvokale fiir 
die europdischen Kausativa durchaus leugnet. Denn die 
karglichen Beispiele, die man fiir europiaische Dehnformen 
anzufiihren pflegt, sind ohne Beweiskraft. Wie abg. 
chvalitt von chvala abgeleitet ist, so das angebliche Kausa- 
tiv plaviti.von plavi. Lat. sopire fallt seiner Flexion nach 
aus dem Rahmen der Kausativa heraus und bei w@déopat 
stimmt es mit der Bedeutung nicht. 

So bleibt fiir das in offner Silbe erscheinende @ der 
arischen Kausativa nur die Brugmannsche Erklarung iibrig. 
Sie ergibt sich mit um so zwingenderer Notwendigkeit, 
als von einem Morenersatz keine Rede sein kann. 

Nicht besser ists um die 3. Person Sing. Perf. 
Akt. bestellt. Auch hier erscheint, von Jdabhiiva ab- 
gesehn, Dehnung nur bei a@ in offner Silbe, ohne dass 
Parallelen auf europdischen Boden zu _ finden © waren. 
Denn yéywve und 076 beweisen nicht, was sie sollen. 
Ist doch 076 aus *BeBdwe die regelrechte ungedehnte 
Vollstufenform des Perfekts zum langvokalischen Prasens 
Inf. dia béa. 

Auch beim Perfekt besteht so wenig wie beim Kausativ 
die Méglichkeit, mit dem Prinzip des Morenersatzes zu 
operieren. Auf der andern Seite dagegen gibt de Saus- 
sures Theorie iiber den Perfektablaut in Verbindung mit 
Brugmanns Gesetz eine glatte Erklarung der in der 1. und 


Vol. xxiv.] Ein Ablautproblem der Ursprache. 47 


3. Singularperson des aktiven Perfekts auftretenden Lauter- 
scheinungen. 

Endlich der Aorist. Bechtel nimmt hier Dehnung des 
Wurzelvokals im unthematischen Aorist an. An sich 
ware diese Annahme mit dem Prinzip des Morenersatzes 
sehr wol zu vereinigen, wenn man den unthematischen 
Formen 4ltere thematische zu Grunde legte. Aber der 
Tatbestand rechtfertigt Bechtels Auffassung nicht. 

Was die indischen Formen drat dvadt yat anlangt, so 
hat Bartholomae sie, wie ich glaube mit guten Griinden, 
dem s-Aorist zugewiesen, vgl. IF. III 1 ff. 

Von europaischen Formen sind nur die litauischen 
durchsichtig genug, um zur Entscheidung herangezogen 
werden zu kénnen. Die litauischen e-Priterita zerfallen 
in zwei Klassen: 

a) mit schleifend betontem é: bére; 

b) mit getossenem ¢: hée. 

Die erste Gruppe scheidet aus, da ihr Zirkumflex die 
Annahme der Dehnstufe verbietet, vgl. Bartholomae aO. 

Fiir die zweite Kategorie hat Bartholomae aO. den 
s-Aorist als Ausgangspunkt zu bestimmen versucht. 

Eine weitere Dehnstufenbildung sieht Bechtel im ari- 
schen Passivaorist, vgl. aind. ava@cz = avest. avaci. Hier 
aber findet sich von Langen nur a in offner Silbe; das 
Verhaltnis von Linge und Kiirze ist also das gleiche wie 
beim Kausativ und bei der 3. Sing. Perf. Akt. Bartholo- 
maes Versuch (IF. III 5) die Doppelheit dt@pi und ddar%i 
durch Ansetzung schleifender Betonung zu erklaren, ist 
abzuweisen : erstlich wissen wir gar nicht, dass diese in den 
genannten Formen bestanden hat; zweitens wirkt sie im 
Indischen iiberhaupt nicht kiirzend in der von Bartholo- 
mae niaher dargelegten Weise, wie das unstreitig schleifend 
betonte ams dartut. 

Da der indische Passivaorist nur in der dritten Person 
Sing. auftritt, da er ausserdem im europdischen Verbum 
seines Gleichen nicht hat, so ist vielleicht eine ganz 
abweichende Deutung berechtigt. Sie hat Prof. Osthoff 
versucht; ich verdanke sie seiner miindlichen Mitteilung. 


48 . Wilhelm Streitberg. 5 [1893. 


Er geht: von dem Vokalismus der Form aus. Dieser ist, 
wie gesagt, der gleiche wie beim Kausativ und der 3. 
Sing. Perf., dh. bei Bildungen, die idg. 0, das mit e im 
Ablaut steht, aufweisen. Deshalb. nimmt Osthoff auch 
hier o-Stufe an. Dann aber ist die Form keine Aoristform 
mehr. Uberhaupt keine Verbalform, sondern ein Nomen, 
das dem Verbalsystem eingegliedert worden ist. Ein 
Verbalnomen auf idg. -2z, wie es deren im Griechischen 
gibt. Dann erklart sich auch die passive Bedeutung, die 
nicht indogermanisch sein kann, da das Indogermanische 
kein Passiv besessen hat; es erklart sich die eigentiim- 
liche dussere Form, die des fiir die 3. Sing. charakte- 
ristischen ¢ entbehrt; es erklart sich schliesslich die 
Beschrankung auf die 3. Person Singularis. 

So bleiben allein die Nerschicdnen Kategorien des 
s-Aoristes iibrig. 

Wenn man vom si$-Aorist absieht, der nach allgemeiner 
Ansicht ein Kontaminationsprodukt ist, so bleiben im Indi- 
schen der s- und der 2z5-Aorist als alte Bildungen bestehn. 
Beide erscheinen auch auf europdischem Boden: aind. 
dvakiam = lat. véxt = abg. vésit; dvédisam = nbdea (wobei 
der Suffixablaut -as-:-es- zu beachten ist). 

Der indische s-Aorist hat in den starksten Formen 
Vrddhi, der z3-Aorist schwankt zwischen Guna und Vrddhi; 
Guna iiberwiegt. 

Woher die Dehnung beim s-Aorist? Die Antwort kann 
sich der Leser nach dem vorausgegangnen selber geben. 

Das indogermanische Aoristsuffix hat drei Ablautstufen: 
die Vollstufe idg. -es- und die beiden Schwundstufen idg. 
-as- und -s-. In der letzten ist der Vokal véllig verloren 
gegangen, das Wort also um eine Silbe armer geworden. 
Das Prinzip des Morenersatzes muss in Wirksamkeit treten. 

Wie die einzelnen Suffixstufen urspriinglich verteilt ge- 
wesen sind, lasst sich nicht mehr kontrolieren. Nur soviel 
steht fest, dass der Indikativ Sing. Akt. gedehnte Voll- 
stufe besessen hat. Er muss also den Wortakzent auf 
der Wurzelsilbe getragen haben; bei ihm muss die Null- 
stufe des Suffixes, -s-, urspriinglich zu Hause gewesen sein. 


Vol. xxiv.] Ein Ablautproblem der Ursprache. 49 


So erklaren sich unmittelbar alle Dehnformen offner, d.h. 
kurzer, Wurzelsilben, wie aind. dvakiam dnaisam dSrausam. 
Bei den Verben mit geschlossner, also langer Wurzelsilbe, 
wo Dehnung nicht méglich ist, darf man unbedenklich die 
Vrddhierung als Analogiebildung auffassen, z. B. araikiam 
nach dudisam, drautsam nach dsrausam. Auf europaischem 
Sprachgebiet kann man wegen der Kiirzungsgesetze die 
Lange nur dann nachweisen, wenn die Wurzel auf ¢ + 
Verschlusslaut ausgeht. 

Ist die gegebne Erklarung der Vrddhierung im s-Aorist 
richtig, so miissen die beim zS-Aorist bestehnden Verhilt- 
nisse die Probe darauf bilden. Das Schwanken zwischen 
Vrddhi und Guna, je nach der Form des Wurzelauslauts, 
ware kaum geeignet ein sicheres Resultat zu liefern, wenn 
es nicht einen festen Punkt gabe: die Behandlung eines 
a vor Verschlusslaut. Es ist urspriinglich unver- 
langert. Es stehn sich also gegeniiber anditam und 
anayisam. Dieses Verhaltnis tragt den Stempel der Alter- 
tiimlichkeit: dort wo ein Silbenverlust nicht stattgefunden 
hat, bleibt die Kiirze des a erhalten; dort wo die Silben- 
zahl verringert ist, wird @ gedehnt. — 

Ich hoffe, der Beweis ist erbracht, dass sich alle dehn- 
stufigen Kategorien durch das Prinzip des Morenersatzes 
erklaren. Ich iiberlasse dem Leser, weitre Konsequenzen 
hieraus zu ziehn. Nur auf einen Punkt sei mir zum 
Schluss noch hinzuweisen gestattet. Bewahrt sich das 
Dehnungsgesetz, so ist damit die Modglichkeit gegeben 
scharfer als bisher zwischen einsilbigen und zweisilbigen 
sogen. Wurzeln zu scheiden. Verba wie idg. és¢z konnen 
nicht von einer zweisilbigen ‘Wurzel’ ese- gebildet sein, 
wahrend umgekehrt bei Substantiven wie idg. wédgs die 
Einsilbigkeit des ‘Wurzelstamms’ erst die Folge einer 
Reduktion ist. Man sieht also, konsonantische und vokali- 
sche Flexion sind nach wie vor zu scheiden, wenn auch 
das Gebiet der ersten vielfach zu Gunsten der zweiten 
eingeengt werden muss. 


50 Hermann Osthoff. [1893. 


IV. — Dunkles und ' helles \ 1m latetnischen. 


By Pror. HERMANN OSTHOFF, 


HEIDELBERG. 


NacupeEm die wirkungen, welche der gemein- und urindo- 
germanische vocalablaut im historischen vocalismus der eiii- 
zelsprachen hinterlassen hat, in den letzten jahren immer 
genauer ermittelt worden sind, ist es zeitgemasse aufgabe 
der sprachwissenschaft, mehr und entschiedener noch als 
friiher das augenmerk auf diejenigen vocalwandelungen zu 
richten, die ihren entstehungsgrund in lautvorgangen des 
einzelsprachlichen lebens selbst gehabt haben. Eine solche 
erscheinung ist das lautgeschichtliche problem der latei- 
nischen sprache, welches, den gegenstand meines vortrages 
vor dem internationalen sprachforschercongress bildend, hier 
in kiirze zur darstellung kommt.! 

Im lateinischen ist die lautverbindung e/ von einer 
modification ergriffen worden, die mit erscheinungen, fiir wel- 
che die germanische grammatik den terminus “ brechung” 
hat, vergleichbar ist: urspriingliches e/ wurde in 
weiterem umfange lateinisch zu o/ (uw) gebrochen, 
so dass hier teilweise ein zusammenfall der drei alten ablaut- 
stufen indog. e/ und indog. of = lat. of (ul), sowie indog. 
Z > lat. of (ul) sich ergeben hat. Die tatsache an und fiir 
sich ist von seiten der heutigen sprachforscher nicht vdllig 
unbeachtet geblieben; doch hat man einerseits noch nicht 
gesehen, innerhalb welcher bestimmten grenzen sie sich 
halt, andererseits sind, soweit man sich um regel und 


1 Es sei bemerkt, dass die sache, da sie als einzelnes glied in eine zusammen- 
hangende kette von mir angestellter laut- und ablautgeschichtlicher untersuchungen 
sich einreiht, an anderem orte abermals und mehr in extenso mich beschaftigen 
wird, 


Vol. xxiv.] Dunkles und helles \ im lateinischen. 51 


gesetzmassigkeit bemiiht hat, nur unzulangliche aufstellungen 
erzielt worden. 
Wir versuchen, folgende regeln zur anerkennung zu bringen: 


I. Urspriingliches e/ erlitt die brechung zu lat. 
of (ul), wenn unmittelbar darauf einer der dunklen 
vocale urlat. @, 0, « folgte; jedoch e/ blieb unverandert 
vor vocalen, wenn diese @ oder Zlaute waren. Es _ ist 
iiberall hier und im folgenden darauf zu achten, was die 
in der urlateinischen periode bestehende lautliche gel- 
tung der hinter der liquida stehenden vocale war; von der 
historisch vorliegenden form, wie sie vornemlich durch die 
vocalschwachungsgesetze entwickelt war, muss abgesehen 
werden. Beispiele sind fiir . 

A. of (ul) aus el vor ataut: 

oliva, olivum, entlehnt aus édai(¢)a, érar(e)ov und auf 
diese griechischen quellformen zuriickgehend vermittelst 
der zwischenstufen urlat. *ol/aiva, *olatvom. Die ent- 
lehnung ist anerkannt; naheres iiber die zeitliche datierung 
derselben weiter unten. Die nebenformen olea, oleum 
diirften wol durch erneuten einfluss von seiten der gr. 
éXaid, €\acov hervorgerufen sein, nachdem nemlich letztere 
in jiingerer zeit diese ihre digammaverlustige lautgestalt 
erlangt hatten und indem zugleich eine suffixangleichung 
an echtlateinische baum- oder pflanzennamen und mit sol- 
chen in verbindung stehende productbezeichnungen, wie 
laurea, picea, pinea, vinea und pineum, vindceum, linteum, 
mitwirkend im spiele war. 

Volaterrae aus etrusk. Velasrt. 

‘gula aus *gela, vgl. ahd. anfrk. ké/a, ags. ceole ‘kehle.’ 

B. of aus e/ vor o-laut: 

olor m. ‘schwan,’ gen. o/ér-is: gr. éA@p-to-s ein wasser- 
vogel, vgl. auch mir. e/a, corn. elerch ‘schwan.’ 

holus (olus) n. ‘griinkraut, gemiise’ = klruss. zelo, poln. 
ziolo n. ‘kraut.’ Daneben alat. helus, helusa Paul. Fest. 
und helitérés in glossen fiir olitorés; die erklarung des 
doppelvocalismus ergibt sich auf grund der anzunehmenden 
alten flexion holus, gen. *heler-ts. 


52 Hermann Osthoff. [1893. 


molo ‘ich mahle’ < *melé = air. melim. Hier war der 
vorauszusetzende durch ausgleichung beseitigte dltere flexi- 
onszustand molo, molimus, molunt, conj. molam u. s. w. 
und daneben *melis, *melit, *melitis, imper. *mele, *melito, 
part. *melent-. Das got. asachs. ahd. malan, aisl. mala 
diirfte nur scheinbar fiir ein grundsprachliches praesens 
mit o-vocalismus *mo/é sprechen, es mag got. mada speciell 
germanische umbildung eines alten jodpraesens *malja = 
lit. maliz sein. An die von Bartholomae, Brugmann u. a. 
vorgeschlagene zuriickfiihrung des lat. mo/o auf ein aorist- 
praesens indog. *mllo ist gleichfalls nicht zu denken. Denn 
im italischen ist nicht o/, ov, nach dem was man gewohnlich 
bis jetzt annimmt, sondern a@/, a7, wie ja auch im keltischen, 
die vertretung der vorsonantischen tiefstufenformen indog. 
1, rr; vgl. lat. palea : lit. pelai, abulg. pléva < *pelva ‘spreu,’ 
lat. salix = air. sail ‘weide’ ; gr. édikn, lat. caro ‘fleisch’ 
umbr. karu ‘pars’ ; gr. xeipw xépya ahd. scéran, lat. 
paréns : lit. periu ‘briite,’ varix ‘krampfader’ : ahd. wérna 
weérra ‘varix,’ lat. varu-s = lit. wira-s ‘finne’ : lat. vermt-s 
‘wurm’ und 4dhnliches von mir an anderem orte beizu- 
bringendes. 

Bei volo ‘ich will,’ volunt 3. plur. lasst sich fiir die 
samtlichen verbumsformen mit der durch den optativ velim 
zunachst in den géesichtskreis geriickten wurzelablautstufe 
wel- auskommen ; inwiefern auch fiir solche wie volt (vult) 
3. sing., dariiber naheres weiter unten. 

C. of aus ef vor wlaut: 

Hierher gehért co/uber ‘schlange,’ wenn es nach Havet 
und Keller als alte entlehnung auf gr. yéAuvdpos beruht. 

Dass volvo ‘ich walze rolle,’ wofiir alter mit diaeresis 
dreisilbiges volud, nebst dem zubehor volitus, volimen, 
volucra, tn-volicrum auch im wurzelvocalismus die gleiche 
basis *we/u- wie die griechische verwandtschaft éAvtpoy, 
éXvyos haben mag, ist unbezweifelbar. Morphologisch 
etwas ferner diirfte das gr. eiAiw, von dem eidvpévos, 
elAdwa ihrerseits beeinflusst sind, liegen: eAdw wol mit 
einem nasalsuffix aus *eAdviw. . 

D. of aus e/ vor urspriinglichem a (indog. “schwa”): 


Vol. xxiv.] Dunkles und helles \ im lateinischen. 53 


Dies vermutlich zunachst in volumus 1. plur., so dass 
ein *welamos hinter ihm zu suchen ware, gleichwie vo/o 
I. sing. von *we/é ausging (s. 0.). 

Von lat. columen, columna ist wahrscheinlich, dass sie 
die gleiche wurzelstufe mit ce/sws und lit. Aeliz ‘ich hebe’ 
enthalten, aus dem grunde nemlich, weil columen, columna 
in der bildung sich mit fegumen tegumentum und mit gr. 
TEAaULwV, Tepauwv a-Tépauvos vergleichen lassen. 


II. ef blieb erhalten vor e- und 7-lauten: 

A. vor folgendem e: 

Lat. celeber < *celes-ri-s; vgl\. fiunebri-s, fénebri-s. Ur- 
spriingliche bedeutung war ‘ gainge, gangbar, viel begangen.’ 
Das zu grunde liegende neutrale nomen *ce/-es- ‘begehung’ 
gehorte wurzelhaft zu gr. xéA-evfo-s und lit. ké/za-s kelj-s 
‘weg, strasse.’ 

Betreffs ce/er ‘schnell’ zur seite des gr. «éA-ns m. ‘renner,’ 
‘jachtschiff’ mag hinsichtlich der stammbildung an grie- 
chische adjectiva wie @anepd-s, parepo-s sich ankniipfen 
lassen, so jedoch, dass das verhialtnis des lat. cel]-er-t-s ZU 
dem gr. -epo-s als ein solches der weiterbildenden suffixalen 
ableitung betrachtet wird, ahnlich wie bei szmz/-z-s : ouanro-s, 
humil-t-s : y@apanro-s, agtl-i-s : aind. ajird-s u. dgl. mehr. — 
Fiir ce/ox m. f. ‘schnell segelndes schiff, jacht’ sollte man 
*colox erwarten; es ist wol, wie auch andere annehmen, 
zufolge von entlehnung das gr. «éAns, mit suffixanbildung 
jedoch an lat. vélox, navis velox Vergil. 

In sceler-is gen., sceler-a plur., sceles-tu-s, sceler-are ist 
scel- lautgesetzlich; die einzige -o/-form *sco/us nom.-acc. 
sing. stand zu isoliert gegeniiber den vielen mit sceles-, 
sceler- da, als dass sie gegen die ersetzung durch analo- 
gisch entwickeltes sce/us hatte widerstandskraftig sein k6n- 
nen. Giinstigerer existenzbedingungen erfreute sich holus 
(s. 0.), schon allein wegen des mangels so haufig gebrauchter 
ableitungen mit -es-, -erv-, wie dort sceles-tu-s, sceler-dtu-s, 
sceler-dsu-s. 

Die wunderliche alte erklarung von elementum aus den 
liquiden 4, m, m hatten neuere etymologen, Havet und 


54 Hermann Osthoff. [1893. 


O. Keller, nicht aufwarmen sollen. Am besten ist, was 
iiber den ursprung des wortes Leo Meyer mit heranzie- 
bung von aind. auu-sh adj. ‘fein, diinn, sehr klein,’ duzman- 
n. ‘das kleinste stiick’ gelehrt hat. Das zweite -e- von 
elementum hat man auf gleiche linie mit dem zwischen- 
vocal von gr. jye-pov, xnde-pov, aveyo-s, yevé-Twp ZU 
stellen; hatte an seiner stelle ein dem griech. -a- von 
Te\a-wV, Tepa-wwv, dem aind. -2- von jani-man- n. genau 
entsprechender vocal gestanden, so ware *olumentum ent- 
sprungen, wie columen aus *celamen (s. 0.). 

B. vor folgendem 2: 

Das frappanteste beispiel ist der optativ der wurzel we/- 
‘wollen’ : lat. velim, velis, velit, velimus u. s. w. neben 
indic. volo (s. 0. Ss. 52), volumus (s. 53). Als sonstige 
zeugnisse diirften hier noch in betracht kommen 

Lat. felzx f. ‘farnkrant’ und seine ableitungen /e/ic-ula, 
-atus, -onés,; in vermutlich mundartlicher lautvariation dane- 
ben flix. Verwandtschaft mit fol-ia-m ‘blatt’ ist wol 
nicht zu verkennen; die ableitung -zr wie in den pflanzen- 
benennungen J/arix, salix : gr. édix-n (s. 0.). 

Velinus, name eines sees im sabinischen und adjectiv zu - 
dem stadtnamen Ve/za (s. u.). 

Velitrae ; vgl. das morphologisch abweichende volsk. Veles- 
. trom ‘Veliternorum.’ 

Als ausnahmen, die unserem lautgesetze sich nicht zu 
fiigen scheinen, sind noch velut und gelw hier zu beriick- 
sichtigen. | 

Ist vel-ut, vel-utt aus verhaltnismassig spater zusammen- 
riickung seiner beiden einzelbestandteile hervorgegangen, so 
mag die lautgestalt des ve/ in vocalischer hinsicht unab- 
hangig von dem nachfolgenden w- in wz¢(z) geblieben sein, 
indem vorher das wirken eines nichtpalatalen vocals auf 
den vorausgehenden nexus -e/- abgeschlossen war. , 

In der sippe gelu, gelidus, gelare kann wenigstens dem 
adjectiv auf -zdus lautgesetzmassig die e/-form zugekommen 
sein; das gleichgebildete umbr. kaleruf calersu ‘callidos, 
AevKopeTwTrovs’ hatte auch von hause aus palatalvocalischen 
anlaut seines dem lat. -zdus entsprechenden ableitungsbe- 


Vol. xxiv.] Dunkles und helles \ im lateinischen. 55 


standteils. Doch kommt auch in betracht, dass ge/u selbst 
oder ge/u-s masc. als mit dem suffixe -ew- geformtes nomen 
von hause aus in den obliquen casus teilweise die e-hoch- 
stufige stammform hervortreten zu lassen hatte, den loc. 
sing. aber mit dehnstufigkeit als *ge/ew > geli dat. 

Fiir chronologische datierung des lautgesetzlichen 
wandels von e/ zu of vor nichtpalatalen vocalen bietet sich 
zunachst der anhaltspunkt dar, dass o/iva, olivum zeigen: 
die entwickelung des o/ hatte statt, bevor an die stelle von 
-ai- in nicht erster wortsilbe sich durch vocalschwachung 
-z- geschoben hatte. Darnach k6énnte sich vermuten lassen, 
dass zur zeit des wirkens der brechungsregel iiberhaupt 
noch in weiterem umfange der ungeschwachte vocalismus 
der nach der 4lteren wortanfangsbetonung nicht haupt- 
tonigen silben bestanden habe. Und hieraus wiirde weiter 
zu folgern sein, dass dann auch in diesen nicht haupt- 
tonigen silben ein e/ also von dem einfluss der nachfol- 
genden vocale je nach der lautqualitat dieser afficiert wurde 
oder nicht. Z. b. ein *-cel6 = ahd. ilu, air. celim ‘ich 
verhehle’ koénnte mit od- zusammengesetzt eben nach 
unserer brechungsregel in *oc-colé, daraus oc-culo, iiber- 
gegangen sein; als die streng lautgesetzliche flexionsweise 
ware oc-culo, oc-culunt, oc-culam, aber *oc-ctlis, *oc-cilit 
u. S. W. vorauszusetzen. 

Bestatigung findet das vorstehende durch die vocalisation 
urspriinglicher e/-formen wie Szculus = XiKedos, scopulus = 
oxotrexos, famulus = osk. famel, catulus = umbr. katel, 
porculus : lit. parszel-t-s ‘ferkel’; nebula = gr. vedédn. 
Diese bilden mit Szczlia, familia, porcilia im grunde den- 
selben lautlichen contrast, wie volo, volunt mit velim, velts. 
In Sicilia, als entlehnt aus Sixedia, stand von hause aus 
sonantisches -z- hinter dem -/+; aber -y- = consonans -2- 
hatte in familia = osk. famelo fiir *famelyo u. dg). natiirlich 
dieselbe wirkung auf den vor der liquida stehenden vocal, 
woriiber naheres weiter unten. 

Fiir die nachtonigen silben diirfte aber doch eine weitere 
fassung der regel, als die, dass -e/- vor den nichtpalatalen 
vocalen in -o/- > -ul- iibergegangen sei, sich empfehlen ; 


56 Hermann Osthoff- [1893. 


denn nach den lehnwortern wie jpessulus = mdacanos, 
crapula = xpaitradn, scutula = oxvtadn, spatula = oratary 
erweist sich ja -#/- hier auch als das substitut eines ur- 
spriinglichen -a/. 

An der hand der formen der.dialekte osk. famel, umbr. 
katel, tigel ‘dicatio’ u. ahnl. ersehen wir auch, dass die 
vocalbrechende wirkung eines von nichtpalatalem vocale 
gefolgten 7 vollends nicht uritalisch, sondern eine im 
speciellen sprachleben des lateins aufgekommene erschei- 
nung war. ; 

Alle diese datierungsversuche sind aber offenbar nur 
bestimmungen nach relativer sprachchronologie. Durch 
oliva, olivum sind wir jedoch auch in die lage versetzt, 
mit einer annahernden jahreszahlangabe den zeitpunkt zu 
treffen, nach welchem die verwandlung von e/ in of unter 
den erwahnten bedingungen sich zugetragen haben muss. 
Der 6lbaum ist zur zeit der Tarquinischen k6nige von 
Grossgriechenland zu den Rémern verpflanzt worden, nach 
dem chronisten Fenestella bei Plinius nat. hist. XV § I 
bis zum jahre 173 der stadt unter Tarquinius Priscus 
den Lateinern fremd geblieben; vgl. O. Weise d. griech. 
worter im lat. 132 f. Die entlehnung von *elazvd, *elatvom 
fallt also in diese zeit, die phonetische umgestaltung der 
wortformen in *o/azva@, *olatvom mithin noch spater. 

Es erhebt sich die frage nach der phonetischen 
auffassung des lautprocesses, dass e/ in der stellung 
vor den a-, o- und wz-lauten in lat. o/ (u/) iiberging. Die 
antwort liegt nahe, dass in solcher stellung die liquida 
den dunkleren klang als 7 hatte, vor den palatalen 
oder hellen vocalen @ und z dagegen helles /’ gesprochen 
wurde. Das fiihrt auf ein durchaus analoges verhialtnis 
der verteilung der beiden klangfarben des f, wie es 
bekanntlich auf baltischem und mebhreren orts auf 
slavischem sprachboden ganz regelmassig herrscht, wo 
“je nach der beschaffenheit des folgenden vokales” sich 
“hartes” und “weiches” / gegenseitig ablésen, jenes vor 
den sogenannten “dunklen” dieses vor den “hellen”’ 
vokalen seinen platz hat. Im litauischen z. b. gilt 7 vor 


Vol. xxiv.] Dunkles und helles \ im lateinischen. 57 


a, 0, u, u, daneben “2” d. i. 2’ vor e und 7; entsprechendes 
im lettischen, ferner in dem russischen, polnischen und 
einigen anderen slavischen sprachen. gl. Schleicher lit. 
gramm. § 10, 2 s. I9 f., Kurschat gramm. d. litt. spr. 
§§ 80 ff. s. 26 f., Bielenstein lett. spr. § 47 I 87 f, 
Miklosich vergleich. lautl. d. slav. spr.? (vergleich. gramm. 
I) s. 475. 

Die weitere fiir das latein insbesondere sich erhebende 
frage, welche der beiden klangqualitaten die altere gewesen 
sei, ob /! sich vor a, 6, # in 7 verdunkelt oder anfangliches 
Z vor é und Z zu /' verdiinnt worden sei, wird man, glaube 
ich, zu gunsten der prioritat des 7 zu entscheiden 
haben. FEinmal erklart sich bei dieser annahme am ein- 
fachsten die beteiligung des @ an den das o/ erzeugenden 
lagen, wenn das a wirklich nach der iiblichen auffassung 
als die neutrale mitte der vocallinie, gleich weit von der 
palatalen mundstellung des z wie von dem anderen extrem, 
dem labialismus des w, abliegend, geltend darf. Sodann 
aber kommt dieser annahme stiitzend zur hilfe, wenn sich 
zeigen lasst, dass auch vor consonanten, und zwar vor 
den verschiedenartigsten und zum teil gegen die labiale 
articulation ganz indifferenten, der wandel von e/ zu ol, 
ul tatsachlich auftritt. : 

Es ist e/, beziehungsweise in nicht erster wortsilbe — 
zufolge mundartlicher farbung vereinzelt auch in der an- 
fangssilbe — durch 2z/ vertreten, in vorconsonantischer stel- 
lung nur unter zwei umstanden sicher lautgesetzlich 
erhalten: einerseits, wenn der nexus -/y-, sodann, wenn 
die geminata -/+ auf das e folgte. 


III A. e erhalten vor -dy-: 

Beispiele dafiir sind zunachst nur melior melius compar., 
Velia (s. oben s. 54); dazu kommend jedoch mit der 
erwahnten lautmodification mzlzum ‘hirsen’ und fila 
‘linde,’ gemass ihrer vergleichung mit gr. pedrivn, mredéa 
‘ulme, riister.’ Hier finden aber auch ihren platz jene 
familia : famulus, porcilia : porculus (s. 55). Wenn in 
melior, familia u. dergl. fiir das -y- zuvor silbisches -z- 


58 Hermann Osthoff. [1893. 


eingetreten war, wie es in Sicilia = Xwxedi@ wol immer 
bestand, wiirde es fiir die beispiele dieser art gar keiner 
besonderen regel iiber ¢ vor -/y- bediirfen. 

B. e erhalten vor -d-: 

Ein -// war im lateinischen durch die mannigfaltigsten 
assimilationsprocesse entsprungen ; ihm mag ja bekanntlich 
-ln-, -ls- und wol auch -/d@-, andererseits -d/-, -n/-, -rl- zu 
grunde liegen. Welcher herkunft aber auch -// in jedem 
einzelnen falle sein mag, immer bewahrt ein davor ste- 
hendes ¢ seine lautnatur unverandert. Als beispiele dienen, 
unter beschrankung auf dasjenige e//-, in welchem sicher 
ein indog. e/- enthalten war: cedla, fell- ‘galle,’ me/l- ‘honig,’ 
pellis, telliis, vellus und die praesentia pello und vello; 
insbesondere aber ve//e inf. und vellem conj. imperf. 

Im einklange damit steht es, dass auch -a//-, wenn es der 
vocalschwachung unterliegt, den weg der e-farbung ein- 
schlagt in fefellt perf., re-fello.comp. zu fallo. Die sonder- 
stellung eines -a/ + -+, welche darin liegt, tritt hervor bei 
der vergleichung der verdumpfenden behandlung eines -a/,, 
dem anderweitige consonanz folgt, in con-culcare, in-sultare, 
in-sulsus u. dergl. 

In der geminata -//- herrscht die “‘diinne” aussprache 
auch auf einigen der romanischen sprachgebiete, im spani- 
schen, ratoromanischen, ferner siiditalienisch und sardisch ; 
wahrscheinlich doch auf grund ererbter lautverhialtnisse der 
lateinischen muttersprache. Daher ist ja auch im spani- 
schen 4 dazu gekommen, als der graphische ausdruck fiir 
mouilliertes 7 zu dienen, z. b. in batalla, maravilla. 

Wie ist es zu erklaren, dass -/+ gemass seinen wirkungen 
auf vorhergehende vocale sich als eine lautverbindung aus- 
weist, in der irgendwie das helle 7’ enthalten war, wenn 
anders dunkles 7 urspriinglich der alleinige lautwert der 
liquida im lateinischen war? Ich weiss dariiber nur eine 
vermutung beizubringen. Geminata ist, wie so oft, nichts 
anderes als consonantenlange. So gut nun lange vocale 
ganz gewohnlich im laufe der zeit durch unahnlichwerdung 
ihrer zwei moren diphthongiert zu werden pflegen, ebenso, 
sollte man meinen, miisste entsprechendes gelegentlich auch 


Vol. xxiv.] Dunkles und helles \ im lateinischen. 59 


einem langen consonanten widerfahren kénnen. Nehmen 
wir an, -2f sei also allmahlich zu -7/’- diphthongiert worden, 
so konnte mit dieser entwickelung der keim zu einem 
fiirderhin sich einfindenden -¢'/’- gegeben sein; auf die 
diphthongierung folgte wiederum eine monophthongierung, 
die einsatz- und absatzmora der langen liquida assimilierten 
sich wieder, aber unter vorwiegen des klangcharakters des 
schlusselements. Hier lage folglich der grund, warum 
velle, vellem sowol unter den den e-vocalismus schiitzenden 
formen anzutreffen sind, wie ve/zm, im gegensatz zu vo/o, 
volunt. 


IV. Vor allen iibrigen consonanten, ausser -y- 
und einem zweiten -4, hat die lautgruppe e/ eben- 
falls die verdumpfung zu o/ > w/ erlitten. 

Es ist hier geratener, die abfertigung der scheinbaren 
ausnahmen der regel im voraus vorzunehmen.  Solche 
sind: celtis oder celthis f. eine afrikanische art des lotus, 
nur bei Plinius, ce/tzs f. ‘meissel des steinmetzen,’ celtium 
‘schildpatt’ bei Plinius, alat. me/tom i. q. meliorem Paul. 
Fest., spatlat. spelta ‘spelz’; helvus; celsus. 

Von diesen entfallen wol ohne weiteres das erstere ce/tis 
(celthis) und spelta, als vermutlich fremdsprachliche pflan- 
zennamen ; sfe/ta ist zudem augenscheinlich nur ein spat 
aufgekommener ersatz fiir die dem alten latein gelaufigen 
ehrwiirdigen kulturausdriicke far, ador, far adoreum. Auf 
celtium ist auch nicht viel zu geben, da bei Plinius nat. 
hist. VI § 173. IX § 38 die lesart zwischen dieser form 
und chelyon schwankt; zusammenhang mit gr. yédus, 
yerovn, xédvov auf dem wege der entlehnung wol in 
jedem falle nicht abzuweisen. 

In. celtis ‘meissel,’ das wol mit cu/ter ‘messer’ und gr. 
Ko\drrw, xokaTTHp wurzelhaft zusammenhangt, ferner in 
dem alat. me/tom und in helvus hindert nichts die synkope 
eines -e- nach der liquida anzunehmen; bevor dass -e- 
unterging, kénnte es zur erhaltung des e-vocalismus der 
wurzelsilbe mitgewirkt haben. Dann wiirden *ce/-e-tz-s und 
*mel-e-tom morphologisch ihre ankniipfung an griechischen 


60 Hermann Osthoff. [1893. 


bildungen wie yév-e-ci-s, véu-e-oi-s, beziehungsweise é\-e-Té-s, 
oxed-e-To-5 finden; *helewo-s stiinde zu germ. *7e/-wa-z = 
ahd. asadchs. gé/o (gen. gélwes), ags. jeolo ‘gelb’ und zu 
lit. ze/-wa-s ‘griinlich’ in demselben verhaltnis der suffix- 
abstufung, wie gr. ion. xeve(¢)o-s kypr. xeveugo-v zu xev(¢)d-s. 

Lat. pelvi-s war pélvi-s, nach der bei 4lteren dichtern 
vorkommenden messung mit dreisilbigkeit und “ diarese” 
péluis; es fallt also ausserhalb des bereichs unserer laut- 
regel. Gegen die vorgeschlagene auffassung von /elvu-s 
aber kénnte noch eingewendet werden, dass ein indog. 
*3helewo-s schon ins uriateinische in der form *fe/ovo-s 
aufgenommen werden musste, da man ja den wandel von 
altem hetero- und tautosyllabischem ew in ow in die 
uritalische sprachperiode zu verlegen pflegt. Aber zu 
einer so hohen datierung des ow aus ew ist trotz der 
teilnahme aller altitalischen dialekte an dieser erscheinung 
kein durchschlagender grund vorhanden. In dieser erwa- 
gung nehmen wir auch keinen anstand an der erklarung 
des gen. sing. ge/é#s aus einer urlateinisch zunachst noch 
vorhanden gewesenen -ew-form *geleus; vgl. oben s. 55. 
Man braucht demnach auch das archaische Leucesze des 
Saliarliedes nicht in der iiblichen weise scheel anzusehen. 

Was endlich celsu-s, ex-celsu-s anbetrifft, so ist vielleicht 
die vermutung nicht zu kiihn, dass hier ein contaminations- 
product vorliege, an dessen zustandekommen einerseits ein 
geminatabehaftetes adjectiv *ce/lo-s < *cel-no-s und anderer- 
seits ein participiales *cu/so-s sich beteiligten. Gerade bei 
der wurzel se/- ‘sich erheben’ bieten nominale bildungen 
mit -#- suffix, durch die das hypothetische *ce//o-s zu 
stiitzen ware, ungesucht sich reichlicher dar: lat. colli-s < 
*col-ni-s, im litauischen Ad/-na-s ‘berg’ und besonders die 
adjectiva lit. 4i/na-s ‘erhaben,’ pra-kil-ni-s ‘stattlich, an- 
sehnlich.’ : 

Fiir unsere regel nun, dass e/, ausser vor -y- und einem 
weiter hinzutretenden -/, sonst vor consonanten im latei- 
nischen zu w/ verdumpft wurde, lassen wol eine anzahl 
von belegen sich anfiihren. Doch ist vorsicht bei der 
wertschatzung der einzelnen beispiele allerdings geboten ; 


Vol. xxiv.] Dunkles und helles \ im lateinischen. 61 


denn da auch die ablautstufen indog. of und 7 vorcon- 
sonantisch im lateinischen durch z/ vertreten sind, so 
bleibt mehrfach die méglichkeit offen, dass eine zu einer 
ausserlateinischen e/-form gehaltene lateinische w/form mit 
jener nicht dieselbe wurzelvocalstufe gehabt, sondern zu 
ihr im ablautsverhaltnisse sich befunden habe. Folgendes 
mag unter dem vorbehalt dieser cautel hier in betracht 
kommen : 

ulcus nN. = gr. &dkos n., aind. areas n. ‘hamorrhoiden.’ 
Die griechische form sollte *é\xos lauten; sie hat wol 
den spiritus asper durch volksetymologische verkniipfung 
mit €\«m ‘ziehe, zerre, reisse’ bekommen. 

culmen n. < *cel-men, zu der wurzel el- ‘sich erheben, 
ragen’ gehoérig; den neutren mit dem suffixe -men- eignete 
bekanntlich mit vorliebe von alters her die e-hochstufige 
wurzelgestalt. 

pulmentum < *pel-men-to-m : umbr. pel-mn-er gen. ‘ pul- 
menti,’ welches “a pu/mento non longius distat quam feg- 
minis a tegumento’’ (Biicheler). 

ulmu-s f£. = ags. ahd. mhd. é/m m. ‘ulme’ (engl. e/m-tree, 
ahd. mhd. é/m-boum). Allerdings ist eben dies ein fall 
der verschiedenen ablautsmoéglichkeiten, da herleitung des 
ulmu-s aus */mo-s an ir. /em und solche aus *o/mo-s an aisl. 
alm-r den erwiinschten riickhalt finden k6nnte. 

Den participien pzlsu-s und volsu-s (vulsu-s) gibt man 
immer am besten die gleiche wurzelablautstufe wie den 
praesentien ello, vello. Auch fer-culsus wird sich zu per- 
cello im grunde nicht anders verhalten haben, als /a/su-s 
zu fallo, salsu-s zu sallo; nur dass der voraussetzbare 
identische vocalismus bei per-culsu-s, per-cello, als zusammen- 
gesetzten formen, nicht. notwendig der der e-stufe gewesen 
zu sein braucht, ganz wol auch *fer-calsu-s, *per-callo dahin- 
ter stecken mégen. Die vertreter der vulgaten ansicht, 
dass pulsu-s und volsu-s ein -zl- (-ol-) aus indog. -/ enthalten, 
gemiass der alten tiefstufigkeit des wurzelablauts der -Zo- 
participien, beriicksichtigen zu wenig den charakter dieser 
-soformen als anerkannter massen jiingerer analogiebildun- 
gen. Es lasst sich im allgemeinen der satz aufstellen, 


62 Hermann Osthoff. [1893. 


dass iiberhaupt die participia auf -sw-s, zu mindesten aber 
die nicht lautgesetzlich aus ehemaligen -fo-formen  ent- 
wickelten unter ihnen, keinen selbstandigen wurzelablaut 
mehr innerhalb des verbalsystems vertreten ; in den weitaus 
meisten fallen schliessen sie sich einfach der vocalisation 
des zugehorigen praesens an. Und wie nun z. b. mersu-s, 
tersu-s vorliegen, so werden auch *felso-s zu pello, *velso-s 
zu vello gebildet worden sein; die vocaldifferenz hier ent- 
sprang erst secundar durch das wirken unseres brechungs- 
gesetzes, das pudsu-s und volsu-s ins leben rief. 

Besonders beachtenswert diirfte mu/su-s ‘mit honig 
angemacht,’ mulsu-m n. ‘weinmet’ sein : dessen verhaltnis 
zu mel mell-ts gestaltet sich doch erst vdlliger gleich dem 
von salsu-s ‘gesalzen’ zu sa@/ sal-zs, wenn jenes eben auf 
*melso-s zuriickgeht. 

In volt (vult) 3. sing. praes. ist das genaue ebenbild von 
lit. (pa)-welt ‘er will’ zu erblicken. Wenn man dann auch 
voltis (vultis) 2. plur. aus *veltis herleitet, so stellt sich 
damit erst ein vollstandigerer parallelismus der flexionen 
von volo, volt, voltis, volunt < *vel6 u. s. w., velle, vellem 
und andererseits von fero, fert, fertis, ferunt, ferre, ferrem, 
sowie ¢0, tt < *ett, itis, eunt, tre, trem heraus: das latein 
hat bei diesen urspriinglich der “athematischen” wurzel- 
classe angehérigen praesenssystemen iibereinstimmend die 
é-hochstufige starke stammform des sing. act. des praes. 
indic. verallgemeinert. 

Uber die affection oder nichtaffection eines im wortaus- 
laute stehenden -/ durch unsere brechungsregel will ich 
hier nur ein paar zerstreute andeutungen geben. Sie 
kniipfen an die auf -e/ ausgehenden worter ve/, mascel und 
semel an. Es ist aus allgemeinen griinden wahrscheinlich, 
dass diese drei formen irgend eine der bedingungen ent- 
hielten, unter denen nach unseren regeln II (s. 53 f.) und 
III (s. 57 f.) die brechung zu unterbleiben hatte. 

Uber den ursprung der conjunction ve/ ‘oder’ stehen 
sich im wesentlichen zwei ansichten gegeniiber. Nach der 
vulgaten durch Brugmann begriindeten theorie ist es 
injunctivgebilde und als solches auf *ve/-s beruhend. 


Vol. xxiv.] Dunkles und helles \ im latetnischen. 63 


Dagegen spricht von vocalischer seite nichts, da *vel-s 
wol zuniachst friihzeitig durch assimilation der lautgruppe 
-ls zu *vell werden musste und also dann gemination, wie 
in velle, vellem, vorlag; ein *vol < *vols < *vels wire 
mithin nicht zu erwarten. Andere, wie Leo Meyer, Whar- 
ton und Skutsch, bevorzugen jedoch die erklérung aus 
einer imperativform *ve/e, die vorsonantisch und auch als 
sogenannte “schnellsprechform’’ vorconsonantisch zu vel 
verkiirzt worden sei. Auch das vertragt sich mit unserem 
lautgesetz: die regel iiber -e/- vor palatalem vocale kame 
zur anwendung. Skutsch fiihrt fiir letztere ansicht und 
gegen die Brugmann’sche deutung das fehlen von zeug- 
nissen des *ved/ in der plautinischen metrik an. Vielleicht 
spricht in demselben sinne insbesondere die zusammen- 
riickung vel-ut, vel-uti, die zwar nach dem oben s. 54 
bemerkten nicht so friihzeitig in dieser form bestand, dass 
-el- durch -z- hatte in -o/- gebrochen werden k6énnen, jedoch 
wol immerhin alt genug ist, um bei etwaigem ursprunge 
des vel aus *vel-s die form *ve//-ut erwarten zu lassen ; 
velut ist ja auch schon plautinisch, und Plautus’ sprache 
kennt das von Biicheler nachgewiesene ¢err-uncius mit 
terr- aus *ters- ‘dreimal.’ 

Als vertreter der synkopierten nom.-sing.-bildung wie 
osk. famel, umbr. katel, tigel gilt fiir das lateinische 
bekanntlich famul bei Ennius und Lucretius. Nun rivali- 
siert aber mit diesem famu/ hinsichtlich der frage der 
lautgesetzlichkeit das ein paar mal bei Probus angefiihrte, 
einmal auch inschriftlich als eigenname vorkommende 
mascel = masculus. Da es die form auf -e/ ist, die aus 
dem system herausfallt, so hat wol famu/ seinen vocalis- 
mus von famulo- in den obliquen casus bezogen ; folglich 
hatte mascel den anspruch, eine erklarung auf lautgesetz- 
lichem wege zu fordern. Eine solche ergibt sich bei dem 
ansatz der entwickelungsreihe *smascel-s > *mascell > mascel 
und der annahme, dass auf der durch *masce// vertretenen 
zwischenstufe der e-laut durch die nachfolgende geminata 
geschiitzt worden sei. 

Hat man das zahladverb seme/ mit Jakob Wackernagel 


64 Hermann Osthoff- [1893. 


aus *sm-méli herzuleiten und mit got. mé/ ‘zeitpunkt,’ 
unserem -mal in etn-mal u. s. w. zusammenzubringen, so 
wiirde es hier unerwahnt bleiben miissen. Aber diese 
deutung ist doch gar zu unsicher; viel probabler bleibt 
die altere auffassung, dass seme/ irgendwie mit szmu/, altlat. 
semol semul, umbr. sumel ‘simul’ in naherer verbindung 
stehe. Allerdings wird man seme/ und simul nicht als 
schlechthin identisch betrachten diirfen. Die lautliche dif- 
ferenz in der schlusssilbe klart sich auf, wenn seme/ auf 
*semell < *semel-s zuriickgebracht wird: wie 467-s, ferner 
ter < *ter-s (vgl. oben s. 63 terr-uncius), quater < *quater-s, 
so enthielt auch das erste zahladverb urspriinglich das fiir 
diese formenkategorie charakteristische -s; ebenso ja auch 
griech. da£ in iibereinstimmung mit 68i-s, tpi-s, TeTpaxi-s 
u. s. w. Wenn szmul, wie ja meistens angenommen wird, 
der alte doppelganger von szmz/e, neutrum zu szmzlt-s, war, 
so muss jene kurz- oder “schnellsprechform” auf so 
friihzeitiger synkope an dem _ grundgebilde *seme/(z) be- 
ruhen, dass das abgestossene -2 hier bereits verschwunden 
war zu der zeit, als unser lateinisches brechungsgesetz in 
wirksamkeit kam. Ein gleiches gilt fiir facul, difficul = 
Jacile, difficile. 

Der beobachtung, dass e/ sich unter gewissen umstanden 
normal in lat. o/ (w/) verwandele, ist vor uns am niachsten 
Havet mém. de la soc. de linguist. V. 43. 46 anm. 
gekommen. In einigen die negative seite der hier behan- 
delten lauterscheinung betreffenden bestimmungen kommt 
der franzdsische sprachforscher zu dem gleichen ergebnis 
wie wir : in der erkenntnis, dass e erhalten bleibe einmal 
vor //, wie in vellem, mellis, sodann vor 2, in velim, melior 
u. dergl. Die dritte derartige bestimmung bei Havet, 
dass dies auch nach ¢ und g geschehe, ist nicht stichhaltig, 
da die dafiir angefiihrten belege celer, celeber, scelus, gelu 
und ce/sws anderer auffassungsweise zu unterwerfen sind, 
wie sich uns im vorhergehenden im einzelnen gezeigt hat. 
Das wahre wesen der erscheinung hat Havet besonders 
insofern verkannt, als ihm entging, dass. die verwandlung 
oder nichtverwandlung von e/ zu of durch eine lautliche 


Vol. xxiv.] Dunkles und helles \ im lateinischen. 65 


doppelnatur des lateinischen / bedingt war, die ihrerseits 
wiederum von der phonetischen beschaffenheit des unmit- 
telbar auf die liquida folgenden lautes, insbesondere von 
der urspriinglichen natur nachfolgender vocale, abhing. 

Den unterschied der beiden klangqualitaten des 7 kannten 
in dunklen umrissen auch schon die alten grammatiker. Es 
kommen hier vornemlich die zeugnisse des Plinius bei 
Priscian I § 38 H. = gramm. lat. II 29, 8 K. und des 
Consentius gramm. lat. V 394, 30 K. in betracht. Was 
diese beiden gewahrsmianner als den “diinnen”’ laut, “exi- 
lem sonum,” der liquida anmerken, im gegensatz zu dem 
“plenum” des Plinius, dem “pinguius” des Consentius, 
das entspricht im wesentlichen unserem hellen /; iiber- 
einstimmend ist ja auch bei beiden die angabe, dass der 
“diinne laut” in der geminata // herrsche, wofiir sie die 
beispiele 2//e, Metellus, Alia anfiihren. Bei anderen gram- 
matikern, nemlich bei Servius, Pompejus und Isidor, kehren 
im grunde dieselben distinctionem der verschiedenen aus- 
sprache des 4, wie in den berichten des Plinius und Con- 
sentius, wieder, wenngleich zum teil unter anwendung einer 
abweichenden terminologie. Von neueren haben Corssen 
‘und Wilh. Meyer-Liibke das, was Plinius, Consentius und 
genossen iiber / und seine wechselnde klangqualitat haben 
sagen wollen, im ganzen richtiger erfasst, als der phonetiker 
des latein Seelmann, der auch hier, wie sonst seiner gepflo- 
genheit gemiss, in die lehren der alten grammatiker allerlei 
diesen gewiss fern gelegene moderne lautphysiologische weis- 
heit hineinzuinterpretieren sich abgemiiht hat. 


66 : Paul Shorey. [1893 


V.— On the Implicit Ethics and Psychology of Thucydides. 


By PAUL SHOREY, 


PROFESSOR IN THE UNIVERSITY OF CHICAGO. 


THE main interest of the History of the Peloponhesian War 
does not lie in the incomparable vividness of the narrative, 
nor even in the tragic drama of the pride and fall of imperial 
Athens, and the pity of this suicide of the. Hellenic race in 
its culminating age. Fascinating as is the mere story, its 
chief attraction for us consists in the fact that it is the 
embodiment of a subtle and consistent, if one-sided, philoso- 
phy of life ; that it is, to adapt a phrase of Carlyle, a portion 
of human history penetrated and informed by the spirit of 
the man Thucydides. This Thucydidean criticism of life I 
propose to study in its two chief aspects, which for conven- 
ience I will designate as (1) ethical positivism, (2) intel- 
lectualism. 

The fundamental assumption of this ethical positivism is 
that the nature and conduct of man are strictly determined 
by his physical and social environment and by a few elemen- 
tary appetites and desires. Around this primitive core of 
human nature society and convention have wrapped sheath 
upon sheath of decorous pretence — ethical, social, religious. 
The naive man is duped by this moral drapery, he accepts 
the word for the deed, the alleged motive for the true, and 
rarely, if ever, penetrates to the underlying realities. The 
wise man is not so deceived. He has looked into the work- 
ings of his own heart, he has studied human nature in the 
revealing light of war, pestilence, and revolution, and, how- 
ever well draped the figures he meets in his daily walk, his 
penetrating imagination discovers the naked man beneath. 
Such is the conception of human life everywhere suggested 





Vol. xxiv.] Jmplicit Ethics and Psychology of Thucydides. 67 


when not explicitly affirmed by Thucydides. The first 


axiom of this doctrine is that human nature remains always 
essentially the same, and that it cannot be permanently 

















restrained or moulded by the artificial conventions of law, 
and religion.! = 

It is on this belief that he bases his conception of history 
as philosophy teaching by example. He commends his work 
to the favorable judgment of those who desire to have an 
accurate knowledge of the past and so forecast the future 
which from the nature of man will resemble it.? 

The atrocities of the revolutions of Corcyra are such as 
occur and always will recur while the nature of man remains 
unchanged. (III. 82.) The Athenians, so their envoys at 
Sparta declare, were constrained to accept and maintain their 
invidious empire by motives resistless to human nature, ambi- 
tion, gain, and fear. It has always been the rule that those 
should take who have the power and those should keep who 
can, and no man possessing this power ever stayed his hand 
for abstract considerations of justice. This basic human 
nature the Athenians have indulged with great moderation.* 
I do not blame the aggression of the Athenians, says Hermo- 
crates. It is human nature everywhere to dominate those 
who submit. “We hold the customary beliefs about the 
gods” (the Athenians declare at Melos, V. 105), and we 
know for a certainty that men by an inevitable law of their 
nature dominate when they can. We did not promulgate 





1 The connection of cynicism and the doctrine of necessity in Thucydides is 
not accidental. In Machiavelli’s first work on the revolt in the Val di Chiana, 
he appeals to Roman history for the solution of a problem in present politics: 
“. . . Perché gli uomini in sostanza sono sempre gli stessi ed hanno le medesime 
passione: cosi quando le circostanze sono identiche, le medesime cagione portano 
i medesimi effetti, e quindi gli stessi fatti debbono suggerire le stesse regole di 
condotta.” 

27. 22. xara 7d dvOpmmeov by Thucydidean usage means more than “in all 
human probability,” as Professor Jebb renders it, Hellenica, p. 266. Plato, Crito 
46 E 50a ye 7’ dv@pwrea, is not a parallel. 

31. 76 xpnoduevoc TH avOpwreig pice. For the ethical suggestions of this 
phrase, cf. Aristoph. Nubes 1078 xp 77 pice, cxipra, yéda, rouse under aloxpdr* 
Isoc. Areopagit. 38 éxvobvras TH pice xpjcOat. 

4 TV. 61. 


ee ae nl 








r 
| 
| 


| 
| 
L 


68 Paul Shorey. SS 7 B53; 


this law, nor were we the first to profit by it. We found it 
in operation and shall leave it for all futurity.1 It is in 
seasons of pestilence and revolution that all disguises are 
thrown off and this indomitable brutality of man is most 
plainly displayed. Neither fear of God nor law of man 
could check them, he says of the plague at Athens. Human 
nature prevailing over all laws is his summary of the condi- 
tions at Corcyra.?" 

The contempt of Thucydides’ alma sdegnosa for this aver- 
age elemental human nature is hinted in many a_scornful 
phrase. Man is naturally fickle,? boastful,* envious,’ ungrate- 
ful, and selfish,® elated by success, yet unable to bear pros- 
perity.7’ The multitude are prone to magnify the unknown, 
and remote,’ intolerant of painstaking accuracy,’ and easily 
seduced by false glitter. Their judgments are swayed by 
mere words," their beliefs determined by their desires,!? and 
their moods shift with their changing conditions.® 

But we look for something more philosophic than these 


1 This is generally rendered “ We opine that the gods, and we know that men, 
rule when they can,” etc. The sentence, if critically studied, is, as Dionysius 
says, ducelkacros kal Tots wavy Soxovow eurelpws Tod dvdpds €xewv, but in spite of 
the co-ordination with re, I am inclined to take 7yovmefa in the first member 


‘absolutely: “We believe in the gods as a matter of opinion, and we know for a 


fact that men,” etc. It is not in accordance with Thucydides’ mental habit to 
argue that the gods rule when they have the power, and Dionysius in his close 
paraphrase ignores this thought. He says: drt rd mév Getov SbEy YeyveoxKovow 
drravres. 

2 r&v vduwv kparjoaca % avOpwrela pvors III. 84. The phrase is in Thu- 
cydides’ manner, even if the paragraph be spurious, 

3 Cf. II. 65 Sirep pire? Suchos racetr, etc. Cf. IV. 28; VI. 63; VIII. 1. 

4 7d dvOpwreov Kou @des V. 68, 

5 TI. 35; Il. 45; VI. 16 rots wev dorots pOovetrac pica; II. 64 doris 3 eri 
peyloros rd éripbovoy apBdver 6pO&s Bovdeverar, 

6 TI. 40. 4; VI. 16 GAN worep SvoruxodvTes ob mpocaryopeudueda ; VII. 89. 3. 

TIIT. 39. 

8 \VIL 11 Ta yap 51a wrelorou mdvres toper Oavuatdueva, 

®T. 20; VI. 54 dxpiBés oddev Aéyovras, etc. 

10]. 22; IV. 108. 4 épodrxa; V. 85 éraywyd; VI. 8. 2. 

VI. 34 mpds Ta Neyoueva Kal al yrOmat loravrac. 

12 TIT, 3 pettov pépos véuovres TE mh BovrAecOar adyOH elvar; IV. 108 6 dé wh 
mpoclevrat hoyioug@ abroxpdropte Siwhetobat, 

18 TI, 54; I. 140. 1; III. 82, 2; cf. also IV. 61. 


Vol. xxiv.] Jmplicit Ethics and Psychology of Thucydides. 69 


isolated disparaging utterances. We want _a_ systematic 
ethical terminology based on a psychological analysis of the 
chief springs and motives of human action. The nearest 
approach to this is to be found in the speech of Diodotus 
on the affair of Mitylene, III, 45. “All men are naturally 
prone to error (he sdys in substance), and there is no law 


‘that will keep them from it. Legislators have run through 
the list of possible penalties to no effect, and we must invent 
some more awful terror than the fear of death if we expect 
to bridle human nature. At one extreme of human con- 
dition poverty and necessity inspire reckless daring, at the 
other license begets grasping greed on insolence and pride; 
and so the various accidents and conditions of life acting 
with fatal necessity on the various tempers of men lure them 
on to danger. And in addition to these impulses, hope and 
passionate desire are everywhere operative for harm, the 
one leading, the other following, the one devising enter- 
prise, the other whispering promise of success,’ — anticipa- 
tions of the unseen future yet more potent over men’s 
minds than dangers plainly seen.* 

“ Fortune too, contributes her part to exalt men’s spirits, 
and by the unexpectedness of her aid often induces them 
to venture with inferior resources— more especially states 
in so far as they contend for the highest stakes, freedom, 
or imperial dominion, and the individual acting with a multi- 
tude is more prone to an irrational overestimation of his 
powers. In short, it is impossible (and the supposition of 
the contrary is a mark of the utmost simplicity) to restrain 
by law or any other deterrent force any strong bent of 


























1 Cf. Bacon’s saying: “There is no passion in the mind of man so weak but 
it mates and masters the fear of death.” 

21 cannot accept Classen’s text nor his interpretation here. The scholiast 
rightly says: “dpyh 6 rpémos.” dpy7 though construed with é&d-youer is felt with 
évvruxla. For the conjunction of gvyruxla:, yvéun, and épy%, cf. III. 82. For 
the thought in dynxéorov rivds xpelocovos, cf. Emerson’s words: “. . . Temper 
prevails over everything of time, place, and condition, and is inconsumable in the 
flames of religion.” 

* Cf, TVa66..4; VI..15. 3,2. 

4 Cf. V. 87 Gv dpare; V. 103 pavepal— ddaveis; V. 113; VI. 9. 3. 


70 Paul Shorey. [1893. 


human nature.’’! If we add qiAoTipia? and girovxia, perti- 
nacity, or “ persistive constancy,” to the positive promptings 
here enumerated, and supplement voyos and géB8os by the 
restraining principles of aicydvy and édeos, we shall have a 
‘nearly complete list of Thucydidean motives. Every phrase 
in the speech of Diodotus is of typical significance for the 
whole history, and every term demands a commentary. This 
épws retains nothing of the associations of that “thirst in all 
men’s nature named épas,” which fills so large a place in 
Greek literature. It is simply the master passion, or the 
passion which for the moment has mastered the mind. — 


eo» , 9” , ’ ” 
ed & ich, Srouv Tis TUyxaver ypelay Eyov 
To0T éoO éExdotw petlov 7) Tpoiav éreiv, 


says Menelaus in the Andromache. — It is the pws which the 
tyrant soul of the Republic (573 A) establishes on the throne 
as its bosom’s lord; the desire of which Diotima says (Symp. 
205 D), To péev Kehararov éeott Taca 7 TOV ayabav ériOupia 
kal Tod evdaipovety Oo péytatos Te Kai Sorepds epws TravTi. 


“Speaking broadly, all desire of good things and of happi- 


LIMIT. 45.4 aX’ 7 wey wevla dvdyxy Thy réd\uav wapéxovea, 7 5 efovcla UBpe 
Thy wreovetlav kal ppovfwari, ai d¢ &ddac Evyrvxlac épyy Trav avOpdrwy, ws 
éxdorn Tis Karéxerat br’ dvnxéorou Tivds Kpelocovos, éEd-youow és Tods Kivdbvous. 
n Te éXmls kal 6 Epws ért wavrl, 6 per tyovpevos, 7 -5é épewouévn, kal 6 pev Thy 
ériBorhy éxppovritwy, 7 dé rhv edrroplav THs TUXNS UroTOcioa TheicTa BAdrToOVEL, 
kal dvTa dpava xpeloow éorl T&v dpwpévwy Sevdv. Kal % TUX €x’ adrois obdéev 
Edaccov EvuBaddrerac és 7d éralperv: ddoxirws yap torw bre wapicrapévy Kal éx 
TGv brodecorépwy kivduvevery Tivd mpodyer Kal obx Fooov rds wodes, dow wept Tdv 
peylorwv, édevdeplas 7 AdNwy dpyjs, kal werd wdvrwv Exacros ddoylcrws él 
whéov Te abrov €ddEacevr. adwdds Te ddUvarov kal rodA7s ednbelas, boris olerar THs 
dvOpwreias picews dpuwpuérns mpoOtuws Tr mpatar dworporhy Twa txew 7 vouwry 
loxte 7 AAW TY devG. With this résemé of Thucydidean psychology it is inter- 
esting to compare the poetic description in the Timzeus of the mortal soul, “sub- 
ject to terrible and irresistible (dvayxata) affections, — first of all pleasure, the 
greatest incitement to evil, then pain, which deters from good,—confidence and 
fear, two foolish counsellors, anger hard to be appeased and hope easily led astray. 
These they mingled with irrational sense and /ove (passionate desire) that at- 
tempteth all things, according to necessary laws, and so framed man.” (69C D, 
Jowett. ) 

2 III. 82. 8. 7d giddripov dyhpwv pdvow II. 44. 4. 


Vol. xxiv.] Jmplicit Ethics and Psychology of Thucydides. 71 


ness is the chiefest and cunning lure of love to every man.”’ 
Like the Heracleitean @uyos, it buys its will at the price of 
death. It is éwvyespntys dravros in Plato’s phrase, — éxdpor- 
rity Thy émiBornv, as Thucydides puts it. Thucydides usu- 
ally employs the weaker synonym opp), reserving the tragic 
intensity of épws for the fatal passion of Athens for the expe- 
dition against Sicily.1 But whether exalted and animated by 
desire? or goaded by necessity and _ intolerable humiliation,® 
men’s acts are too rarely determined by a cool, logical cal- 
culation of the chances of success. Their judgments are 


affected by their tempers.‘ Toda is is frequently Gdoyearros 
(III. 82; VI. 59). For when opp hath fallen upon or épas 


taken possession of the soul, éAmis enters in to heighten 
confidence and blind to the risk of failure. 

The Greeks seem to have been particularly exposed to 
the temptations of the over-sanguine temperament, and their 
moralists are inexhaustible in warnings against its illusions. 
“From Zeus there cometh no clear sign to men: yet, never- 
theless, we enter on high counsels and meditate many acts ; 
for by shameless hope our bodies are enthralled, but the 
tides of our affairs are hidden from our fore-knowledge,” 
says Pindar (Nem. XI. in fin., Myers). “For that hope 
whose wanderings are so wide is to many men a comfort, but 
to many a false lure of giddy desire,” sing the chorus of the 
Antigone (615 Jebb). And similar is the lesson which the 
speakers in Thucydides constantly inculcate in a more bitter 
and cynical tone. “Intelligence,” says Pericles, “relies not 
so much on hope, which is strongest when all else fails, as 
on estimates based on existing resources by judgment, whose 























1Cf. IV. 4 rots orpariéras ...dpun éwérece; VI. 24 kal Epws évérece rots 
mao dpolws éxrdedoa; VI. 13 und... dvedpwras elvac rv drbvrwy, 

2 éwalpecOar xépdec III. 38; édwide III. 45; Evvécews dyGu III. 37; 7H vl 
VII. 41; Hdovq I. 84; edruxig I. 120, etc. 

3 rapottve, mapoztver Ba: I. 67, 1. 84, VI. 88, V. 99; TO dvayKaly VI. 56. 

4 VIII. 2. 2 épyGvres xplvew ra mpdyuara; Il. 21 dv dxpodcOa ws Exacros 
wpynro; IV. 108 BovAhce: xplvovres doade? 7 mpovola dcpadet; I. 122 ebopyy- 
tws. Cf. also VI. 13 éwiOuula pev édXdxuoTa Karopbotvra:, mpovolg 5é mheiora; 
Il. 11 Kat of Aoywru@ eAdxicTA Xpwpeve Ovug wreicta és Epyov Kabloravra; 
II, 22 dpyq7... youn. 


‘ye oe Paul Shorey. [1893. 


forecasts are surer.”! The feeble who put their trust in the 
spendthrift hope (the Athenians warn the Melians) discover 
her perfidy only when she has left them nothing for their 
dear-bought knowledge to guard.2 This disparagement of 
hope is frequently accompanied by an allusion to the pro- 
verbial uncertainty of the future,’ — the surprises of war,‘ the 
paradoxes of fortune.® 

“You Athenians (say the Spartan envoys, IV. 17) will not 
abuse your success at Sphacteria like fools unaccustomed 
to the vicissitudes of fortune, who ever reach out in hope 
of more because their present good luck has come as a 
surprise.” Similarly, Hermocrates, urging on the Sicilian 
States peace and union in the face of Athenian aggression 
(IV. 62), warns those who expect to profit by a prolonga- 
tion of their dissensions: And if any one bases expectations 
of advantage on the justice of his cause or his superior 
might, let him not expose himself to reverses that will griev- 
ously disappoint his hope ... . for righteous Vengeance does 
not necessarily prosper because deserved, nor is strength 
secure because it is full of hope. 

“When rational grounds of hope fail, men resort to the 
unseen, to oracles and prophecies,’’ Thucydides says with 
cold contempt, or its place is taken by stubborn persistency 
in a course once determined upon. This pertinacity is char- 
acteristic of the eager Athenian temperament. As emulous 


1 JI. 62. Ido not accept Classen’s suggestion that réAua is the subject here. 
For the antithesis of é\mls and brdpxorra cf. VI. 31.6 émi peyloryn édwldc Tov 
pedrévTwv mpds Ta Urdpxovra érexeip7On, and also V. 111 éAmcfdueva .. . brdp- 
xovra; VI. 9. 3. 

2 V.-103. 

$1.42. 2; Uk 42s¥ 62. 45-1435 - VI.'9. 

4 III. 30.4; 1.122 Fxeeora 6 mbdeuos ert pynrots xwpet; I]. 11; VII. 61; 1.78; 
Vi t02. : 

5 VIII. 24 év rots dvOpwrelos Tod Blov mapaddyors, etc.; I. 140. I, 2. 

6 Jowett’s “let him not take his disappointment (sc. at the frustration of his 
hope by my words) to heart” is doubtful. r@ wap édmlda opddreoOau, cf. VII. 
66, seems to refer to an actual reversal of expectation by the event. The sentence 
is a curiously worded threatening admonition characteristic of Hermocrates. Cf. 
VI. 78: “ And if he prove to have erred in judgment, he may live to bewail his 
own misfortunes and wish to be envying my prosperity again.” 


Vol. xxiv.] Jmplicit Ethics and Psychology of Thucydides. 73 


thirst for fresh glory,! it built up and maintained their em- 
pire? As “persistive constancy” it appears in the bull-dog 
tenacity with which they held on to Aegina in the face of an 
overwhelming combination of enemies (I. 105) and in the 
proud boast, “The Athenians never yet withdrew from any 
siege from fear of any.” * As blind, presumptuous folly‘ it 
wrought their final ruin at Syracuse.® 

We have already noted the impotence of the fear of God 
or the law of man to control these active promptings of 
human nature, and we shall find generally a touch of irony 
in Thucydides’ allusions to the checking and restraining 
principles. ‘‘ Pity should be reserved for equals,” says Cleon 
in his speech on Mitylene,® and sweet reasonableness or 
indulgence (émveieva) should be shown to those who are 
likely to prove conformable‘ in the sequel. They are danger- 
ous feelings for an imperial city to entertain towards infe- 
riors. And his opponent, too, is careful to insist that he 
would not have the decision of the Athenians influenced in _ 
the least by pity or equity. ‘Do not let the Plataeans melt 
your hearts, O men of Sparta,” the Thebans cry, “ by appeal- 
ing to good deeds that are ancient history now. Degenerate 
virtue can claim no remuneration for the thing it was.” 
(III. 67.) And the Athenians peremptorily bid the Melians 
base their arguments solely on the real purposes and power 
of the contestants, and not on any unreal moral conventions. 
Even _where Thucydides’ ethical language is not distinctly 
cynical, it is singularly lacking in warmth and depth of feel- 


ing. He frequently indulges in sneers at the illusions of 














1 gidovixla wept Tod wréovos H5n Kadod. 

21. 70 nv dpa Tod cal welpg cpadGow dvredricavres Edda éxdijpwoav Thy 
xpelav. Cf. IV. 55. 3. 

3 V. 111; cf. IL. 64 da 7d rats Evudopats uh elxecy, and III. 16. 

4 uwpla piroukdry IV. 64; cf. rév rdvrwv dreplorro rapa 7d wxay I. 41. 

§ VII. 28. 

6 III. 40. Jowett’s “Mercy should be reserved for the merciful” is a miscon- 
ception. Vide si tanti Classen’s note. For the idea that justice obtains only 
between equals, cf. III. 9; V. 89. 

7 This is the best single word to convey the associations of éirjdeos here. 
Cf. éwerndelws I. 19, I. 144, and éwirniela V. 82. 


74 Paul Shorey. [1893. 


poetry, patriotism, and the mythical fancy.1 He habitually 
speaks of virtue in a hard, external way as something to be 
acquired, professed, husbanded, exchanged, I had almost said 
bought and sold.? 

A similar moral insensibility is to be noted in his em- 
ployment of the words @epazrevew, edrper&, Sixasov, avdpa- 
yaGia, etc., etc.% 

A good illustration of Thucydides’ tone in these matters 
is his treatment of the specially Greek notion of aids, that 
delicate sensitiveness to the disapprobation of our fellows 
that sometimes approaches very nearly to the modern idea 
of self-respect.* It is, perhaps, hardly an accident that 
Thucydides, except in one passage (I. 84. 3), everywhere 
substitutes the coarser term aicyvvy or To aicypov for the 
more distinctly ethical aié#s. The implication is that aides 
is a rational motive only when it takes the form of intolerable 
constraining shame. At the time of the plague (II. 51) 
_those suffered most who had a reputation for virtue to 
keep up; for from very shame they were unsparing of them- 
selves. That is the tone. One should deal with a powerful 
enemy in a spirit of sweet reasonablenéss and virtue (say 
the Spartan envoys, IV. 19), for be will be more likely to 
keep faith from very shame. But this sense of shame is 
mere folly when cherished as a Quixotic sense of honor 
by the weak. ‘What is that word honor? Air. A trim 
reckoning,” the Athenians declare in substance at Melos. 
And on those who pertinaciously follow its lure it brings 


1II. 41. 4; V. 41. 2 é56xec wwpla elvat radra, of the combat for Thyrea in 
Hat. I. 82; VI. 83 od xaddcerovpeba; I. 21 7d wh pvOGdes . . . dreprécrepov; 
I. 10. 3, 4. 

2]. 123 ras dperds xracOar; II. 51 ol dperfs re dvrimoovuevror; I. 33 pépovca 
és Tovs wohdovs dperyv; VI. 11 Sbgav dperfs weer Gow; VI. 54 érerhdeveay aperhy 
kal tvveowv; IL. 40 Thy dperhy droddcwrv; IV. 19 dvrarododva dperhy. . 

3 E.g. VI. 79 decdig...7d Sixawv... Geparedcere; III. 56 7d... Evudépov 
. . . Oeparevovres; I. 39 7d edmperés Tis Sixns; III]. 64 él T@ éxelvwy xaxg 
dvipayabiay mpovOerGe; Il. 63 dmxpayyocivy dvdpayabliera:; III. 40 éx rod 
dxivdvvov dvipayablier@ar. Cf. also the use of dverlg@fovoy to debase the 
moral currency in I. 75, I. 82, VI. 83, VIII. 50, and the similar employment of 
évyyveun I. 32. 5, III. 40, IV. 61. 5, VIII. 50. 2. 

* Cf. Gildersleeve on aldec@évres ddxdy Pindar. Pyth. IV. 173. 


Vol. xxiv.] Jmplicit Ethics and Psychology of Thucydides. 75 


the greater*dishonor of impracticable folly (V. 111). In 
short, a nice sense of honor is simply one of the many 
perturbing emotional forces that are the cause that men so 
rarely bring an unimpassioned judgment to bear on the 
complicated game of life. 

This brings us to what may be called the intellectualism 
of Thucydides, —his constant preoccupation with the part 
played in human life by the conscious calculating reason.? 
“The moral and the intellectual,’ says Professor Jowett, 
“are always dividing, yet they must be reunited and in the 
highest conception of them are inseparable.” In Homer 
we are happily unconscious of this opposition, —the true 
man is aya0ds Kai éxéppwv, and to “know lawless things” 
is to do them even as to know good things is to be just 
according to the reasoning of the Socrates of the Gorgias. 
In Thucydides we are never allowed to forget the antithesis. 
Plato endeavors to reunite the severed halves of our nature; 
and Aristotle by his formal distinction between the ethical 
and the intellectual virtues recognizes from the point of 
view of common sense the impracticability of the Platonic 
ideal. ‘‘We must not permit the wicked to give the name 
smartness to their unscrupulousness,” says Plato (Theaetet. 
176 D), “for they glory in the reproach.” ‘Most men,” 
says Thucydides (III. 82), “more easily submit to be called 
clever knaves than honest simpletons;* they glory in the 
one epithet and blush at the other.” 

There is a seeming injustice in attributing to Thucydides 
this feeling of “the many.” But his protest is couched in 
language half contemptuous: “ Simple-mindedness, a chief 
element of nobility, was quite laughed down.” And the 
entire history is pervaded by a most un-Platonic antithesis 
between the just and the profitable; a most un-Platonic 
association of cwdpoovvn and dayaGia, and a constant exalta- 


1 Cf. supra, p. 6; VIII. 27. 

2 Chez Thucydide partout oi les idées paraissent elles priment les sentiments. — 
Girard. 

8 bgov xéxAnvrac does not mean “are oftener called,” as it has been taken, It 
must be construed by the analogy of pgov pepe VIII. 89. 


76 Paul Shorey. [1893. 


tion of unscrupulous intellect.1 The nomenclature of this 
intellectual principle is noteworthy. Thucydides does not 
use godds with its earlier suggestion of skill and its later 
connotation of the higher wisdom. He does not employ 
the dpdovnois and ¢poveiy of Plato and the dramatists 
with their moral and religious coloring, nor vods with its 
speculative associations. His words are: yveun, mind, 
judgment; &veows, understanding, the intelligence that 
penetrates shams; Aoyiouds and its paronyms, the calcu- 
lating reason.” 

- His most characteristic laudatory epithet, applied to Archi- 
damus, Themistocles, Theseus, Pericles, Hermocrates, and 
Phrynichus, is ov« a£vveros, not unintelligent, they could 
see through a mill-stone. When c@dper is added, it denotes 
judgment, moderation, discretion, prudence unclouded by 
|passion, rather than any distinctively moral excellence.* 
And the most unpardonable insult, the most stinging im- 
putation, to a Thucydidean personage is the suggestion that 
he is deficient in penetration or dull in perceptions.* ‘Do 
not suppose that we would insult your intelligence by 
attempting to instruct you,’ say the Spartan envoys at 
Athens (IV. 17), “our words are only a reminder.” The 





1 E.g. 1. 42; V. 893: 1.68; ITI. 37; III..56; III. 44. 4. 

2Cf. Il. 11. 7; 40. 3; IV. 10; IV. 108. 4, etc. For éxdoylferGar cf. II. 40. 3; 
IV. 10, etc. I will not, with Professor Jebb (Hellenica, p. 302), add d:dvoa; for 
didvora in Thucydides means usually “ purpose,” or mind and temper generally. 
It rarely is used to denote the intellectual power in especial as in III. 82. 3 rod 
kavodoGat Tas diavolas. Neglect of this nicety lias, I think, led Professor Jebb 
into error in his interpretation of VI. 11.6 xpy 5€ uh mpds Tas Téxas TOv évarriwy 
éralpecGat dda Tas Siavolas Kparjcayras Gapoetv, which he cites together with 
I. 84. 3 in support of the sentence: “In a trial of human forces the chances 
baffle prediction, but superiority in ideas (didvorac) is a sure ground of confidence.” 
But the words really mean: We should feel confident only when we have subju- 
gated the minds (broken the spirit) of our opponents, —i.e. made them feel that 
they are beaten. The context and the use of the aorist are sufficient confirmation 
of this reading; but if more is wanted, cf. II. 87 ob6é Sixasov ris yueouns Td wh 
Kara kpdros uxnOév,éxov dé Tia év abt@ avtidoyiay, THs ye Eyudopas TE dwroBdvri 
aduBrdverGa. Cf. VI. 72. 

31. 80. 2. Cf. I. 84. 2, I. 80, V. 101. Cf. cwppocivy ... aBovdria I. 32; 
VI. 6. 2; VIII. 24. 5 éowppbvncay .. . mapa Td dopadécrepov mpata. 

* dvalcOnros VI. 86; I. 69; I. 82. 


Vol. xxiv.] Jmplicit Ethics and Psychology of Thucydides. 77 


frequency of similar oratorical precautions! and phrases like 
Evvicews ayouopa and ris Evvicews peratroveicOar (I. 140) 
testify to the intensity of this feeling. A Corcyrean audi- 
ence, like an audience of the Italian renaissance, would 
certainly have sympathized less with the avaic@nros Othello 
than with the £uverds Iago, 67. amdrn mepuytyvopmevos Evvécews 
ayoucpa traperapBave (III. 82).2 “We Athenians,” Pericles 
boasts (II. 40), “can all either originate or at least judge 
political measures.” ‘What each of you most desires,” 
says Cleon (III. 38), here, as often, showing us the seamy 
side of the Periclean ideal, “is to be able to speak himself, 
or, failing that, to vie in cleverness with the speakers in 
the readiness with which you apprehend, or anticipate, and 
applaud their points, however slow you may be in fore- 
seeing practical consequences.” It is only those who have 
a saving distrust of their own intelligence,® like the slow- 
witted Spartans, who will admit that they are duaGéorepar 
Tov vopev.* 

The empire of what our ignorance calls chance® reduces 
this power of the intellect to a comparatively humble réle.§ © 

The course of human events, especially in war, is full 
of baffling surprises. The wise man is at the best like 


1Cf. IV. 126; V.9; V.69 in fin. Cf. also IV. 10 pndels . . . Evverds BovrAécOw 
Soxetv elvar éxroyitduevos. ‘ 

2 But cf. the manlier language of Brasidas, IV. 86. 6. 

8 dricrodvres TH €& EavT dy tvvéce III. 37. 

* III. 37; 1. 84. 3. \ 

5 Submep Kal Thy TUxnv boa bv rapa Abyor tvuBD eldOaper alriaoba I. 140. 
Cf. Anaxagoras apud Diels Doxographi 326a (rv tuxnv) ddndov alrlay dvOpo- 
mivy hoyioug. This more nearly represents-Thucydides’ feeling than Professor 
Jebb’s “inscrutable dispensation of a divine Providence.” The phrase tixy é« 
Tov Oelov (V. 104, 112) is not used absolutely, but with a verb, éA\acowoerOat, 
woven, with which the éx is connected at least as closely as it is with the noun. 
It is not réx7y in general, but the special favor of heaven, the last straw at which 
the despairing Melians vainly clutch. Similarly, “they are matters not for rea- 
soning, but for resignation” is too unctuous for the cold severity of @épey re xph 
Ta Te Sayudma dvayxatws of II. 64. It is rather Oupdy évl orhOecor piddv Sapa- 
cavres dvdyxy—let determined things to destiny hold unbewailed their way. 
For additional copious but undiscriminating references on tUx7, cf. Classen, Einl. 
LVIII. 

6 IV. 62. 4; I. 84. 


78 Paul Shorey. [1893. 


Themistocles, trav peAdXOvT@Y . . . apiotos eixactys. The 
lamp of his intelligence illumines but dimly a few steps in 
front of his feet, but only a fool (d&dveros) or a charlatan 
(III. 42) will affirm that he knows of any other light cast 
upon the unseen future save that thrown upon it by reason 
and rational discussion (Aéyos). The sensible man will not 
wish to resemble the herd who when expectations based 
on visible tangible realities fail them turn in their extremity 
to the invisible, to prophecies and oracles, and other de- 
lusions that lure men with hopes to their ruin, neglecting 
the human instrumentalities that might still save them from 
the worst.1 He knows that he cannot control fortune as 
he can his own resolutions, and so is prepared to make 
reasonable concessions in the hour of success.2_ He knows 
that the malignity of chance and the illogical logic of events 
may defeat the best laid plans,? and that no human achieve- 
ment is secure against change and decay. And so he 
accepts the strokes of human adversaries with courage and 
those of the higher powers with submission to the inevitable.® 
Still more baffling to the wise man’s sagacity is the dis- 
simulation of his fellows. The naive man believes what he 
is told and suspects nothing. On emerging from this naiveté 
he passes to the opposite extreme (Plato, Phaedo, 89 D E). 
He looks always for the dessous des cartes, and the antithesis 
of the real and the apparent becomes the chief category of 
his thought. This is the attitude of the personages of 
Thucydides, who are never weary of ‘distinguishing the word 


1V. 103. Professor Jebb’s paraphrase misrepresents the feeling of this 
passage: “This, however, he would affirm —that such resources are not to be 
tried until all resources within human control have been tried in vain.” This is 
a distinctive Socratic or Platonic thought —I do not believe that it can be found 
in Thucydides. I do not wish to seem to split hairs, but shades of meaning are 
as worthy of observation as niceties of syntax, and it is as important that our 
quotations should be strictly relevant as it is that our accents should be correct. 

*1V.64>) 1 190.9, 48 

81. 140 évdéyerar yap Tas Euugopds Tay mpayudrwy obx Focov dualas xwpF- 
gat 7 kal rds Siavolas Tod dvOpwmrov. VIII. 24. 

#1I. 64. 

5 Jbid, and II. 44. 1. 


Vol. xxiv.] Implicit Ethics and Psychology of Thucydides. 79 


from the deed,! the pretext from the motive,? the specious 
or plausible from the actual or true. Readiness to con- 
ceive suspicions and quickness to anticipate an injury are 
throughout regarded as marks of a superior intelligence.* 
But as Plato says, the cleverness of the over-suspicious 
man is really a low superficial cunning. He quite loses his 
bearings in the society of large, true natures. The really 
difficult thing is to discriminate, to know when to trust 
and when to distrust. For unreasonable suspicion is as 








stupid _as_ naive credulity. Moreover, as Thucydides ob- 
serves, universal distrust overreaches itself. The unhappy 
Greeks of this age had become so perfect in this fatal logic 
of suspicion that they could find a flaw in any argument 
that promised assurance of security in another’s pledges, 
and so being unable to confide were compelled to forestall.® 
In these contests brutal dullards who from self-distrust 
struck at once, got the better of the finer wits who, relying 
on the ingenuity of their combinations, contemptuously 
bided their time.’ This self-defeat of the power of the 








1 Passim. 

21. 23 rhv adnbecrarny mpbpacw ddavecrarny 5é byw; III. 86 rpopdce .. 
Bovrduevn; VI. 6. 1; VI. 33 mpdpacw ... 7d 5é ddnOés; VI. 76 rpopdce — 
diavolg. 

3 E.g. mpdoxnua I. 96; III. 82. 4; V. 30. 2; I. 37. 4 7d edmperés Aorovdor 

. . mpoBéBrAnvra; VIII. 66 Hv Todro edrperés mpds Tovs whelouvs; I. 39. 2; III. 
38. 2, etc. Cf. in Tacitus the use of obtentui; sub obtentu; praetendere; ob- 
tegere; ut ea specie; specie honoris; speciosa verbis; re inania aut subdola, and 
similar expressions. 

4 IIL. 82. § xal brovojoas ere Secvdrepos; cf. III. 43. 3. 

5 Republic, 409 B C D. 

6 III. 83. The sentence here paraphrased has been strangely misunderstood 
by Classen and others through failure to appreciate the Greek point of view. An 
argument or a speech is a combat of wit in which he who fails to convince is 
worsted. Cf. III. 37 rév re det Neyoudvwy és 7d Kowwdv weprylyverOar; III. 42 
6 uh weloas dévverwrepos Av Sééas elvar, etc.; Plato Phaedr. 272 B 6 wh weOduevos 
kparet. He who says, “I don’t believe you,” has the better of him (Jowett). 
Now everybody at Corcyra was superior in his reasoning to any considerations that 
held out hope of security; which in Thucydides’ implicit manner becomes “ was 
superior in argument for (to, towards) the hopelessness of security.” Cf. further 
IV. 108 Aoyicug abroxpdrop: SwOetoGar; Il]. 11 7d 5é dyriwadov déos pdvov 
miro, etc. Cf. the complaint of Diodotus, III. 43. 3, of Athenian ee 

T III. 83. 35 Ill. 37. 4. Cf. La Rochefoucauld, maxim 129. 


s 


80 Paul Shorey. [1893. 


intellect suggests its abdication, and so, as an alternative 
to the dominant Ionian ideal, Thucydides depicts for us 
that of Sparta, — self-restraint in place of expansion, dis- 
cipline and caution rather than the free play of the intelli- 
gence: “We are not cunning in useless matters. We think 
the wits of our neighbors as keen as our own. We do 
not expect by ratiocination to forecast the caprices of for- 
tune. We do not base our hopes on the blunders of our 
opponents. We hold that man does not differ much from 
man, and that he is best who is trained in the severest 
school.”! This is the Spartan theory of practice. Thucy- 
dides pronounces no judgment. Truly, as Pindar says: 
todTo 8 apdyavov evpeivy 6 te viv év Kal TedXevTa PéptaTtov 
' avdpt tuxeiv. But the imperishable interest of the history 
lies chiefly in its incomparably vivid presentation of the 
struggle between these two conflicting ideals of human life.? 

To this prevailing intellectualism it would be possible by 
the exercise of a little ingenuity to trace the special minor 
characteristics of Thucydidean style and idiom, carefully 
noted by critics and editors from Dionysius to Classen 
and Jebb. E.g. the archaic poetical diction, the bold meta- 
phor,? the abuse of antithesis,t otiose periphrases,® and 
pointed pedantic discrimination of synonyms® the loose 


1]. 84 freely paraphrased; cf. III. 37, 38. 

2Cf. I. 70; I. 84; II. 39, 40, 46. 

3 To few examples in Blass Att. Bered. 12, Pp. 211, add vrovior adrovoulay VIII. 
64; puyds Te yap elu THs TOv ekeXacdvTwy wovnpias VI. 92; mpocelovres PbBov 
VI. 86; Il. 53 rhv 75n Kareyngiruérvny chav éemixpeuacOjva. Cf. III. 40 and 
1.18. VI. 18 rv rode... tplverOal re abrhy wepl aithy wowep cai Gddo TH; 
VI. 36. 2 dws TE KowG PbBw Tov opérepov éerndrvydiwrrat; VI. 41 ols 6 wédemos 
ayadrerar; VI. 18. 3 Kal obk otuv juivy TapsederOar és bcov Bovdbucba dpxev. 

* \oyw—Epyw; olxetos— dddérpuos II. 39; III. 13. 5; I. 78; IV. 98. 3; I. 
70. 6; t6ov— kody III. 14; Il. 43. 2; doxodoa . . . patvouévn I. 32; paxpay— 
éyyiGev III. 13; III. 64. 5; IV. 36. 2; IV. 61. 3, 8; VI. 76. 2, 3; III. 38. 4. 

5 Dionysius De Thucyd. 29, 32. Cf. the mannerism of 7a wap tudr I. 69; Ta 
Tv wodewv III. 82; ra Tis dépyfs II. 60; 7a Tod wodeuov V. 86; ra THs TUXNS 
IV. 55; 74 awd rod xaracrpwyaros VII. 70. 

6 I. 69. 6; I. 122. 4; II. 62. 3; ILI. 39. 2; IIL. 82. 4; VI. 76.3; Blass I. 219. 
Dionysius De Thucyd. Judic. 46” re r&v dvoudrwr ftyynows dupbrepoy cogutTixh 
kal dmeipoxanos. 


Vol. xxiv.] Implicit Ethics and Psychology of Thucydides. 81 


anacoluthic structure conforming rather to the implicit logic 
of association than to the explicit logic of formal grammar ;1 
the vrocxvAia or wanton variation of the syntax of functionally 
parallel clauses and adverbial phrases ;? the fondness for 
litotes and suggestive pregnant uses especially of the ad- 
verb,® the passionate desire, as it has been put, to compress 
a book into a chapter, a chapter into a paragraph, and a. 
paragraph into a sentence. Also the deliberate preference 
for the abstract generalizing vague expression over the con- 
crete definite, and the forcing of Greek idiom in this direc- 
tion, as illustrated by the quasi-philosophical use of idéa,° by 
the substitution of abstract nouns or neuter adjectives and 
participles for verbal forms of expression,® by the gener- 
alizing use of the neuter participle or adjective,’ and the 


1 See the list in Boehme’s index, s.v. Anakoluth, Accommodation, Erginzung, 
Subjekt, Uebergang, Verschmelzung, Wechsel, etc. Note especially the use of 
tovro Spay I. 5.2; II. 49.5; IV. 19.4; adrd dpav = rodeuetv IV. 59. 2; I. 31. 2; 
III. 10. 6; V. 49.4; VI. 83. 1 = vaurixdy Krd. mrapexduevo, 

2 Blass I. p. 215; I. 2; I. 9.1; I. 49. 3; I. 82. 1 mare wédenoy dyav Sndodr- 
tas un’ ws émirpépouev; Il. 27.2 katd re Td... sat dita kal 8rt; V. 9. 4, 6; 
VITS2.:23 NITA57. ¥. 

8 As Shaks. Cor. I. 1, what he hath done famously. Ct IV. 100. 4 lotca ore- 
yards, “through a covered passage”; I. 92 dverixAjrws; II. 64. 2 dvaykalws, 
“with close-lipped patience for our only friend”; II. 65. 8 éXevdépws; III. 40. 1 
dvOporlvws; III. 40. 4 kuupdpws; III. 56. 7 kepdadréws; IV. 18. 4; olrives 
Tayaba és duplBorov dopards eGevro; IV. 62. 3, 4 xaderds, Sixalws; VI. 11. 3 
éxelyws; cf. I. 77 and III. 46; V. 91 xpnoluws; VI. 87. 5 dmrpayysvws. The. 
‘grammatical indexes ignore this usage, and give only commonplace instances of 
és with implied motion. . Thucydides’ use of litotes hardly needs illustration. Cf. 
ox Hooov, obx nKicra, odk éddocous, obk dtdveros, ov5é 4dUvarTos. 

47d wepacOa df édaxlotrwr dvoudrwv miere onualvey mpdyuara Dionysius 
De iis quae  Thucyd. 2. 

5 E.g. 9 abr lida éxetvd te Zoxov cal ra évOdde viv repdvra VI. 76. 3. 
Cf. III. 62. 2; and eidos in VI. 77. 2, etc. 

6 Cf. Blass I. p. 213, Dionysius De iis quae Thucyd. propria sunt V., Classen 
Einleitung LXXX. A good example is II. 64 doris 5° ért peyloros 7d éripbovor 
AauBdver, where besides the avoidance of the passive form’ we gain the thought 
that it is in any case impossible to escape envy, and therefore the sage will choose 
to be envied for something worth having. Cf. supra, p. 3; cf. also riv obkére 
éravaywyv VII. 34 and similar expressions III. 95, V. 35, and V. 50. 

7 Blass I. p. 214, Classen Einl. LXXX. The neuter undoubtedly does, as 
Classen says, give a body to the abstraction, but the natural Greek would in the 
majority of instances have avoided the abstract form altogether. 


82 Paul Shorey. [1893. 


use of a generalizing personal relative clause in loose appo- 
sitional exegesis of a preceding generalization, expressed or 
implied.? 

It would, I say, be a very interesting but somewhat fanci- 
ful undertaking to trace these minor traits of style to their 
source in the dominant qualities of Thucydides’ mind. If 
we followed the lead of Dionysius, we should account for 
most of them by the writer’s conscious desire to display 
his own ingenuity and startle and subjugate his reader’s 
intelligence.2, The more approved modern view is that in 
Jebb’s words, “we see a vigorous mind in the very act of 
struggling to mould a language of magnificent but imma- 
ture capabilities.” To this view we all except Mr. Mahaffy 
incline. But I think few of us can read Dionysius’ analysis 
of the Corcyra passage or of the Melian dialogue with- 
out being shaken in our faith. mepiméppacta: mpos ovdév 
avayxaitov is his illuminating condemnatory phrase. Do the 
periphrases and the contortions of structure and the affected 
nicety in the employment of synonyms add anything to 
the real weight of the thought? Do they result from the 
struggle of a powerful intelligence with an unformed idiom, 
or are they added, for general literary gorgeousness (as 
Mark Twain would say), by a conscious and perverse art? 
There is enough truth certainly in the disparaging view 
to make all who have struggled with Thucydides enjoy 
Dionysius’ amusing account of how “he spent the whole 
twenty-seven years of the war in ‘upsetting’ the style of 
those eight books and filing and polishing each one of 
his parts of speech; now expanding a word into a phrase, 
and now condensing a phrase to a word, and at one time 
expressing a verbal idea by a substantive, and again turning 


1TI. 44.1 7d & edruxés, of Av, etc.; Il. 62. 4 xaradpévnors dé Os Gy; VI. 14 
kal Td Kah@s dptac Toor’ elvac Os Av; VII. 68.1; VI. 16.3; IV. 18. 4 cwppdrwv 
5¢ dvdpG@y olrives, etc.; III.45. 7; V.16.1. Similar is the use of the relative with 
ellipsis to motivate or expand a preceding suggestion. I. 40.2; I. 68.3; I. 82.1; 
TIT. §5. 3; Ill. 39; VI.-61. 13. [Vi a6migeede, gq. 2; I. gg. 2; IV.-92. 2; “VI. 
68.1. Cf. Jebb on O.C. 263. 

2 éwirernieuxas . . . twa diadddéy Tods Addovs cvyypadgets, Dionysius De 
Thucyd. 51. 


Vol. xxiv.] Jmplicit Ethics and Psychology of Thucydides. 83 


the substantive into a verb; and perverting their use so 
as to make appellatives of names, and names of appella- 
tives; active verbs of passive, and passive of active; and 
interchanging singular and plural, and predicating mascu- 
lines, feminines, and neuters of each other to the utter con- 
founding of the natural sequence of the thought” (De 
Thucyd. Judic. 24). 

Quite apart from the contortions of the style, the sympa- 
thetic student experiences a sense of strain in reading 
Thucydides. The actors in the drama never relax the ten- 
sion of their intellectual faculties.1 We are in a world of 
analysis and logical relations in which nothing, to borrow 
Professor James’ phrase, is given over to the “effortless 
custody of habit.” We are constantly called upon to weigh 
evidence, balance probabilities, divine motives, and to com- 
pare or contrast human characteristics and faculties, national, 
typical, or individual? We are required to forecast the 
probabilities of the proverbially uncertain future in the light 
of the entire record of the past at every crisis of the action, 
and, whenever the power of God or fortune makes fore- 
cast foresworn, as Pindar hath it, we are expected to feel 
a shock of surprise at the illogical logic of events and the 
paradoxes of human life.? evAoyos, dAoyos, eixés and their 
synonyms and paronyms occur on every page. The chief 
concern of every speaker is to show that his own course 
of action, whatever the actual event, was logical, plausible, 


17d... dvrirerdx Oar dddAfrAs TH veep III. 83. 1. 

2 TA 90e area; 1. 14t; 1..142; IL. 87. 4; UL: 89; Tl. 37.23 TIL. 56. 3 
IV. 10.5; V.9.1; V.14; V.16; IV. 40; V. 75; V. 105. 3,4; VI. 11; VI. 17. 
13° VL. 183 VL 68.35. VI. 69, 3;- VI.. 72:3; VI. 773. VI. Sax: 33) VIL. gegen. 
14. 2, 4; VII. 48. 4; VII. 34. 7; VII. 61-64; VII. 66-68; VII. 71; VIII. 96. 
Observe the frequency of dvrimanos, lodpporos, dyx panos, dvririOévar, and other 
compounds of dyrl, as dvremiBovdedoa, dvTiveddfoa, and the weighing in the 
argumentative balance of éumecpia and uedérn and dice and didax7p and edbpuxla 
and édveo.s and duaGla. Ionian and Dorian, land power and naval power, etc., etc. 

SEIT. 3203; EV. 22.35 TT. 65. 12, 13; IV. 39.33 IV. 403, VIL) 42.,2;, VIE. 
28; VII. 34. 7; VII. 55; VII. 62. 4 wefouaxety dd ray vedv; VIII. 25. 5; 
VIII. 66. 5. Observe also the frequency of rapa dtvauy, rapa yvdunv, rapa 
dbtav, mapa byov, wodds 6 mapddoyos, TH adoKkATy, és Tobvavtloy—mepiécry I. 
120. 5; és Todro mepiéorn 7 TUX IV. 12. 3, etc. 


84 Paul Shorey. [1893. 


consistent,! and when formal disputation ceases, men argue 
still in the forum of their own minds, and abstractions are 
personified to continue the debate.” 

Thucydides himself, in one of the few passages where he 
betrays a personal interest, goes out of his way to defend 
at length the cwdpocvvn, that is, the good judgment of the 
Chians in their treacherous revolt from Athens (VIII. 25). 
et 6€ Te ev Tois avOpwrreiows Tod Biov Tapaddyos éopadrynoav 
—why their error, shared by the best minds of the time, 
was quite excusable. Similar is the feeling underlying his 
eager defence of the justice of Pericles’ forecasts of the 
future.® 

In conclusion, it would be an interesting if elusive inquiry, 
to ask how much of this disputatious, analytic, antithetic, 
cynical manner was due to the fashion of the new rhetorical 
dialectic, how much to the disintegration of popular morality 
under the stress of war, how much is the real expression 
of the mind and heart of Thucydides. The rhetoric of the 
time was responsible for much. It is impossible to accept 
Jevon’s critical dictum that Thucydides is no stylist, but 
rather a perpetual demonstration that there is a higher art 
than that of concealing art—the art of dispensing with it. 
And there are many exceptions to be taken to Jebb’s state- 
ment that the student of Thucydides always has the con- 
solation of knowing that he is not engaged in the hopeless 
or thankless task of unravelling a mere rhetorical tangle. 
Thucydides is doubtless rich in ideas—@omep é« mnyijs 


1 J. 32. 3; IV. 87. 3; V. 104; V. 105. 4; VI. 79. 2; VI. 85 dvdpt 5¢ rupdvvw 
9 Tore apxhy éxovoy ovder Adoyov & re Evudépor. 

2 VIII. 24 wal rods "A@nvalovs yoOdvovro 08 abrods Rerckt yeores . 2. @s Ov 
ravu wévnpa opdv BeBalws ra rpdyuara ely; Il. 87. 3 obd€ Sixarov THs yvouns Td 
uh Kara kpdros wuxnOév,éxov dé twa év aire arridoylay. Cf. the Euripidean 
subtlety of cal yap 6 uh pnOels Abyos rots GI Fxoverw airlay av wrapdoxa ws el 
€XEXOn cwrhpws av Hy III. 53. Cf. Eurip. Suppl. 298 otro: cwrds’ efra péupouat 
wore | Thy viv cwrhy ws éovyhOn xaxds. Cf. 1.140 phd év dyir abrois alrlay 
brodimnode, etc. Cf. also the curious subtlety of VII. 66.3; I. 36.1; VI. 78. 1-3; 
VI. 79; V. 86; IV. 92. 2; VIL 34. 73°4V. §5. 33: IL. 64. 6; V. 90; II. 8 4, 
repeated IV. 14. 2. 

3 II. 65. 








Vol. xxiv.] Jmplicit Ethics and Psychology of Thucydides. 85 


mrovcias atreipoy TL ypHpa vonudtwy Kal. évOvpnpdtwr, says 
Dionysius. But the expression is almost always elaborately 
tortured for effect.: Often what we take for a new substan- 
tive thought is merely an ingenious variation on a common- 
place theme. Often periphrases that are apparently wrapped 
around a kernel of profound suggestion are found empty 
when unfolded. Irrelevant distinctions of synonyms abound. 
In place of real antithesis we are frequently put off with 
the verbal form of an antithesis,! and speech after speech is 
wound up with an aphorism that under scrutiny turns out to 
be a pompous truism.? More matter with less art, we cry. 
And these faults, to judge by almost the only strictly con- 
temporaneous writer of prose, — Antiphon, — were character- 
istics of the formal rhetoric of the time slightly exaggerated 
by Thucydides. Dionysius cites Antiphon, together with 
Lysias and Andocides, to prove that Thucydides’ style was 
not that of his contemporaries. But this is very undis- 
criminating criticism. Antiphon can narrate the murder 
of Herodes as simply and lucidly as Thucydides the attack 
on Plataea. But when he puts on the buskins of formal 
argumentation, we can hardly distinguish his gait from that 
of the historian? . 
Still more difficult is it to apportion the responsibility 
for the cynicism of the history between the historian and 
his time. The theme was certainly disheartening enough. 
A writer would need great nafveté or the support of a 
transcendental faith in order to retain any moral illusions 
while chronicling the affairs of Melos, Plataea, and Corcyra, 
the butcheries of Mycalessus, Mitylene, Scione (IV. 122, 
V. 32), the treacherous murder of the Spartan Helots (IV. 
80), the trick of the Sicilians at Egesta, the impudent 


1 A notable instance in II. 42 r&vde 5¢ ovre rhovTy Tis Thy Erc Gwbdavew Tpo- 
riphoas éuadaxicOn ovre revias édmldt, Ws kav Ere Seag@uywr abrhy rhovrhoesey. 

21. 34. 3 6 yap é\axlioras ras perapedelas éx rot xapiferGac Trois évavrios 
AapuBdvwy acopadécraros ay diaredoin; III. 48. 2 dots yap eb Bovdeverar mpds 
tovs évavtious xpeloowy écriv pet Epywv loxvos avolg éri6v; V.9.4; III. 30. 4; 
V. 111. 4; Il. 11.9; IL. 64.6; VI. 14 xal 7d xad@s Gptac rotr’ elvac Os av Thy 
warplia wpedhoyn ws rreicra 9 éxwy elvar undery BrAdyp. 

3 Tetral. I. T 3; Herod. 7, 73, 92-93, 84. 


86 Paul Shorey. [1893. 


knaveries of Alcibiades, the clever stratagem of Phrynichus 
(VIII. 50), the negotiations between the Peace of Nicias 
and the battle of Mantinea, the machinations of the revolu- 
tionary party of 411, and the various minor treasons and 
atrocities that darken these pages.!. And there is little evi- 
dence of any such triumphant faith in Thucydides. Classen, 
it is true, brackets him with Aristophanes as a high-minded 
castigator of the immorality of his age, and extracts a wealth 
of moral and religious truth from his unimpassioned narra- 
tive. But the more critical Jebb is obliged to put a great 
strain on the text in order to discover one or two edifying 
aphorisms, such as that justice is the common good and 
is identical with true self-interest ;? or that we ought to 
receive the inscrutable dispensations of heaven with resig- 
nation ;° and is at last forced to fall back on the oft-quoted 
sentence about simple-mindedness and true nobility, and the 
two-edged argument of the “naked repulsiveness in which 
he exhibits the right of might.” We cannot, it is true, 
attribute to Thucydides himself all the cynicism of the 
Thebans at Plataea, of the partisans at Corcyra, of the Athen- 
ians at Melos, or the shameless euphemisms of the xandoi 
kaya0oi of the oligarchical party at Athens,* but there is 


1Cf.. eg. TL. 67.45 EL yan; IL 79. 25: TIE 320.3; SHI. 113.6; EV. 23; FV. 
76.2; VI.61; VI. 74; VII. 48. 2; VIII. 93. 3; VIII. 56.2; VIII. 89.3; V. 76; 
I. go; I. 107. 4. ; 

? He cites V. 90 and 1. 41, which should be 1. 42 76 re yap tuudépov ev G apy 
ris €XdxioTa duaprdvy uddora érerat, which will certainly bear no more moral 
meaning than is given it by Crawley’s “ the straightest path is generally the best.” 
The kotvdy dyabdv of V. 90 is not abstract justice, but that reasonable forbearance 
towards the vanquished and the weak of which the Melians warn the Athenians 
that they too may one day stand in need. 

8 II. 64. 2, on which Boehme, Einleitung XVIII., naively remarks: “Es geht 
ein Zug tiefen religidsen Ernstes eben so entscheiden durch das Werk, als dasselbe 
durchweg von echt sittlichem Geiste erfiillt ist.’ Cf. supra, p. 12. - Similarly 
Classen, ed. 1879, Einleitung LVIII. 

* VIII. 47 od rovnpig od5é Snuoxparig; VIII. 53 ef uh wodcrevoouer cwhpore- 
orepov; VIII. 65 Kat dddovus rivds dverernicious . . . kpida avidwoav; VIII. 66 
evOds €x rpdmov rivds éwirndelov reOvixec; VIII. 68 6 pévror dav rd rpayyua 
turOels . . . Avtipde Rv dvhp ... Trav cad éavrdy dperp Te ovdevds Sevrepos; 
VIII. 69 “EdAnves veavicxor ols éxpavro ef ri wov Séou xeipoupyety; VIII. 70 ot 
€ddxouv érirHdecor elvar bretacpeOFvat. 


Vol. xxiv.] Implicit Ethics and Psychology of Thucydides. 87 


little evidence in his writings of any generous indignation 
at them. The declaration that Nicias least deserved to suf- 
fer what he did, on account of his habitual practice of all 
conventional virtue, conveys quite as much irony or sense 
of dramatic contrast as moral affirmation. We learn else- 
where that Nicias was excessively devoted to religiosity, 
and that sort of thing (VII. 50. 4), and there is an intense 
Sophoclean irony in the statement that he had resolved to 
leave behind him, if possible, the name of a man who had 
never brought disaster upon the state, as well as in the 
repeated malicious allusions to his good fortune? Vive 
pilus moriere pius seems to be the moral. Thucydides 
merely chronicles, he does not himself indorse, the pious 
attribution by the Spartans of their failure in the first period 
of the war to their violation of their treaty obligations.* 

The impression made by the whole history is that the 
writer’s mind was subdued by what it worked in. Only once 
or twice does he let fall a word of pity, as pe{wm 7} Kata 
Saxpva, of the sufferings of the Athenians at Syracuse 
(VII. 75), or ovSevds . . . Hocov drodvpacba: akiw of the 
butchery at Mycalessus (VII. 30). Elsewhere the repressed 
feeling finds vent in such loaded and contorted phraseology 
as KaTa TavTa yap TavTas viKnOévtes Kal ovdév Odiyor és 
ovdev xaxorrabncavtes TravwreOpia 8) TO NEeyouevov Kai Telos 
Kal vies Kal ovdev 6 TL ovK amr@deTo (VII. 87); or maod Te 
isa xatéotn Oavdtov Kal. . . ovdev 6 Te ov EvvéBn Kal Ere 
mepattépw (3. 81). Sometimes, also, feeling is displayed by 
the brief pregnant suggestion of a startling dramatic con- 
trast: eg. ta pev cata Llavoaviay tov Aaxedarpoviov Kai 


1 VIL. 86 8:a Thy racay és dperhy vevomouérvnv éxirjdevow. That racav and 
vevouopévnv go with émirtdevorv does not affect the sense, which is not “ der 
Gewissenhaftigheit gemiss eingerichtet” (Boehme), nor “das durch Gesetz und 
Herkommen geregelte Streben nach dem Edeln” (Classen), nor quite “ he lived 
in the practice of every virtue” (Jowett), nor precisely “ his exact attention to 
every religious duty” (Crawley). 

2V.16. VI. 17 wat 6 Nixlas edruxhs Soxet elvar. Cf. VI. 23. 3 dre Adora 
TH TUXD Tapadods EuauvTdy Bovouac Ex ei. 

8 VII. 18. On Thucydides’ attitude towards the religious opinions of his time, 
see two good pages in Jevons’ History of Greek Literature, 336, 337. 


88 Paul Shorey. [1893. 


Qewicroxréa Tov “AOnvaiov AXapmpoTtatous yevouevous TaV Kal” 
éavtovs ‘EdXAjvev odtas érereUTycev (I. 138), and Ta wer KaTa 
THY peyadnv otpateiav ’A@nvaiwv xal tov Evppdyov és 
Aiyurtov otras érerevtnoev (I. 110), or, best of all, the 
allusion in the midst of the horrors of the break up of 
the camp before Syracuse to the magnificent description 
of the splendors of the embarkation at Athens... . do 
olas NapTrpeTNTOS Kai avYNpaTOS TOD TPHTOU &s Olay TEAEUTIY 
Kal tatrewvoTnta adgixto.! Even when his own feelings are 
most strongly enlisted, the expression of them is checked 
and embarrassed by his deep-seated fear of the spirit of 
blague, as tyrannous in ancient Athens as in modern Paris. 
His contempt of sentimental expansion (uaxpnyopeiv év 
e(ddo1, etc.) returns upon himself and destroys the sources 
of genuine feeling. 4) madatds apetas el Tis dpa Kai éyiveTo 
axovovtes émixracOnre the Thebans say with a cold sneer 
to the Spartans sitting in judgment on the men of Plataea. 
We make no fine speeches about our merits as the over- 
throwers of barbarians, the Athenian envoys protest at 
Camarina. The habit-of utterances like these makes it im- 
possible for Thucydides to relieve his feelings by free expan- 
sion of Nicias’ last words in the supreme crisis of Athens. 
The smile of an imagined cynical reader? stays his pen, and 
in place of what might have been the most moving speech 
in the history, we have the cold, indirect report: a\Xa Te 
Aéyov bca ev TH ToLvovT@ Hdn Tod Kaipod dvtes avOpwrrot ov 
mpos TO Soxelvy Tin apyaoroyeiv®? gurakdpevor elrrovev av: 
“With other remarks that at such a crisis men would not 
spare from fear of seeming to fall into old-style sentimental 
commonplace ’’’ —the most pathetic words in the entire eight 
books when interpreted in the light of the spiritual history 
of the time and the writer. 


1VII. 75; cf. VI. 32. Cf. also III. 113. 

2 Cf. I. 73. 2 ra 5@ Mydcxd . . . ef kal d¢ Bxdov padrov Ecrar det rpoBadropé- 
vous. 

3 Cf. Isocrates’ use of dpxata, Orat. III. 26, and IV. 30. 


Vol. xxiv.] Attraction in English. 89 


VI.— English Words which hav Gaind or Lost an Initial 
Consonant by Attraction. 


SECOND PAPER. 


By CHARLES P. G. SCOTT. 


In a previous paper, publisht in the Transactions for 
1892, pp. 179-305, dealing with English words which hav 
gaind or lost an initial consonant by Attraction, I set forth 
the largest class (§ I), namely, those involving the gain or 
loss of initial 2, in words preceded by (I) the article az or a, 
(II) the dativ article then, (III) the possessiv mzne or thine, 
(IV) the negativ zone, (V) the conjunction az, (VI) the 
preposition zz (an, on), and (VII) inflexiv z. . 

In this paper I take up the next largest classes, with some 
smaller, dealing first with those in which particles ar con- 
cernd (articles and pronouns), as a sequence to the similar 
classes before treated, and then with the rest in the alphabetic 
order of the consonant affected. The same abbreviations ar 
used as in the former paper. 


§ II. Initial T gaind. 


VIII. Cases involving the article that. The final ¢ of the 
article chat, also thet, a pronunciation now recognized only in 
dialect (1847 Lowell, Biglow Papers, etc.), but common 
everywhere in the unemphatic use, and often so speld in 
ME., is in some instances attracted to the noun, leaving the 
article in the usual and therefore more stable form (he. 
(Compare the article then and them, dativ, similarly reduced 
to zhe, in the instances given before (TRANSACTIONS, xxiii. 
279-287), and hereafter, § III, p. 108). The cases of attracted 
zin this sort, ar not many, and all but two ar of limited use. 


go Charles P. G. Scott. [1893. 


1. Effigies. That effigies became the feffigies. Effigies was once 
very common, in the sense of ‘likeness,’ ‘picture.’ It came to be 
regarded as a plural, and a new singular effgie, effigy, in dialectal use 
*effige, effiy (from effigies ef'i-jiz taken as *effiges ef'ij-ez) arose. 


(a) And as mine eye doth his effigies witnesse, 
Most truly limn’d, and liuing in your face... 
- 1623 SHAKESPEARE, 4. Y.Z. ii. 7. 193 (F! p. 194). 
Efij. A likeness—a strong likeness. “‘He is the very Zfij of his 
father.” Evidently from the Zfegies [sic] used the century before last 
for picture or portrait. 1823 Moor, Suffolk Words, p. 118. 
(6) The teffigies and counterfait. 1610 Honours Academie, ii.9. (H. p. 856.) 


2. Even, contracted cen. In Scotch use thet een has become éhe 
zeen, this evening. Compare ¢he day, this day, to-day. Compare 
also éeen for at een (IX. 10), and see good een, good den, etc. (XVII). 


(46) But thinks I, chaps, ye’re aff your eggs for ance, gif ye ettle to come on 
us the ’teen at unawares. Saint Patrick, i. 168. (1882 Jam.) 


3. Harbinger. Zhat harbinger makes (the) tarminger. 


(a) Let me alone, for the king’s carminger [read harbinger ?] was here; 
He says the king will be here anon. 
1594 4 Knacke to Knowe a Knave (Dodsley, ed. Hazlitt, vi. 567). 
(6) Tarminger. Warbinger. A corruption. 
1847 HALLIWELL, p. 852 (without reference). 
[This may refer to the above passage. Hiazlitt’s edition is a thing to be 
abhord. } 


4. Hayloft. Zhat or thet hayloft, became the *tayloft, taylot, tal- 
lut, tallat, tallet, tollet; one of the most permanent aberrations of 
this class. For the reduction of the second element /o/¢ to -40/, -/az, 
-Jet, compare the Somerset cocklawt for cockloft (1825 JENNINGS, 


p. 31). 


(a) An haye house or loft: an haye mowe or rieke: a place where hay lieth. 
Feentle. [1580 haze. ] 1573 BARET, Alvearie, H. 15. 
(4) Zaylot. Glocestershire word; meaning an 4ay-/oft. At first, no doubt, 
they said 2% ‘aylot for in the hay-loft; and then converted the whole 

into a substantive, calling a hay-loft by that name. 
1777 Gent. Mag., Aug., p. 373. 

Tallet (i.e. top-loft). A hay-loft. Exm. 1790 GROSE, Prov. Gloss. 

Tollet, a hay-loft. 1804 DuncuMB, Herefordshire Gloss. (E.D.S.), p. 63. 

Tallet,s. The upper room next the roof; used chiefly of out-houses, as 
a hay-¢adlet. 1825 JENNINGS, Somerset Gloss. p. 74. 

Tallet. (Tal, Brit. tall; high.) The upper room next the roof; applied 
chiefly to a stable, as a hay-éallet. Som. Hants. 

Why, you must know that the puggen end of the linney neist to the peg’s- 
looze geed way and was ruseing down: maester was staunding by ¢he 
tallot whan the cob-wall sluer’d away all to wance and made such a sture 
that a come heal’d in brist and grute. 1837 Devonshire Dialogue, p. 3. 

1839 HOLLowAy, Gen. Dict. of Provincialisms. 

Taillit. A hayloft. West. “When the prisoner came in he was watcherd, 

which shewed he had not been all night in the fa//it.” 1847 HALLIWELL, 


Vol. xxiv.] Attraction in English. gI 


Why coud not a “ watcherd”” man hav been all night in the hay- 
loft? Because haylofts ar usually dry, not to say dusty; and one 
comes not out of them, after a night’s sleep, ‘ wet-shod.’ Watcherd 
is for watched (H.), and this is for watshod, dialectal form of wet- 
shod, ME. wetschod, wete-shoed. — 


Wolleward and wete-shoed [wo-werie and wetschod C] went I forth after. 
1377 LANGLAND, Piers- Plowman (B), xviii. I. 
Tallet,s. A hayloft. West. Any upper room with a lath window instead 
of glass. Last. 1857 WRIGHT, p. 942. 
Tallut. The hayloft. 1868 HUNTLEY, Gloss. Cotswold Dialect, p. 65. 
I determined to sleep in “he ¢a/lat awhile, that place being cool and airy, 
and refreshing with the smell of sweet hay. 
1869 BLACKMORE, Lorna Doone, xxxi. 
Tallet. The hayloft over a stable. 
1891 CHOPE, Dial. of Hartland (E.D.S.), p. 115. (Tallat, p. 18.) 


5. Heft. TZhat heft is the source of (the) “ft, and this of ‘the 
verb “7. 


(a) Heft... . An effort, a heaving. 1775 ASH. 

(4) Teft or Heft, v. to judge of the weight of anything by poizing it with, or 
in, the hand. “ 7eff this, wul ye?” See HEFT. 

1825 BRITTON, Beauties of Wiltshire (E.DS.), p. 45. 

Teft. The same as He/2, q.v. 1847 HALLIWELL. 


6. Hovel. That hovel, the(t) hovel, became (the) tovel, or tuffold. 


(a) In, fellow, there, into #h’ Houel; keep thee warm. 
1623 SHAKESPEARE, Lear iii. 4 (F! p. 298). 
(6) Tuffold, or Tovel. This means an hovel in Derbyshire, where they first 
said in fovel, i.e. in the hovel; and then by mistake took fove/ to be 
the substantive, for hovel. 1777 Gent. Mag., Aug., p. 373- 


7. Iron. That ierne cross appears as (the) Herne cross. 


(6) Tierne cross (in Somner’s Antiq. of Canterb., pp. 11, 169), is the irom cross. 
1777 Gent. Mag., Aug., p. 373- 


8. Old, dial. auld, aud, owd, etc., ME. old, ald. That or thet 
old hen and shat or thet old \aw ar subject to the same new law. 


(2) The owd hins cackled in the yard, 
For we forgot to feed ’em! 
1806 BLOOMFIELD, 7he Horkey, in Wild Flowers, p. 35. 


(4) [ The] towd hen, the old hen, was a popular name for the eagle of the 

lectern in Chester Cathedral. 1882 PALMER, Folk-Etym., p. 570. 

(a) Be olde lage. c1230 A Bestiary, \. 293 (in Old Eng. Misc., E.E.T.S., p. 10). 
fe hald{e) (var. e alde, Je olde \aw. 

¢ 1300 Cursor Mundi (Cotton MS.) (E.E.T.S.), 1. 116. 


(4) Forr patt nass nohht onn3zness Crist, 
Patt talde \a3zhe stode. ¢ 1200 ORM, Ormulum, |. 18196. 


92 Charles P. G. Scott. [1893. 


g. One. The word one in its various forms, one, dial. ane, and oa, 
dial. ae, ME. on, oon, an, and 0, 00, a, AS. @n, has run, both alone, 
and with its alternativ ofher, a long course of riot after the shifting 
that, thet, the. That one, thet one, dial. thet ane, ME. that on, oon, 
an, is divided she tone, dial. the toon, the tane, ME. the ton, toon, tan. 
The form with x lost, az or thet 0, or oe, dial. ae, becomes in like 
manner Zhe 20, the toe, dial. the fae, or tea. 

Examples ar innumerable. I giv here only a few of those that 
involv “he(¢) one; other examples of ¢he(¢) one ar given below in 
connection with examples of ¢he(?) other. 


(4) There was nother power ne ryche, - 
Who that beheld hem both, 
Fayrer neuer more ne cowde say, 
That knew ¢he Zoon of the children tway 
Bote be colour of here cloth. . 
¢ 1430 Amis and Amiloun, |. 92 (Weber, Metr. Rom. ii. 373). 
My desteny is for to dey a shamful dethe, I trowe, 
Or ellis to flee — che fon must be, none other wey I knowe. 
¢ 1502 The Nutbrowne Maide (Arnold’s Chronicle, 
repr. 1811, p. 198; Child, Ballads, iv. 146). 
Ther-for he ton of us shall de this day. 
a 1550 Hunting of the Cheviot (Child, Ballads, vii. 32). 
And therefore it was misliked in the Emperor Nero, and thought vncomely 
for him to counterfet Alexander the great, by holding his head a little 
awrie, and neerer toward ¢he tone shoulder, because it was not his owne 
naturall. 1589 [PUTTENHAM], Arte of Eng. Poesie (repr. Arber), p. 302. 


’ So especially in phrases, the(*) one part, the(?) one side, the(t) one 
haf. 


Tapart. Of the one part. 1847 HALLIWELL. 
Tonpart. Of the one part. 1847 HALLIWELL. 


These doutless refer to old passages not quoted. 


Now he setteth his hat on the toe side, and commeth sailing in like a shippe 
in a tempestuous tide. 1609 7he Man in the Moone (Wright, p. 966). 
[This is different from “che fo side, the right hand side” under which 
Wright puts it. Compare a¢o side, a to side, IX. 7.] 
There’s twa o’ them faulded unco square, and sealed at ¢he éae side. 
1816 Scott, Antiguary, xv. 
There is neither wark nor the very fashion or appearance of wark, for 
the tae half of thae puir creatures; that is to say . . . cannot employ 
the one moiety of the population. 1818 Scott, Rob Roy, xxvi. 


to. Other. Zhat or the tother, dial. ither, oor, ME. that, thet 
other, becomes the tother, dial. the tither, the toor, ME. the tother, 
sometimes shat tother. 

(a) Upon fet ofer dai.  ¢ 1258 Meidan Maregrete, \. 221 (E.E.T.S.), p. 40. 


pat oper. 


¢ 1300 Cursor Mundi (E.E.T.S.), 1. 83. (See full quotation below.) 


Vol. xxiv.] Attraction in English. 93 


Thorgh me men gon, than spak “hat othir [the other, Wright] side, 
Unto the mortal strokis of the spere. “_ 
¢ 1374 CHAUCER, Parl. of Foules, |. 134. 
(4) Swanborow his sister, Helfled the tother. a 1300 Havelok the Dane, \. 411. 
: For in pis loue scho failes neuer, 
And in Jat toper [var. Je toher, 2 mss., Sat oper, t mss.] scho lasts euer. 
¢ 1300 Cursor Mundi (Cotton MS.) (E.E.T.S.), 1. 83. 


The tother thinge that we may se. ¢ 1325 Eng. Metr. Hom., p. 11. 
We wil sle be giltif and late Je tother [var. the other, fe other, pe oper, pe 

opere| goo. ¢ 1400 Gamelyn (Six-Text), 1. 822. 
On Ze todir syde. ¢ 1430 Yerk Plays, xiii. 1. 51 (p. 104). 


The ¢ofer was cledde in a cote alle of clene siluer. 
¢ 1440 Morte Arthure (E.E.T.S. 1865), 1. 3335. 
(Other examples 1. 233, 2208, 2530, 3855, etc.). 
A’ the tothar syde. 
a 1550 The Hunting of the Cheviot (Child, Ballads, vii. 38). 
Ouir man, said scho, vnto the tother two. 
1552 LYNDEsAY, Testament of the Papyngo (E.E.T.S), |. 1182. 


The phrases ¢he(¢) one, the(t) other, ar most common in associa- 
tion, as opposits or alternativs: shat, thet one (0, ae, etc.) . . . that, 
thet other (ither, etc.), appearing also as the fone (to, fae, etc.) ... 
the tother (tither, etc.). To show how the use shifted, I arrange my 
quotations this time in mere chronological order, without classification. 


(a) (6) He spused Jat an [var. Je tan, fe toon), Nachor Je tofer. 
¢ 1300 Cursor Mundi (Cotton ms.) (E.E.T.S.), 1. 233. 


That nolde spare for kin ‘hat o kosin that other, 
So the fend hem prokede uch man to mourdren other. 
¢ 1312 Poem on the Times of Edward II. (Camden Soc. 1839), p. 343. 
The tan was man, the tother wif. ' 
¢ 1325 Eng. Metr. Hom., ed. Small, p. 156. 
Per is an eddre pet is y-hote ine latin aspis, bet is of zuiche kende bet hi 
stoppeb Set om eare mid erpe and Jet ofer mid hare tayle bet hi ne 
yhere bane charmere. 1340 MICHEL (tr.), Ayendite (E.E.T.S.), p. 257. 
Sem sobly Jat on, Jat oper hy3t Cam. ° 
 € 1360 Cleanness, |. 299 (Early Eng. Allit. Poems, E.E.T.S., p. 45). 
pire both are hydir brought, 
fe tone Moyses, Je todir Ely. _¢ 1430 York Plays, xxiii. 1. 137 (p. 189). 
And seyd to him, “ Mi leue brother, 
Kepe thou ¢hat on, and Y that other.” 
¢ 1430 Amis and Amiloun, |. 319 (Weber, Metr. Rom. ii. 382). 
So grete a multitude that they coverde the ysz fra the faa bank to the 
tother. a 1500 (?) A/S. Lincoln A,, i. 171 f. 19. (H. p. 844. 
This indentur made betweii Johfi Bolle thelder armerer, and J. Bolle the 
yonger grocer citezens of London, of “hat one partye, and Johfi de 
Castro . . . om that other partye ... 
¢ 1502 Arnold’s Chron. (1811), p. 111. 
There ben two dyfference of perspectyves, the one is pure, separate of 
erthlynesse, and the ¢other is spotted by the same and myxed. 
¢ 1532 Dewes, /ntroductorie for to lerne French (1852), p. 920. 
That xxx of the principall men of “e Za clan sal cum with othir xxx of 
the tothir clan. 
1536 BELLENDEN, tr. Boece, Hist. (in Scott, Fair Maid of Perth, Note 2 F). 
He winkth with fhe fone eie, and lokth with she éother. 
1562 Heywoon, Proverbs and Epigrams (Spenser Soc.), p. 33- 


94 Charles P. G. Scott. [1893. 


That with spitefull obrayds and uncharitable chaffings alweiz they freat, 
az far az any whear “he fom can héer, see, or smell the toother: and 
indeed at vtter deadly fohod. 

1575 LANEHAM, Letter from Kenilworth (N.S.S. 1890), p. 17. 

Armathor ath to the side [read at the to stde| O a most dainty man... . 

And his Page atother [read at other] side, that handfull of wit. 

1623 SHAKESPEARE, Z.Z.Z. iii, 1. (F! p. 131.) 

Tak 3e the taine, and I the fother. 

1602 LyNDESAY, Thrie Estaits (E.E.T.S.), 1. 2214. 
The tane was buried in Maries Kirk 
And the tither in Marie’s quire; 
Out of the tane there grew a birk, 
And “he tither a bonny brier. 
@ 1824 Fair Fanet (Child, Ballads, ii. 92). 

[Similar stanza in Sweet Willie and Fair Annie, ii. 139. J 

The ghaist gae Rab a kick wi’ the tae foot, and a kick wi’ the tother. 

1816 SCoTT, Aztiguary, ix 

The tane gies up a bit, and she tither gies up a bit. 

1818 Scott, Rob Roy, xiv. 

By the grace of Mercy, the horse swarved round, and I fell aff at the tae 
side as the ball whistled by at che tither. 

1819 Scott, Bride of Lammermoor, xxiv. 


In many cases the / disappears from one or the other of the two 
terms, leaving simply the one opposed to the tother, or the other 
opposed to ¢he zone. And ¢he may disappear, or may not hav been 
used, before one. 


ene enne hi honge in one half, for to don him teone, 
And on bi Jat ofer half and ihesuc heom betweone. 
a 1250 Passion of our Lord, \. 439 (Old Eng. Misc., E.E.T.S., p. 49). 
Be oone ys heuy and rede, Ze tofer is li3t and no3t bittere. 
@1350(?) ms. (cited by Way, Prompt. Parv., p. 94, note). 
He schal hate ooe and loue ¢he tothir. ¢ 1382 WIcLIF, Luke xvi. 13. 
The oon halfe of the sayd forfeiture to be unto the kyng our souereyn 
lorde, and that other halfe to be unto hym or theym of his subgettis . . . 
[etc.]. 1489 Statutes of Henry VII. (Caxton, facsim.), p. 12. 
Sir Gawaine tooke the lady by ¢he ove arme, 
Sir Kay tooke her by Zhe éother. 
a 1650 Marriage of Sir Gawaine (Child, Ballads, i. 38). 
You Glasgow tradesfolks hae naething to do but to gang frae the tae end 
o’ the west of Scotland to the tther. 1818 Scott, Rob Roy, xxviii. 
It’s a common case — the ae half of the warld thinks te tither daft. 
1824 Scott, Redgauntlet, viii. 


Then the 7 disappears from both terms, leaving she one, the other. 
This is the now establisht usage. 


The one wes callit the tre of lyfe; 
The other tre began our stryfe. 
1552 LyNDEsAY, Zhe Monarche (E.E.T.S.), 1. 743. 


The ane has taen him by the head, 
The z¢her by the feet. 

a 1803 Lord William (Child, Ballads, iii. 21). 
Said, “‘ Saw ye ever a fitter match 


Betwixt the ane and ither ?” 
@ 1827 Earl Richard (Child, Ballads, iii. 276). 


Vol. xxiv.] Attraction in English. 95 


The initial ¢is actually «gaind’ in ‘one and “other when it is used 
without “te. Zone so used is confined in dialectal use; ‘other is 
common also in colloquial use. 


Had not the Angell thither directed the Shepheards; had not the Star 
thither pointed the Magi, neither Zone nor ¢othir would ever there have 
sought Him. a@ 1626 Bp. ANDREWEs, Sermons, fol. p. 110. (P. p. 589.) 

Sayes one to tother, what quoine hast? 

1611 ROWLANDS, Knave of Clubbs. (Wr. p. 778.) 
From One [house] he dates his Foreign Letters 
Sends out his Goods, and duns his Debtors: 
In fother, at his Hours of Leisure, 
He smokes his Pipe and takes his Pleasure. 
1733 Prior, A/ma (Poems, 5th. ed., i. 93). 

Ton and Tother ; as, do you take on, and I'll take ‘other ; meaning the 
one and the other. 1777 Gent. Mag., Aug., p. 373- 

Tuther, pron. The other. Zutheram, Tuthermy, pron. The others. 

1825 JENNINGS, Somerset Gloss., p. 77. 

Toor. Tother; the other.... TZother. The other. 1847 HALLIWELL. 

Tutheram. The others. West. 1847 HALLIWELL. 

Tarrawan [tother one: Negro-English in West Indies]. 

1848 HARRISON, Zng. Lane: See: p- 117. 

Ton tother. One another. Derd. 


The change is complete when “he neither appears nor can be sup- 
plied before Zone or sother, as when a different particle, as ¢his, my, 
his, etc., is used. 

This same éother thing. . 
1599 PoRTER, 7wo Angry Women of Abington. (Hazlitt, vii. 328.) 
My ae best son is deid and gane, 


And my tother ane I'll ne’er see. 
The twa Brothers (Child, Ballads, ii. 356). 


But I’se gang hame, and finish the grave in the tuning o’ a fiddle-string, 
lay by my spade, and then get my ‘other bread-winner [his fiddle], and 
awa to your folk, and see if they hae better lugs than their masters. 

1819 Scott, Bride of Lammermoor, xxiv. 


It is to be noted that ¢he one and she other wer early contracted to 
thone . . . thother, \ater printed ¢h’one, th’other. 
The kynge to haue ¢hone halfe of euery of the sayed forfeytures And the 
partie that wylle sue ¢hother halfe of the same. 
1489 Statutes Hen. VII. (Caxton, facsim.), p. 28. [Similarly on p. 29.] 
As ye may see in Chaucer and Lidgate, “#’one writing the loues of Troy- 


lus and Cresseida, ¢/’other of the fall of Princes. 
1589 [PUTTENHAM], Arte of Eng. Poeste (repr. Arber), p. 80. 


As the article she, th’, is in some dialects pronounced # and 7, 
‘and combines with the noun, /o¢her may be in some instances for #h’ 
other instead of (the) sother. 

The ¢hat in these phrases is not to be taken as the strong demon- 
strativ, but simply as the article, the neuter article hat, thet coming 


96 Charles P. G. Scott. [1893. 


to be used concurrently with “Ze as masculin and feminin also— 
apparently first before nouns beginning with a vowel. 


The erthe sal tai do to rift And up out of the sted to lift; 
The deuele out sal be fordreuin Of that erthe that sal be reuin; 
Ber thair bodis in at air... . 
¢ 1325 Signa ante Fudicium, quoted in Eng. Metr. Hom., ed. Small, p. xii. 
Pat oyle [var. Jo oyle, hat oyle, pe oile). 
¢ 1300 Cursor Mundi (Cotton ms.) (E.E.T.S.), l. 1394. 





IX. Cases involving the preposition az. P 

Cases in which the final ¢ of the preposition a¢ before a 
word beginning with a vowel or % has gon over to the follow- 
ing word, the remaining @ being then in some cases lost. 

The first cases involv locativ surnames, which ar to be 
compared with those which hav attracted the x of the article 
then. (See my previous paper, TRANSACTIONS, xxiii. 280-284). 


1. Ash. A? then ashe, atten ashe, atte ashe, at ashe, results in 
the surname Zash. Compare Wash from the same source (TRANS- 
ACTIONS, Xxili. 282). 


(6) Tash ... At ash. 1633 CAMDEN, Remains, p. 123. 


2. Asp, dial. as. Some one dwelling a#e aps, ‘at the asp,’ gave 
rise to the surname Zafps. Put him afen aps, and he becomes Mr. 
Vabbs (TRANSACTIONS, xxiii. 282). 


(6) Revd. Mr. Richard 7afps. 1783 LEMON, Eng. Etym., List of subscribers. 


3- Elm. John atte E/me became John 7Ze/me, as John atten Elme 
became John Ve/me (TRANSACTIONS, xxiii. 282). 
This indentur made betwin T. D. of Oxford Aldirman on y one pty and 


John Ze/me of the same brewar on the other party wytnesseth (etc.). 
1502 Arnold’s Chron. (1811), p. 109. 


4. Heath. One dwelling afte hethe is supposed to be the ances- 
tor of persons whose ‘ dental formula’ is Zze¢h. (P. p. 561.) 


5. Well. There were many living afe welle or atte welles, ‘at the 
well’ or ‘near the spring.’ Some of their descendants ar named 
Atwell; some ar reduced to Twells. 


* 


Vol. xxiv.] Attraction in English. 97 


(a) Wretyn on Candilmas Day, in hast, a¢ Welles. 
1489 Paston Letters, ed. Gairdner, iii. 347. 
Thom’ 4¢ Welle carte. 1482 (?) Jd. iii. 292. 
(4) Twells. As we have the name A/wells or Atwell, one has certainly reason 
to think that 7we//s is a crasis for At Wells. 
1777 Gent. Mag., July, p. 322. 


The next two cases ar also locativ, but do not involy surnames. 


6. Home. Az home, ME. at hom, atom, once a tom. A British 
servant, with equal disregard of veracity and orthoepy, without an 
aspiration for either truth or home, wil report his master or mistress 
not a tome. 


(a) Atom his [var. at hom is] hire pater noster biloken in hire teye. 
@1250 A Lutel Soth Sermun,\.67. (Old Eng. Misc., E.E.T.S., p. 190.) 
Pe were betere habbe bileued @¢ om ban icome me to fonde; 
Li doun, bu ert ouercome: ic wole on pe stonde; ‘ 
pu mi3t telle z¢ om hou bu were vnder a maidenes honde. 
¢ 1300 Seinte Margarete, |. 180 (E.E.T.S., p. 29). 


Here it is one of the devils who ought to hav staid at home. To 
them, surely, “ there’s no place like home.” 


(4) Ergo, he nis not alwey a¢ om [a tom, ms. V] among ow freres. 
1362 LANGLAND, Piers Plowman (A), ix. 20. 
A good gosse iche hav a foome. 
1542 BoRDE, /ntrod. to Knowledge (cited in Spec. Cornish 
Dial., EE.TS., p- 84). 


7. One, reduced to 0. Ato side became sometimes a vo side. 


(4) A £0 side. 1575 LANEHAM [see quot. below]. 
Muscles collateraux. Two muscles in the mouth, one bringing the tongue, 
the other drawing the Larinx, @-to-szde. 1611 COTGRAVE. 


8. Other. Aft the(?) other (side) is found asa. . . toother. 


Thearfore thus, with fending & proouing, with plucking & tugging, skrat- 
ting & byting, by plain tooth & nayll a Zo side & toother, such expens 
of blood & leather waz thear between them, az a moonths licking (I 
wéen) wyl not recoouer. 


1575 LANEHAM, Letter from Kenilworth (E.E.T.S. 1890), p. 17. 


The next two cases refer to time. 


9g. Erst, aterst. ME. at aresie, at arst, becomes ¢arst. 


(2) Bot ay be redye in araye and a¢ areste ffoundene. 
c 1440 Morte Arthure (E.E.T.S. 1865), l. 311. 
(3) Tho ¢arst began Godrich to go 
Upon the Danshe, and faste to slo. 
a 1300 Havelok, |. 2688. (H. p. 852.) 


98 Charles P. G. Scott. [1893. 


1o. Even. <A/ even, is contracted to at cen, and is then reduced 
to “en. 


(a) And up thai wol ate eve Into a tree lest thai by nyght myscheve. 
1420 Palladius on Husb. (E.E.T.S.), i. 613. 
Thane syr Arthure, At euene at his awene borde auantid his lordez. 
¢ 1440 Morte Arthure (E.E.T.S. 1865), l. 1593. 


Wretyn in hast, the secund Sunday of Lent by candel light a¢ evyn. 
1461 MARGARET PasTON in Paston Letters, ed. Gairdner, i. 544. 
At midday and a¢ een. Larl Richard (Child, Ballads, iii. p. 400). 
Sae near Sabbath a¢ e’en. ; 
1818 Scott, ob Roy, xxii. [Az e’en is common in the Waverley novels. ] 
(4) Wow, Jamie! man, but I’d be keen 
Wi’ canty lads like you, a wheen, 
To spen’ a winter Fursday ‘een. 
1788 PICKEN, Poems, p. 98. (1882 Jamieson.) 


11. All. The phrase a¢ ai, especially in the negativ, zo¢ at all, 
is commonly pronounced @ /a//, and in childish or dialectal use 
simply ¢a//. I hav heard the emphatic negation, “Taint, #a@Z7/” 

(6) [Narrow Romic:] :djzwab dzhek(t)ta(qw)ta bekaws'moolk/ —nota 

*toml\. : 
[Broad Romic:] dyuwab jekt(t)ata bekos:mouk/ nota taol. 
1877 SWEET, Handbook of Phonetics, pp. 114, 115. 


[This the common polite question, ‘Do you object to tobacco-smoke?’ 
with the usual polite lie, ‘ Not at all.’] 


These cases ar different from those in which #, reduced to 7?’ 
before a word beginning with a vowel, is written with that word, as 
in ME. #0 ehken, AS. 10 edcan, ‘to eke,’ for addition, besides (see 
Matzner, ii. 7): in ME. &@ eve, reduced to “ve; and in ME. “% 
before an infinitiv beginning with a vowel, as /azyse for fo arise, topon 
for 40 open, etc. : 





X. Cases involving the pronoun z¢. The final ¢ of z¢ goes 
over to the following word, the unaccented vowel being lost. 

1. It is becomes #s, usually written ’#s or ’¢ zs. Common in 
spoken English, and in verse. Emerson unhappily affects it in prose. 


2. It is not, contracted 77 zs m’t, becomes 7¢ ain¢ or ant, and so 
taint, written also apostrophically ’saint, ’tain’t, ta’nt, etc. 


Ta’nt, Taint, Ti’n’t. Contractions of “it is not.” “No éaint.” 
1854 BAKER, Northampt. Gloss. ii. 328. 


Vol. xxiv.] Attraction in English. 99 


3. It has or it hath, provincially unified with the first person and 
the plural form, z¢ have, or it ha, becomes fave, or fa. 


T’ave. It has,...it have...or,it hath. Za is our common word for #¢, 

1823 Moor, Suffolk Words, p. 420. 

4. It will, provincially z2¢ wu//, it ool, colloquially z¢ ’7/, becomes 
twill (twill), ’¢ wid?, dial. twool, tool, tull. 


Tull. It will: also twool or ta wool. See Ta and Wool. 
1823 Moor, Suffolk Words, p. 450. 


5. It will not becomes fwont, ’twon't. 


6. It would becomes ’fwould, dial. 'tood. 
Praps ’tood. 1853 Spec. in Gloss. Glouc. Dial. (E.D.S.), 1890, p. 209. 





XI. Cases involving saznt. 

Cases in which the final ¢ of sazw¢ in names of saints and 
churches, or places named after them, or persons named from 
such places, has gon over. to the name itself, that beginning 
with a vowel or % or w. 

Note first that sazzt, ME. saint, saynt, seint, seynt, also 
_ fem. and sometimes masc. saznfe, etc. (also in variant form 
sant, sanct, whence mod. sauntz), was in ME. often reduced, 
even before a name beginning with a vowel, to sazu, sayn, 
sein, seyn, sen (occasionally spelled sayne, etc.). I mention, 
omitting authorities, Sain Bede, Sain Benet, Sen Benett, 
Seyn Cutbert, Sain Denis, Seyn Edward, Sain Gregorie, Sain 
Jam, Sain Jerom, Sain Jon (very common), Sex Lauerauns, 
Sain Louk, Sayn Mark, Sayne Martyne, Sain Matheu, Sain 
Peter, Sain Poule, Seyn Savour, Sain Symeon, Sain Thomas. 

This saz, sein, sayn, seyn, shortend to sen, sufferd further 
alteration to szz. This explains the “ queer” pronunciation 
of the surname written Sz _/o/n, and pronounced in England 
Sinjon (sin'jun). The name should be written etymologically 
Saint Jon, or rather Sain Jon, or, recognizing the actual] muta- 
tions of mortality, Sexjon, and now Siujon. The same reduc- 
tion, partly due to old French usage, appears in the surnames 


100 Charles P. G. Scott. [1893. 


Sample, Simple, Sempill, Semple, Simpole, representing the 
ME. Sain Paul (Poule, Powle); in Simbarb, Simbarte, 
Symbarbe, formerly Sembarbe, ME. Seyntbarbe, representing 
Seinte Barbe (Sancta Barbara), who gave name to St. Barbe 
in Normandy ; in Semper, Simper, with more than apostolic 
right of succession representing S¢. Peter (St. Pierre); in 
Seymour, equivalent to St. Maur (Sanctus Maurus); and in 
Sinclair, less happily Szxkler, equivalent to St. Clair (Sancta 
Clara). 
I find the reduced form Saz stil in provincial use. 


San Jam Pear. The Green Chiswell Pear, usually ripe about the 25th 
July (St. James’s Day) is so called. At Altrincham, they have a fair 
called Sazjam Fair on July 25. That day is almost proverbially wet. 

1877 LEIGH, Cheshire Gloss., p. 175. 


Seeing how loosely attacht the final consonant of saznt 
was, we ar prepared to see it slide over to the name it pre- 
cedes, if that begins with a vowel; to see Saint Ann become 
Sain Tann, then with the ¢ restored where it ought to be, 
and also left where it ought not to be, Sazut Zann. And so 
it happend with Ann and Abb and other names in the bead- 
roll of attracted saints. 


1. Saint Abb or Ebb. Saint Abd’s or £Od’s church has become 
Saint Tabd’s or simply Zadd’s (P. p. 571), also Stadés. 


(a) Mary Aéchirch, diocis London, patron master of Seynt Laurence Pulteney. 
1502 Arnold’s Chronicle (1811), p. 251. 
Mary Apchirch. 1502 /d. p. 76. 
The beacon at Saznt Aiichaed: 1816 Scott, Antiguary, c. 8. 
(4) St. Tabbe. St. Ebba was the famous prioress of Coldingham ... lc. 
(p. 123), also Fuller, Worthies in Rutland. 

1777 Gent. Mag., Aug., p. 373- 

Saint Ebbes [in Oxfordshire, pron. locally] Séadds. 
1883 Hop, Gloss. Dial. Place-Nomenclature, p. 58. 


2. Saint Aidan, or Azthan, or Athan, appears as Saint Tathan. 
(a) (6) S. Tathan, St. Athan or Aithan. Memorial of Brit. Piety, append., 

P. 40. 1777 Gent. Mag., Aug., p. 373- 
3- Saint Alchmund or Alkmund becomes Saint Talkmund. 


(a) (6) Talkmund. St. Alkmund’s church in Derby is commonly called 
Talkmund. 1777 Gent. Mag., Aug., p. 372. 


Saint Aldates. See Saint Olds. 


Vol. xxiv.] Attraction in English. 101 


4. Saint Alfey, A/phey, Alfege, Alphege, Alphage, AS. Aifhech, 
appears in the possessiv as Saint Talphes and Talfas. 


(a) Saint Alphay at Crepilgat. 1502 Arnola’s Chron. (1811), p. 77: 
Alphey wythin Crepilgate, deocis London, patron deane of Seinte Martyn 
the Graunte, the sine, 1502 /d. p. 247. 


The first instance of the attracted form shows it as a surname: 


(4) Item as for Zad/fas.... Caly hadde ben at hem, and desired to carye 
up Za/fas on his owen cost and yeve hem goode wages.... And 
Margaret 7a/fas seide tome... 

1452 (?) Paston Letters, ed. Gairdner, i. 247. 
The xxx day of July was bered in Sant 7alphes in Crepullgatt, masteres 
Parston. 1562 MACHYN, Diary (Camden Soc. 1848), p. 289. 


5. Saint Andrew becomes Saint Tandrew, and so simply Zan- 
drew, Tander. Compare Dandrie, Dandie, Dandy, from Andrew 


(p. 131). 


(a) Saint Andrew. ¢ 1305 St. Andrew (E.E.T.S.), p. 100. 
Saynt Andreu. 1340 MICHEL (tr.), Ayenbite of Inwyt (E.E.TS.), p. 12. 
The bisshop of Seint Andre. 

$4 1306 Execution of Sir Simon Fraser (Child, Ballads, vi. 276). 


Item in y chirche of Saint Andrew is M. yere of pardon. — - 

1502 Arnold’s Chron. (1811), p. 152. 
(6) Tander, Tandrew. Corruptions of St. Andrew, who is looked upon by 
the lace-makers as their patron saint, as St. Crispin is considered the 
patron of shoemakers. The 30th of November, the anniversary of this 
saint, is, or rather was, kept by lace-makers as a day of festivity and 
merry-making; but since the use of pillow lace has in a great measure 

given place to that of the loom, this holiday has been less observed. 
1854 BAKER, Vorthampt. Gloss. ii. 326. 


6. Saint Ann becomes Saint Zann. 


(a) Seynt Anne. ¢ 1300 Cursor Mundi, \. 154. [See below.]} 
Of blissed seywt Anne, moder to our lady. 
¢ 1485 Killing of the Children,\.1. (Dighy Mysteries, N.S.S., p. 1.) 
Abbey of Saint Anne, on the Tour-Hyll, Whit Monkis. 
1502 Arnola’s Chron. (1811), p. 259. (Seynt Anne, p. 75.) 
I sweir to Yow, sir, be Sanct Ann! 
1602 LynDESAY, Thrie Estaits (E.E.T.S.), 1. 878. 
(4) O[f] Toachim and of sant tanne (var. seynt anne, saint ane, Seynt Anne). 
¢ 1300 Cursor Mundi (Cotton MS.), 1. 154. 
Say quhat 3e will, sirs, be Sauct Tan / 
1602 LyNDESAY, 7hrie Estaits (E.E.T.S.), 1. 3029. 


So Saint Ann's chapel became Saint Tann’s Chapel, Tann’s 
Chapel, and finally Zurnchapel (P. p. 567) —as if a chapel for 


per “ert persons ‘ turnd’ eae the true faith. 


102 Charles P. G. Scott. [1893. 


7. Saint Antolin or Antholin’s church has become Zanjéolin’s. 
(P. p. 571-) 


(a) Seynt Antolyns. 1502 Arnold’s Chron. (1811), p. 77. 
Antelyne in Bogerowe, diocis London, patrone deane and chapiter of 
Poules, the decis. 1502 Arnold’s Chron. (1811), p. 247. 


8. Saint Antony, or Saint Anson, besides suffering the indignity 
of having his name speld wrong, Aushony, has had it reduced to 
Tantony. 


(a) Saint Anton. ¢ 1325 Eng. Metr. Hom., p. 69. 
Seint Antonis [chirche}. 1502 Arnold’s Chron. (181 1), P- 75. 
Saint Anton’s well shall be my drink. 

21724 Waly, Waly, but Love be Bonnie (Child, Ballads, iv. 134). 


The name of Antony is used to describe a bell, a cross, and a pig, 
and in these familiar uses sometimes appears as Fan‘ony. 
(1) Saint Antony(’s) bell becomes Fantony bell. 
(a) St. Anthony’s dell, hung about the necks of animals. 
1765 Lord Hailes. (1808 Jam.) 


(4) He had to,sell the Tantonie bell 
And pardons therein was. 4@ 1765 Godly Songs. (Jam.) 


Baie cites “Fr. éanéan, ‘ the bell that hangs about the neck 
of a cow,’ etc., Cotgr.,” and doubts “if this has any relation to St. 
Anthony.’’] 


Tantony. The small bell over the church-porch or between the chancel 
and the nave: the term is also applied to any small hand-bell. “ Ring 
the ¢antony”’ is evidently a corruption of S4 Anthony, the emblem of 
that saint being a bell at his tau-staff or round the neck of his accom- 
panying pig. 1854 BAKER, orthampt.. Gloss. ii. 328. 


(2) Saint Antony(’s) cross becomes Tuntony cross. 


(3) Saint Antony(’s) pig became Tantony pig, of the same family 
as a WVantony grice, for which see TRANSACTIONS, xxiii. 189. Saint 
Antony enjoyd the personal attendance of a pig, who claimd the 
title of a ‘page’; a slovenly valet, one would think, but then, we ar 
told, some of those eremitical saints, whether troglodyte or stylite, in 
cave or on pillar, wer too. particular about godliness to care much 
for the next thing to it. 

(a) I haue behest a pygpe to Saynt Antony, voto nuncupavi. 

1519 HoRMAN, Vulgaria. (Way, Pr. P., p. an) 


Behald in every kirk and queir . . . 
Imageis maid with mennis hand, . . . 
Sanct Anthone sett vp with ane soow, 
Sanct Bryde, weill caruit with ane koow. 
1552 LyNDESAY, 7he Monarche (E.E.T.S.), 1. 2280-2306. 


i 


Vol. xxiv.] Attraction in English. 103 


The gruntill of Sanct Antonis sow” 
Quhilk buir his haly de//. 
1602 LynpEsay, 7hrie Estaits (E.E.T.S.), 1. 2099. 


The proverbial comparison, ‘ to follow one like S¢. Anfony’s pig,’ or 
‘an Antony pig,’ later ‘a Tantony pig, refers to the alleged docility 
of the animal. We ar stil cald upon to wonder at “learned pigs.” 
I take Stow’s account to be a fable: 


The Officers (charged with oversight of the Markets in this City) did 
divers times take from the Market people, Pigs starved, or otherwise 
unwholsome for mans sustenance: these they did slit in the eare. One 
of the Proctors for St. Anthonies [Hospital] tyed a Bell about the 
necke, and let it feed on the Dunghils, no man would hurt, or take it 
up: but if any one gave to them bread, or other feeding, such would 
they know, watch for, and daily follow, whining till they had somewhat 
given them: whereupon was raised a Proverbe, Such an one wil follow 
such an one, & whine as it were an Anthonie Pig. ; 

1633 Stow, Survey of London, p. 190. 
(4) Lord! she made me follow her last week through all the shops like a 


Tantiny pig. 1738 SwIFT, Polite Conversation, i. 

A Tantony pig. 
1736 DRAKE, Zboracum, p. 315 (1777 Gent. Mag., Aug., p. 372). 
Tantony Pig. 1820 WILBRAHAM, Cheshire Giloss., p. 89. 


(4) Saint Antony(’s) pouch becomes Tantony pouch. 
Tantonie pouch. 1632 LILLY. (H. p. 850.) 


g. Saint Audry, Awdry, Audrey, Awdrey, Audery, ME. Awdry, 
Awdrey, Awdre, a popular reduction, by successiv ecthlipsis of zt 
and 4, of the AS. <4eldryht, of which the regular modern Eng. 
form would be *Z¢heldright. The AS. was Latinized “¢heldrytha, 
Ai theldritha, Atdilthryda, Etheldrytha, Etheldreda, whence ME. 
Etheldrede, mod. Etheldred. St. Audry, St. Aipeldryht, founded a 
monastery on the Isle of Ely. 


(2) 673. Her Ecgbryht Cantwara cyning forbferde and by geare wes senod 
zt Heorot forda, and Ste 4 Zeldryht ongon pet mynster et Elige. 
¢ 900 (?) A.S. Chron. (Parker MS.), ed. Earle, p. 36. 
And bongedone god mekelyche & seynt Awdrey. 
¢ 1420 Vita S. Etheldredae Eliensis,\.925. (Horstmann, 
Altengl. Legenden, 1881, p. 303.) 
Porwe goddus grace & pis blessude virgyn seynt Awdre. c 1420 Jd. 1. 935. 
[So Seynt Awdrey, \. 768, 967; Seynt Awdry, |. 1052, 1070; Seynt 
Awzdre, \..414, 899, 947, 958, 1117; Awdre, |. 506, 1029.] 
Of pis blessude virgyn seynt Etheldrede. 
¢ 1420 /d.1. 587. [So Etheldrede, \. 137, 141, 483, 582.] 
Wretyn att London on Seynt Awdryes Daye, anno E. iiij‘t xvij®. 
1477 Paston Letters, ed. Fenn, ii. 248; ed. Gairdner, iii. 195. 
Audry, Sax. Itseemeth to be the same with LZtheldred, for the first 
foundresse of Ely Church is so called in Latine histories, but by the 
people in those parts, S. Audry. 1637 CAMDEN, Kemaines, p. 93. 


104 Charles P. G. Scott. [1893. 


Saint Audry came to be sometimes Sain(?) *Zaudry, * Tawdry, 
Taudery. 


(6) Taudery, for St. Audery, (Etheldred.) 
1692 CoLes, Eng. Dict. (See full quot. below.) 


On and after Saint Audry’s day, there was held, in the Isle of Ely, 
a great fair cald Saint Audry’s fair or Audry-fair, and, we may 
assume, * Zawdry-fair. 


(a) Audry-Fair in Cambrid[g ]eshire. 
1692 COLEs, Eng. Dict. (See full quot. below.) 


Lace sold at Saint Audry’s fair or *Tawdry-fair was cald Saint 
Audry’s lace, or Tawdry- (Tawdrie-, Taudery-) lace. The origin of 
Tawdry- coming to be forgotten, it was taken as an adjectiv, appella- 
tiv of this kind of lace, and of other cheap finery ; something gaudy, 
but not neat. 


(a) Seynt Audries [misprinted Andries] lace, cordon. 
1530 PALSGRAVE, p. 269. 
(4) Binde your fillets faste, 
And girde in your waste, 
For more finenesse, with a ¢awdrie lace. 
1579 SPENSER, Shep. Cal., April. 
Come, you promis’d me a ewdry-lace, and a paire of sweet Gloues. 
1623 SHAKESPEARE, W.T. iv. 4. 253 (F p. 293). 
Tawdry lace, (i.e.) Astrigmenta Fimbriz seu Fasciole emtz nundinis 


Fano Sancte Etheldredz celebratis. 1671 SKINNER, Etym. Angi. 
Taudery, for St. Audery, (Etheldred.) Zaudery dace, bought at Audry- 
Fair in Cambrid[g]eshire. 1692 CoLes, Eng. Dict. 


The explanation in the following passage is inaccurate. 


Tauary, garish, gawdy, with Lace or mismatched and staring Colours: A 
Term borrow’d from those Times when they trick’d and bedeck’d the 
Shrines and Altars of the Saints, as being at Emulation with each other 
upon that Occasion. The Votaries of St. Audery (an isle of Ely 
Saint) exceeding all the rest in the Dress and Equipage of their Altar, 
it grew into a By-word, upon anything very gawdy, that it was All 
Ti hi fe as much as to say, all St. Audery. 

1737 Canting Dict., App. to Bailey, vol. ii. 


The use of sawdry, taudry, as a mere adjectiv began about the 


middle of the seventeenth century. Hence ‘sawdriness (1670), and 
tawdrily (1736). See examples in Richardson. 


I came from the exchange where I saw a flock of English ladies buying 
taudry trim’d gloves. 1674 HOWARD, Luglish Monsieur. (Wr. p. 946.) 
This adjectiv sawdry (in a dialectal pronunciation speld sardry — 


cf. Sc. arns for awns) has come, thru the notion of “ cheap and nasty,” 
to hav a moral application : 


Vol. xxiv.] Attraction in English. 105 


Arthur to Doll Is grown bobbish and uxorious, 
While both she and Huncamunca tipple, talking /awary. 
@ 1825 Doodle and Noodle ( Universal Songster, i. 401). 
Tardry. Immodest. Last. 1847 HALLIWELL. 
From the adjectiv, or directly from sawdry-/ace, came the noun 
tawdry, lace or other finery. 
Of which [coral] the Naides and the blue Nereids make 
Them /awdries for their necks. 
1613 DRayTON, Polyolbion, ii. 46. (C.D.) 
Once more, and finally, from Zawdry (*tawdery, taudery), regarded 
as formed from */awder, with the adjectiv suffix -y, was evolvd the 
verb de-¢awder, to dress in a tawdry or gaudy style. 
Go, get ye home, and tricke and defawder yourself up like a right city lady. 
1688 Mrs. BEHN, City Heiress. (Wr. p. 203.) 
Thus hath the name of a sainted lady come to ae cheap 
vulgarity. Sancta Etheldreda, ora pro nobis ! 


10. Saint Austin, whom we now call Saint Augustine, has become 
reduced as a surname to Zustin, as well as to Sustin. 


(a) ber stod sein¢ [var. sein] Austin. ¢ 1200 LayAMon, Brut, |. 29551. 
Seynte Austyn. 1297 Kobert of Glouc., p. 235. (Wr. p. 13.) 
Thurrow Goddes helpe and Sentawsden, 
The spere anon he toke to hym. 
¢ 1435 Torrent of Portugal, p. 44. (H. p. 721.) 
(4) Tustin. 1889 Philadelphia Directory. 


Saint Ebb. See Saint Add. 


1z. Saint Ellen, the earlier form, without aspiration, of Saint 
Helen (ME. Eline, Elyne, AS. Elene, LL. Helena, Gr.‘EXém). The 
modern /He/en is a restored form, after the Latin. 


Saint Ellen’s or Helen’s church has become Ze//in’s. (P. p. 571.) 


(a) I swer by Seynt Elyne. 
¢ 1300 Richard Coer de Lion;\. 771. (Weber, Metr. Rom. ii. 33.) 
Sentt Elyne.... Seint Eline.... Seint Elyne.... Sent Eline. 
¢ 1400 Legend of the Three Kings (ed. Wright in 
Chester Plays, i. pp. 289, 300, 301). 
Priores of Seynt Helyns. 1502 Arnold's Chron. (1811), p. 251. 
You owe me ten shillings, 
Say the bells at St. Helens. 
1783 Gammer Gurton’s Garland. (Northall, p. 399.) 


12. Saint Etha appears as S¢. Zhetha or St. Teath. 


(a) (6) St. Thetha or St. Teath. St. Etha was a Cornish Saint. 
1777 Gent. Mag., Aug., p. 373: 


13. Saint Helen. See Saint Ellen. 


106 Charles P. G. Scott. [1893. 


14. Saint Isay or Hsay, the old English name of the prophet 
now known in the imperfect Hebrew form of /saiah (pronounced 
variously ai-zé'a, ai-zé'-ya, ai-zai'a, ai-zai'ya, ai-za'ya), appears to be 
the source of S¢. Zizzy. The ME. forms ar /say, /saye, Lzaye, sae, 
Ysay, Ysaye, Ysate, Esay, Hsate, Esaii, from OF. Lsaie, Lsaie, 
LL. Zsaias, Gr. ‘Hoaias, Heb. Yesha‘yéh. 


(a) Saint Vaaye, ¢ 1325 Eng. Metr. Hom., p. 48. 
Lsay.... Esait. ¢ 1430 (ms. 1582) Chester Plays, i. 155, 159,... 156. 
As Moyses sayd, and /say, Kyng- David and Jeromy. 

¢ 1450 Towneley Myst., p. 73.. [Also Jsae, p. 92; say, pp. 93, 129, 145. ] 

Now may I trost be techeyng of /zave in scryptur. 
c 1485 Mary Magdalene, \. 697 (Digby Myst. (N.S.S.), p-81). 

(4) Young was the lass, a servant at St 77zzy, 

Born at Polpiss and bred at Mevagizzy. 

a 1847 4 Western (Cornish) Eclogue. (H. p. xii.) 


15. Saint Ive appears as Saint Tive (Zyve). 


(a) At y feste of Seint Jue xxv. marc. 1502 Arnold’s Chron. (1811), p. 101. 
As I was going to S¢. Jves 
I met a man with seven wives. 
a 1800 Riddle (Halliwell, Mursery Rhymes, p. 53)- 
Sanctus Ivo erat Brito, 
Advocatus sed non latro, 
Res miranda populo. Zfizgram (source forgotten). 
Nine days I fell, or thereaboits; and had we not nine lives, 
I wis I ne’er had seen again thy sausage-shop, S¢. /ves / 
Had I as some cats have, nine tails, how gladly I would lick 
The hand and person generally of him who heaved that brick! 
1871 CALVERLEY, Sad Memories (in Fly Leaves). 
(4) Hit is an old said saw, I swere by seynt Tyve, 
Hit shalbe at the wyves will if the husband thryve 
@1500(?) The Enchanted Basyn, \. 21 (Jamieson, Pop. Ballads, 1806, i. 273). 


16. Saint Olave, ME. Olof, Oluf, Olef (gen. Oloves, Olovis, 
Olevis), later reduced to *Owle, Olye, appears as St. Towle, Toole, 
Toly, Tooley. Hence the name of Zooley street, famous for its “three 
tailors,’ who, we ar told, once met, and signd a petition beginning 
“We the people of England.” But it seems that one of the three 
tailors was a grocer, and that only one of the two remaining had a 
shop in Tooley street. See WV. and Q., 21 Jan. 1888, PP 553 1891, 
Brewer, Historic Note-Book, p. 885. 

(a) Saint Oloves church in Southwerke. 

1459 Paston Letters, ed. Gairdner, i. 462. [Also 1462 Seynt Oleffes, 
Id. ii. 112; 1465 Seynt Olevys, and Seynt Olovys, ii. 240, 260. 
Seint O/of in Siluer Stret. Seint Olof inthe Jury. Seint Olof at Crouchid 


frier. Seint O/of in Southwarke. 
1502 Arnold’s Chron. (1811), p. 76. [Similarly O/u/, four times, p. 253.] 


Vol. xxiv.] Attraction in English. 107 


(4) In saynt Towlles in the OW Jury. 
1551 MACHuyYN, Diary (Camden Soc.), p. 6. 
The parryche of ens, Tops in Sowthwarke. : 
1556 /d. p. 118. [So, sant Towlys, p. 221; Sant Towllys, p. 303. 

Take Saint Tooles Parish. : , Be Ps 9°3-) 

1604 Meeting of Gallants at an Ordinarie (Percy Soc.), p. 11. 
Saint Oly into [Saint] Toly. 1637 CAMDEN, Remaines, p. 123. 
Tooley Street, Tooley Bridge, Tooley Corner, all in Southwark. 

1777 Gent. Mag., Aug., p. 373- 


To the same source we may refer the surname Zooley (1889 
Philadelphia Directory, etc.), and probably also Zovode. 


17. Saint Old’s, the local reduction of the name of Saint Aldate’s 
church at Oxford, is stil further alterd to Saint Toles or Stolds. 


(a) (6) St. Tole. St. Aldate’s church, or St, O/d’s at Oxford, is vulgarly called 
St. Tole’s. Poynter, Oxon. Acad. p. 109. 1777 Gent. Mag., Aug., p. 373- 
Saint Aldates [in Oxfordshire, pronounced locally] Sto/ds, ‘ 

1883 Hope, Gloss. Dial. Place-Nomenclature, p. 58. 


18. Saint Omer, sometimes given a ‘locus classicus’ as Saint 
Homer, appears also as Saint Tomer. 


(a) Also, ther is on comythe eu’y Markett daye ffro Seynt Omerys to Caleys. 
1473 Paston Letters, ed. Fenn, ii. 150; ed. Gairdner, iii. 95. 
[Also Seynt Omers, 1477 Jd. ed. Fenn, ii. 252, 253, 254; 
ed. Gairdner, iii, 202, 203, 204. 
Seynt Homers worstedde, demy ostade. 1530 PALSGRAVE, p. 269. 
(46) S. Tomer, De [Sto,| Audomaro. 
1637 CAMDEN, Kemaines, p.151. (Also 1777 Gent. Mag., Aug., p- 373-) 


To this source we may perhaps refer the surname Zoomer. 


1g. Saint Osith, in the popular form Sant Osy, in the genitiv Saint 
Osy’s (ME. Seynt Osyes), appears as Saint Tows, in the genitiv 
Saint Tooses. Ido not find Saint * Zoosy, but it probably existed, 
as the source of the surnames Zousey, Zoucey. The surname Zowse 
appears to come from the form Saint Zows. 
(a) Seynt Osyes in Essexe. 
1473 Paston Letters, ed. Fenn, ii. 142; ed. Gairdner, iii. 92. 
(4) St. Tooses. 
a 1604 R. HAL, Life of Bp. Fisher [first pub. 1665 as by Thomas 


Baily], p. 88. (Gent. Mag., Aug., 1777, P- 373-) 
St. Osyth into Saint Tows. 1637 CAMDEN, Kemaines, p. 123. 


20. Saint Owen may be the eponym of some named Zowz, tho 
that surname is in most cases of the more obvious origin. 
(2) S. Owen. De S. Audoeno. 1637 CAMDEN, Remaines, p. 151. 


(4) Town. This sirname, I imagine, may be corrupted of St. Owen, who 
occurs in Camden, p. 151. 1777 Gent. Mag., Aug., p. 373- 


108 Charles P. G. Scott. [1893. 


21. Saint Winnol, in fuller form Saimt Winwaloe, appears as 
Saint Twinnol. 
(2) Wéinnol-far, the great horse-fair now held at Downham Market; so called 


from having been originally granted to St. Winwaloe’s Priory, at Were- 
ham. Cf. the proverb concerning the weather in the first three days 


of March — 
First comes David, Then comes Chad, 
Then comes Winnol, Blowing like mad. 


1858 (1840) SPURDENS. Suppl. to Vocab. East Anglia (E.D.S.), p. 85. 


There is a town in France, near Dunkirk, named Bergues-Saint- 
Winoc. 


(6) S. Twinnel, ie. St. Winnoc. Ibid. [Memorial of Brit. Piety, Append.], 
p. 48. 1777 Gent. Mag., Aug., p. 373- 


The same change of such saint names occurs in Italian and Spanish, 
Santo Elmo, for example, becoming San?’ Elmo, San Telmo. 

The opposit change, whereby a name beginning with Z loses that 
initial after Sazmz, occurs in French, where Saint Audard stands for 
Saint * Taudard, Saint Theodard (P. p. 518). 





XII. An isolated case of ‘obscure attraction appears in the 
following: J wot well, Sc. I wat weel (ai wat wil) becomes 
atweel (a-twil), and so tweed (twil). 


Atweel, at well, adv. Truly, assuredly; from 7 wot weel; that is, Z wot 
well, oss. It is sometimes abbrev[iated] to ’ Tweed. 
1866 JAMIESON. 





§ III. Initial M gaind. 


XIII. In one instance the final m of the article zhem, 
originally only dativ singular and plural, but later extended to 
all cases and now used in provincial speech as a demonstrativ, 
equivalent to the literary ¢/ose, has gon over to the following 
noun, just as the z of the parallel form ¢/en has done in some 
other cases. See § 1, II. A. (TRANSACTIONS, xxiii. 279-287). 


1. Adze, Ads, having a final s (z) sound, is sometimes taken as 
a plural, and then requires the plural demonstrativ — them ads, just 
as we hear in the cross-roads grocery them molasses. Then them ads 


Vol. xxiv.] Attraction in English. 109 


is divided as the mads, and so mads occurs without #e. We hear of 
a mads in Connecticut. 


In my boyhood, in western Connecticut, I learned to know a common 
carpenter’s tool as a “ madz,” and I think most of the fellow-workmen 
of my father, who was a carpenter, used that name. ... The word 
on account of its form was looked on as a plural, and always took a 
plural verb, as I remember the usage. 

1893 E. H. Bassitt, in Dialect Notes, p. 278 (with explanation as above). 


On the other hand, the printer’s types once with fine irony turnd 
“the masses of the people” into “them asses of the people.” 





§ IV. Initial R gaind or lost. 


A. Initial R gaind. 

Cases in which a word has permanently gaind an initial r 
by attraction from a preceding word do not appear. But the 
possibility of such a change, and therefore the possibility of 
the reverse process, which comes next in order, namely, the 
loss of initial 7, which is to be proved in the next section, is 
shown by the three casual examples I shal cite. The first is 
a doutful reading, the seeond a mere pun, the third a popular 
blunder ; but all ar phonetically allowable, and therefore help 
to support the next cases. 





XIV. Cases involving the possessivs our and your. 


1. Anchor. Our anchor (ancor) may be read our rancor. 


Our ancor is come back. 1606 MARSTON, Sophoniséba, I. ii. 76. 


Here Bullen conjectures ovr. rancour. See London Academy, 
1893, Aug. 12, p. 131. 


2. Oar. A perpetrated pun shows that your roar may be used, 
coram populo, for your oar: 


“I wish you monkeys would quit your everlasting chattering,” exclaimed 

the Lion. “ What do you want to put in your roar for?” asked a giddy 

young Simian. 1892 Puck's Library (Oct.), p. 15. 

Thus do even comic papers justify their existence. Would that 
serious papers had equal reason ! 


IIo Charles P. G. Scott. [1893. 


3. The nawwab or viceroy of Bengal who took Calcutta in 1766 
was cald Siraj-ud-daula, ‘the lamp of the State.’ He appeard in the 
British newspapers of the time in the guise of a good old English 
knight, S’x Roger Dowler. He also gave name to Sir Roger Dowlas, 
one of the characters, an East Indian proprietor, in Foote’s play, 
The Patron. (P. p. 557.) 





B. Initial R lost. 

XV. Cases involving the possessivs our, your, etc. 

It is known that the familiar names Richard, Robert, Robin, 
Roger or Rodger, hav long appeard in homely use as “ick or 
Hich- (Hichcock), Hob, Hobbin, Hodge; but the manner of 
these changes, and the reason, hav never been given, nor, so 
far as I know, even considerd. The changes ar generally 
taken as a matter of course. Yet they ar of a strange sort, 
and pique curiosity. There was a reason. 

One might conjecture (and what etymologist wholly refrains 
from that gentle exercise ?) that the initial 7, as at one time 
strongly “trild,” or “old,” rold one day into the strongly 
aspirated 4, But x was not always strongly trild, was not 
always strongly aspirated, and the interchange has no physio- 
logical basis. Moreover, why should the interchange take 
place in these few household names, and in no other words? 
Mere infantil variation may be conjectured as the cause; but 
why, again, should it affect only these few household names ? 
A more definit cause must be found. 

The cause was evidently one existing outside of the names 
Richard, Robert, Roger, themselves. In other words, it was 
interference, not “phonetic decay” or internal change. The 
names must hav been used, and used often, in collocations 
that affected the stability of the 7 I find these conditions 
in the use of these names after the possessivs our, your, their, 
her, and the obsolete (ME.) here, hire, her, hir, ‘their.’ As 
the parents of a child spoke, individually, of mzne Ann, mine 
Ed, etc., and so, later, of my Man, my Ned, and as a neighbor 
talking to a parent would speak of thine Ann, thine Ed, 
and so, later, of ¢hy Nan, thy Ned (see my previous paper, 


Vol. xxiv.] Attraction in English. III 


TRANSACTIONS, Xxiii. p. 295-301), in like manner the parents, 
together with the brothers and sisters, would speak of our 
Ann, our Ned, etc., and the neighbors, referring to them, of 
their Ann, their Ed (ME. also her(e) Anne, her(e) *Edde, 
etc.), or, speaking to them, of your Ann, your Ed, or, speak- 
ing of the mother, of er Ann, her Ed, etc. The form mzne 
Ann (ME: myn Anne) changed to my Nan because there was 
a form my to rest on, and because final 2 was a shifting 
letter inclined to change sides—the Eng. v paragogic; our 
Ann remaind stable because there was no known short form 
to help the change. So mine Richard, etc., past to my Rich- 
ard, etc., and then remaind stable. But in our Richard, our 
Robert, etc., your Richard, your Robert, etc., there is a weak 
point, the final -r before initial ~. In the easy homely utter- 
ance of these possessiv terms, the two adjacent r’s would 
tend to merge into the second », as if ou Richard, you’ Rich- 
ard, etc.; but because there was no short possessiv ow’, you’, 
alredy existing, parallel to my for mzne, the r that survived, 
as I explain it, was the first, the possessivs our, your (ME. 
oure, youre, etc.) being too stable to yield their final element. 
Hence the result was our (R)ichard, our (R)obert, our (R)oger, 
or, taking the curt colloquial forms of the names, our (A)zck, 
our (R)ob, our (R)odge, ME. our(e) (R)icke, our(e) (R)obbe, 
our(e) (R)oge. This leavs ’/ck, ’Ob,’Odge, ME. *’ Icke, * Obbe, 
*’Oge, to represent the names concernd. As used after the 
usual possessivs, their weak form would not be noticed; 
when used alone, a feeling of their insufficiency, the absence 
of familiar masculin names beginning with z and o (because 
Osborn, Osmund, etc., wer not familiar), combined probably 
with a tendency to conform these unaspirated names to cer- 
tain wel-known masculin names alredy aspirated (Henry, also 
Herry, Huge, Howe, Hugon, Huggin, Hutchin, etc.), led to 
their appearance as Hicke, Hobbe, Hoge, and so later Hick, 
Hob, Hodge. That ’Icke, ’Obbe, ’Obbim did exist is proved 
also by their appearance with attracted d in Dick, Dod, 
Dobbin, etc. See later, XVIII. 1, p. 128, etc. 

For the change of *’/cke, *’Obbe to Hick, Hob, compare 
the Scotch Halbert for Albert, whence the abbreviated forms 


112 Charles P. G. Scott. [1893. 


Hab, Habbie, forms also probably due in part to variation 
from Hob, Hobbie. 

For the use of our, your, their, her, and especially of ozr, 
in the way indicated, no proofs ar needed. The use is 
abundant to the present day. 


Our. A term implying relation. Our Thomas, Thomas belonging to 
our family. Var, dial. 1847 HALLIWELL. 


That the diphthong in our may be slighted and even 
reduced to nothing, is proved by the frequent occurence, in 
the sixteenth century and later, and probably earlier, of dy7- 
lady, berlady, byrlakin, for by our lady, by our *ladykin. 

I giv below such early quotations tending to establish 
the process of change here set forth, as I hav found; but 
from the nature of the case, contemporary proofs ar meager. 
The changes took place in household speech and so wer less 
likely to get into record. They took place in conversation, 
involving chiefly the first and second person, and so wer less 
likely, again, to get into record, except in representations of 
dialog, or in familiar letters, forms of literature scantly cul- 
tivated in the period in which the changes took place. Even 
when “plays” came to be written, they delt chiefly with 
Herod, Pilate, Mahound, and other ancient worthies, and 
little with Hick and Hob who gaped at the plays, and for 
whom alone, let us pretend, the “comic business” was got- 
ten up. Yet traces of Hick and Hob, of Jack and Jill, of 
Mack and Moll and Maud, do occur very early, and they 
abound in the later colloquial records. 


1. Richard, ME. Richard, Rychard, from OF. Richard, Richart, 
also Ricard (Sp. Pg. Ricardo, It. Riccardo, ML. Ricardus), from 
OLG. Richard, OHG. Richart, MHG. Richard, G. Reichard. The 
name exists also in the unassibilated form Rickard (see below). It 
does not seem to hav sufferd the supposed change or loss of its 
initial consonant, that change taking place chiefly in the homely 


abbreviated forms. 
We ben at on acord 
To wende with thee, Rychard our lord. 
¢ 1300 Richard Coer de Lion, \. 1370. (Weber, AMetr. Rom. ii. 55.) 
The best tresore had Richard our kyng. ¢ 1300 /d. 1. 3316. (Weber, ii. 129.) 
When the name of her Richard she knew! a 1843 7he Maid of the Inn. 


a Dit 


Vol. xxiv.] Attraction in E nglish. 113 


2. *Rich for Richard, by dropping the apparent suffix -ard (cf. 
Rob for Robert, often Roberd; Rodge for Rodger, Roger; Walt, 
Wat, for Walter, Water). Examples of this *ich ar not found 
except as a written abbreviation Ach., but it is evidenced by the 
unassibilated form Ack (see next), and by its diminutiv Richie (which 
is also found as a surname Richey, Ritchie), and it may indeed exist 
inthe surnames Rich, Riche, Ritch, which may be only in part due 
to the adjectiv rich. . 

Our (your, etc.) *Rich became, with the loss of initial x and the 
supplial of the initial aspiration, ovr *Hich. This Hich- exists in 
the surnames Hitch, Hichcock, Hitchcock, Hitchins, etc. (See my 
previous paper, TRANSACTIONS, Xxiil. 231.) 

(a) *Rich. ' a@1600(?) [See Hichcock, Hitchcock, \.c.] 

Richie Storie [a ballad]. a 1800 (Child, Ballads, viii. 255). 
Richard Moniplies . . . Richie. 1822 Scott, Fortunes of Nigel. 


(6) *Hich-, *Hitch- (dim. Hichcock, Hitchcock]. a 1600 (?) (See above.) 
fitch [surname]. 1891 NV. Y. City Directory. 


Hichcock, Hitchcock, has a history of its own. In my previous 
paper, just mentiond, the forms Mitchcock and Lithcock, in the 
ballads, ar shown to be errors for Hitchcock. 


3. Rickard, ME. *Rickard, Ricard (late AS. Ricard) from OF. 
Ricard, ML. Ricardus. This unassibilated form of Richard appears 
to hav existed from an early period. It exists now in the surnames 
Rickard, Rickards, and Rickardson. 


4. Rick, a variant of Ach, abbreviation of Richard; or a direct 
abbreviation of Rickard. In ME. it usually appears as Ac., a written 
abbreviation of Richard or Ricardus, but it must hav existed also as 
a spoken abbreviation, Ac, giving rise, in our Rick, to Hick, ME. 
LHicke, Hikke, Hykke, a common name. Jick is not derived from 
Isaac, as some say (Bardsley and others). 


(a) Which box she delyvered to Ric. Call... Ric. can tell you of the 
gydyng of the cofer with other boks. . .. And ic. hath the copes of 
them. ... AndIand Ac. informyd hym. 

1465 Paston Letters, ed. Gairdner, ii. 187 and 238. 
Sir John Fogge, Ric Hault for ther suster and me. 
1473 Paston Letters, ed. Fenn, ii. 142; ed. Gairdner, iii. 92. 
[Here “ Ric” may stand for the written abbreviation “Ric.,” but it is 
printed without a period in both the careful editions cited, and may wel 
represent the actual spoken abbreviation. | 
Rick [nickname for Richard }. 1853 DiIcKENS, Bleak House. 


Hence the surnames Fick, Ricks, Rix, Rickson, Rixon, 


114 Charles P. G. Scott. [1893. 


The name Ack, derived in this manner, became very common, 
much like Zom or Fack. It appears first in the fourteenth century. 


(6) Hikke [Hicke C] be hakeney print and Hogge pe neldere. 
1362 LANGLAND, Piers Plowman (A), v. 161. 
[Also Hikke, var. Hicke, v. 172, 183, 185.] 


A! Hicke Heuyheed! hard is thi nolle 
To cacche ony kunnynge but cautell bigynne! 
¢ 1399 [LANGLAND], Richard the Redeless, iii. 66. 


Hycke Scorner. Enprynted by me Wynkyn de Worde. 
¢ 1525 (Wynken de Worde). (Dodsley, ed. Hazlitt, i. 144; 
Lowndes, 1834, p. 993.) 
Hick Scorner, the titular character in this interlude, became pro- 
verbial for a reckless scoffer : 3 
Zeno beeyng outright all together a stoique, used to call Socrates the 
scoffer or the Hicke-scorner of the citee of Athens. 
1564 UDALL, tr. Erasmus Apophth., Preface, sign. xxv. b 
(Nares,? p. 442). [Also Hicke skorner, id.]} 
This fleering frumpe is one of the Courtly graces of hicke the scorner. 
1589 [PUTTENHAM], Arte of Eng. Poeste (repr. Arber), p. 201. 


Hick at length came to be used as a common name for a fellow: 


A kind of gamball called the haltering of Hix.[Hick’s] Mare. 
1585 HicGINs (tr.), Momenclator, p. 298. (H. p. 448.) 
That not one Azck spares. - 
1655 Acad. of Compl. (1713), p- 204 Sei ) P- 442). 
That can bulk any 4zck. 1655 /d. (1713)- 
Hfick is not now much used as a given name; but it occurs as a 
surname, //ick, Hicke, Hickie, Hickey, also in genitiv form Aiicks, 
flix, Hickson, Hixon. 
How fick gave rise to Dick is an other story. See XVIII., below. 


5. Robert, ME. Robert, Roberd, Robart, Robard, from OF. Robert, 
Robers (It. Roberto, ML. Robertus, Rothbertus), from OLG. Rod- 
braht, OHG. *Hruodpreht, *Hruodpert, Hruodbert, Hrodebert, 
MHG. Ruodpert, Ruoprecht,G. Ruprecht, Rubrecht, Robrecht, Rupert, 
Robert. The AS. Hrodbert (Chron. an. 1050) follows the Continental 
form; the vernacular form would hav been *H700derht. Hence the 
surnames Robert, Robart, Roberts, Robarts, Robartes, Robertson, 
Roberson. 


(a) Sire Robert the Bruytz furst kyng wes ycore. 
¢ 1306 Execution of Sir Simon Fraser (Child, Ballads, vi. 276). 
[In an other stanza cald Kyung Hobde (see Hod, p. 117).] 
And Robert [ Robyn, B.C.] the ribaudowr for his rousti words. 
¢ 1362 LANGLAND, Piers Plowman (A), vii. 66. 


pe kyng was hote kyng Xoderd ; 
Neuer man wiste him ferd [var. aferd]. 
¢ 1400 Roberd of Cisyle, 1.9. (Horstmann, 
Altengl. Legenden, 1878, p. 209.) 


Sie A 


Vol. xxiv.] Attraction in English. 115 


Robert, probably only because of the mere similarity of sound, 
became associated with vodter, ME. robbere, robbour, robbur, often 
varied with the form roderd, robard. 

Competenter per Xodert, robbur designatur : 
Et per Richard riche hard congrue notatur; 
Gilebert non sine re gi/ur appellatur; 
Gefrei, si rem tangimus, in jo frei commutatur. 
a 1300 Harl. ms. 978, in Pol. Songs, p. 49. (Ellis, Z.2.P. p. 462.) 


Robert [Roberd C] the Robbour [rodbbere B, ryfeler C] on Reddite he 
lokede. ¢ 1362 LANGLAND, Piers Plowman (A), Vv. 242. 

Ac Xobert Renne-aboute shal now3te haue of myne. 
¢ 1377 LANGLAND, Piers Plowman (B), vi. 150. 


Robert was sometimes used a common appellativ : 


[Cain to Abel: ] 
Goo, iape be, robard iangillande. 
¢ 1430 York Plays, vii. 47 (p. 36). 
Because of this association, or because of some actual robber of 
that name, the term Roder?t’s men arose in the fourteenth century to 
designate a set of vagabond thievs who wer more definitly described 
as drawlatches and wasters. They wer also cald Roberts knaves. 
Robert was more familiar in the diminutiv form odin; and Robin 
Hood, that archer good, may owe his prenomen, or else-his infirmity 
of grammar respecting meum and tuum, to the mere popular ety- 
mology which made Rodert a robber. 
Bidders and beggers faste a-boute eoden.. . 
In glotonye, God wot, gon heo to bedde, 
And ryseth vp with ribaudye, this Roberdes knaues + 
Sleep and sleu3the suweth hem euere. 
¢ 1362 LANGLAND, Piers Plowman (A), prol. 40-45. 


[And] ryzt as Robertes men raken aboute, 
At feires, & at ful ales, & fyllen be cuppe, 
And preche} all of pardon, to plesen the puple. 
¢ 1394 Pierce the Ploughman’s Crede (E.E.T.S.), 1. 72. 


From our (R)obert (Robart, Roberd, Robard) may be supposed 
to come our *Oéder¢ or * Obard (which last I take to be represented 
by the surname Oddard), and hence Hobert, Hobart, Hoberd, Hobard, 
all found in ME. records. All examples of these latter forms may 
of course present the other Hobart, Hobard, etc. (modern also 
Hubbard), from OF. Hobart, a variant of Hudert, but it seems 
likely that with this name of OF. origin has been merged the name 
thus developt in England from Rodert. Evidence is meager. 

With this Hodert hav been associated more or less vaguely, but 
perhaps in part with real justification, the various forms which ar 
represented by or involvd with the word hobdledehoy, namely hodlede- 


pe 


116 Charles P. G. Scott. [1893. 


hoy (1540 Palsgrave), Hodberd de hoy (1580 Tusser), hober de hoy 
(1678 Ray), etc. They hav a curious history which I can not narrate 
here. = 


6. Robin, ME. Rodin, Robyn, Robyne, Roben, Robene, is from OF. 
Robin, a diminutiv of Rodert, which, tho etymologically Ro-dert, was 
supposed to consist of Rod- + -er/, a mere termination. It is the 
source of the surnames Rodin, Robins, Robbins, Robinson. 


(a) Til Robyn the ropere weore rad forte a-ryse. 
¢ 1362 LANGLAND, Piers Plowman (A), v. 180. 
And Robyn [1362 Robert se the rybaudoure, for his rusty wordes. 
7 LANGLAND, Piers Plowman (B), vi. 75. 
But I can rymes of Robyn Hod and Randolf erle of Chestre. 
¢ 1377 LANGLAND, Piers Plowman (B), Vv. 402. 
Bothe Bette the Bakere and Aodyn Rede. 
1468 Coventry Myst. (1841), p. 131. 
Robene sat on gud grene hill. 
¢ 1475 HENRYSON, Rodene and Makyne (Child, Ballads, iv. 245). 
Robin became so familiar as to be applied in rustic personification 
toa common bird. Robin Redbreast was at first a name parallel to 
the imagind names Jack Whitehead, or Tom Bluenose. Robin is 
the real personal name, and redbreast is a predicate adjectiv ele- 
vated to a surname. In present use we hav reduced Rodin Redbreast 
to robin-redbreast, and employ either vodin or redbreast alone as a 
~ name for the bird. An other name for him is Rodin Ruddock, reduced 
in provincial use to robin-riddick (1825 Jennings, Somerset Gloss. 
p- 64). 
I find odin applied to inanimate figures. 


The twa cross-legged figures that the callants used to ca’ Robin and 
Bobbin, ane on ilka door-cheek. 1816 SCOTT, Antiguary, xvi. 


Of Hobin, Hobbin, I find no early record. It emerges in the six- 
teenth century, and exists in the present surname Hodin, Hobdin, 
Hobbins. 


(4) Hobbin, ah Hobbin ! I curse the stounde. 
1579 SPENSER, Shepheards Calender, Sept. 
I wote ne, Hoddin, iy I was bewitcht. 1579 fa. 


7. Rob, abbreviation of Rodert. Hence the surnames 2odé, Rodds, 
Robson. Of Rob I hav come upon no examples in ME., except as 
a written abbreviation. Yet it must hav existed also as a spoken 
abbreviation. It appears as such in the patronymic surname Rodson 
(1450). It gave rise, in the manner before explaind, to Hod (ME. 
Hob, Hobbe); unless we ar to explain Hod as a direct abbreviation 
of Hobsin. From Hod arose the surnames Hodds, Hodson. 


Vol. xxiv.] Attraction in English. 117 


Tob, as a familiar abbreviation of Rodsert, appears to be stil so 
used in England. In the United States, Rodert, when abbreviated, is 
always either Rod or Bob. The last form od grew out of infantil 
pronunciation — Rod, Wob, Ob, Bob. 


(2) ob. Newton lymebrenner . . . and Rodert Bery. 
1470 Paston Letters, ed. Gairdner, ii. 404. 
(4) Sire Robert the Bruytz furst kyng wes ycore: 
He mai everuch day ys fon him se before — 
Nou kyng /odde in the mures yongeth 
For te come to toune nout him ne longeth. 
¢ 1306 Execution of Sir Simon Fraser (Child, Ballads, pp. 276, 277). 
Jak Chep, Tronche, Jon Wrau, Thom Myllere, Tyler, Jak Strawe, 
Erle of the Plo, Rak to, Deer, et Hod Carter, Rakstrawe. 
a 1400 On the Slaughter of Archbishop Sudbury ; in Polit. Poems, i. 230. 
Hobbes wit. 1451 in Paston Letters, ed. Gairdner, i. 192. 
John Hodbis . . . Jon Hobdys. 1465 Jd. ii, 209. 
Why in this Wooluish tongue should I stand heare : 
To begge of od and Dicke, that does appeere, 
Their needlesse Vouches: custome calls me too’t. 
1623 SHAKESPEARE, Cor. ii. 3, 123 (F! p. 12). 
Call’d in of Dandrie, Hod, and Jock. 
a 1650 The Raid of the Reidswire (Child, Ballads, vi. 133). 
Thoo’s a good lad, my //0dd, that teeak sike care. 
1685 Yorkshire Dialogue. (M. C. F. Morris, 
Yorkshire Folk- Talk, p. 101.) 
Much water slides past the mill that //oé Miller never wots of [p. 187]... 
ob Miller of Tuyford [p. 189] [ascribed to the period 1187-1190]. 
1825 Scott, Betrothed, xxvii. 
Hob,s. Bob or Robert. North. 1839 HOLLOWAY, Gen. Dict. Provincialisms. 


Hob soon came to be used, like Ack, as a general appellativ for 
any common fellow, a rustic. 


Other 4obdis 3e hadden of hurlewaynis kynne. 
1399 [LANGLAND ], Richard the Redeless, i. go. 
Le pied gris. The Hos, Clowne, Boore, Hind; so called, of his euer- 
dustie, or durty shooes.. .. Pied-gris: m. A clowne, 40d, hinde, or 
boore of the countrey. 1611 COTGRAVE. 
A hob or clown. Rusticus. 1677 HOLYOKE. 


This sense arose in part from the use of od with some descriptiv 
term to form a feignd person’s name— Hob Clunch, Hob Hansom, 
Hob Lob, used originally like Piers Plowman or Hodge Plowman, 
but soon reduced to the rank of a general appellativ. 


Flobclunch. Promos and Cassandra, iii. 2. (H.) 
Poore unbegotten wether beaten Qualto, an 4od-hansom man, God wot. 
rs 1583 Philotimus. (Wr. p. 246.) 
The rustical 4od/obs of Cretes, of Dryopes and payneted clowns Agathyrsi 
Dooe fetch theyre gambalds. 1582 STANYHURST, Zn. iv. 150. (D.) 
The draffe of the carterly Hod/obs thereabouts. 
1593 NASH, Lenten Stuff. (D.) 


The familiar use of Hod in these ways led to its use in the names 
of some homely games ; which I must here omit. 


118 | Charles P. G. Scott. [1893. 


8. Robbie, Aoddy, diminutiv of od, or, rather, directly substi- 
tuted for Rodin. Compare Colle, Colly, for Collin, Colin, diminutiv 
of Cole for Nicol, originally WVicolas, misspeld WVicholas ; also Scotch 
corbie for corbin, a raven (corbin, var. rauen, rauyn, occurs ¢ 1300 


Cursor Mundi, E.E.T.S., |. 3332). 


(a) Robert [so cald by Miss Wardour]... . adie [so cald by Ochiltree]. 
1816 Scort, Antiguary, xlii. 


Hence, in the way before indicated, the form /704dy, which is in 
form from Hob (fHobbe) + dim. -e or -y, but in fact probably 
directly substituted for Yoddin, variant of Robbin, Robin. 


(4) The Laird’s Jock ane, the Laird’s Wat twa, 
O Hobie Noble, thou ane maun be; 
Thy coat is blue, thou has been true, 
Since England banish’d thee to me. 


Now //odze was an English man, 
In Bewcastle-dale was born and bred: 
a 1784 Fock o’ the Side (Child, Ballads, vi. 82). 
Hobie Noble [a ballad so named]. @ 1784 in Child, Ballads, vi. 98. 


According to Sir Walter Scott, Hodie or Hobdie in this and other 
instances is a familiar form of Halbert: 
albert, or Hobbie, Noble appears to have been one of those numerous 
English outlaws who, being forced to fly their own country, had 
established themselves on the Scottish Borders. 


1802 Scott, Minstrelsy of the Scottish Border ii. 90 
(Child, Ballads, vi. 98). 


But it will be observd that, greatly to his credit 


Ffobie was an English man, 
In Bewcastle-dale was born and bred, 


and Halbert appears to hav been peculiarly Scotch (it is supposed 
to be an aspirated form of 4Mert). There is a Scotch form Haddie, 
which may, with equal propriety, be taken as a variant of HYoddie for 
original Roder?, and as a diminutiv of Halbert. 
A young man called Halbert or Hobbie Elliot. 
1816 Scott, Black Dwarf, ii. [Hobbie Elliot is a prominent character 
in the story. In ch. x. he calls himself Hod ZWiiot.] 


Habbie Gray (p. 168) . . . Halbert Gray (p. 169). 
1819 Scott, Bride of Lammermoor, xxivi 


9g. Roger, also Rodger, ME. Roger, Rogger (= D. Rogier, G. 
Roger), from OF. Roger, Sp. Pg. Rogerio, It. Ruggiero, ML. Rogerus, 
from OLG. Rodger, OHG. Hrodgér, Hruadger, Ruodiger, MHG. 
Riiediger, Riiedger, G. Riidiger. The AS. Rogcer (Chron. 1076) is 


Vol. xxiv.] Aitraction in English. 119 


borrowd. Hence the surnames 2oger, Rodger, Rogers, Rodgers, 
Rogerson. 
Our Roger may be the source in part of Oger, found as a ME. 
name, and extant in the surname Oager. 
(a2) Kyng Roger spak fyrst above. 
¢ 1300 Kichard Coer de Lion, |. 1689. (Weber, Metr. Rom. ii. 67.) 


Roger (var. Rogger). 
¢ 1386 CHAUCER, Cook’s Prol. (Six-Text, A., ll. 4345, 4352, 4356-) 


Roger was once much more familiar than it is now. Like the 
other names here treated it became a common appellativ for an 
animal — in this case a ram (H. p. 689). 

The form Oger is found very early, and is in part at least of Old 
French origin: OF. Oger, Ogier, ML. Ogerus, i a Udgerus, 
from OLG. Odger, ODan. Odger, Icel. Oddgeir. eet 


Alan fil. Oger, E. Roge fil. Oger, E. Oger, fil. Oger. GG. 
@ 1500 in Bardsley, Eng. Surnames, p. 580. 


“Oger the Breton” is mentioned in Domesday Book (f. 364 b). 


10. Rodge, ME. oge, is short for Rodger, Roger, ME. Roger. 
For the abbreviation, compare ich for Richard, Rob for Robert, 
Walt, Wat for Walter, Water, etc., above. Ido not find modern 
examples of Rodge, unless there be one in Ruage. 

Our (your, etc.) Rodge has become our Hodge (ME. Hilige, 
Hogge, Hoge). Hence the surnames Hodge, Hoge, Hodges, Hodge- 
kin, Hodgson. 


(a) Rage fil. Oger. @ 1500 in Bardsley, Eng. Surnames, p. 580. 


I find the diminutiv *Rodgecock, parallel to the fuller form Roger- 
cock, and the source of *Hodgecock, which is itself the source of 
nodgecock (see TRANSACTIONS, xxiii. 233). 


Stephen Rogekoc. a 1500 in Bardsley, Eng. Surnames, p. 591. 
Tis our Me and I think he lies asleep. 
1599 PorTER, Two Angry Women of Abington. 
(Dodsley, ed. Hazlitt, vii. 308.) 


The form Hodge has attaind celebrity. 


(4) Hikke the hakeney mon and Hogve [Hughe B, Houwe C] the neldere. 
1362 LANGLAND, Piers Plowman (A), v. 161. 
If euere sithe I highte Hogge [var. Hoge] of Ware. 
¢ 1386 CHAUCER, Cook's Prod. Ce A. 4336. 
[Cald Roger, var. Rogger, in ll. 4345, 4352, 4356. ] 
A turne-broche, a boy for Hogge at Ware. 
¢ 1430 LyDGATE, Minor Poems, p. 52. (H. p. 895.) 


120 Charles P. G. Scott. [1893. 


John son of Hodge [ Hogge, p. 237, Roger, p. 241] Ratcleff. 

1452 Paston Letters, ed. Gairdner, i. 244. 
Hoge. Rogerus, nomen proprium. 1483 Cath. Angl. p. 187. 
[Perhaps the first appearance of the word in a dictionary. ] 


I know small difference herein, Hodge brother, 
And I (Hugh) know as littell in the tother. 
1562 HEyYwoop, Proverbs and Epigrams (Spenser Soc.), p. 65. 


Sat pesyng and patching of Hodg her man’s briche. 
1575 STILL, Gammer Gurton (Old Pl. ii. 12). (Wr. p. 737-) 


Old Hodge Bacon and Bob Grosted. 
1664 S. BUTLER, Hudibras, pt. II. iii. 224. 


Like Hod for Robert, and Hick for Richard, and like Jack, Hodge 
came to be used as a familiar term for any fellow, especially a coun- 
tryman, and is now familiar in the political nomenclature of England 
as a general appellativ for a farm laborer. 

These Arcadians are given to take the benefit of euerie Hodge. 
1587 GREENE, Menaphon, p. 58. 
No hodge plowmaninacountrie. @ 1600 NASH (in Greene’s Works, vi. 21). 


“ Not much in ’em either,” quoth perhaps simple Hodge ; 
But there’s a superstructure. Wait a bit. 
1871 CALVERLEY, Fly Leaves: The Cock and the Bull. 


11. Rod, short for Rodger, Roger, or Rodge. Compare Geordie for 
Georgie. Hence the surnames Rod, Rodd, Rodkinson, Rodman 
(1891 WV. Y. City Directory). But these may be in part from an 
other source. 

Parallel to Rod, existed Hod, either developt from our Rod, or 
shortend directly from Hodge. 

Hod, ME. * Hodde, is not found in that form in the ME. period ; 
but I suspect it exists in the name Hud, Hudde, which I find in the 
York Plays and the burlesque Zurnament of Totenham, in just the 
atmosphere suited to Hod and Hod. The change of vowel is not 
unparalleld ; compare hod and hud, hod and hud. 


i, Pas. We! Hudde! 
ii. Pas. We! Howe! 

i. Pas. Herkyntome!... ’ 
i. Pas. We! Colle! 

iii. Pas. What care is comen to pe?... 
i. Pas. Whe! Hudde! be-halde into the heste! 
A selcouthe sight ban sall hou see... 
i. Pas. We! no Colle! nowe comes it newe i-nowe. 


¢ 1430 York Plays, xv. ll. 37-39, 46, 54 (pp. 119, 120). 


The editor prints “hudde!” “howe!” “colle!” without capitals, 
as if they were mere interjections ; but they ar obviously the names 


Vol. xxiv.] Attraction in English. 121 


of the shepherds, Hod, Howe (a form of Hugh), and Cod/, otherwise 
Cole, the abbreviation of Colin, Collin. 
“T make a vow” quod //udde, “T shalle not leve behynde.” 
a 1500 The Turnament of Totenham (Child, Ballads, viii. 107). 


(So Hudde (bis) and Hud on p. 112. In Harl. ms, different: “I wow 
to God, quoth Herry,” etc.). 


Hence *Hodcock, in a nodcock (see TRANSACTIONS, xxiii. 233). 


12. Roddy, diminutiv of Rod for Rodge or Rodger, Roger, may be 
the source of the surname Oddy or Oddie, in which the aspirate 
supplied to Hodédy, etc., does not appear. Roddy, Oddy, Oddie ar in 
the Mew York City Directory. 


(4) Daniel Fortesku, Alisaunder ody. 1460 Paston Letters, i. 522. 


Other names in R- existed in ME., Ralph (ME. Rauf, Raufe, 
Raaf, Raaff, Raff, whence the modern British pronunciation Rafe), 
and others, but none wer so familiar, it appears, as to hav sufferd the 
changes set forth above. 

The use of Sir before Richard, Robert, Roger, tho very common, 
would hardly be familiar enough in household use to affect the form 
of the names; but it may be supposed to hav assisted the change 
begun by an other cause. Perhaps the ~ in MJas¢er also helpt the 
change: Master Richard, Master Robert, Master Rauf, etc., wer in 
constant use. ; 

The possibility of the interchange of Rod and Hob, Robert and 
Hobert, etc., is indicated by a mistake made. by an advertiser of 
bicycles who heds some words of praise for his machine from a rival 
dealer, “‘ Praise from Sir Rupert.” (Lvening Star, Washington, 
D.C., Sept. 22, 1892; part II. p. 1, col. 2.) If Str Hubert can 
become Sir Rupert, even by mistake, the reverse process is possible. 


For all old things there ar more causes than one. 


Felix qui potuit rerum cognoscere causas. 
¢ 31 B.C, VERGIL, Georgics, ii. 490. 
I am willing to ad, as a possible additional impulse to the change 
of Rob to Hob, etc., the endless succession of cunabular infants who 
hav been floord by the initial 7, and hav pronounced Rod as ’O% or 
Wob.. Ex ore infantium et lactentum — 


I think the examples and analogies I hav given go far to 
establish the case. It is true, the evidence is meager. But 


122 Charles P. G. Scott. [1893. 


evidence may be meager and yet conclusiv. In philology it 
is not the fact that evidence is meager, which condemns a 
proposition ; it is the fact that it is inconsistent with other 
evidence of a stronger kind. Many important propositions 
rest, and rest firmly, on a meager muster of facts. A ‘law’ 
may be derived from a fact. (I hav known a ‘law’ to be 
derived from no facts at all.) But we desiderate a plurality 
of facts. Attraction in philology, as in physics, must be 
general. It must work everywhere, unless counteracted. 





XVI. If the kind of attraction and resultant loss of one 
yoccurd in our Rod, etc., as here supposed, it would also 
probably occur in other cases involving a like collocation of 
final and initial The-entire absence of such other cases 
would tend to throw dout on the change supposed in our Rod, 
etc. On the other hand, reflection wil show that cases in 
which a word ending in v occurs before an other beginning 
with », and occurs so often as to be current, and so liable to 
the change supposed, must be comparativly rare. And being 
by supposition current, the words must form a familiar phrase 
or compound. I can furnish three examples: 


1. Reck. ever reck, in the dialectal form ze’er rack, ‘never 
mind,’ appears in the Cumberland dialect as meer ak, a form so 
remarkable that the glossarist explains it erroneously as ze rack, as 


if Latin ne cures, ‘reck not.’ 
(a) Recche, care. They use the word rack or reck in the North parts of Eng- 
land at this time for fo cave. Hence never rack you is the same as take 
you no thought or care. 7 
1724 HEARNE, Gloss. to Rob. Glouc. (ed. 1810) (E.D.S.), p. 87. 
(4) Neer ak, C., never mind. Ray says, ‘To rack or reck’ to care, never rack 
you, ice. take you no thought or care. In that case it should be ‘ne 


rack,’ never care, never care [sic]. 
1878 DICKINSON, Cumé. Gloss. (E.D.S.), p. 66. 


2. Ree. A peculiar case of the loss of initial by absorption 
with a final x in the preceding word, appears in the name of a church 
in Southwark. It was the church of Saint Mary over she ree, that is, 
‘over the river,’ vee, speld also rhee, rhe, rte, being a rare or doubtful 
word, of which I find no early record. 


Vol. xxiv.] Attraction in English. 123 


Even to this daie in Essex I have oft observed that when the lower 
grounds by rage of water have been overflowen, the people beholding 
the same have said, All is on a rhe, as if they should have said, All is 
now a river. 1587 HARRISON, Deser. Eng. (H. p. 681.) 

Rhe. The course of water, and the overflowing of it. 

1847 HALLIWELL [to introduce the above quotation]. 

Ree, a river or flood. “All is in a ree,” that is overflowed with water. 
Essex, 2 1847 WRIGHT. 


Over the ree came to be written overtheree, overthere, and, with 
omission of the article, over ree, over rhee. ‘The tuching 7’s then 
melted into each other, and over prevaild, leaving Overee, Overe, 
Overie, Overy as an apparent proper name, in the possessiv form 
Overes, Overies, Overus ; the church being now St. Mary Overy. 


(a) Prior of Seit Mary Ovirthere. 1502 Arnold’s Chron. (1811), p. 248. 
Mary, ouer the ree in Southwerke, a priorye of Chanons. 

1502 Arnold’s Chron. (1811), p. 252. (Sim. twice more, p. 252.) 

Priour of Saint Mary Outheree. 1502 Arnold’s Chron, (1811), p, 258. 

I will not heere give notice how far they are deceived, which call the 

aforesaid church [Saint Marie over Rhee] by the name of Saint Mary 

Auderies, or Saint Mary ouer Isis or Ise. 1618 Stow, Survay, p. 24. 

A faire Church, called S. Mary over the Rie, or Overy, that is over the 

water. 1633 STOW, Survey, p. 450. 

(4) Seint Mary Ouerey Priory. ¢ 1502 Arnold’s Chron. (1811), p. 76. 

Saint Mary Overes. 1555 MACHYN, Dzary (Camden Soc.) (1848), p. 96. 

At Saint Mary-ouerus. 1604 The Meeting of Gallants (Percy Soc.), p. 28. 

S. Mary Oueries Chirch. 1618 Stow, Survay, p. 48. 


3- Riddle, a sieve. We find (a) *haver-riddle, a sieve for haver 
or oats, speld (4) haveridi/, a ME. form given by Halliwell (. 438), 
without a reference. 





§ V. Initial D gaind or lost. 


A. Initial d gaind. 

The cases of attraction now to be shown hav been hitherto 
wholly unnoticed. They arose in household or colloquial 
speech, and involv extremely common household or colloquial 
words, good and old. 


XVII. Good. This adjectiv, as a part of conventional formulas of 
greeting and farewell, originally prayers or precations, in some 
instances spred over to its noun, to which the @ thus extended 
became attacht. Thus Good even, contracted good cen, gooden, 
goden, became good deven, contracted good den, godden. 

The full form of the precation was God give you a good even. 


God gyve you a good evyn. Dieu vous doynt bon vespre. 
1530 PALSGRAVE, p. 867. 


124 Charles P. G. Scott. [1893. 


It soon became contracted: *God gi’ you good een, God ye good 
den, God-dig-you-den, Godgigoden, Godigoden, Godigeden, Godigodin. 
So God give you good morrow, contracted to God ye good morrow. 


Rom. ... Godden, good fellow. 
Ser. Godgigoden. I pray, sir, can you read? 
1599 SHAKESPEARE, AX, and F. ® 2. 55 (Q? p. 14; F! p. 55). 
Nur. Ispeake no treason. 
Cap. O Godigeden [1623 Godigoden]. 
1599 SHAKESPEARE, &. and F. iii. 5. 173 (Q? p. 67; F' p. 70). 
Nur. God ye goodmorrow, Gentlemen. 
Mer. God ye goodden [1623 gooden], faire gentlewoman. 
Nur. Is it good den [1623 gooden]? 
1599 SHAKESPEARE, &. and F. ii. 4. 95 (Q? p. 395 F! p. 62). 
Clo. God dig-you-den all, pray you which is the head Lad dy? 
1623 SHAKESPEARE, J.Z.Z. iii. I. (F! p. 130). 


For the contraction of Ged give you a good even to Godigoden, 
compare the contraction of God be with. you to God be wi’ you, God 
buy you, goodbye, goodby. 


God be with you, a dieu soiez. ¢ 1532 DEWES, Jntroductorie (1852), p. 919. 
Good b'w’y, gentlemen. 
1594 A Knacke to Knowe a Knaue. (Dodsley, ed. Hazlitt, vi. 553.) 
God buy you; fare you well. 1623 SHAKESPEARE, Ham. ii. 2 (FI p- 259). 
God ’uy, my Lord. 1623 SHAKESPEARE, s Hen. VI, iii. 2 (F1 p. 108). 
Good boy / with all my heart. 
1646 SUCKLING, Ballad upon a Wedding, |. 120. 
Bwye, interj. Bye! adieu. This, as well as good bye and good-bwye, is 
evidently corrupted from God be with you; God be wi’ ye, equivalent to 
the French 2 Dieu, to God. Awye, and good-bwye, are, therefore, how 
vulgar soever they may seem, more analogous than dye and good-bye. 
1825 JENNINGS, Somerset Gloss. p. 28. 


But the better abbreviation of God give you good even is obviously 
good even. It became the prevalent form, often contracted good een. 
So, later, good evening. 


(a) Good evyn, bon vespre. 1530 PALSGRAVE, p. 867. 
Good evenyng, bon vespre. Good evyn, bon soir. 
¢ 1532 DEwEs, /ntroductorie (1852), p. 918. 
Lul. Good euen to my ghostly confessor. 
1599 SHAKESPEARE, R. and F. ii. 6, 21 (Q? p. 45). 
The pawky auld carle came o’er the lee 
Wi’ many goode’ens and days to me. 
@ 1700 (?) The Gaberlunzie-Man (Child, Ballads, viii. 98). 
As I came by the Lowden banks, 
They bade gude e’en to me. 
a 1800 (?) Young Benjie (Child, Ballads, iii. 301). 
I crave your forgiveness, Master George, and heartily wish you good even. 
1822 Scott, Fortunes of Nigel. 


The spreding of the @ appears in the change of good even, to good 
deven (devon), and of good cen to good den, god-den, godden. 


Vol. xxiv.] Attraction in English. 125 


(6) Gud devon, dame, seyd he; 
Sir, sche seyd, welcum you be. 
¢ 1440 Sir Amadas,\. 110, (Weber, AZetr. Rom. iii. 248.) 


Gentlemen, good den, a word with one of you. 
1599 SHAKESPEARE, &, and 7. iii. 1. 34 (Q? p. 46; F! p. 64). 
See also the three quotations from the same play, above; and 
elsewhere in the same author. 
When the Spaniard saith God keepe you, a good houre haue you, God 
giue you health: and the Englishman God den or good euen, and the 


other like, I allow it for good salutation. 
1623 MINSHEU, Dialogues in Sp. and English, p. 49. 


The following passage is written as of date about 1620: 


Propera pedem, O Geordie, and god-den with you. ’ 
1822 Scott, Fortunes of Nigel, xxix. 
Godden. Goodeven. North. 1847 HALLIWELL, 
XVIII. Old. The way in which the adjectiv o/7 has affected the 
_ form of certain familiar household names seems never to hav been 
noticed before. O/d, beside its literal sense (1) ‘advanced in years,’ 
‘ aged,’ was and is extremely common in other uses; (2) ‘relativly 
{advanced in years,’ ‘senior,’ as a father compared to his son; 
(3) ‘long known,’ ‘familiar’; hence implying on one side special 
interest or affection, as o/d fellow, old boy; on the other, contempt 
or dislike, as o/d fogy, old scoundrel; the two phases being curiously 
mixt in the familiar names of the devil, Old Harry, Old Scratch, etc. 
In many cases o//, truly meaning ‘ aged,’ implies also ‘long known,’ 
‘familiar,’ and indicates thus a degree of interest or affection : 
Old King Cole was a merry o/d soul 
And a merry o/d soul was he. 
@ 1845 in HALLIWELL, Vursery Rhymes, p. I. 
Indeed, o/7 in colloquial use is so frequent that it becomes in many 
cases completely void of meaning. 
Old, adj. This word is constantly applied to anything or anybody without 
any reference to age. 
1887 PARISH and SHAW, Dict. Kentish Dial. (E.D.S.), p. 111. 
We ar now prepared to understand how o/7 might in careless 
unletterd use .affect a following name. Evidence is abundant, that 
old, in the familiar uses above mentiond, especially of frendly or 
contemptuous familiarity, was very common in the thirteenth and 
fourteenth centuries, as before and since, in connection with house- 
hold names ; and my proposition is, that when it thus recurd before 
a name beginning with a vowel or 4, there was a tendency to carry the 


126 ‘Charles P. G. Scott. [1893. 


final element of this naturally long-drawn word (0 vowel + Z liquid 
or semivowel + @ sonant) over to the name itself. And so, I think, 
Old Hick, whether the aged Hick of fourscore, or the senior Hick 
of forty seen walking with the junior Hick of five, or the jolly middle- 
aged Hick cald “ old” because he was, in the current phrase “ popu- 
lar with the boys” (sc. boys of forty-five or fifty or more) — O/d@ 
Hick, without the aspiration O/d ’Zck, became: O/’ Dick or Old Dick, 
and hence, when the epithet was not used, simply Dick. So Old Hob 
would become Old Dob, Old Hobbin would become Old Dobbin, 
Old Hobby Old Dobby, and Old Hodge probably Old Dodge. 

In dialectal form o/d was and is also ould, oud, owd, auld, aud, 
awd, aad; and Old Hick would be variously ould, oud, or ow’ Dick, 
auld, aud, aw’ or aa’ Dick. The ow’ or aw’ or aa’ may hav become 
confused with the @ in names like John a Nokes, John a Styles, etc. 
(see TRANSACTIONS, xxiii. 283), and perhaps in some cases with the 
indefinit article a, and so would drop out. Compare audacious, 
dialectal owdacious, outdacious (Tennyson, Village Wife): an auda- 
cious fellow in, dialectal speech a audacious fellow, becomes at last 
a dacious fellow (dactous, Peacock, Gloss. N. W. Lincolnshire). So 
an occasion, a occasion, becomes a@ ’casion, a ’cayshun (Holderness 
Gloss.). 

One poor Highlander, on his deathbed, is even said to have contemplated 

_the possibility of finding whisky in the next world. To the minister 

who had been trying to give him some idea of heaven he said: “ But, 

sir, will there be any whusky in heaven?” “Oh, no, Donald, there 

will be no occasion for that.” ’Caston or no ’casion,” said Donald, 
‘it wad be but dacent to have it on the table.” 

1893 Davip PryDE, Pleasant Memories of a Busy Life. 

(London Academy, 23 Sept. 1893, p. 251.) 

The record of the development of Dick, Dobbin, etc., from Hick, 
Hobbin, etc., is incomplete ; but there ar reasons for this incomplete- 
ness. The development took place in household, rural English 
speech of the thirteenth and fourteenth centuries. It would seldom 
find record; and the records of homely English speech of that 
period ar scanty. This homely English crops out now and then in 
Chaucer and Langland, and in pieces before their time; but it is 
not conspicuous until the plays or ‘ mysteries,’ which belong mostly 
to the fifteenth century. My examples then ar mostly late, and ar 
chiefly illustrations, not evidence ; but it can not be douted by any 
one verst in Middle English, that the homely and colloquial phrases 
here illustrated by examples of the fifteenth and later centuries, 
existed long before. 


Voli xtiv.] Attraction in English. 127 


For the use of o/7 before personal names in general, I hav many 
random quotations from the earliest times down. I take room here 
only to mention some of the names, annexing only the chance dates : 
Oulde Abraham (¢ 1360), ould Addom (1814), auld Aiken (1816), 
auld Ailie (1816), old Arthur (1602), Antele the olde (¢ 1440), 
Austyn fe olde (1362), auld Davie (1818), auld Downie (a 1700), 
auld Ellieslaw (1816), auld Elspeth (1815), auld Gibby (a 1700), 
old Hugh (1602), old Jacob (1816), old Jervie (1816), old John 
(1623), old Johnnie (1816), oulde Josephe (c 1430), auld Paull 
(1602), auld Rab (1816), old Richard (1600), old Roger (1823), 
ald Roger (1785), auld Saturne (1552), etc. 

I hav many quotations also illustrating the use of o/d7 before common 
nouns having a personal and often contemptuous reference ; as au/d 
doited carles (1816), auld companjeoun (1602), auld deevil (1816), 
old dote (¢ 1450), auld gowk (1816), ould hagge (1598), old hag 
(1816), auld hellicat (1816), auld hystoricience [historicians] (1552), 
auld crippled idiot (1816), old mon (a 1250), awlde mene (¢ 1440), 
old rogue (1816), oulde vylarde (¢ 1430), old wyfe (¢ 1425), etc. 

How easily the @ of o/d could wander off may be seen from the 
fact that in both English and Low German use o/@ in colloquial or 
dialectal speech often loses the d entirely. English 0/7, not alone in 
negro speech, becomes o/’, o/e, and in Low German o// is in inflec- 
tion usually o/. In Old Friesic we find besides a/d, old, auld, the 
forms a/ and of So North Friesic w//, for u/d (1837 Outzen, 


P- 375): 


Oold, alt, R. AS. eald, E. old, H. oud. Wenn dieser Wort am Ende 
verlangert wird, so wird by uns in der Aussprache das @ gemeinlich 
ausgestossen. De Olen, statt Oolden, die Alten, die Aeltern, die Vor- 
fahren. De Ole, der Vater, die Mutter. Seven mit der Olen: Mutter 
mit 6 Kindern. 1768 Bremisch-Niedersichsisches W Grterbuch, iii. 262. 


For the reduction of o/7 before a noun to o/’, o/e, even from early 
times, there is abundant evidence. It is found in Friar Geffrey (Gal- 
fridus) of the fifteenth century and in “ Uncle Remus” of the nine- 
teenth. 


Olde, or elde. Antiquus, vetus, veteranus, senex, grandevus, annosus 
(veteratus, P). [Next entry is:] Oé, for-weryd, as clothys, and other 
thyngys. Vetustus, detritus. 1440 Prompt. Parv. p. 363. 

[Hence old clothes, ol’ clo’es: in the mouths of street pedlers: o/’ clo’, 
even 0’ clo’.] : 

Bang! went queen’s-arm, o/e gander flopped 
His wings a spell, an’ quorked an’ dropped. 
1848 LOWELL, The Two Gunners (Biglow Papers, p. 164). 


128 Charles P.. G. Scott. [1893. 


This loss of @ after a liquid is quite ancient. I find go/ for gold 
(a 1300 Havelok, |. 357), don for lond (id. 1. 340). 

We are now prepared to enumerate the names which involv, as I 
believ, the conceald operation of Attraction from odd. 


1. Hick, a familiar form of Richard. (See before, p. 113.) 

Old Hick, owd Hick, awd Hick (Ick), old Dick became ol’ 
Dick, ow’ Dick, aw’ Dick, and so simply Dick. Hence the diminu- 
tivs Dickie, Dickon (see below), and the surnames Dick, Dicke, in 
the possessiv form Dicks, formerly Dickes, Dykkys, Dykys, speld also 
Dix, formerly Dixe,; with the filial addition, Dickson, speld also 
Dixon, Dixson, formerly Dicson (1375 Barbour, Bruce), Dikson, 
Dykson, Dyxon (1474 Paston Letters, iii. 174), Dyxson (1479 
id. 258). 

(4) Peter Dicke, Thomas Fitznell, sherefs; the x. yere [sc. of king John: 

namely 1208]. ¢ 1502 Arnold’s Chron. (1811), p. xx. 
Dick o’ the Cow. ai 596 Dick o the Cow [title] (Child, Ballads, vi. 69). 
And, Dick, she dances sucha way. 

1646 SUCKLING, Ballad of a Wedding. 

“ Hae ye ony tidings? Hae ye ony speerings, Hobbie? — O callants, 
dinna be ower hasty,” said o/d Dick of the Dingle. 

1816 Scott, The Black Dwarf, viii. 


There’s Dick, who sold wine in the lane, 
And old Dickey himself did not tope ill. 
¢ 1825 Writing and Reading (Univ. Songster, i. 74). 


2. *Hickon, in surnames Aicken, Hickin, Hickins, diminutiv of 
Hick (see above), also *Higgon, ME. Hegon, in surnames fiiggin, 
Figgons, Higgins, Hyggins). Old * Hickon became old Dickon, speld 
also Diccon, Dicken, ME. Decon, with variant Diggon, ME. Degon. 
From Dickon, Dicken ar derived the surnames Dickens, Dickins 
(formerly Dikkins, Dicons), Dickinson, Dickenson (formerly Dickon- 
son, Dyconson, Dykynson), Digginson, Digison, parallel to Hickins, 
Figgins, Higginson. 


(a) Barow and Hegon and all the Lord Moleynys men that wer at Gressam. 
. And ther xuld no mor com with him but egon and on of his owyn 
1450 Paston Letters, ed. Gairdner, i. 110, I11. 

(6) The while the Degonys (ms. dogonys) domes were so endauntid. 
1399 [LANGLAND], Richard the bibacaies iii. 351. 


Tyll Degon and Dodyn that mennys doris brastyn, 
And were y-dubbid of a duke ffor her while domes 
Awakyd ffor wecchis and wast that they vsid, 
And ffor her breme blastis buffettis henten. 
1399 [LANGLAND], Richard the Redeless, iii. 362. 
[Degon and dobyn, evidently Diggon and Daeddin, both common names for 
country bumpkins, here used in contempt of the upstarts who used to 
burst in men’s doors, and rob them. 
1886 SKEAT, note l.c., vol. ii. p. 302.] 


Vol. xxiv.] Attraction in English. - 129 


And ay he sange in fayth decon thou crewe. 
@ 1529 Sapa Bowge of Court. 
[This song is again mentioned in Why come ye not to Court. 
1790 Ritson, Ane. Songs, li.] 
Diggon Davie! I bidde her god day; 
Or Diggon her is, or I missaye. 
1579 SPENSER, Shepheards Calender, Sept. 1. i. (and 10 times more). 


**T’ll speak him fair,” he said, “as au/d Dickon advised me.” 
1816 Scott, The Black Dwarf, viii. 


3. Hob. Od Hod would, by the process I hav described, result 
in old Dob; for which evidence exists in the surnames Dodds 
(formerly Doddes, Dobbis, Dobbys), and Dobson (formerly *Dodde- 
son, Dobyson), along side of Hodds and Hobson. 

(4) There was a man and his name was Dod 


And he had a wife and her name was Mob. 
@ 1845 HALLIWELL, Vursery Rhymes, p. 75. 


4. Hobbin, ME. * odin, * Hobyn, familiar form of Rodin, Robbin. 
See p. 116. Hence the surnames Hodin, Hobdbin, Hobbins. 

Old Hobbin would giv, after the manner above shown, o/d Doddbin, 
and so simply Doddin. Company Dobdy from Hobby (below). 
Hence the surnames Doddin, Dobin, Dobbyn, Dobbins, Dobbinson, 
Dobinson, parallel to Hobbin, Hobbins, etc. 

(4) Tyll Degon and Dodyn that mennis doris brastyn. 


1399 [LANGLAND ], Richard the Redeless, iii. 362. 
[See quot. under Dickon, above. ] 


_ Dobbin came to be a familiar name for a horse ; it is often con- 
joind with the very adjectiv o/¢ to which the name Doédin owes its 
initial consonant. 


Dobbin my philhorse. 1623 SHAKESPEARE, J. I. ii. 2 (F p. 168). 
My o/d Dodbin stands in the little stable beside the hencoop.... Take 
Dobbin, and do not forget to leave your own horse instead. 
1825 Scott, Peveril of the Peak, xxiii. 
Horses’ names. ... Dick, Dodézn, Doctor . . . Nob. 
1889 PEACOCK, Manley and Corringham Gloss. (E.D.S.), p. 279. 


Hence dodsin as a common noun, ‘an old horse.’ 
Dobbin. An old jaded horse. 1847 HALLIWELL. 
Dobbin, a familiar name for a horse. 
1875 NODAL and MILNER, Lance. Gloss. p. 107. 
Hence dodésin, a timber cart. 


Dobbin, s.— A timber cart. Dobdin wheels, the very high wheels of the 
same. 1877 LEIGH, Cheshire Gloss. p. 63. 


130 Charles P. G. Scott. [1893. 


5. Hobbie, also speld odéy, diminutiv of Hod, in fact an alterd | 


form of Hobéin. See above. Old Hobby, for which I hav no 
example at hand, may be the source, in the way above shown, of 
dobby, dialectal daudy, ‘a silly of@ man’ (H.), ‘a fool’ (H.), also 
‘a kind of spirit’ (H.)—the last sense going to confirm the con- 
nection here asserted with Hod, Hobdy, which names ar often applied 
to spirits or goblins. Doddy also exists as a surname, Dobby, Dobbie, 
Dobbey. 


(46) Dobby. A fool, a childish old man; also, a sprite or apparition. orth. 
1790 GROSE, Prov. Gloss. 
Dobby. 1822 IRVING, Bracebridge Hall, ii. 183-6. (H. p. 307.) 
Dobby. A fool; a silly old man. Also a kind of spirit. Vorts. The 
dobbies seem to be similar to the Scottish Brownzes. 
1847 HALLIWELL, p. 307. 
Dauby. A fool. Northumé. 1847 HALLIWELL, p. 293. 

6. Hodge, a familiar form of Roger, Rodger. See p. 119. 

From Old Hodge may come the surname Dodge. 

I find “ dodge, a dog (Alleyn Papers, p. 32)” in Halliwell, an 
assibilated form of ME. dogge. This can hardly be the source of 
the surname Dodge, as the name of dog or hound was felt to be 
strongly opprobrious, and however freely applied to a man, it would 
not, like fox, wolf, ud/7, and other animal names felt to be in some 
way complimentary, admit of a humorous or complimentary inter- 
pretation and thus enter into general use. Hence the surnames Dag, 
Hound, or Hund, never common, ar now practically extinct. 


(a) Old Hodge Bacon and Bob Grosthed. 
1664 S. BUTLER, Hudibras, pt. I. iii. 224. 

(4) Dodge [surname]. 1891 NV. Y. Directory. 

7. Hod, a shortened form of Hodge, as Rod is of Rodge, for 
Rodger, Roger. From old Hod, or from Hod by conformity with the 
perversion of Hodge, may have come Dod, known as a surname, 
Dod, Dodd, also in patronymic form Dodds and Dodson. The 
diminutiv Hoddy occurs as a surname. The terms hod, hoddy, dod, 
doddy, run thru a remarkable series of words meaning ‘something 
short or squat,’ as a short person, a snail, etc., but these words ar 
partly associated with other roots, and it can not be safely asserted 
that they ar derived from the name Hoddy or Hodge. 

The surname Dodson, so unhappily familiar to Mr. Pickwick, may 
be partly from Davidson, as Daud is a contraction of David. 

The next name I consider does not appear to have been common 
in the Middle English period, but it was in use, and it underwent 
the same kind of change. 


Vol. xxiv.] Attraction in English. 131 


8. Andrew, dialectally or formerly Androw, Andro, Andre, ME. 
Andrew, Andreu, OF. Andreu, Andriu, Andrieu, Andre, LL. 
Andreas, from Gr. ’Avdpéas ; in popular form *Andrie, Andie, Andy. 
Hence by influence of o/d, as in the previous cases, or perhaps by 
mere conformity, the forms Dandrie, Dandie, Dandy. We may 
suppose, as a contributing cause, the riming variation of the name 
Andy by children and nurses — Andy Dandy, Handy Andy, etc. 


(a) Androwe, Andreas. 1483 Cath. Angi. p. 9. 
Andro... Androw. 
1552 LyNDEsAY, The Monarche (E.E.T.S.), 1. 4639, 4750, 4790, etc. 


So other quotations for Andrew under Saint Andrew, IX., p. 101. 


(4) Call’d in of Dandrie, Hob, and Jock. 
¢ 1650 The Raid of the Reidswire (Scott, Minstrelsy, 4 Re 
Child, Ballads, vi. 133). 

In the small village of Lustruther in Roxburghshire, there dwelt in the 
memory of man, four inhabitants called Andrew or Dandie Oliver. 
They were distinguished as Dandie Eassil-gate, Dandie Wassail-gate, 
Dandie Thumble, and Dandie Dumble. 

1815 Scott, Guy Mannering. Note 5. 

Dandy Dinmont (xxii. p. 101 — first mention). ... Dandie Dinmont 
(xxiv. p. 109). ... Dandy (xxiv. p. 109). “Mr. Andrew Dinmont 
(xxvi. li.).. . . Andrew Dinmont (xxxvi.). [Dandie is the usual 
spelling throughout the book. 1815 Scott, Guy Mannering. 


In this Dandy, a familiar form of Andrew, I find the hitherto 
undiscoverd origin of the common nouns dandy and dandiprat. 
That dandy and dandiprat ar connected, there should be no dout. 
But the connection is peculiar. Dandiprat is at least three centuries 
older than dandy. I therefore treat it first. 

Dandiprat, in other forms dandyprat, dandy-prat, dandie-prat, 
dandeprat, also in two parts dandy prat, dandy pratt, is found in two 
senses. In a personal sense it signifies ‘a little fellow, a dwarf, an 
urchin,’ and fs used generally in contempt. It is often attributiv. 


Yet as the giantes pawes pat downe dandipratts, 
So shall we put downe these dandiprat brag brattes. 
1556 J. HEywoon, Spider and the Flie. 
(MY. & Q., 29 July, 1893, p. 82.) 
A cockney dandiprat hop-thumb. 1582 STANYHURST, tr.  Pncid, iv. 349. 
Nano, a dwarfe, or dandiprat [1611 and 1659 ad “a twattle’’]. 
1598 FLorio., 
The vile dandiprat. 1607 BREWER, Lingua, iii. 3. (Richardson.) 
Vn manche d’estrille. A dwarfe, elfe, dandiprat, low scrub. 
1611 COTGRAVE. 
A Dandiprat or Dwarfe, ex. B. Danten, i. ineptire, & Praete, i. sermo, 
nugee, fabulze: Solent enim Nani ad sermocinandum ineptiores esse. 
1617 MINSHEU. 
The smug dandiprat smells us out. 
1622 MASSINGER, Virgin Martyr, ii. 1. (Richardson.) 
A dandiprat or dwarfe, v. Enano. 1623 MINSHEU, Span.-Eng. Dict. p. 284. 


- 


132 ‘Charles P. G. Scott. [1893. 


A Dandeprat, ora dwarfe. Een dwergh, oftedwerghsken. 1648 HEXHAM. 


Sometimes with lacings and with swaiths so strait, 
For want of space we have a Dandiprat. 
1653 in BULWER, Artificial Changeling. 
“Tt is even so, my little dandze-prat— but who the devil could teach it 
thee?’ “Do not care about that,” said Flibbertigibbet. 
1821 Scott, Kenilworth, xxvi. 
Little Jack Dandy-prat was my first suitor. 
@ 1819 in HALLIWELL, Vursery Rhymes, p. 92. 


In the second sense, by the record apparently older, it is “a small 
silver coin current in England in the sixteenth century” equal in 
value according to one statement (1600) to three half pence. The 
name was not official, but popular. > 


I coyle with money, I trye the currante from the badde. Jesluys.. . 
Coyle out the dandyfrattes and Yrisshe pence: eslisez les dandyprattes 
et les deniers dIrlande hors de la reste. 1530 PALSGRAVE, p. 498. 

The king’s grace went over with a ten thousand men to conquer all France, 
and spent haply ari hundred thousand pound, of which he saved the 
fourth part in the dandypra/s, and gathered at home five or six hundred, 


or more. 1530 TYNDALE, Practyse of Prelates (Parker Soc., 1849), 
p- 306. (Spelling modernized.) 

Dandiprat. 1542 R. RECORDE. (NV. & @Q., July 29, 1893, p. 82.) 

Danaiprat (a coin). @ 1600 in ELLIs, Orig. Letters, ser. iii. vol. i. 


(Oliphant, Mew Eng. ii. 385.) 
And for such stuffe passe not a Dandy Pratt. 

@ 1600 (?) Dialogue between Comen Secretary and Felowsy (see Beloe's Ss 
Anecdotes, vol. i. p. 890; Gent. Mag., 1819, Part II. pp. 7, 8; 
Gent. Mag. Libr., 1884, p. 142). 

3 halfe-pence maketh 1 Dand?prate. 
1600 T. HILL, Arithmeticke, i. 13. (C.D.) 
Shall I make a Frenchman cry O! before the fall of the leaf? Not I, by 
the cross of this Dandyprat. 
1602 MIDDLETON, Blurt Master Constable, ii. 1. (C.D.) 
A Dandiprat or Dodkin, so called because it is as little among other 
money as a Dandiprat or Dwarfe among other men. 1617 MINSHEU. 
K. Henry the seventh stamped a small coine called Dandy prats, & first, 
as I read, coyned Shillings. 1637 CAMDEN, Remaines, p. 188. 
A Dandeprat, or a dodkin [erroneously explained as] Een kleyn man, 
ofte manneken. [See quot. 1648, above. ] “1648 HEXHAM. 
Scant worth a Dandeprat. Triobolaris homo, homo trioboli. 
1677 HOLYOKE. 
A small silver coin, struck by Henry VITI., of little value, called a dandy 
pratt. 1819 Northampton Mercury, April 17 (in V. & @., 
8th ser. iv., Aug. 19, 1893). 


Mr. Henry H. Gibbs has brought out a statement from Mr. Head 
of the coin department of the British Museum that there was no 
such coin of Henry VII. A further statement from the same source 
is thus exprest by Mr. Wroth. 

We can only suppose that it [the coin called dandiprat] was some small 
coin of the Tudor period. The 2d. piece (half groat) of Henry VII. 


has a small head of the king-on it (so, also, however, has the shilling of 
the same reign), and the silver penny of Henry VII. has a small seated 





Vol. xxiv.] Attraction in English. 133 


figure of the king on it. Your ingenious explanation that the name 
dandiprat was given because of the small head, or the small figure on 
the coin, is therefore possible. ... I am rather inclined myself to 
believe that dandipfrat was merely suggested by the small size of the 
coin. ... 1893 W. WRoTH, in V. & Q., 8th ser. iv., Aug. 19, p. 153. 


Without laying stress on the forms dandy pratt, dandy prat, 
dandy-pratt, dandie-prat, as cited, which imply a name of two origi- 
nal terms, I am inclined to conjecture that in the reign of Henry VIL., 
to whom the first coinage of dandiprats is ascribed, there existed, 
probably in London, or in some other place where he would be often 
seen by the populace, a dwarf named Andrew Prat or Pratt, familiarly 
known as Dandy Prat; that his name past into popular speech, like 
Zom Thumbé, as a synonym for smallness of size ; that when the little 
coins wer issued, they wer cald at first in jest Dandy Prats, as it was 
the passing custom a few years ago to call anything huge of its kind 
a Jumbo, after the great elephant of that name, and as the name 
of Daniel Lambert, the big man, was at one time used as a general 
term for anything big. 

That dwarfs at the time in question and later wer frequent objects 
of popular notice, needs no demonstration. For the naming of a 
coin after a man (tho for other reasons) compare aschison, harring- 
ton, Harry noble, louis. Whether the supposed Andrew Prat thus 
etymologically excogitated, really existed, I must leav undetermind. 
I may note that Pratt is a common surname in the sixteenth cen- 
tury, as it is now; and that Prat was then a common spelling. 


- 


A mery Play betweene the Pardoner and the Frere, the Curate and Ney- 
bour Pratte. 1533 HEywoop [Title] (1860 Halliwell, Dict, Old 
Plays, p. 188; Dodsley, ed. Hazlitt, i. 198). 


I find “ Pretty Pratt” used as a playful or fond address to a little 


boy, early in the fifteenth century. Perhaps he was a relativ of little 
Dandy Pratt. 


Howe! Prittie Pratte, my messinger ! 
Come heither to me, withouten were. 
¢ 1430 (ms. 1592) Chester Plays, i. 173. 


Here we hav all the conditions for the rise of dandiprat as a 
humorous popular name for a small coin ; and a popular name it was, 
much like our modern greenback, shinplaster, etc. 

Dandy, as applied, half kindly, half in contempt, to a trim little 
fellow, a fop, does not appear on record before this century. It has 
nothing to do with the French dandin, of which the sense (“a mea- 
cocke, noddie, ninnie, a hoydon, sot, lobcocke; one that knowes not 


134 Charles P. G. Scott. [1893. 


how to looke, and gapes at every thing he knowes not”) (1611 Cot- 
grave) is far from that of dandy. I think dandy is either due to 
dandiprat taken as ‘a little fellow,’ hence ‘a trim little fellow,’ ‘a 
fop,’ and hence ‘any man conspicuously neat in his dress’; or is an 
other, independent, use of the personal name Dandy for Andrew. 
When it was coming into literary use it was regarded as a “slang 
word”; which implies that it was of popular, perhaps local or anec- 
dotal, origin, and was current some time before appearing in print. 

Origin of the word Dandy: This term, which has recently been applied 
to a species of reptile very common in the metropolis, appears to have 
arisen from a small silver coin, struck by King Henry VIL., of little 
value, called a dandy pratt; and hence Bishop Fleetwood observes, the 

term is applied to worthless and contemptible persons, 
1819 Northampton Mercury, April 17 (in MV. & @Q,, 

8th ser. iv., Aug. 19, 1893). - 

Dandy was and is used also as an adjectiv of admiration. It used 
to be common in songs. It is now rife in popular speech, to express 
any kind of approbation for which an exact descriptiv does not sug- 
gest itself. At the World’s Fair in Chicago, where this paper was 
red, I heard a Western farmer, in the Horticultural Building, express 
his admiration of a dozen diverse things by the same comprehensiv 
formula, ‘‘ Ain’t that dandy /” Anything that meets approbation as 
being neat, fine, nice, is ‘a dandy.’ 

The Alert was agreed on all hands to be a fine ship, and a large one: 


“A crack ship.” —“ A regular dandy,” etc. 
1840 R. H. Dana, 7wo Years before the Mast (1842), p. 205. 


9g. Woman. A clear case of the development of an initial @ from 
the @ of a preceding o/d, is the following: Az old woman, the old 
woman, my old woman, his old woman, ar common phrases, the 
possessiv especially in rural or low use. A pesant, a costermonger, 
a jockey, wil speak of ‘ my old woman,’ meaning either his wife or 
his mother. In dialectal use woman often becomes oman, coman 
(commonly written ’oman), aspirated hooman; and old 'oman, old 
‘ooman has become in many cases old dooman. 

(2) By your pore bede oman and cosyn, Alice Crane. 


¢ 1455 Paston Letters, ed. Gairdner, i. 343. 


Lua. For shame, o’man [read ’oman]. 

1623 SHAKESPEARE, A/. W.W. iv. 1 (F! p. 53). 
Euans. O’man, art thou Lunatics? 1623 Jd. (F! p. 54). 
Lua. By yea, and no, I thinke the o’man is a witch indeede: I like 
not when a o’man has a great peard; I spie a great peard vnder his 
muffler. 1623 /d. iv. 2 (F! p. 55). 

“ How ar’ry jung umman,” sez a, “how dost do?” 
1846 Spec. Cornish Prov. Dialect, p. 24. 


Vol. xxiv.] Attraction in English. 135 


Oman. A woman. Var, dial. 1847 HALLIWELL. 
Hlooman, The common pronunciation of Woman in many of our vil- 
lages. 1854 BAKER, Northampt. Gloss. i. 335. 


Ooman, a woman, 1881 SMITH, /sle of Wight Words (E.D.S.), p. 24. 


The term o/d woman is of course of innumerable occurrence. 


He has bin . . . greeuously peaten as an o/d o’man. 
1623 SHAKESPEARE, J/.W.W, iv. 2 (F! p. 56). 
?Ooman. My old ’ooman is the usual term used by an old labourer in 
speaking of his wife. 
1876 South Warwickshire Provincialisms (E.D.S.), p. 130. 
* My ole ’oman done gone en tuck mighty sick,’ sezee. 
18381 J. C. HARRIS, Uncle Kemus, p. 76. [Sim. p. 46.] 
Awild,.—* Awld”’ is specially used as a term of familiarity, or even endear- 
ment. Thus a man would say of his wife, “ My awld ’ooman ’ool hev 
dinner jus ready vor us.’ 
1888 LowsLrEy, Berkshire Words (E.D.S.), p. 43- bee also p. 20. ] 


(4) Dooman. Awoman. Var. dial, 847 HALLIWELL. 
Dooman,@wooman; only used when preceded by o/d <a ’ dooman, old 
oman. 1881 SMITH, Js/e of Wight Words (E.D.S.), p. 9. 


Pure, nice, excellent. ‘She’s a pure wold dooman, 
1881 SMITH, Jsle of Wight Words (E.D.S.), p. 26. 
’Ooman.— Woman. When awld precedes *ooman the d is carried on, and 
’00man is sounded dooman. 
1888 LowsLEy, Berkshire Words (E.D.S.), p. 121. 


In some examples we hav o/d dooman, with two mi) just like old 
Dick. 

It is a beutiful and touching custom of the profession to which I 
hav the honor to belong, after the citation of a dozen incontrovertible 
causes of action, or offering seven inexpugnable answers to a plead- 
ing, to go right on with an other as a “further and separate” cause 
of action, or answer, as the case may be; to the complete confusion 
of the other side, until it does the same, m.m. Let me do the like 
in this freer forum. 

In further evidence that the initial D in Dick, Dickon, Dobbin, 
Dobby, etc., is derived by attraction from o/7/, I cite together two 
significant facts: first, that these names and the parallel forms “ick, 
fickon, Hob, etc., and the original of Hob, namely Robert, ar often 
used with reference to the devil or to less malignant demons or gob- 
lins ; and secondly, that o//7 as a term implying at once antiquity and 
contempt or familiarity, has for centuries been a stock epithet of the 
devil. The proofs collected for this purpose I must omit; but I 
may use them on an other occasion. 

It may be taken as a rule in philology and indeed in all sciences 
involving time, that when a cause is obscure, there is more than one 
cause. While I believ that the initial ¢@ in question originated mainly 
by attraction from o//, it is probable that the process was assisted by 


136 Charles P. G.- Scott. [1893. 


a similar attraction from good. I hav shown how the d of good even, 
good cen, spred over to the noun. It is clear that the same might 
happen with good before personal names. That good was a very 
common epithet, often merely conventional, before personal names, 
is wel known. 

I omit the numerous examples showing the frequency of this use 
of good in Middle English — good boy, good brother, good cosin, good 
em (‘uncle’), good mother, good nece, good sir, good sister, good son, 
etc. So good John, good Mary, good Mawdleyn, etc., and good alone. 


I think it probable that further research along the line thus opend 
would reveal other personal names and surnames beginning with @ 
thus derived from odd. 

An instance of the running together of a particle ending in @ with 
a word beginning with #4, and the consequent absorption of the 4 
appears in the following: Noah sends out the raven, saying: 

Pou arte ful crabbed and al thy kynde, 
Wende forthe bi course I comaunde pe, 
And werly watte andySer be wynd, 


Yf bou fynde awdir lande or tree. 
¢-1430 York Plays, ix. |. 213, p. 52. 


[Read ‘and werly watte [it would be originally zwzte], and hyper be wynd,’ 
that is, ‘and warely wit (carefully observ), and hither turn thee.’] 


XIX. God. The medieval Englishman, as wel as the medieval 
Frenchman and the rest of medieval Europe under the fostering care 
of the Roman church, was extremely apt in that kind of piety which 
consists in the frequent utterance of the name of God and the names 
of saints. As the intervals of piety wer fild up by the utterance of 
profanity, which employd precisely the same vocabulary (as Mr. 
Smallweed used the deprecations of the litany, “from battle and 
murder, and from sudden death,” as the handiest source of impre- 
cations he could think of), the medieval Englishman had much occa- 
sion to use his Maker’s name. This led to economy. In the first 
place, God was assimilated to a following word where convenient : 


1. God wot, ‘God knows,’ was in ME. assimilated to god. dot, 
god dote, goddot, goddote, godote. 


Goddot, Goddoth. a 1300 Havelok .... (Often.) 

“Nai, goddot,” said bat felun [var. for-sof Fairf.; omitted in other 2 mss.]. 

¢ 1300 Cursor Mundi (Cotton MS.) (E.E.T.S.), 1. 773- 

Godote [var. Goddote, God wat] saidioseph ... . ¢ 1300 Jd. 1. 4473. 

[So godote, var. goddote, |. 4491, sim. 4612; goddote, var. god dote, godote, 
god woote, |. 3729; god dote, var. godote, 1. 15983; etc. ] 


a 


Vol. xxiv.] Attraction in English. 137 


Compare with this the fact that God (’s) will has become Goddil in 
some places: 


“ Goddil for God’s will, in Yorkshire and Lancashire. 
1843 WAY, Prompt. Parv., p. 201, note 2, 


2. God yield you, that is, ‘God pay you,’ ‘God reward you,’ was 
once very common. It fel into God ’ield you, God ’ild you, and 
this was sometimes expanded into God dild you. 


(@) God yelde the, frend. ... ¢ 1374 CHAUCER, 7%. and Cr. i. 1055. 
The highe father of heaven I praie 
To eylde you your good deed to daie. 
¢ 1430 (ms. 1592) Chester Plays, i. 169. 
God yeld you, brother, that it so is 
That thou thi hyne so wold kys. ¢ 1450 Towneley Myst., p. 48. 


To begyn. God yeld yow for my hatys [hats]. 
1469 Paston Letters, ed. Gairdner, ii. 355. 
God-ild-you. ¢ 1600. (H. p. 407.) 
How you shall bid God-ey/d vs for your paines. : 
1623 SHAKESPEARE, Macbeth i. 6 (F! p. 134). 
(4) King. How do ye, pretty Lady? 
Ophe. Well, God dil’d you. 1623 SHAKESPEARE, Ham. iii. (F! p. 273). 
Blessing and cursing ar often associated in Scripture and they 
may result in similar changes of speech. But I pass the cursings by. 
In the history of drot, drat, and similar forms, there is a kind of 


_ Attraction which must here be left with a mere mention. 


B. Initial D lost. 


XX. Deal. An instance of the loss of initial @ by attraction to 
or absorption in a preceding final @, appears in @ good deal, often 
pronounced @ good ’eal. So a@ great deal (often assimilate to @ 
gread deal) is often heard as a great ’eal. I hav heard this pro- 
nunciation from all classes of people. Children use it also; and I 
hav been askt by a child who had used or heard the phrase a good 

"eal, “‘ What is ’eal?” 


A bousand soulis ber-in bai bren. 

“ Alas,” sayd Poule, “here is gre¢ deel!” 

¢ 1426 [AUDELAY], Zhe X/ Pains of Hell, \. 51. 
(Old Eng. Misc., E.E.T.S., p. 212.) 


It is warryed a@ grete dele. 1450 Towneley Myst., p. 32. 


XXI. Dish. An instance in which initial @ in the second ele- 
ment of a compound has been absorbd in the final @ of the first ele- 
ment, is *stand-dish, a dish or standing receptacle for ink. As the 
thing was clearly a stand (we call it now an :7k-sfand), and was not 


138 Charles P. G. Scott. [1893. 


always clearly a dish, the second element yielded to the first, and 
dish took the form of the suffix -isk. I hav not yet found a single 
instance of the full form *s¢and-dish; nor any instance of standish 
earlier than the end of the sixteenth century. 


(4) Pausing a while ouer my s¢andish, I resolued in verse to paynt forth my 
passion. 1592 NAsH, Pierce Penntless, p. 5. 
Calamaio, Calamaro . . . a standish, or a pen and inkehorne. ...  Scrit- 
toio, a standish, an inke-horne. 1598 FLorRIo. 
Cabinet d’Allemagne. A kind of standish ; or a small cabinet seruing 
for, or hauing in it, a standish. 1611 COTGRAVE. 
A box-standisk. Cabinet d’Alemaigne. 1632 SHERWOOD. 
An Ink-horn, or a standish, Atramentarium. 

1677 Compleat Clerk, opp. sig. Vvv. 

The massive silver zz4standish which stood full before her. 
1819 Scott, Bride of Lammermoor. 





§ VI. Initial S gaind or lost. 


A. Initial S gaind. Cases in which a final s has spred 
over to the beginning of the next word, producing two s’s, 
or drawing the final s out of its place. The words from 
which the Attraction takes place ar horse, zce, any and what 
with the ME. possessiv suffix, 4zs and other possessivs, yes, 
and probably other words. Among the nouns affected ar 
courser, a dealer in horses, zck/e, an icicle, yok/e, an icicle, 
and 47m, kind or sort. The cases ar curious and involvd, and 
call for considerable illustration, 


XXII. One case involving horse. 


1. Courser, early mod. Eng. also coursar, coarser, corser, ME. 
corser, corsser, corsere, cursoure, cursure, coresur, from OF. courtier 
(F. courtier), a reduced form of couretier, courretier, corretier, 
couratier, coratier, curatier, Prov. couratié (Roquefort), It. curar- 
tere, ML. reflex corraterius, coraterius, coratiarius, curaterius, a 
broker, agent; OF. courater de chevaux, a horse-trader ; in which 
use the word was taken over into English. The OF. courater, cora- 
tier, curatier, represents a ML. *curatarius, equivalent to the usual 
L. and ML. curator, which is in OF. *cureor, *coreour, cureur, and 
curatour, a factor, agent, tutor, curator ; cureur de chevaulx, a horse- 
trader. Cowurser, in the first instance I find, occurs in the general 
sense of ‘trader.’ 


\ 


Vol. xxiv.] Attraction in English. sthgas) ST 


pei ben corseris & makers of malt, & bien schep & neet & sellen hem for 
wynnynge, & beten marketis, &c. 
¢ 1380 WiciiF, Eng. Works (E.E.T\S.), p. 172. (C.A., p. 77+) 


In the next instance it refers to a horse-keeper or groom, 


’ 


Foles with, hande to touche a corser weyveth; 
Hit hurteth hem to handel or to holde. 
¢ 1420 Palladius on Husbondrie (E.E.T.S.), p. 135. 


The next courser we meet is a young man who steals a colt and 
offers to sel it for “ten mark of sterlynges.” He says it is a mighty 


fine colt : 
* For noon of all thy elderynges 
Hadde noo swych in stabele; 
For emperours sone, or for kynges, 
Hyt ys profytabele,” 


The buyer acts in the regular modern fashion, and is duly taken in ; 


Florent answerede to the corsere : 

“ Me thyngeth thou louest hyt to there [read dere]; 
Sterlynges ne haue I non here, 

As thou gynnest craue: 

Here beth ten pound of florens clere; 

Wylt thou ham haue 


For that colt that ys so bold?” 
The corser seyd, “Tak me that goold: 
To no man schuld hyt be sold 
Half swych a chepe.” 
He tok the florens all vntold; 
Away he lepe. 
¢ 1430 Octovian Imperator, ll. 807-821, (Weber, Metr. Rom. iii. p. 191.) 
Corsoure of horse. Mango. 1440 Prompt. Parv., p. 94. 
Wyth hem they toke stedys sevyn, 
Into Almayn they can ryde; 
As a coresur of mekyll pryde 
He semyd for to bee. 
¢ 1450 E7/ of Tolous, 973-978. (Ritson.) 
[This is Halliwell’s (and Wright’s) “ coresur, a courier,” for which no 
example is given. Halliwell gives also “corretier, a horse-dealer,” 
without reference. This is merely the OF, form, not a ME, form.] 


The corsser hathe his palfrey dy3t 
All reydy for to sell. 
¢ 1460-70 The Good Wyfe Wold a Pylgremage, \. 47 
(E.E.T.S.), 1869, p. 40. 

And if Bayard be onsolde, I pray yow late hym be made fatte ageyns the 
Kynge come in to the contre, what so ever I pay for the kepyng of 
hym, and I schall wete how goode a corser I schall be my selfe at my 
comyng in to the contre, be the grace of God, who have yow in kepyng. 

1489 WM. PasTON, in Paston Letters, ed. Gairdner, iii. 354. 

Corsers of horses by false menys make them loke fresshe, 

1519 HORMAN, Vulgaria. (Way, Pr. P., p. 94.) 

A corser is he that byeth all rydden horses, and selleth them agayne. 

1523 FITZHERBERT, Boke of Husbandry, sign H. 2. (C.A., p. 77-) 

Courser [coursar, p. 210] of horses, covrtier de chevaulx. 

1530 PALSGRAVE, p. 209. 


140 «(°° Charles P. G. Scott. [1893. 


He can horse you as well as all the corsers in this towne: il vous scayt 
aussi bien monter que tous les courtiers de chevaulx en ceste ville. 


1530 /d. p. 588. 


The word early underwent a corruption from corsere to *cosere, 
cosyr, coyseyr. (See further, p. 142.) 


Hic mango, a cosyr. c 1450 Nominale. (Wright, Vocad.? 684, 1. 40.) 
A Coyseyr of hors, Mango. 1483 Cath. Angl., p. 77. 


The phrase courser of horses (corsoure of horse, 1440 Prompt. 
Farv.), a translation of the OF. courtier de chevaulx (Palsgrave), 
couratier de chevaux (Roquefort), subsequently took the form of a 
compound, horse-courser. ; 


Mango. ... An horse coarser that pampreth and trimmeth his horses 
for the same purpose [‘to sell them the deerer ’]. 
1565 COOPER, Thesaurus. 
He that letteth horses or mules to hire: a mule-letter [1 580 muletter] : 
an horse courser: an hackneyman. Veterinarius. . . mango. 
1573 BARET, Alwearie, H. 602 (1 ee H. 650). 
When horsecorsers beguile no friends with Iades. 
¢ 1576 GASCOIGNE, The Steele Glas (repr. Arber), p. 79. 
A horse corser, a hackney man, a horse rider, a horse driver, a cariour, or 
a carter. 1578 FLorio, Second Frutes, p. 43. (C.A., p. 77-) 
Cozzone, a horse-courser, a horse breaker, a craftie knaue. 1598 F LORIO. 
(Also horse-courser, s.v. palafreniere; horse-coarser, s.v. scozzonato.) 
Courratier de chevaux. A horse-courser. 1611 COTGRAVE. 


I omit many later quotations. The word disappeard from general 
use about the middle of the seventeenth century, and was tuckt away 
in the dictionaries. 

Horse-courser, by a spreding of the final s of horse to the next 
element, came to be pronounced and written horse-scourser, horse- 
Scorser, horse-scoorser. 

(4) Mango equorum, qui emit equos et permutat distrahitque. Maquignon. 

An horse scorser : he that buyeth horses and putteth them away again 
by chopping and changing. _ 1585 Nomenclator. (N. p. 775.) 
Corratier: m. A Broker; also, a horse-scoorser. 1611 COTGRAVE. 
Maquignon: m. A hucster, Broker, Horse-scourser, cousening Marchant. 
1611 COTGRAVE. 

A horse Courser, or horse scourser. 1617 MINSHEU (under courser, p. 103). 


A horse courser. ... A 4orse scourser. : 
1632 SHERWOOD (under ¢ and s respectively). 


Nares and others say horse-courser “is corrupted from horse- 
scourser” (N2p.775), but the fact is the other way, as above shown. 
I hav not found an instance of scourser alone. 

When all is said, there is usually more to follow; and the tale of 
the horse-courser is not yet ended. Horse-courser, with its increast 
form horse-scourser, ment simply a ‘horse-keeper,’ but it came to 


Vol. xxiv.] Attraction in English. 141 


“mean also a ‘horse-trader’ and so courser, beside its proper sense 
of ‘keeper’ or ‘groom’ came to be used separately in the sense of 
‘trader,’ ‘exchanger.’ And courser being treated as a nativ noun 
of agent in -er (-e7" from AS. -eve), in sted of an imported noun of 
agent in -er (-er,” from OF. -éer, L. -a@rfus), was then naturally referd 
to a supposed verb course, which accordingly crept into use, along 
with the verbal noun. I mention first the compounds horse-coursing 
and horse-scoursing, ‘horse trading’ : 


(a) Horse-coursing. 1616 BEAUMONT & FLETCHER, Scornful Lady. c. D.) 
1708 CoLes, Zng.-Lat. Dict. 

(4) Horse-scoursing. 1611 and 1673 COTGRAVE(s.v. couretage and courratage). 

Horse-scoorsing. 1611 and 1673 /d. (s.v. corratage). 


But the simple verb appears much earlier, namely, in the fourteenth 
century (first recorded in the verbal noun), in the senses of ‘ groom,’ 
or ‘train’ (?) and ‘trade, exchange, chop.’ , 

The first record of this pseudo-primitiv verb appears in the verbal 
noun coursing, ME. corsing, with the attracted form scoursing, early 
mod. Eng. scorssyng, and the equivalent abstract noun in -ery, 
*coursery, ME. corserie, coserie, meaning ‘ trading,’ ‘ merchandizing,’ 
‘ trade.’ 


(a) This ‘catel gat he wit ec {usury ], 
And led al his lif in corstng [Camb. ms. cursying), 
For he haunted bathe dai and niht 
His okering, sine he was kniht. 
¢ 1325 Eng. Metr. Homilies, ed. Small, p- 139. 
(4) Scorssyng or exchangyng, eschange. [Not in the list of verbs.] 
1530 PALSGRAVE, p. 268. 
(a) It semeb, pat alle doyng in bis mater is cursed corserie of symonie, zevynge 
pe sygne of holy, ordris for temperal drit. 
@ 1384 WycLiF, Select Works, II1. 283. 
It come neuer of knyghthede, knaw it 3if hyme lyke, 
To carpe of coseri, whene captyfis ere takyne. 
¢ 1440 Morte Arthure (E.E.TS., 1865), 1. 1581. 


The verb itself appears, in the first quotation I hav, in the sense 


of ‘groom,’ ‘manage,’ or ‘ train.’ 


Here be the best coresed hors 
That ever yet sawe I me. 
¢ 1500 A Lytell Geste of Robyn Hode (Child, Ballads, v. 62). 
[This is Halliwell’s “ coresed, harnessed.” ] 


Here belongs the modern provincial verb horse-course, to beat, as 
it were ‘to groom,’ ‘curry down,’ with a subaudition of ‘ horsewhip.’ 


Horse-course. To beat. “I'll hoss-course ony o’ yon lads I find ony 
moore o’ my otcherd.” “It wo’d hav been a vast sight better to hev 
gen him a good horse-coursin’ an’ not to hev hed no justice to do aboot 
it.” ©1889 Peacock, Manley and Corringham Gloss. (E.D.S.), p. 278. 


142 Charles P. G. Scott. [1893. 


The sense ‘exchange,’ ‘ barter,’ ‘chop,’ ‘truck,’ ‘swap,’ stil pre- ~ 
vailing in provincial English, appears in the sixteenth century, the 
forms being (a) course, coarse, corce, etc., and (6) scourse, scoorse, 
Scoarse, Scorse, scorce. 


(2) Scozzonare, to breake a coult. Also to coarse or change a horse. Scoz- 
zonato, broken a coult, coarsed a horse. 1598 FLorRIo. 
Scozzonare, as Cozzonare [i.e. ‘To break and tame coltes, to play the 
horse-courser’], also to coarse or truck horses or colts as Jockies are 
wont to do. Scozzonato, broken or tamed, also trucked or coarsed with 
a horse-coarser. : 1659 FLORIO, ed. Torriano. 
Corce. To chop, or exchange. 1847 HALLIWELL. 
(6) Scorssyng. 1530 PALSGRAVE, p. 268. [See above.] [Not in 1570 
Levins, or 1573 Baret, or 1580 Baret, in either form. ] 
Changer ... scorse. 1593 HOLLYBAND, Dict. (H. p. 713.) 

This done she makes the stately dame to light, 

And with the aged woman cloths to scorse. 

1591 HARINGTON, tr. Ariosto, Ordando Furioso, xx. 78. (N.? p. 774.) 


See examples from Drayton and Spenser in Nares, ed. Halliwell, 
PP- 774, 775- 


Barater. ... To trucke, scourse, barter, exchange. 1611 COTGRAVE. 
Changer. To change .. . exchange, interchange, trucke, scoorse, barter, 
chop with. ‘ 1611 COTGRAVE. 


So elsewhere in 1611 Cotgrave: Scoarse (s.v. harder) [1673 
scoorse|. Soin 1617 Minsheu (scourse), 1632 Sherwood (scourse), 
ete: 

Scourse is stil common in provincial use. 

Scorce, v. to exchange. ’Tis Somers. [Exmoor] too; Gent. Magaz. xvi. 


Pp. 407. 1736 PEGGE, Alphabet of Kenticisms (E.D.S.), 1876, p. 45. 
Scorse. To exchange ... still in use. 1847 HALLIWELL. 


This is a long course ; but the end is not yet. The word course, 
together with its by-form scourse, early sufferd the loss of the 7, 
appearing as coyse (cotse, cose) in 1483, and as *cose, in the derived 
cosyr for corser, and coseri for corserie, as early as 1440 (see pp. 140 
and 141). In modern provincial use we find (@) couse, coase, cose, 
coce, Coss, cotss, cots, and (6) scouse, scoase, stose, scoace, scwoce, 
SqUuoace, SCOSS. 

(a) To coyse, alterare, & cetera; vbi to chawnge. 1483 Cath. Angl., p. 77. 

To cope or coase. Cambire. 1573 BARET, Alvearie. 
To coce, cambire. 1570 LEVINS, Manip. Vocad., 155, 1. 17. 
Couse. To change the teeth. Warw. Formerly to exchange anything, 

as in the Relig. Antig. ii. 281. 
1847 HALLIWELL. [H. does not give cose.] 

For the forms coase, cose, coce, coss, coiss, cois, see 1808 Jamison, 
1889 C.D., 1893 N.E.D. 


Vol. xxiy.] Attraction in English. 143 


(4) Scoase. ...« Scoarse or scoace, 
1746 Exmoor Courtship (E.E-T.S.), pp. 78, 152. 


Scorse, or scoace. To exchange. Exm. 1790 GROSE, Prov. Gloss. 
Squoace. To truck or exchange. Somerset. 1847 HALLIWELL. 
Scorse, or scose. To exchange; probably from the fact of discoursing 

previously to the exchange. 1853 Cooper, Sussex Gloss., p. 72. 
Scoss. To barter; to exchange. 1857 WRIGHT. 


Scoase (skoa’us), vb. To exchange. ‘I’ll scoase horses with you.” 
1887 PARISH and SHAW, Dict. Kentish Dial. (E.D.S.), p. 136. 
The loss of medial 7 in this position, with its recognition in writing, 
at so early a period, is somewhat remarkable, but it is not without 
parallel. I find just the same loss of 7 in a word of similar phonetic 
form, namely courser, a warhorse, early found as couser, and current 
in Scotch as couser, cooser, cusser. 
Kyllede cousers and couerde stedes. 
¢ 1440 Morte Arthure (E.E.TS., 1865), 1. ae 
[Read coursers? Cf. coursere, \. 2166. 
For ye ken a fie man and a cusser fearsna the deil. 
1815 Scort, Guy Mannering, xi. 
“ Whisht, man, whisht,” said the king; “ ye needna nicher that gait, like _ 
a cussery at a caup o’ corn, een though it was a pleasing jest, and our 
ain framing.” 1822 Scott, Fortunes of Nigel, xxxi. 
The same loss occurs in the dialectal coose for course, and discoose, 
for discourse, and appears at equally early date in scace for scarce, a 
form known from the beginning of the fifteenth century to the present 
time. And indeed, according to Mr. Sweet and other British authori- 
ties, the 7 in such a position is in all modern English words, as spoken 
in London, totally lost. The same is true of all black and most white 
speech in the southern United States. Mr. Sweet pronounces course 
as “kaos,” in C.D. notation kés, riming with sauwce. In the South, 
they pronounce it cose (kOs), riming with dose. 





XXIII. Two cases involving ice. 


1. Ickle, an icicle. This word has had a checkerd career. Tho 
stil in provincial use, in various forms, it is in literary English no 
longer recognized as an independent word. It exists unrecognized 
in icicle, properly zce-ick/e, but now so speld as to simulate a diminutiv 
in -i-cle. Jckle, especially in this compound, has undergone strange 
mutations. One is shown in my previous paper, TRANSACTIONS, 
xxiii. 240. There ar really two words concernd, ick of nativ 
(Anglo-Saxon) origin, and yok/e of Scandinavian origin. Both hav 
been affected by Attraction. 


144 Charles P. G. Scott. [1893. 


Ickle, speld also iccle, with varied vowel eccle, ecle, with sonant 
guttural igg/e, aigle, is from ME. zkel, ykel, thyl, tekyll, ykle, ycle, ekel, 
assibilated ychele, from AS. gice/, an icicle, in zses gice/, ‘ice’s ickle,’ 
ickle of ice, *zs-gice/, ‘ice-ickle,’ icicle, and the poetical, each once- 
occurring, cy/e-gicel, ‘ *chill-ickle,’ ickle of cold, i.e. winter, h7im- 
gicel, ‘*rime-ickle,’ ickle of rime or frost, Az/de-gicel, ‘ *war-ickle,’ 
hailstone, and the adj. activ grcelig, ‘*ickly,’ icy. It is cognate with 
Frisian sokkel (in tsjokkel), jockel, jogel, jikel (in its-jikel), gael 
(1837 Outzen, Géossarium, p. 143), an icicle, and with Icel. jokul/, 
an icicle, also ice, in mod. Icel. especially a glacier, and hence common 
in local names, the original sense ‘ icicle’ being quite lost (1874 
Cleasby) ; Norw. jokul, dial. jukud, jukel, an icicle, also a glacier, 
is-jokul, icicle (Aasen) ; Sw. dial. zk&e/, an icicle, Dan. obs. or 
dial. ege/, egle (1841 Molbech, Dansk Dialekt-Lextkon, p. 97), huus- 
egel, ‘*house-ickle ’ (1833 Molbech, Dansk Ordbog, i. 470). The 
word is perhaps a derivativ, probably diminutiv, of the simple form 
seen in Icel. jak, a piece of- ice, broken ice (1874 Cleasby and 
Vigfusson, s.v., and p. xxxii). 


Un esclarcyl, an ychele. ; 

¢ 1300 Bibelesworth Gloss. (Wright, Vocad.) i. 161.) 

Thyl (tehkyll, W.).  Stiria. _ 1440 Prompt. Parv., p. 259. 

Thowe, of snowe, or yclys or [of?] yce. Resolucio, liquefaccio, degelacio. 

. . . Thowyn, as yce or ober lyke (or yelys, S.). Degelat, resolvit, Cath. 
1440 Prompt. Parv., p. 492. 


Lkle. ¢ 1460 Nominale ms. (H. p. 473-) 
A nykle [an ykle]. @ 1500 Med. Cant. (Way, p. 259.) 
Ichles, stirize. 1570 LEvINS, Manip. Vocab. (E.E.T.S.), 125, 6. 


Os cowd os iccles. 
1750 COLLIER (“TIM BoBBIN”), Lancashire Dial. .(P. p. 632.) 
Iccles. Isicles. North. 1790 GROSE, Prov. Gloss. 
Ickles, Isicles, water ickles, stalactites [sic]. 
1828 [CARR], Craven Gloss. i. 241. 
Iccles. Icicles. North. ... Also, spars in the form of icicles. 
: 1847 HALLIWELL, p. 472. 
Eecle. Anicicle. Salop. 1847 HALLIWELL, p. 329. 
Iggle and aigle, an icicle. 1848 Evans, Leicestershire Gloss. (P.p. 186.) 
Ickle. Another name forthe zcicle. 1854 BAKER, Vorthampt. Gloss. i. 349. 
Stiff us zccles. 1857 SCHOLES (Lanc. Gloss., p. 165). 


The Kentish form aguadbod, an icicle (1790 Grose, 1847 Halli- 
well), I take to be *zcklebob (tickle + bob), dial. *ickabob, which, 
written down by some profound Latinist, became aguadod, apparently 
one of those “hybrids” which used to disturb purists. 

In the compound form ice-ick/e, icicle, the word has run an other 
course: A.S. *isgice/, ises gicel, ME. isykle, ysse-ikkle, ysekele, ise- 
ick, hyshykylle, isezekelle, izekelle, etc., mod. E. isickle, isikie, isycle, 
tsicle, icicle. 


Vol. xxiv.] Attraction in English. 145 


* Js-gicel. ¢ 1000? (Bosworth, ed. Toller, PP: 474, 602— no ref.). 
Stiria stillicidia, zses gice/. 
¢ 1000 AELFRIC, Gloss. (Wright, Vocad.? 117, 14.) 
per as claterande fro be crest be colde borne renne3, 
& henged he3e ouer his hede in hard ysse-ihhles. 
¢ 1360 Sir Gawayne and the Green Knight (E.E.T.S., 1864), 1. 721. 
And thanne flaumbeth he [the holygoste] as fyre on fader and on filius, 
And melteth her my3te in-to mercy as men may se in wyntre 
Ysekeles in eueses [ /sy4/es in euesynges C] thorw hete of the sonne, 
Melteth in a mynut-while to myst and to watre. 
1377 LANGLAND, Piers Plowman (B), xvii. 227. 


Styrium, Aysehyhylle. a 1500 Vocab, Roy. ms. 17 C. xvii. (Way.) 


LTyshykylle ought to satisfy those who love rhyme and ‘fongue and 
programme and through and other nimious spellings. 


An Jzekelle (Jsezekille A), stirium, stiricus. 1483 Cath. Angl., p. 198. 
Droppe of yse called an zszk/e, whych hangeth on a house eaves or 
pentisse. Stiria. 1552 HULOET, Adecedarium. (Herrtage, C.A., 198.) 
Gouttes gelées. /szcles, 1611 COTGRAVE. 
Her tresses of gold... were now growne more white then thistle downe, 
the zsickles of frozen ice, or the white mountains snow. 
1635 R. JOHNSON, Zom a Lincoln. (Thoms, Zarly Prose Rom. ii. 94.) 


The form icicle, ice-ick/e, in its earlier and more correct spelling 
*ise-ickle, isickle, istkle, isicle, tsycle, underwent the spreading of the 
sibilant, and appeared as ise-sickle, yse-sycle, ice-sickle, as if from ise, 
ice, + sickle. 


The longe yse sycles at the hewsys [eaveses] honge. 
¢ 1520 Cytezen and Uplondysh-man (Percy Soc., 1847, x xxii. 3). 
For it had snowen, and frosen very strong, 
With great ysesycles on the eves long, 
The sharp north wynd hurled bytterly 
And with blacke cloudes darked was the sky. 
¢ 1520 COPLAND, The Hye Way to the Spyttell Hous, \. 15. 
(Early Pop. Poetry, ii. 9.) 
Bruosina, Bruosa, a flake of ise, a hoare-frost, an zsesichle. 
1598 FLorio. [Ed. 1611 has “ Brudésa as Bruosina,” but omits 
Bruosina by accident. Ed. 1659 (Torriano) is quite different. | 
Ciondolare, to droppe or thaw, to hang downe dangling, as zce-sickles 
[1611 ice sickles, 1659 ice-sickles|. Ciondoli, ice sickles [1611 ice sikles, 
1659 ice-sickles], danglings, labels. Ciondoloni, dingledangle, hanging 
downe, as ice-sickles [1611 ice sickles, 1659 om. ]. -1598 FLorIo. 
[Sim. ise-sickles, ise sickles, ice-sickles, ice sickles, etc. s.v. diaccuioli, 
ghiacciuoli, scoladura, scolature, stillecchio, in the three editions. ] 
lis-dacken, /se-sick/es. lis-droppels, lis-kegels, ofte kegels, se sickles 
hanging downe house ewsings. 1648 HEXHAM (sim. 1658, 1678). 


Hence sickle came to be used alone, in the sense of ‘icicle.’ It 
occurs in one of those modernized ballads which lug in “ Phoebus ” 
and “ Flora.” ‘ Phoebus”! what a name—to appear in a Robin 
Hood ballad ! 


146 Charles P. G. Scott. [1893. 


When Phcebus had melted the ‘ stch/es’ of ice, 
With a hey down, &c., 
And likewise the mountains of snow. . . 
a 1795 Kobin Hood and the Ranger (in Ritson, Robin 
Hood’s Garland, 1795; Child, Ballads, vy. 207). 
The form sickle thus developt was confused with the sickle of 
harvest ; and 7zce-stckles wer thought to be so cald because they wer 


sharp or pointed. A happier likeness is exprest by the term daggers : 


Daggers. Icicles. So called from their pointed appearance. 
1854 BAKER, Northampt. Gloss. i. 171. 
This is a plain provincial use. Tennyson’s use is individual, and 
merely allusiv : ; 
The daggers of the sharpened eaves. 1850 TENNYSON, Jn Memoriam, cvi. 


In keeping with dagger is the other provincial name daglet (dag + 
-let): 
Dagilets.... Icicles. Wilts. 1847 HALLIWELL. 


A similar notion lies in Johnson’s definition of the word : 
Icicle. A shoot of ice hanging down. 1755 JOHNSON. 


The same notion of something pointed or tapering appears in the 
other Teutonic names for an icicle, which I mention for comparison : 
Sw. zs-pigg, ‘*ice-peg’; Dan. zs-tap; G. eis-tapfen, ‘*ice-tap’; D. 
ws-kegel, ‘ *ice-pin’ (see ait in C.D.); Eng. dial. zce-candle 
(Halliwell). 

The present notion in regard to zcic/ is reflected in its spelling 
and its common pronunciation (ais'-i-cl in sted of ais'ic'’l), namely, 
that it is a diminutiv of zce, like particle, diminutiv of part. This is 
formally stated in some dictionaries : 


Icicle. Dim. of ice, that is, a small body of ice. 
WILLIAMS, Readable Dictionary. 
[The next entry is iceberg, which is a large body of ice!] 


The following is from a recent British dictionary of considerable 
pretensions : 


cle (L. culus, a dim. termination), also cule, ule, el or le, en, kin, let, et or 
ot, ling, ock, y or ze, which form nouns and signify “little”; diminu- 
tion: examples —icic/e, a little conical mass of ice; canticée, a little 
song; animalcz/e, a very little creature (etc.). 
1881 STORMONTH, Etym. and Pron. Dict. (6th ed. “ revised,” 
Edinb.), p. 768. 


Imprest with the belief that the -zcZe of icicle was a diminutiv suffix, 
one poet has formd a parallel diminutiv with the suffix -Z¢, namely 
tcelet: 


Vol. xxiv.] Attraction in English. 147 


Whilst each tree bereft 
Of.leaves, did like to virgin mourners stand 
Cloathed in white vails of glittering zce/ets. 
1659 CHAMBERLAYNE, Pharonnida. (Wr. p. 590.) 


2. Yokle, the other word for ‘icicle,’ ME. yok/e, is not from the 
AS. gice/, but from the Scandinavian cognate, Icel. 7ékull, now jokul, 
jukul, jukel, above mentiond. Yok has disappeard from use, 
except as disguised in the compound ice-shackle, ice-shoggle (for 
*ice-yokle), as shown below. ; 


Stiria est gutta frigore concreta pendens guttatimque stillans, a yo%/e. 
1500 Ortus Vocab. (Way, Prompt. Parv., p. 259, note 1.) 
[Halliwell’s entry, ‘ Yo&de, an icicle,’ probably refers to this. ] 


Parallel to ice-ick/e, icicle, there existed a form containing the other 
term yokle: namely *ice-yokle, ME. *is-yokel, from ts, ise, ice, + yokle. 
(a) An izokelle [read iszokelle?], stirium. 5 

1483 Cath. Angl. (cited by Way, Prompt. Parv., p. 259, note 1). 
[This is a different reading from that which appears in the printed edition 


of the Catholicon Anglicum (E.E.T.S. ed. Herrtage, 1881), where we 
read “ An izehelle ([sezehille A)”: see above. ] 


As 1ice-ickle, ise-ickle, developt a sibilant in its second element 
by attraction from the first, so *¢ce-yok/e, *7s-yokel, has done: s-y 
becoming sh, speld sh, sch, just as in the pronunciation of issue 
(is'yu > ish'u), Aassion (pas'si-on > pas'syon > pash’on), etc. Hence 
the form zse-schokill, ice-shoccle, ingeniously alterd, in the plural, to zce- 
shackles, as if ‘ fetters of ice’ ‘icy fetters’—a common figure in poetry. 


(4) Furth of the chyn of this ilk hasard auld 
Grete fludis ischis, and styf zseschokillis cald 
Downe from his sterne and grisly berd hyngis. 
1513 (pub. 1553) Douctas, Virgil, 108. 30. (Jam. 1808.) 
Over craggis and the frontys of rockys sere 
Hang great yse schokkalis lang as ony spere. 
1513 (pub. 1553) Douctas, Virgil . . . (Craven Gloss. i. 241). 
Ice-shackles, Icicles— May not this word be derived from shackle, the 
wrist, as a shackle of ice. Though icicles vary in their dimensions, 
they certainly frequently resemble the wrist in rotundity. 
1828 [CARR], Craven Gloss. i. 241. 
Ice-shoccle. An icicle. 1849 TEESDALE Gloss., p. 67. (Also 1847 
Halliwell and 1857 Wright.) 


Ice-shoccle also appears in an other alterd form, 7ce-shoggle, ice- 
shogle, simulating shog, shake (ar not icicles often shaken down?) ; 
and ice-shoggles takes on a verbal or diminutiv form zce-shogglings. 


But wi’ poortith, hearts, het as a cinder, 
Will cald as an zceshogle turn! 
1805-06 J. NicoL, Poems, ii. 158. (Jam. 1808.) 
Ice-shoggle, an icicle. 1825 BROCKETT. (Whence in 1847 Halliwell, and 
1857 Wright misprinted ice-skoggle. ) 
Ice Shoglins or Ickles, icicles. 1855 [RoBinson], Whztby Giloss., p. 90. 


148 Charles P. G. Scott. - [1893. 


XXIV and XXV. The next two classes involv a gain of 
initial s by attraction from a preceding possessiv or genitiv 
-s or -es. It occurs in certain loose compounds or phrases 
involving kzz (ME. kin, kyn, gen. kinnes, kynnes, often kines, 
kynes, and contracted kins, kyns), ‘kind,’ constructions in 
which we now use &zzd, which is in fact, in these construc- 
tions, a variant of £zz with excrescent d (as in hind for hine, 
pound for poun, sound for soun, etc.), but confused with kznd, 
ME. kinde, kynde, tkynde, AS. gecynd, nature. The develop- 
ment was probably thru the genitiv forms £zznes, often 
written zxes, with consequent lengthening of the vowel, 
kines, whence, with the loss of the inflexion, ize, kin, then 
by confusion with éxd, kind, nature, the modern znd, sort. 

The phrases with kz, genitiv kinnes, kins, kin, which I 
hav noted and illustrated by numerous quotations, ar a// kz, 
any kin, many kin, no kin, other kin, some kin, this kin, what 
kin. I giv for comparison the forms of all these, but giv 
quotations only for the two which show the attraction of the 
genitiv s, namely any kin and what kin. 


(1) Ad kin, ME. genitiv alles kinnes, al Bnd alle kyns, al kyns, 
alle kynez, al kyns, alle kyn, al kyn, alkyn. 

(2) Any kin. The ME. genitiv anyes kinnes, enyes kennes, AS. 
@niges cynnes, appears sometimes as any skynnes. 

(2) Zyf by wyl rejo[isse] more 

In enyes kennes bynges. ¢ 1315 SHOREHAM, Poems, p. 95. 

The genitiv of the atecus is here used as in other instances — 

enies monnes, ‘ of any man,’ etc. 


LokiaS hweSer exzes monnes sare beo iliche mine sare. 
c 1175 Lambeth Hom. (E.E.T.S.) 121. 


(4) & pyne yow with so pouer a man, as play wyth your knyzt 
‘ With any skynnez countenaunce, hit keuerez me ese. 
¢ 1360 Syr Gawayne and the Green Knight (E.E.T.S.), 1864, 1. 1539. 

(3) Many kin, ME. genitiv many kyne; mod. many kinds (of). 

(4) (Vo kin, ME. gen. nanes kynnes, nones cunnes, no kynnes, 
no-kynnez, non kyns, no kyns, nokyns (mod. dial. neeakins) ; also with 
loss of inflexiv -s, no-kyne, nakyne, no kyn; mod. no kind (of). 

(5) Other kin, ME. gen. othres kynnes, mod. dial. otherkins. 

(6) Some kin, ME. gen. sum kyns, somkyns, sumkyn, somkyn ; 
mod. some kind (of). 


Vol. xxiv. ] Attraction in English. 149 


(7) This kin, ME. gen. this hyn. 

(8) What kin, ME. genitiv what kinnes, or by conformation, 
*whats kynnes, appears as what skynnes. With loss of the inflexiv 
-s, it appears as what kin, what kyn, quat-kyn, what-kynne. 

(a) What-kyn folk so per fare, feche3 hem hider. 

¢ 1360 Cleanness (Early Eng. Allit. Poems, E.E.T.S.), 1. 100. 


Why what-kynne thyng art pou, 
pat telles pis tale to me? ¢ 1430 York Plays, v. 52 (p. 24). 


(4) In what skynnes maner so hyt be wro3t, 
Dedly synne hyt ys forthe broght. 
¢ 1400 Myrc, J/nustructions for Parish Priests (E.E.T.S.) 1. 210. 


Take gode hede on hys de-gre 
Of what skynnes lyuynge bat he be. 
¢ 1400 Myrc, /ustructions for Parish Priests (E.E.T.S.), 1. 1637. 


In this what kin, gen. what kins, lies the explanation of a form 
hitherto misunderstood — devidkins. 
And of every handfull that he met 
He lept ouer fotes thre; 
“ What devilkyns draper,” sayd litell Much, 


“ Thynkyst thou to be?” 
¢ 1500 A Lytell Geste of Robyn Hode,\. 292. (Child, Ballads, v. 57.) 


Devilkyns is not a form of devilkin, from devil + dim. -&in, nor 
is it, so explaind or otherwise, the source of dickens. The significant 
phrase is what kyns draper, ‘a draper of what kind’; devi/ is merely 
the common term of emphasis used by impatient persons who ar 
conscious of a lack of the finer kind of rhetorical ability, and there- 
fore feel obliged to invoke, for the requisit intensity, something 
wholly irrelevant, like devz/, or “something hot,” like Ae//. This is 
the explanation of modern profanity. Actual swearing, the intended 
invocation of supernal or infernal powers, is almost extinct. The 
gost of Hamlet’s father now gets no satisfaction in the cellarage, 
except in the cellarage under the stage. 


Gho. Sweare. Ghost cries under the Stage. 
1623 SHAKESPEARE, Hamlet, i. 4. (F! p. 258.) 





XXVI. Case involving “zs. 


1. Noll. An other instance in which the possessiv -s has gon 
over to the following noun is probably presented by the provincial 
snowl, the head. This was probably, at first, Ais now or his noll: 


150. Charles P. G. Scott. [1893. 


then his snow/ or ’s nowl, and so snowd even when an other possessiv, 
or none, precedes. ; 


(a) What good can the great gloton do with . . . Azs mo/l toty with drink? 
@ 1535 Sir T. MorE, Works, p. 97. (Wr. p. 118.) 
(6) Snowl,s. The head. 1825 JENNINGS, Somerset Gloss. p. 71. 





XXVII. Case involving yes. 


1. Ma’am. In yes, the common affirmativ response, before a word 
admitting initial sibilation the s is often spred over, so as to begin 
the next word. This ‘pronunciation is not, however, recognized, 
except now and then in humorous writing. I find examples in one 
writer : 

“ Yes, sma’am—sir, I mean,” said she, as she went downstairs. 


1885 F. R. STOCKTON, Rudder Grange, p. 56. So also p. 69. 
“Ves, sm’am,” said Pomona. 1885 Jd. p. 48. 


On the other hand yes, séry is often written in novels yessir, 
implying a rapid pronunciation, with one s. 





XXVIII. Similar cases. 


There are many other cases involving a shifting of initial or final s, 
for which there is here no room. They ar not important. 

The ease with which s may spred or be duplicated appears from 
the following otherwise trivial examples : 


Vacation is taken in hot Summer, and the temples of learning open in 
Autumn because it’s school weather. 1893 Philadelphia Times, Sept. 
When Madeline was asked if she would have her new gown cut after the 

latest style, she said she’d just as sleeve as not. 
1893 Boston Transcript, July. 


The Public Ledger and Transcript of Philadelphia, in July, 1893, 
sacrificed some of the space which it daily devotes to “ athletics” 
and lists of persons admitted to the hospitals with broken legs or 
contused heds or other infirmities of wide public interest, to admit a 
pun on “ Just as Siam,” in neat allusion to the troubles in Siam, and 
to a wel-known hymn. 


Vol. xxiv.] Attraction in English. 151 


B. Initial S lost. 

XXIX. Cases in which an initial s, following a final -s in 
the preceding word, is fused with it and so lost, or is mis- 
taken in a compound for a genitiv suffix belonging to the 
preceding element, and is transferd as such. 


1. Several. An instance of the loss of initial s by its absorption 
in a preceding possessiv ’s, appears in evera/, sophistically written, in 
the plural, everhills, also contracted erri/s, a field or enclosure, 
originally a severa/, or allotment of common land to an individual 
proprietor. His several, Fohn’s several, seems to hav become Ais 
everal, Fohn’s everal, and so everal emerged as an independent 
form. 


(2) Of late he’s broke into a several 
Which doth belong to me, and there he spoils 
Both corn and pasture. : 
1600 Sir Fohn Oldcastle, iii. 1 (Sternberg, p. 92). 
Truth lies open to all; it isno man’s severa/. (Patet omnibus veritas; 
nondum est occupata.) 
1641 JONSON, Discoveries (Works, p. 742). 
Some are so boysterous, no severa/s will hold them, but lay all Offices 
common to their power. : 
1648 FULLER, Holy and Profane State, p. 234. (P. p. 113-) 
(6) Several. Frequently corrupted into everhills, errils, etc. A field or 
enclosure. 1851 STERNBERG, Vorthampt. Gloss., p. 92. 


2. Skep, also skip, skepe, *skap, a basket, also a beehive made of 
twisted straw ; especially in dee-skep, bee-skip, which appears to hav 
been taken sometimes as a dee’s cap. 


(a) A bee-skip, a bee-hive. 
1691 Ray, South and East-Country Words (E.D.S.), p. 91. 
A bee-skep. In Scotland . . . a bee-hive. 
1823 Moor, Suffolk Words, p. 352. 
Bee-skep. A beehive made of rushes or straw. 
1892 M. C. Morris, Yorkshire Folk- Talk, Gloss. 
(4) Skep. A basket without a lid, with short handles. “A bushel skep.” 
“A bee skep.” In Scotland, the latter is, I believe, used for bee-hive. 
I have seen, but I forget where, this Scottish saying, “my head is 
bizzing like a dees cap,”” which is probably the same word. 
1823 Moor, Suffolk Words, p. 352. 


3- Slip, ME. sUippe, slyppe, AS. slyppe; also slop, ME. sloppe, AS. 
sloppe: the second element in the compounds cows&p and oxslip. 
The term s/f, variant slop, AS. slyppe, sloppe, is supposed by Skeat 
to refer to the loose droppings of a cow, and to allude to the growth 
of the plants along fences and roadsides. See the quotation from 


152 Charles P. G. Scott. [1893. 


Lowsley, Berkshire Words, below. Compare the modern slop 
applied to any liquid refuse. I find s/op used, like wash, of the 
shoal water of the sea, next the shore: 


He . . . Gers lawnche his botes appone a lawe watire, 
Londis als a lyone with lordliche knyghtes, 
Slippes in in the s/opfes o-slante to be girdylle, 
Swalters vpe swiftly with his swerde drawene. 
¢ 1440 Morte Arthure (E.E.T.S. 1865), l. 3922. 


(1) Cowstip, also dial. and early mod. E. cowslop, ME. cowslyppe, 
couslyppe, cowsloppe, cowslope, cowslowpe, AS. cuslyppe, cusloppe, from 
cu, cow, + slyppe, sloppe, as above explaind. 


(a) Genim ifig Se on stane wyxd on eorpan j gearwan 3 wudubindan leaf 4 
cuslyppan J oxanslyppan. 
¢ 1000 Lacnunga 42. « (Saxon Leechdoms, iii. 30.) [See also 61 (iii. 46).] 
Nim > ifig be on stane weaxe J gearwan J wudu bindes leaf 7 cuslyppan. 
¢ 1000 Leech-book, III. xxx. (Saxon Leechdoms, ii. p. 326.) 
Nim wudu merce 9 hrefnes f6t 7 wermod niopoweardne cz slyppan, rudan, 
wudu bindes leaf. 
¢ 1000 Leech-book, III. xxxi. (Saxon Leechdoms, ii. 326.) 
Brittannica, cusloppe. ; 
¢ 1000 AELFRIC, Vocab. (Wright, Vocad.? 135, 1. 26.) 
Brittanica, cuslyppe. : 
¢ 1000 Lat. AS. Glosses. (Wright, Vocad.? 361, 1. 23.) 
Glustrum, an°® Cowslyppe. 
¢ 1450 Lat. Eng. Vocab. (Wright, Vocad.? 586, 1. 44.) 
Herba paralisis, an°¢ Cozslyppe. 
¢ 1450 Lat. Eng. Vocab. (Wright, Vocab. 587, 1. 46.) 
Hoc ligustrum, a cowslowpe. 
¢ 1450 Nominale. (Wright, Vocad.? 713, 1. 11.) 
Cowsloppes. 
1584 A Handeful of pleasant Delites (Park’s Heliconia II. pp. 1-6). 
(Littledale, ed. 770 Noble Kinsmen, N.S.S. p. 110.) 


The Cows/ips tall her pensioners bee, 
In their gold coates, spots you see. 
1623 SHAKESPEARE, J/.NV.D. ii. 1. (F! p. 148.) 


And hang a pearle in every cows/ips eare. 1623 Jd. 2d. 


About the middle of the fifteenth century cows%p appears to hav 
been taken as cow’s lip. 


(6) Vaccinium, an a coweslyppe. 
c 1450 Lat. Eng. Vocab. (Wright, Vocad.? 618, 1. 24.) 
Hoc ligustrum, A°* a cowys/lepe. 
¢ 1475 Pictorial Vocab. (Wright, Vocad.? 786, 1. 25.) 


Here @ cowyslepe may hav been taken as ‘a cow’s leap.’ 


Verbascolo odorato, the cozw's-dip [1598 cow-slip, 1611 cowslip). 
1659 FLorio, ed. Torriano. 
Cowslip, ab AS. Cuslippe, Flores seu herba Paralyseos, sic dicti quoniam 
iis Vaccee delectantur, vel, ut aliis placet, & similitudine Labiorum Vacce. 
Doct. Th. H. dictum putat ab odore suavi animz Vaccarum zmulo, 
cujus sc. odor talis est qualem Vacce ore & labiis expirant. 
1671 SKINNER, Etym. Ling. Anglicane. 


Vol. xxiv.] Attraction in English. 153 


This is truly rural. ‘ Sweet is the breth of morn;” and sweet, 
some say, is the breth of a cow. 

(2) Oxslip, ME. *oxe-slyppe, AS. oxan slyppe, ‘ox’s slip’ or 
‘slop,’ in like manner became reduced to ox/ip, and has been mis- 
taken as ‘ ox-lip,’ ‘ ox’s lip.’ 


(a) pip flie oxan slyppan nibewearde 4 alor rinde wylle on buteran. 
¢ 1000 Leech-book, 1. ii. 15. (Saxon Leechdoms, ii. p. 32.) 
[Cockayne translates here “ ox-slip,” elsewhere “ oxlip.”] 
I know a banke where the wilde time blowes, 
Where Oxs/i~s and the nodding Violet growes, 
Quite ouer-cannoped with luscious woodbine, 
With sweet muske roses, and with Eglantine. 
1623 SHAKESPEARE, J/.N.D. ii. 1. (F} p. 150.) 
Paigle, it is of use in Essex, Middlesex, Suffolk, for a cowslip: cowslip 
with us signifying what is elsewhere called an oxs/ip. 
1691 Ray, South and East-Country Words (E.D.S.), p. 88. 
Ox-slip~s. The flowers of Cowslip roots as produced when these roots are 
planted upside down, and with cow-dung or soot around. The manure 
doubtless accounts for the tint produced. 
1888 LowsLey, Berkshire Words (E.D.S.), p. 122. 
(4) Verbascolo, the pettie-mulleyn, of which be diuers kinds, as Verbascolo 
odorato, the cow-slip [1659 cow’s lip], Verbascolo albo, the oxe-lip, 


Verbascolo minore, the primerose. 1598 FLoRIo. 
Brachecuculi, the flowers Cowslips, Paigle, Palsiewort or Oxe-/ipfs [so 
1611; in 1659 Oxelips]. : 1598 FLorIo. 


Oxlips. 1623 SHAKESPEARE, W.7.iv.3. (F! p. 292); 1634 FLETCHER 
(and SHAKESPEARE?), Zwo Noble Kinsmen, i. 1. 10. 
Ox-lips, herb Paralyseos species, 4 similitudine Labri Bubuli in floribus, 
vel, ut Doct. Th. H. divinat, ab odore grato florum, instar Animz seu 
Anhelitfis Bovini fragrante, v. Cows/ip. 
1671 SKINNER, Ztym. Ling. Angl. (Etym. Botanicum.) 
Paigles, Ox-lips. 1692 CoLes, Eng. Dict. 


4. Stang, ME. s/ang, a sting. I find stang in his stang, referring 
to a scorpion’s sting, reduced to /amg, and that in one instance mis- 
taken for sung, tong, the word we ar now pleased to spel songue. 
The tongue never stings except with words, but most persons, I 
think, believ that the pretty little harmless red forked tongue of a 
serpent is his “ fangs,” with which he “stings.” The scorpion’s sting 
is of course as remote as possible from the tongue. The mod. proy. 
Eng. fang, a sting, may be in part a particular use of /amg, a point 
or projecting part. 

(a) It war to lang to mak narratioun 


Off sychis sore, with mony s¢ayg and stound. 
1552 LyNDESAY, 7estament of the Papyngo (E.E.T\S.), 1. 1139. 
My curse upon thy venom’d stang. 1789 BURNS, Address to the Tooth-ach. 
(4) The scorpion for-bare is tang [his stange Fairfax ms., his tunge Gott. ms., 
his tonge Trin. ms. ]. 
Fra bestes bar he lai amang. ¢ 1300 Cursor Mundi (Cotton ms.), I. 
A tange of a nedyr, acus. ¢ 1500 Ms, dict. (H. p. 850.) 
Tang. The sting of a bee, &c. North. 1847 HALLIWELL, 


154 Charlés P. GC. ‘Seote [1803. 


The next two cases arise from the fusion of an initial s with a 
preceding plural suffix, and the consequent loss of the former. 


5. Strickle, ME. s¢rik/en, frequentativ of s¢rzken, intr. go, pass, 
etc., tr. go against, hit, strike, etc. The form s¢zk/en, being, as Pro- 
fessor Skeat first pointed out (Zam. Dict. s.v. trickle), nearly always 
used after or in connection with “ars (teres strikled, teris *strikland), 
came to lose its initial s in the plural s of “eves; hence ¢rik/en, mod. 
Eng. ¢rickle, which has been more or less confused with ¢i// in like 
sense. 


(2) His salte teres strikled doune as reyne. 
¢ 1386 CHAUCER, Prioress’s Tale (Six-Text), 1. 187. 


Thus the Lansdowne manuscript. The Ellesmere and Hengwrt 
manuscripts hav ¢rikled, the Cambridge “¢vekelede. The Petworth 
manuscript has sérzked, the Corpus s¢ryked, the weak preterit plural, 
and the Harleian has stzken, the present plural, of the verb sizken, 
the simplex of the frequentativ st7ik/en. These variations sufficiently 
prove that the original was s¢rik/en,; altho no other instance of 
striklen appears. All later examples hav ¢rik/en, trickle. Tyrwhitt’s 
reading ¢vz//ed seems to hav no manuscript authority. 


(4) With zeris trickland on his chekes. 
¢ 1400 Ywain and Gawain, \. 1558. (Ritson, Metr. Rom. i. 66.) 


Nay, ful of sorowe thou now me seest; 
The ¢eeris trikilen dowun on my face, 
For “ filius regis mortuus est.” 
¢ 1400 Political, Rel. & Love Poems, ed. Furnivall, p. 207, 1. 46. 
The red blode ¢rzé/ond to his knee. 
@ 1500 (?) AVS. Cantab. Ff. v. 48, f. 122. (H. p. 889.) 
To 7rickle downe, destillare. 1570 Levins, Alani. Vocab., 122, |. 7. 
To 7rickil, destillare. 1570 /d. 128, 1. 22. 
He said and from his eyes the ¢rickling teares ran down amain. 
1573 PHAER, Virgil, p. 300. (Wr. p. 61.) 
The christall dew of faire Castalian-springs 
VVith gentle floatings ¢vick/ed on his braines. 
1594 LopGE, Wounds of Civill War (Hunterian Club), p. 57. 
But where found they thée, O holie Anthony . . . testifying thy contrition, 
by thy ¢rickling teares. . 
1596 LopGE, The Divel Contured (Hunterian Club), p. 8, 
The éeares fast trickled downe his face. 
a 1650 King Arthur's Death. (Child, Ballads, i. 44.) 


And so ‘tears trickle’ thru all the ballads. In modern dialectal 
use “rickle has senses which appear to represent the original sense 
‘go’ or ‘glide quietly,’ without reference to tears or other water. 


Vol. xxiv.] Attraction in English. 155 


To Trickle, Trittle, v.a. (To Trickle, to run down in streams or drops.) 
To bowl or rall; as, “77ick/e me an orange across the table.” orf. 

1839 HoLLoway, Gen. Dict. of Provincialisms. 

Trickling, part. Applied to the uncertain scramble of a wounded hare. 

“I seed the hare a ¢rickling along the deitch, through the brimbles 

under the boo of yon wicken.” 1877 LEIGH, Cheshire Gloss., p. 215. 

We'll make shift to ¢rickly down as far as the gate. 
1893 Q[UILLER-CoucH ], Delectable Duchy, p. 54. 


6. Strike, ME. striken (pret. strok), AS. strican (pret. strac), 
go, go in a course, glide, flow ; stil used in these original senses, tho 
they ar probably regarded as developt from the now prevailing sense 
‘hit.’ The ME. s¢riken is said to appear rarely as ¢riken, just as 
striklen appears as ¢riklen (see above). 

(a) pe cwellers leiden swa luderliche on hire lich tet blod barst ut and 

strac adun of hire bodi as stream ded of welle. 
¢ 1200 St. Marherete, p. 5. 


Ase strem bat strikeZ stille. 
¢ 1300 Song, in Spec. Eng., Part II. p. 48,1. 21. (Ritson, Anc. Songs, p. 32.) 


A mous that moche good couthe, as me thou3te, 
Stroke forth sternly, and strode biforn hem alle. 
1377 LANGLAND, Piers Plowman (B), Prol. 1. 183. 


He saide to his sone, “‘ Tak a pike, 
To-night thou schalt with mie strike.” 
““Whider?” seide his sone. 
¢ 1320 The Sevyn Sages, 1.1254. (Weber, Meir: Rom. iii. 50.) 


Here I must pause. There remain eight or nine other classes of 
words which hav sufferd change by Attraction. The total number 
of words affected is small, but it is necessary to consider them before 
stating the philological conclusions to which the study leads. In an 
other paper I hope to conclude the subject, to point out similar 
cases of Attraction in foreign tongues, and to giv an index of all the 
words treated. 


156 William Gardner Hale. [1893. 


VIL. —“ Extended” and “ Remote” Deliberatives in Greek. 


By Pror. WILLIAM GARDNER HALE, 


UNIVERSITY OF CHICAGO, 


I. 


In his Att. Syntax, published in 1843, Kriiger (§ 54, 7, 
A. 2) in effect held that no cases of the indirect deliberative 
subjunctive were found that could not be solved by regarding 
the introductory phrases ov« éyw and the like as equivalent to 
a7vop®. In 1847, Madvig, in his Synt. d. gr. Sprache, § 121, 
included, under the head of the deliberative subjunctive, two 
examples which cannot be so explained (Pl. Symp. 194 and 
Isocr. 4, 44, for which see p. 160 of this paper), though he 
did not discuss them. Aken’s procedure in his Grundziige 
(1861), § 292, is similar; and so is Kiihner’s, as late even 
as 1870, in the second edition of his Gr. Gramm., §§ 559, 3, 
and 394, 5. Professor Goodwin, in his Greek Moods and 
Tenses, second edition (1865), § 65, 3, treats the dependent 
clause in éyeu Oru eizry, on the ground of its affinities, under 
the head of the final sentence, but explains it as “caused by 
the analogy of the common expression od« éyet Stu (or 7/) 
ein, equivalent to ov« oidev Ot ely, he knows not what he 
shall say, which contains an indirect question”; and in a 
note on p. 135 of Felton’s edition of the Panegyric of Isocra- 
tes he expresses the same view with more fulness, making it 
clear that the construction in his view is of deliberative origin, 
although in éyw 6rv elm “all trace of the indirect question 
disappears.” In later editions he continues to hold substan- 
tially the same view. Monro, in his Homeric Grammar (first 
edition, 1882, § 282, and at the same place in the edition of 
1891), placed his statement, “in Attic the idiom survives in 
a few phrases, as éyes dtu eimy,” under the general head of 
“final relative clauses” and the specific head of “the Rela- 
tive of Purpose with the Subjunctive.” 


Vol. xxiv.] Deliberatives in Greek. 157 


Such was the state of opinion, when a note in Professor 
Jebb’s Philoctetes, upon opa@vra .. . ovdéy Evrorov, ody Satis 
apkxéceev, V. 281, brought the matter into formal discussion. 
The beginning of the note reads thus: ‘‘The direct ques- 
tion is ris apxéon; (deliberative, or ‘interrogative’ subjunc- 
tive).”” Ina notice of the edition, published in the Classical 
Review for April, 1891, Mr. Arthur Sidgwick comments as 
follows: “The note begins by treating the sentence as 
an oblique deliberative: but if this be right, as it probably 
is, it should be pointed out that the construction is so 
stretched that the interrogative character is lost. The fact 
I believe to be that the deliberative construction is subtly 
modified, and three stages may be traced as follows: First 
stage, od« Oe botis apKxécecev, ‘I did not know who was to 
aid,’ — truly interrogative and deliberative; second stage, 
ovK elyov Gotis apKécevev, where the interrogative character 
is sliding into the relative; third stage, oddév’ eZyov datis 
«.T.r., Where the relative character of éa7is is established. 
The last usage is what we have here: it is so like the Latin 
final guvz with subjunctive that few readers or commentators 
stop to notice the difference; but it certainly is not that, 
else we could say érepya datis ayyédXov, which we cannot 
do: it is always dyyedel.” Mr. Sidgwick appends the form 
in which he himself would have cast the note: “This is an 
extension of the deliberative construction, used after nega- 
tive sentences, which becomes at last practically equivalent 
to the final.” the 

In the Classical Review for July, 1891, Professor Tarbell, 
in a brief article, speaks of the construction as “the exten- 
sion of the dependent deliberative construction, after certain 
verbs, to relative clauses” ; and, a little later, he cites Soph. 
Phil. 938 (od yap dAxdov ot8 btw A€yw) and Isocr. 21, I (od 
mpopdcews atropa, & hvtiva Aéyw Urép Nixiov Trovtovi) as 
illustrating the transition from interrogative to relative 
clauses, and adds the following: 

“ By regarding d\Xov and mpoddcews as proleptic, we could 
explain the dependent clauses as ordinary indirect questions, 
without detriment to the sense. Most of the other passages, 


158 William Gardner Fale. [1893. 


however, resist such treatment. The cases after éorz (Soph. 
Aj. 514 éuol yap ovKér éotiv eis 6 Te BXérw; Eur. H. F. 
1245 yéuw xaxav 67, KovKér éc@ Sry teOH) are especially 
noteworthy, because here the verb is not even capable of 
being followed by an indirect question. ov« éotuw eis 6 Tt 
Bdér@ grows by an easy transition out of ov« éyw eis 6 Te 
AXér@, understood as meaning ‘I have nothing to look to.’” 

Mr. Sidgwick’s note had dealt only with optatives follow- 
ing a secondary tense. Of the fourteen examples cited in 
Mr. Tarbell’s paper, a few were optatives after a secondary 
tense, but the larger part were subjunctives after a primary 
tense. Before we proceed further, it will be necessary to 
have all these examples before us, together with others that 
have been contributed since. (One example of my own find- 
ing I shall add later.) The arrangement is chronological. 
[Examples in which ody éyw may be interpreted as equivalent 
to ov« oida are omitted. ] 


fol 4 bey J \ + 
To.adta unyavnpat éEevpw@v Taras 
Bpotoicty avtos ov exw oddicn Sto 
THS vov Tapovons THmoVAS aTadXaya.— Aesch. P. V. 469. 


> \ \ > PID \9f > <4 , 
E“ol yap OUKET EoTLV eis 6 TL BAETMH 


mAnVv cov.— Soph. Ai. 514. 


® Ayeves, @ TpOBAATEs, @ Evvovciat 
Onpav dpei@v, @ yaTappayes TréTpat, 
bpiv Tad’, ov yap adXov oid’ 6TH A€ya, 
avakXalopat trapodat Tots eiwOdow, 


of épy’ o mais wp edpacev ob& ’AxiAXéws *-— Soph. Phil. 936. 


e al fal A / 
op@vra pév vads, as Ex@v évavoTorour, 

, , Mv > > v.23 
macas BeBwcas, avdpa & ovdév’ vtoTor, 
ovy dots apKéceev, 0VS Batis VoGOU 


Kapvovtt cuAAaBatTto*— Soph. Phil. 279. 


> 4 , 
ovK éyov Baow, 
ovdé Ti éyywpwv KaKxoyeiTova, 
+ ee rn U 
Tap ® oTdvov avtituTov BapuBpOaT aroxXavcelev aiwatnpov* 


Vol. xxiv.] Deliberatives in Greek. 159 


a \ , e / 2 e / 
ds Tay Oepporatay aiudda KnKkiopévav éEXkéwv 
évOnpov todos nmrlovoe 

pvarros Katevvaceev. — Soph. Phil. 691. 


olmol, Tpodedopat, KovKET eioly EdrlOes, 

a / / > .. 4 

dtrot TpaTropevos Odvatow "Apyelwov diya * 

ovTOS yap Hv mot KaTahuy? cwTnpias. — Eur. Or. 722. 


OH. icye ordp’, ds wn péya Aéyov peifov maOns. 
HP. yéuo xaxdv 8, xovdnér’ Ec O drrou reOH. 
— Eur. H. F. 1244. 
ovdéva yap elyov botis "Apyobev podov 
eis "Apryos adOis Tas éuas erricToXNas 
méprpece. — Eur. I. T. 588. ' 


f.> Ww /, > \ (4 9,7? of fal > 4 
tiv éxov dyunv ayadnv es, ef ST@ KYiTapev aryuias ; 
— Ar. Eq. 1321.3 
ef d€ pyre Sv 6 Te pte GTovot pte ad STov TroreuHnowpev 


éoTl, TAS OVK eK TraVTOS TPOTrOU THY ElpynYNY TroLnTéoV HpiV ; 


— Andoc. 3. 16.? 


LA > a / \ > ” x4 fa) Ayn 4 a 
@aTe ov TovTO SédoiKa pi) ovK éyw 6 Tt dH ExdoT@ TOY 
pirar, av ed yévntat, GAA pn ovK exw ixavods ols 8A. 


— Xen. Anab. I, 7, 7. 


éav pev ody vikdot, Ti Set adtovs AVEeLY THY yépupav; ode 

\ x / 5 »” x ied , c a 
yap av TovXal yépupar @ow Exowmev av Orrot huyovtes pets 
cwbapev. “Eav & ad pels vixdpev, NeAvpévns THs yepvpas 
oby &£ovow éxeivor bro Piywouv: — Xen. Anab. 2, 4, 19. 


def pévtot tois wéArXovow avOpw@rras EEew Sti eioh€épwowv 
eis TO OTEYVOV TOD épyacopuévou Tas ev TO UTralOpw épyacias. 
— Xen. Oec. 7, 20. 
icws otv cal Knguciov avtixarnyopycet, cal er 6 Tu Nyy. 
— Lys. 6, 42. 


(Aéyn is a sure emendation of Bekker’s for Aéyex.) 


1 Cited in the Classical Review for Febr., 1894, by Professor Sonnenschein, 
whose attention had been called to it by Mr. H. Richards. Professor Sonnen- 
schein’s article appeared too late to admit of special discussion here. 

2 Contributed privately by Professor Tarbell. 


- 


160 William Gardner Hale. [1893. 


mpotepov yap ovK éxwv mpodacw éf' Hs Tod Biov Adyov 
Soinv vuvi dca TovTov eiAnga. — Lys. 24, I. 

To.ovtov 00s npuiv mapéd0cav, Mate . . . ExaTépous exe 
éd’ ols pidotimnO@owv. — Isocr. 4, 43 and 44. 


> , > lal » ue. - yd ig \ / / 
ov mpopacews aropa, &v hytwa Aéyw brép Nixiov rovrovi. 
—Isocr. 21, I. 


29\ ” , 2 = e Lol es 3 , e fe] / 
ovdev Ett Svolcer avT@ oTnody TaV évOdde OTLOdY yiyverOaL, 
2h , ” ef 4 ” \ Len 
éav povov xn OT Siaréyntat, dros Te Kai Kar. 


— Plat. Symp. p. 194, D. 


év ovv TH oVvYyKOLUHoEL TOD pev épacTod 6 aKdXacTOS LTS 
” iA / \ ‘ > / 
éyer OTL A€éyn Tpos Tov nvioxov. — Plat. Phaedr. p. 255, E. 


(Aéyyn is a sure emendation of Bekker’s for Xéyeu.) 


> v icf > / 289 al 
ovK elyomev GTou érrtAaBoipweda ovS STov Kpatoipev.- 


— Dem. 35, 25. 


2>Q\ \ 3 , ” p Mean \ \. 2» 
ovdé pev avOparrav Tis Env ert Bovol Kal épyots 
havouevos otropiwo.o Sv avraxos, SvTiv Epoipny * 
— Theocr. 25, 218.1 


In the Classical Review for March, 1892, under the head- 
ing of “The Subjunctive of Purpose in Relative Clauses in 
Greek,” Dr. Earle, of Barnard College, combated Mr. Tar- 
bell’s position, holding, as Monro had held, that the idiom in 
question is descended from a clause of purpose, —a conclu- 
sion shared, without argument, by a writer in the same 


1 Contributed to me privately by Professor Tarbell, who also points out that 
the same construction occurs in Theocr. 16, 68, if we take the subordinate clause 


as relative: 
: Slinuar 8, & rim Ovar dv xexaplopevos EvOw 


avv Moicats. 


Oirep Ppdoworr, which used to be read in Thuc. 7, 25, is generally replaced 
now by érws dpdoworv, the reading of the Vaticanus. 

My colleague, Professor Burton, has pointed out two interesting examples in 
the New Testament, Mark 14, 14: wrod éoriv 7d xarddupd pov, drov Ta waoxa 
pera TOV pabynTtov wou Payw; and Acts 21, 16: dyovres wap’ & tevicOGuev Mvd- 
cwvl rim Kurply, dpxaly uadnry (see also the footnote on p. 186). The latter 
probably touches the extreme point reached in the development of the construction. 


Vol. xxiv.] Deliberatives in Greek. 161 


journal, signing the initials J. D. to articles to be referred 
to in Part II. “The source of the error into which,” as 
Mr. Earle thinks, “ Mr. Tarbell and others have fallen” is 
a confusion among the three meanings of yw, namely, “I 
have,” “I know,” and “I am able,” and a confusion between 
the interrogative ris and the indefinite or general relative 
doris. The examples after éyw in the sense of “I know,” 
as well as those after similar phrases like dunyava, Mr. Earle 
regards as indirectly deliberative. The examples after éyo 
in the sense of “I have,’ and examples after such phrases 
as ov yap aXXov oida, quoted above, and after such phrases as 
KovxéeT eioly €drrides, he considers to be of final origin. With 
regard to a number of examples which he gives, he grants 
that it is difficult to determine whether éyw means “have” 
or “know,” and consequently to decide whether the construc- 
tion is final or deliberative, it being final in the former case, 
and deliberative in the latter. 

Mr. Earle’s specific argument! against Mr. Tarbell’s ex- 
planation appears in a brief passage which I quote verbatim: 

“Soph. Phil. 938: dpiv rad’, ob yap adXov 018 btw Aéyo 
may be paraphrased ov yap dddov éxyo, Kaba olda, btw Aéyo. 
The fact that the antecedent is here expressed seems enough 
to show that there is no relation with an indirect question. 
So Isocr. 21, 1: 


> / > a“ > / e \ / / 
ov mpopdcews aropa, Sv hvTia Aéyw vTép Nixiov Tovrovi = 
éy@ Tpopact KTX., 


where also the antecedent is expressed.” 


1A remark of Mr. Earle’s, not in the nature of an argument, calls for a 
moment’s notice: “ As will be seen by the quotations I have made, more than 
one-half of Professor Tarbell’s examples —‘ of the phenomenon which had not 
been recognized by any previous grammarian ’— have been examined and dis- 
cussed by Kriiger and Goodwin.” I am sure that others, like myself, would 
suppose, if they had before them only the passage which I have quoted, that 
Mr. Tarbell meant to say that no one before himself had recognized the phe- 
nomenon at all. What he actually wrote was “the phenomenon had not been 
clearly recognized, so far as I am aware, by any previous grammarian” [the 
italics. are mine]; and nobody could read the passage in its context without see- 
ing that Mr. Tarbell meant before Mr. Sidgwick, not before himself. 


162 William Gardner Fale. [1893. 


On what I have now quoted from Mr. Earle, three com- 
ments need to be made before we proceed to weigh the 
evidence. 

First, neither Mr. Tarbell nor Mr. Sidgwick (if he be 
meant among the “others’’) can have fallen into error 
through overlooking an “ambiguity” between “ darts, an 
indefinite or general relative,’ and “tis introducing an 
indirect question.” Whether right or wrong in their theory, 
both —as the passages which I have quoted verbatim show 
—recognized as clearly as Mr. Earle himself that, in the 
examples which illustrate what they regard as the extreme 
point of development, the connective is not an interrogative, 
but a relative. In fact, the essential point of their belief is 
that the construction has been extended from the interroga- 
tive form to the relative form. 

Secondly, I must express my dissent from the general 
attitude which is indicated in the sentence quoted from 
Mr. Earle. Nothing is surer than that extensions, and 
even new formations through association, take place in 
syntactical mechanisms, just as they do in word-meanings. 
Mr. Earle would be on no more dangerous ground if he 
were to say that our word “palace” cannot be descended 
from the name of a certain hill in Rome (to go no further 
back in the matter), because our word does not carry with 
it any idea of a hill. The fact that, in the last example, 
the antecedent is expressed, shows only that the dependent 
construction is at any rate no longer an indirect question, 
but not in the least that it did not originate in such a 
question. 

The third comment is that, if Mr. Earle’s argument were 
sound, a similar argument would, in the case of several of his 
examples, destroy his own explanation. Stripped of the indi- 
vidual form which it takes in this particular place, his canon 
would have to read somewhat as follows: If in a given exam- 
ple the full apparatus and full force of a given known con- 
struction do not exist, then the construction found in the 
example can have no relation with that known construction. 
What then becomes of the example in Orestes 722, which 


Vol. xxiv.] Deliberatives in Greek. 163 


Mr. Earle pronounces ‘another instance of the purpose- 
construction”? It runs as follows: xovd«ér eicilv édrides, 
dtrot tTpatropevos Odvatov "Apyeiwv diya, “there are no longer 
any hopes to which I may turn and escape death at the hands 
of the Argives.’ The main sentence denies the existence of 
hopes. But hopes, we might make answer to Mr. Earle, do 
not exist in order that one may escape. Inasmuch, then, as 
the thought, though similar, is not identical with that of the 
purpose clause, there could—on Mr. Earle’s canon—be no 
relation between this example and the construction of pur- 
pose. 

But a positive refutation of this position can, I think, be 
given. Questions like cur gaudeam, in the sense of why 
should I rejoice, are well known. But there can be no doubt 
that in clauses of the common type seen in uzhil est cur gau- 
deas or nthil est quod gaudeas, “there is no reason why you 
should rejoice,” the introductory word is a relative, not an 
interrogative. It is, in fact, an interrogative turned relative,! 
—in other words, an example of a phenomenon exactly par- 
allel, so far as this part of the mechanism is concerned, with 
the phenomenon contended for by Mr. Tarbell. 

This evidence — if it be thought to be evidence —of the 
unvalidity of Mr. Earle’s argument undoes his work ; but it 
of course does not solve the question. 

Now it would seem to me entirely possible and natural 
that such a construction as the one under discussion should 
have. been derived from either of the two sources claimed for 
it. From the deliberative question “to whom shall I speak” 
could easily come the chain of combinations, “I am at a loss 
to whom to speak,” “I do not know to whom to speak,” “I 
have no one to whom to speak,” “there is no one to whom 
to speak,” and even, at the extreme, “there is a man to whom 
to speak,” —a construction not identical with that of pur- 
pose, but near enough to it to remind one of it. On the other 
hand, from the combination “he has built a bridge by which 
we may escape” could easily arise the combination “there is 


1 Such is, in my conception, the simple explanation of a much misunderstood 
construction. 


164 William Gardner Hale. [1893. 


a bridge by which we may escape,” and, finally, “there is no 
bridge by which we may escape.” Or, quite possibly, a more 
direct way could be struck out. Under the feeling “I must 
get me a weapon with which to defend myself,” one might 
say “I have no weapon with which to defend myself.” But, 
in either case, there is an extension. The alternatives be- 
tween which we have to choose are not a plain clause of 
purpose and an extended construction of deliberation, but 
an extended clause of purpose and an extended deliberative 
clause. ; 

What evidence, now, can be found to decide between these 
two rival theories, both of which are antecedently rational ? 

Three kinds of evidence might be looked for, turning upon 
three things which have to be accounted for; namely, (1) the 
range of introductory expressions after which the construc- 
tion occurs, (2) the chronological order in which the various 
classes under this range are found to stand, and (3) the form 
of the construction. 

(1) We are of course not dealing with the obvious cases 
of deliberative questions after verbs like dynyava, atop or 
its opposite evzope, but with the construction after such 
phrases as in ov« €yw ixavovs ols 50, ob yap adXov oida bT@ 
eyo, KovKér eiolv édmides Grrr TpaTropevos Oavarov diya, - 
é£er OTe Aéyn, etc. All of these with one exception (Ar. Eq. 
1321) come under the head of the expression either of the 
existence of a difficulty, or (its opposite) the absence of a 
difficulty (existence of a means). Now this limitation is 
intelligible, if the construction is descended from that com- 
mon form of the expression of difficulty, the question of 
perplexity, which is a shade of the deliberative subjunctive. 
It never got wholly out of sight of its starting-point. It 
could not be used, e.g. after a verb like “I give,” “I send,” 
“T appoint,” for there is in the meaning of these verbs 
nothing to suggest anything like perplexity or difficulty. 
But, on the other hand, it is quite impossible to understand, 
on Mr. Earle’s theory, why, after these and countless other 
verbs which are constantly being followed by purpose-clauses, | 
the construction in question never appears. 


Vol. xxiv.] Deliberatives in Greek. 165 


(2) The second point at which evidence might perhaps be 
found is in the historical relation of the two kinds of expres- 
sions after which, with the exception already noted, the con- 
struction occurs; vzg. the expression of the existence of a 
difficulty (‘I am at a loss,” etc.) and the expression of the 
existence of a means. Not only do 19 examples out of a 
total of 25 belong under the former type, but, with the same 
exception from Aristophanes, all the earlier examples belong 
to it down to the last from Xenophon, — which indeed, as the 
statement of the ecessity of the procuring of a means, is not 
far removed, and forms the transition to the new type (“we 
have no means,” “we must have a means,” “we have a 
means’’). This state of things, while not absolutely impos- 
sible upon the theory that the construction originates in a 
final clause, is yet, upon that theory, improbable, while it is 
precisely what we should look for upon the theory of a 
development from the deliberative subjunctive, as sketched 
in detail above. 

(3) The third possible evidence lies in the form of the 
construction. And here, as it seems to me, Mr. Earle’s 
theory receives its death-stroke. In all the examples of the 
subjunctive that have been cited on either side (with the ex- 
ception of a single one from Hesiod, which I shall presently 
show to be of a different character), the mode is unattended 
by av or xe. Yet, practically speaking, the clause of purpose 
introduced by the relative pronoun (the construction after 
iva or ddpa is of course a different affair) took a subjunctive 
with dv or xe, or a future indicative, the former prevailing 
in Homer, the latter winning an almost complete triumph 
later. The subjunctive in I 459 (=I 287), 


D4 \ \ > 4 oe > ¥ 
ExOoTe, Kal TLV ATrOTLVEWEV HV TLV EOLKED, 
H Te Kal €ooopevorct pet avOpwrroict TéANTAL, 


is said by Professor Goodwin,-M. T. § 568, to be the only 
case of the relative with the subjunctive without «xe in such 
sentences in Homer. The exact force of the mode in this 
example is hard to determine, and I am willing, so far as our 
argument is concerned, that it should be classed as final. 


“* 


166 William Gardner Hale. [1893. 


We may even grant the case o 334, placed under this head 
by Delbriick and Monro, 


pn tis Tor Taxa “Ipou apelvav adXos avacTh, 
ds Tis @ audt Kdpn KexoTTas yepol otiBaphow 
dapatos éextréurrnot popvEas aipate ToAAS, 


though it is possible that the relative clause is only a con- 
tinuation of the expression of warning (‘look to it that 
another does not arise and”’), or that it is only an expression 
of expectation without av! (so Ameis-Hentze). But, even if 
these cases are sound, they are but two? against a very large 
number. 

Now if the construction under examination were a devel- 
opment from the relative clause of purpose, it would do one 
of three things. It would, in conformity with ordinary Attic 
usage, appear with the future indicative for its mode; or it 
would appear with the indicative in some examples and the 
subjunctive with dy or «xe in others; or, at the extreme, it 
would appear with the future indicative in some examples, 
the subjunctive with av or xe in some others, and, just con- 
ceivably in a very small proportion of cases, with the sub- 
junctive without av or xe. Instead, however, of doing any 
of these things, it appears in every case with the subjunctive, 
and, in no case that properly belongs under the head which 
we are considering, with av or xe. This state of affairs would, 
humanly speaking, be impossible if the construction had 
originated in a clause of purpose, whereas it is precisely what 
is to be expected if the clause originated in a deliberative 
subjunctive. ~ : 

It remains to speak of the one case which Mr. Earle 
adduces that does not properly fall under the present head. 
In this example, which is from Hesiod Op. 57 (Seca. kaxov 
@® Kev atravtes | téprewvtat), not only is the introductory ex- . 
pression different from any that we have seen, but the 
dependent clause contains a xevy. These two facts sever it 


1 See a similar subjunctive with xev, in 8 754, p. 170, below. 
2 I see no sufficient reason for regarding the verb in old ris .. . Pavudooerat in 
2 467 as a subjunctive, as Monro is inclined to do. 


Vol. xxiv.] Deliberatives in Greek. 167 


from all the other examples, and reduce it to the ordinary 
construction of the earlier clause of purpose, as commonly 
.treated. Out of such cases as this, had they occurred fre- 
quently, a construction similar to that which has been under 
discussion, only wzthk av, not without it, might indeed have 
been developed, but was not. And, if it had been, it would 
not have been confined within the narrow range to which the 
Extended Deliberative Subjunctive is confined; nor would it 
have stopped at the stage of development represented by the 
subjunctive with dv or «xe, but, like the construction of pur- 
pose itself, would have passed on to the stage seen in the 
Attic construction with the future indicative. 

These objections were urged by myself in a discussion 
which took place at the meeting of the American Philological 
Association in the summer of 1892, at the conclusion of the 
reading of Mr. Earle’s “ Notes on the Subjunctive of Pur- 
pose in Relative Clauses in Greek.” Mr. Earle’s paper was 
printed in abstract in the Proceedings for that year. One 
phrase in it, in the form in which it was publishéd, seems to 
point to a conceivable means of escape from the argument 
which I have given. For this reason, and also because the 
paper gives a clearer idea of Mr. Earle’s conception of the 
prototype of the Attic idiom than did his article in the Clas- 
sical Review, I reprint it here in full. 


The paper contained an examination of the idiom ov« ort (uor), or obk exw, 
8s (Sorts or rel. adv.) and subj. (or opt. aft. secondary tense). The prototype of 
the Attic idiom was sought in Homeric Greek: cf. //. 27, rrr sgq., Il. 19, 355-7; 
Il. 6, 450 sqq., U2. 4, 164, 1. 21, 107 sqg., Od. 6, 201 sqq., Il. 3, 459 sq. Od. 15, 
310 sq., with Soph. Ai. 514 sq., Eur. H. F. 1245, Xen. Anab. 1. 7, 7, Eur. Or. 
722 sq. (For other examples from Attic Greek, see Class, Rey. Vol. VI, pp. 93-5.) 
It was suggested that “the gradual obsolescence of the subjunctive which can be 
traced in Ionic and Attic Greek, in what Weber calls ‘ unvollstandige Finalsitze’ 
with Saws, seems to have gone hand in hand with a similar obsolescence in the 
kindred relative final-clauses” (z.e. relative in the more restricted sense). In 
this process the finite construction of the rel. clause may have been influenced 
by the use of the fut. particip. to express purpose after verbs of motion, a usage 
so extensive in Ionic Greek that in Hadt. viii-ix, which, according to my examina- 
tion, contain ot a single fut. rel. clause of purpose, and no certain instance of the 
ovx éxw, 8,7« constr. with (so-called) final subjunct., we find the fut. part. in all 
17 times. — “ In such a sweeping away of the subjunctive constr. we must seek an 


+ 


168 William Gardner Hale. [1893. 


explanation of a survival as certain as the od« €yw 6,7: (8) constr. appears to be, 
examined from the point of view of historical syntax. It is here that Goodwin’s 
remark is suggestive. If, instead of saying that the construction in question 
“may be explained by the analogy of’ the indirect deliberative, we say that it is. 
to be exf/ained from the essential nature of the subjunct., traced in its develop- 
ment in Homer, and found again, in perhaps still further development, in Attic 
Greek, as a survival, sometimes obscured and confused by the indirect deliberative, 
the similar form of which served to prevent it from sharing the fate of its com- 
panion relative clauses of purpose. If we put the case in this form (pointing out 
in our support the triple ambiguity of €yev and the ambiguity of 8c7:s), we shall, 
it seems to me, be as near the truth as we are likely to get in so subtle a matter.” 
[The writer did not make himself responsible for any particular theory of the 
original meaning of the Greek subjunctive. He does not, however, wish himself 
to be considered as favoring the putting on the same footing, though they may 
both for convenience’ sake be classed as “ final,’ such subjunctives as those which 
are discussed above, and the final subjunctive developed from the independent 
hortatory subjunctive! Cf. Eur. Suppl. 1232, with Soph. Antig. 1332 sq., 1184 sq. ] 


From the fact that Mr. Earle printed his paper, it is to be 
presumed that, in spite of the evidence pointed out, in our 
oral discussion, as afforded by the absence of ay or xe, he 
continued to hold his opinion. If that is so, there are but 
two forms in which it would be conceivably possible to hold 
it. These would be somewhat as follows: (1) The construc- 
tion in question was originally expressed by the bare sub- 
junctive, at a time (this must have been well back of the 
Homeric age; yet no example occurs in Homer) when the 
bare subjunctive was the form employed in the relative clause 
of purpose; but, while other clauses of purpose passed on 
to the stage at which dy or xe was used, and then to the 
stage in which the future indicative was used, these particu- 
lar clauses were held to their earliest form by their close 
resemblance to the indirect deliberative question ; or (2) the 
construction began at a time when the relative clause of 
purpose took the subjunctive wth dv or xe, but, when this 


1 I have endeavored to prove, in a paper printed in abstract in the Proceedings 
of the American Philological Association for 1892, that the subjunctive with dy or 
xe in the relative final clause is not of “hortatory ” origin, but expressed originally 
simply something reckoned upon (“ future” or “ prospective’ subjunctive). This 
view, though not the one put forth by Delbriick, seems to me to be necessary, 
and also to have the distinct merit of bridging the chasm which would otherwise 
exist between the Homeric idiom and the Attic. 


Vol. xxiv.] _Deliberatives in Greek. 169 


clause passed on to the stage in which the future indicative 
was employed, it remained in the subjunctive, and also shook 
off its av or «xe, on account of its close resemblance to the 
indirect deliberative question. In either case, Mr. Earle’s 
phrase ought not to be “sometimes obscured and confused 
by the indirect deliberative,” but “completely obscured and 
overpowered by the indirect deliberative.” The phrase, how- 
ever, is unimportant. What is important is that to. grant 
either of these alternatives is to grant a closer resemblance 
between our examples and the indirect deliberative than be- 
tween our examples and the clause of purpose; and to do 
this is to surrender the case entirely. 

The passage in brackets in Mr. Earle’s paper might per- 
haps at first sight appear to undo my objection to his theory, 
inasmuch as he would not speak of the prototype of the con- 
struction as being strictly “final.’’ It becomes necessary, 
then, to examine the examples which he cites from Homer, 
to see:what is the concrete starting-point which he has in 
mind. They are here presented in an order intended to show 
their relationship ; and I add several (marked by the word 
“additional”’) for further illustration. 


A. 
GAG pow ed O brdbev Kai aw’ jryewov’ ec OXOv drraccor, 
isa / Aa. >: > / 
bs Ké pe Keto’ ayayn*— o 310. 
ADDITIONAL. 


GAN’ dye poe SdTe via Oonv Kai eixoo’ éraipous, 
of Ké pou év0a Kai &vOa Svatrpicowor Kédevov. — B 212. 


B. 


ADDITIONAL. 


keivos & ad mepl Khpe waxdptratos é£oyov addon, 
bs Ké ao é&Svoice Bpicas oixovd aydyntar.—€ 158. 


ADDITIONAL. 


€orat pav, 67 dv adbte pirnv yravewrida eitry.—O 373. 


os 


170 William Gardner Hale. [1893. 


éooetat Huap, OT av ToT OA@AH “IALOS ipyH 

kai [Ipiauos Kat Xads évpperiw I praporo, 

Zeds 5€ oft Kpovidns iwifuyos, aidéps vaiwr; 

autos émicceinow épeyynv aiyida Taow 

Tio araTns KoTéwV. Ta pev EccETat OVK aTéAECTA. 
—A 164. 

éooeTat Huap, 6T av ToT OA@AR INOS ipy 

kal IIpiauos cal Xads evpperiw Ipiapovo.—Z 448. 


ovyx opdas, olos Kal éy@ Kaos Te wéyas TE ; 
\ ek ee met a \ 4 f , 
matpos © elm’ ayaboio, Bea Sé we yelvato unrnp* 
GX’ ert Tou Kai éwol Odvatos Kat poipa Kpatain. 
” A XN x I x Ld = 
écoetast 7) Has 7) SeiAn 7) pécov Huap, 
e , , 9 “ ” > \ 
OmmoTe TLS Kal eweto ”Aper ex Oupov Edntat, 
% 6 ye Soup Barwv 7 aro veuvpndiv dicto.—P 108. 


C. 
oThté wot, audimoror* moce hevyere Pata iSodcat ; 
8 , 7 8 4 , ed > lal 
% wy Tov twa Svopevéwov hacO Eupevar avdparv ; 
> ” > 2 =. % ‘\ / »g\ / 
ovK Ec ovTos avnp Siepos Bpotds, ovdé yévnTar, 
bs kev Dainxav avdpav és yaiav ixntas 
dnuoThta hépwv: para yap Piro abavdroow.— £ 199. 
> \ .7 
ov yap dim 
Uj a / \ > / 
mayxu Oeots waxdpeoot younv “Apxevoradao 
” > > 2 7 > , »” 
éy bec , AA Ere Tov Tis éerrécoetat, bs KEV exnol 
deapata O inpepedéa xal amrompobt miovas aypous. 
— 6 754. 
ADDITIONAL. 
> , 2 > 7 ‘ ’ 

el yap K év vioon ye wapeEeXaonoba SiwKor, 
ovx éoO’, bs Keo EXnot peTadpEVOS OVSE TrapérOn, 
ove el Kev petorriabev ’Apeiova Siov éXavvar.—V 344. 


ADDITIONAL. 
xovpn ‘Ixapio.o, repidpov Inverereca, 
Oapoet, un ToL TadTa peta ppeot chot weXovTar. 
> 4 > ? > / 718) > »,” 4 
ovK Eo odTos avnp, ovd ExoeTat, ovdé yévynTat, 
ds kev Tnrepdy o@ vie yelpas erroice 
fwovtos y° €ueBev xal eri xOovi Sepxouévoro.— 7 435. 


Vol. xxiv.] Deliberatives in Greek. 171 


ADDITIONAL. 

? / 4 4 , > oy 7 
ov yap Tis éativ Os Tapa aipnoerat 

\ \ > Lal 4 , > > , 
Thy anv axpetov Sivayw avr’ Eipucbéws. 


— Eur. Heracleid. 57. 
D. 


as dpa puv Ipiduovo mpoonvda daidipos vids 
Arooopevos erréecowy, apeiAtxTov O Om aKovoev* 
VTE, LN for ATroLWa TipavaKEo Nd ayopeve: 
mpl mév yap IldtpoxXov érotreiv alcipov huap, 
Toppa Ti por tmrepidécOar evi hpect pirrepov rev 
Tpawr, kal troddovs Swovrs Edov 76¢ Tépacca: 
viv § ov« éo@’, bs Tis Oavatov dyn, dv Ke Deos ye 
L / / > lol > . / 
IXiouv mpomdpobev euns ev yepol Barnow, 
\ / tA / > L 2 , / , 
kal wavtov Tpwwr, trépt & ad Iprapord ye maidav. 
—® 97. 
byes & ’Apyelny “EnXévnv cal xrnpal aw’ aby, 
éxdoTe, Kal TLunVY ATOTLVEMEV, Hv TLV EoLKeD, 
H Te Kal éocopevoict peT avOpworroor wéAnTat.—T' 458. 


The examples in group A convey ‘‘ purpose” in the ordi- 
nary full sense of the grammatical term (being expressible in 
Attic by the relative with the future indicative), and thereby 
differ radically in force, as well as in form (namely, in being 
accompanied by «ev) and in the nature of the words which 
they follow, from the idiom we are studying. 

The examples in group B are all-simply prophetic in mean- 
ing (“happy is he that shall... ,” “the day will come when 

. shall’’), which characteristic is wholly absent from the 
idiom we are studying. In one example only, ® I11, is av 
or xe wanting, and—to say nothing of the necessities of the 
meaning —the presence of ay in the parallel passages, © 373, 
A 164, and Z 448, shows that the subjunctive here is of the 
same nature as in the rest. 

In group C we get examples the first of which, but for the 
presence of dy or xe, would seem at first glance to be parallel 


- 


172 William .Gardner Hale. [1893. 


to those we are investigating. But the resemblance is purely 
-superficial. In the idiom which we are studying, the prevail- 
ing type of the earlier examples is the negation of the exist- 
ence, or the assertion of the restriction of the existence, of a 
means for carrying out an act which the speaker wills. “I 
must escape this suffering —there is no means,” “I must 
look to some one — there is' none but: you,” “I must escape 
death — there is no hope,” — such is the cast of all the exam- 
ples before Xenophon, excepting two only, namely, the second 
one from Euripides, and the one from Aristophanes. In the 
former, the idea of impossibility, which in other examples has 
been present by inference, is the sole idea conveyed. The 
later growth of the construction mzght have proceeded along 
this line ; but it did not. The construction broadens in scope, 
in that it comes to be used with the third person as well as 
with the first, in this case having the power of expressing an 
act willed by a third person, not by the speaker. In the case 
of the example Eq. 1321, too, the will expressed is not that 
of the speaker, but is an echo of a command just given by 
the person addressed. But here and elsewhere, in all the 
examples that have thus far appeared, saving only the sec- 
ond from Euripides, the idea of zuz//, of demand, of aim, is 
-always present. The latter example, then, represents merely 
a quickly arrested side-growth, and the typical construction 
may be described as involving the expression of the will of 
the speaker himself or of the subject of the main sentence. 
Now in the subjunctive examples under C (to say nothing 
of the presence of «xev), the act of the verb is not willed by 
any one; while, on the other hand, the indicative example 
which I have added shows how different the after-fate of this 
construction is from that of the one which we are studying. . 

The example under D differs from those under C in that 
the element of will does appear. Achilles says, in answer 
to Lycaon’s prayer for mercy, “there is no man that shall 
escape,” or, in substance, “it is my will that no man shall 
escape.”” But this does not correspond to the examples 
-of our idiom. Orestes, in the example from Euripides, Or. 
722, does not mean “it is my will that I shall not escape 


Vol. xxiv.] Deliberatives in Greek. 173 


death.”” Achilles denies the existence of any such impulse 
as is suggested in ¢vyy. Orestes. expresses, in diya, the 
actual existence of an impulse, and denies only the existence 
of a means for carrying it out. 

As for the example under E, though I have already expressed 
myself as ready to concede it, so far as our present argument 
is concerned, as a final clause, I should not be able to follow 
the thought of any one who found a closer resemblance be- 
tween such a sentence and od mpoddcews amopa, b Hvtiva 
Aéyw than between the latter and ov« arop@ &' Hhvtiwa mpo- 
gdaciv éyw, or between such a sentence and ov yap aAXov 
oid’ 6t@ A€éyw than between the latter and ov« of8 6T@ réyo. 

As for the phrase ‘‘the gradual obsolescence of the subjunc- 
tive . . . in the kindred relative final-clauses,” as Mr. Earle 
puts it, I should like to suggest the greater accuracy of some 
less mechanical phrase, such as “the completion of the pas- 
sage of the form from the older expression of futurity to the 
newer, of which hints had already been given in Homer.” 
And as for the phrase “examined from the point .of view of 
historical syntax,’’—a phrase in itself calculated to array all 
good men on the side of the theory of which it is used, —it 
seems to me simply out of place. The one thing that has 
been lacking has been a rigorous examination of the question 
from precisely this point of view. 

Next, — for the question involves details that call for great 
patience, —I have to add several examples confirmatory of 
the theory of the existence of an Extended Deliberative, and 
to discuss a part of them. 

The following from Gothic and Old English, which I owe 
to my colleague, Professor Blackburn, illustrate incontestably 
the subordination of a deliberative. The connectives are still 
interrogatives ; for, at this early date, wa had not yet taken 
upon itself the relative function. In the case of the example. 
from Old English, it might possibly be urged that the form 
of the Greek has been retained, and that the meaning of ov« 
éyovaowv is “they know not,” but a similar explanation would 
hardly be reasonable in the case of the example from Beo- 
wulf. 


174 William Gardner Hale. [1893. 


And hig nabbap hweet hig eton.— Matth. 15,32. (Greek 
ovx éxovow Tt daywowv, Vulg. et non habent quod man- 
ducent.) 

Nah hwa sweord wege. — Beowulf, 2252 (3). (=non habeo 
quis ensem ferat.) 1 

The following has not appeared in the controversy : 


ddrdov 8 ov tev olda, Ted dv KAUTA TevYea SUw. — > 192.? 


If sound, it is of great importance, as proving, of itself 
alone, the interrogative origin of our idiom. But the pres- 
ence of dv is unexpected, and this fact, together with the use 
of the interrogative, might be thought to throw suspicion 
upon the state of the text. What we ought to think upon 
this point will best appear in connection with the discussion 
of certain examples of a type ov« éotiv Oras with the future 
indicative, —at first sight identical in form, except for the 
mode, with the extended deliberative, — which still demand 
consideration. 

In two of the groups below I add, for further light, exam- 
ples with other tenses of the indicative. 


A. 
ovK éo® Straws der od Sedp’ EXOdvTa pe. — Soph. Ant. 329. 
ovK av yévoito TODO’, Oras éyw AaB@v 
onmEia TOLAadDT, OV have Tovpoy ryévos.— Soph. O. T. 1058. 
an? > BA > ef > > > / 
TOUT ovK écO Grrws ToT Eis me 
Tovverdos é£eus évdixws overdicar. — Soph. Phil. 522. 


ov yep Tis €otiv Os Tapa aipieerat 
THY anv axpetov Svvamiy aT Eijpve dé Eur: Heracl. 57. 


oi & év tals povapyiais dvtes, ovK Exovtes btw HOovicover, 
TaVvTwV ws olov T éati Ta BéXTLCTA TpaTTovot.— Isocr. 3. 18. 


1 The resemblance of the Greek idiom to such examples as nec quid speraret 
habebat, Verg. Ecl. 2, 2, nil habeo quod agam, Hor. Sat. 1, 9, 19, is obvious, and 
suggests an interesting question of parallel growth or imitation. 

2 La Roche, Diintzer, Ameis-Hentze, and Nauck all give the verse as it is here 
printed, though various ways of substituting a relative for the interrogative red 
have been suggested. 


Vol. xxiv.] Deliberatives in Greek. 175 


ov yap Eo brrov 
NOyor axodcat fav trot’ nOéAno’ euav.— Soph. Ai. 1069. 


B. 
ov yap éo8 Straws Tod 
xelvny épeivrers. — Soph. O. C. 1372. 
TavTnv ToT ovK éo@ ws ete COcav yapeis. 
— Soph. Ant. 750. 
ov yap éo8 Sarov pw’ dreis. — Soph. O. T. 448. 
ovk cP Stas ovK et od yevvadas avnp.— Ar. Ran. 640. 
ov yap éo® bras 
68° ov« "Opéotns of 6 tpochavav éué.— Soph. El. 1479. | 
ovK oO bras ov mictov €F bpav TTEpdv 
éEnyay és T6d’ adaos. — Soph. O. C. 97. 
ovK éotiv Straws EV TLL LiKP@ 
décrrotva yoXov Katatravoe.— Eur. Med. 171. 
ovK éotiv Aris TOUT av “EXAnvis yuvn 


étAn 1o?’. — Eur. Med. 1339. 


C. 
ovK éotiv aixiop ovdé pnyavny 6T@ 
mpoTpéwetat pe Leds yeywvica Tade, 
mpv dv yaracA Seoua Nvpavthpia.— Aesch. P. V. 989. 


D. 
vooet dé wor potas aTOXOS, OVS EL 
dpovtidos éyyos 
@® Tis adéEerat.— Soph. O. T. 168. 
TavTn jwerdler vauBartns ovdels Ex@V* 
ov yap Tis Bppos Eotiv, ovS bro TAEWV 
éfeurrornoet Képdos, % Eevooetat. — Soph. Phil. 301. 
Yevdavipuas oe Saipoves Hpounbéa 
KaXovowv* avTov yap ce Set mpopnbéws 
bT@ TpoT@ THaS éexxudaOyncer TUYNs. — Aesch. P. V. 85. 


176 William Gardner Hale. £1893. 


In the examples with the future indicative in the first three 
groups of clauses, the mechanism is: simply a negatived state- 
ment with regard to futurity. The evidence of this is to be 
seen in some of the examples, otherwise of the same nature, 
in which a negatived statement for the present or past is 
used, as in the last under A (‘there is no point in which in 
life he was ever willing to listen to my words”’), or as in El. 
1480 under B (“there is no way in which this is not Ores- 
tes”). In no example is there that feeling of demand which 
we found to be characteristic of the extended deliberative, 
and to be present in every example of that idiom excepting 
one. Even in the sentence from Isocrates under A, which 
might at first sight (especially if one translated by the phrase 
“having no one to envy”) appear to resemble the subjunctive 
idiom, this element is lacking, since the meaning appears to 
be simply “there being no one whose station in life will (=is 
of a kind to) lead to envy.”? | 

Under this general identity of mechanism, however, the 
three groups differ in that the first expresses only an un- 
shaded statement of expected future or actual present or 
past fact (“you will not see me coming hither again”), while 
the second expresses impossibility (“you cannot take the 
city’’), and the third expresses resolve (“ Zeus shall never 
force me to utter this” —the cry of the speaker as he steels 
himself against yielding. Cf. the resolve of Achilles ex- 
pressed in the independent indicative in A 88: od Tis eyed 
favtos Kal él yOovi Sepxopuévoro | coi Koidyns trapa vyvol 
Bapelas xelpas éroice: | cvpravtov Aavadr). 

The examples under D (doubtless many more remain to. be 
noticed) clearly differ from those printed above them in that 
they do express an act demanded by some one. They are 
therefore identical in character with the examples of the Ex- 
tended Deliberative, and might equally well be written with 
the subjunctive. On the other hand, they are precisely such 


1 If any, however, take a different view of this example, they will properly 
regard it as an instance of a side-branch from the idiom seen under D, bearing 
the same relation to that idiom which the example from Euripides bore to all the 
othér examples of the extended deliberative. 


Vol. xxiv.] Deliberatives in Greek. 177 


examples as one would expect to find if one had information 
that the Greek Aad developed an extended clause of purpose 
of the class to which Mr. Earle and J. D. supposed the sub- 
junctive examples to belong. 

That such a development is antecedently perfectly possible 
I have already said. But so is also a development from the 
deliberative future indicative, parallel to that of the subjunc- 
tive examples from the deliberative subjunctive. Examples 
of the independent deliberative future indicative, as in Aesch. 
Ag. 1362 and 1367, are common enough. But examples are 
not rare of the dependent clause, as in Soph. O. C. 1742 
(87rws porovpeO” és Sopous | od éyw), Theocr. 16, 16 (1ddev 
oicetat aOpet | dpyupov), and 17, 10 (amraive ... 1dbev 
apferar Epyov), and Plat. Gorg. 521 B (ov«m &€e bri ypioerac 
avtots. Cf. the corresponding subjunctive in Rep. 368 B, 
atrop® tt xypjowpat, and Euthyd. 287 C, ov« éyw bri ypjow- 
pat). As there is a simple dependent deliberative future 
indicative, so there might also be an- extended deliberative 
of the same mode and tense. : 

Which of these two possible origins is the real one? I see 
no means of reaching an absolute decision. Probability, how- 
ever, appears to point to the second rather than to the first. 
In the case of the subjunctive idiom we have found positive 
evidence for belief in a deliberative origin. Unless there is 
positive evidence forthcoming for a different origin in the 
case of the indicative idiom, it will be sound method not to 
assign it to a different cause. The Greek consciousness of 
the kinship of the force of the subjunctive idiom with that of 
the deliberative subjunctive has kept the mode, in spite of the 
approach of the meaning toward that of purpose, from pass- 
ing over into the mode of purpose, and has kept the formula 
very nearly confined to the expression of the existence or 
non-existence of a difficulty. Within this range of meaning, 
then, within which our indicative examples also are confined, 
the Greek feeling for the quasi-deliberative form was clearly 
stronger than for the quasi-final form, and the Greek mind 
is likely, therefore, in using an indicative, to have fallen 
upon the practically identical deliberative idiom rather than 


178 William Gardner Hale. [1893. 


upon the idiom of the final clause. In default, then, of out- 
side evidence for the final clause, — which, up to the present 
time, I have not found, —I should regard the whole group 
of phenomena, subjunctive and indicative alike, as of deliber- 
ative origin.? 

The examples of the unquestionable dependent deliberative 
indicative which in this last discusssion we have seen (p. 177) 
bear upon the question of the genuineness of the dependent 
ted dv .. . doin Homer & 192. The subjunctive with av 
or xe is, in most idioms, either a forerunner of, or an ever- 
possible substitute for, the future indicative. Even, then, 
if no other examples of the dependent deliberative subjunc- 
tive with ay or xe appeared, this example should not sur- 
prise us. But other cases do appear, as in N 741, évOev & 
av para tracav éridpaccaipeba Bovrnv, |} Kev evi vynecoe 
ToruKAniot Twécwpev, | al x €Bédnor Beds Souevat Kpatos, H 
kev éreita | map vndv Ewpev amnpoves, and I 618, dua & 
not harvouerndiy | dpaccducl’, } Ke vewuel’ ed Huétep’, } KE 
pévopev. (Cf. A 14, where Smws gota: tade Epya after 
dpatwoue? is shown by the context to mean “how these 
things shall be,” not “how they will be,” and therefore 
forms the connecting link with the dependent deliberative 
indicative as seen in Attic.)? 

The text in } 192 is therefore not invalidated by the em- 
ployment of xe. Nor is it invalidated by the use of the inter- 
rogative ted, since such a cast of the sentence, seen already 
above in the case of Old English, forms precisely the first 
stage which one must assume for the development of the 
Greek idiom with the bare subjunctive, which, on indepen- 
dent grounds, we have found to be an extended deliberative. 
The text is therefore sound, and adds strength to a case 
already strong. 

1 Cobet, Collectanea Critica, p. 137, takes the dependent clause in Aesch. P. V. 
85, above, as equalling guo pacto . . ., probably having in mind the interrogative 
force. Dindorf, Lex. Aeschyleum, s.v. bots, uses the phrase “in interrogatione 
indirecta” in citing the same passage. 

2 The general view is that the deliberative subjunctive with dy does not occur 


in Attic. So Baumlein, Untersuchungen, p. 184, and Kiihner, Gr. Gramm. § 394, 
Anm. 4. : 


Vol. xxiv.] Deliberatives in Greek. 179 


I regard the evidence, then, as being conclusive in favor 
of the view held substantially in common by Goodwin, Jebb, 
Tarbell, Verrall (by implication from his note upon Aesch. 
Ag. 620), and Sidgwick, against the view of Monro, Earle, 
and: *JaD.? 

Monro’s view appears only in the sentence already quoted. 
The views of Sidgwick and J. D. appear only incidentally, and 
without argument, in their discussion of the optative idiom, 
and will therefore require no separate consideration. 


II. 


The certainty which has apparently been reached with 
regard to the history of the subjunctive idiom does not 
extend to the history of the optative idiom seen in ov« éo® 
draws AéEatpu, etc., to which we now pass. The result of our 
examination will be, I think, the disproof, in any case and 
finally, of one of the two contending views.. The other view 
will be shown to have strong arguments in its favor, but a 
certain difficulty—not necessarily finally insoluble —will still 
remain. 

Madvig appears not to discuss the optative idiom. Kriiger 
(II. § 54, 3, 8) assumes the omission of dy in eight examples 
from the Attic stage (see also I. § 54, 14, A. 4). G. Wolff 
(Rh. M. 1863, pp. 602-6) inclines to emend all independent 
examples by inserting ay or reading a subjunctive. Goodwin 
(M. and T. § 50, 2, N. 1, of the editions between 1860 and 1890, 
and § 241 of the new edition) regards the dependent examples 
as “conditionals” or “potentials.” without av; and this was 
the generally accepted theory, down to the year 1881. 

The participants in the debate which began with that year 
have been Mr. Sidgwick, Professor Jebb (in his note on Soph. 
O. C. 170, and Appendix I.), Mr. Verrall (in his note on Aesch. 
Ag. 620), Mr. Jerram (in the note on v. 52 in his edition of 
the Alcestis of Euripides), Mr. Earle (Classical Review for 
March, 1892), and a writer in the same journal signing the 
initials J. D. to articles in the numbers for December, 1892, 
and March, 1893. Professor Wecklein has also expressed his 


180 William Gardner Hale. [1893. 


opinion in a review of Mr. Sidgwick’s Choephoroi in the Berl. 
Phil. Wochenschr. for 1885, p. 487 seq., and, incidentally, in 
the Jahresbericht LX XI. (1892), p. 181. 

Mr. Sidgwick’s views were first put forth in his school 
editions of the Agamemnon (1881) and the Choephoroi (1884), 
Appendix I. The substance of them may be seen in the 
following passages, which in substance are common to both 
books : 

“Now, it must be plain, considering these examples all in 
a lump, that what they vary from is not the optative with ay», 
but the interrogative subjunctive, or, as it is usually called, 
the deliberative. The subjunctive might be substituted for 
the optative in all these instances: and in the first two it is 
usually so read, though against the best MS. authority. 

The difficulty then is this: not why dy is omitted, for the 
sentences are not conditional; but why the remote form [op- 
tative] is used instead of the przmary form [subjunctive], when 
the sentences are all of a primary character. 

The answer is that the optative expresses the remoteness, 
not as usual [e.g. in past final, or past indefinite, or past delib- 
eratives] of pastness, but of possibility: the instinct is to 
express by optative something more out of the question than 
the subjunctive would have expressed. 

Thus, e.g., in the third instance tis catdoyn ; (see p. 182) 
would be good Greek, but the question of restraining Zeus’ 
omnipotence would seem to be more treated as a practical 
one: the optative puts it further off, as a wild impossibility. 

Or again, in Ar. Plut. 438 mot @uyn would be in ordinary 
circumstances the expression, and so the older editors all 
read it: but @vyo., the MS. reading, and the right one, is 
the exclamation of supreme terror, treating escape as in the 
last degree unlikely.” 

The examples cited by Mr. Sidgwick in these appendices 
were from Attic poetry, with the exception of two from 
Plato. In the appendix to his edition of the Oedipus Colo- 
neus (first published in' 1885), Professor Jebb expresses his 
agreement with Mr. Sidgwick’s main point, but very carefully 
sifts the alleged examples. In Soph. O.C. 170 he would read 


Vol. xxiv.]  _ Deliberatives in Greek. 181 


&On, and in Ar, Plut. 438 déyy, on the ground that the appar- 
ent meaning demands the subjunctive. In the case of most 
of the examples which he cites from Attic prose (Lys. 31, 24, 
Dem. 21, 35, Plat. Gorg. 492 B, Euthyd. 296 E, Antiph. 1, 4) 
he points out an easy way in which av might have acciden- 
tally dropped out, and expresses himself as inclined to adopt 
the theory of such a loss in the case of all the examples. 

Mr. Verrall, in his edition of the Agamemnon, 1889, accepts 
Mr. Sidgwick’s view (see ad v. 625), as does also Mr. Jerram, 
in his edition of the Alcestis of Euripides, in the following 
note, ad v. §2: “ For pedo, the opt., we should expect porn, 
the subj., after the primary tense ov« éort. But the optative 
is used to express something farther removed from possibility, 
‘Is there absolutely no chance, ete. ?’” 

The next reference to the matter is at the conclusion of 
Mr. Earle’s article in the Classical Review for March, 1892, 
in these words: “If the MSS. are to be trusted, we some- 
times have the optative of purpose, instead of the subjunctive, 
after primary tenses. Cf. Eur. Alc. 112, Aesch. Prom. 291, 
Ch. 72.” In the Jahresbericht of the same year (loc. cit.), 
Professor Wecklein incidentally again expressed the belief 
which he had expressed before in the review already referred 
to, that the constructions in dispute were potentials. In the 
Classical Review for December, 1892, J. D. argued the case 
against Mr. Sidgwick’s remote deliberative and for the poten- 
tial. Mr. Sidgwick replied in the March number for 1893, 
holding to his view with regard to the examples from the Attic 
playwrights, but adding that Professor Jebb had convinced him 
that the prose examples ought to be given up, on the ground 
that ‘‘the omission of dv is much more likely to be an over- 
sight of the scribe in these cases.” To some of the objections 
which in this article Mr. Sidgwick urged against J. D.’s view, 
the latter made answer in the Classical Review for October, 
1893, and Mr. Sidgwick, having seen the article in proof, 
added a few comments. 

Before we can proceed to weigh the arguments which have 
been put forward in this discussion, it is necessary to have 
before us, in classified form, all the examples that have thus 


7S 


182 


William Gardner Hale. [1893. 


far been produced. I add a considerable context in the cases 
in which the exact shade of meaning might otherwise not be 


clear: 


XO 


OI. 
AN. 


The Independent Construction. 


TOANA ev ya TpEpeEr. 
devva SeipaTov ayn, 

, / wt ke / / 
movTial T ayKadat KYwddraV 
avtaiwy Bpototct 

, a XA / 

TradGovet* Bractovat Kal Tedaixypsot : 
AapTrades Treddopot 
TTava Te Kal TedoBa- 
ova KavepoevtT av} 

> , , 
aiyidwv dpdcat Koro * 
arn brréptorpov avdpos dpdvnua tis A€éyou 
Kai yuvatkav dpeot TANMOVOY 
TavToAmous Epwras 
araiat cuvvopovs Bporav ; — Aesch. Cho. 585. 


reav, Zed, Sivacw tis avdpav 
irepBacia Katacyot ; 

\ wf)? oe e lal + Sb. 3 / > ? f 
Tav ov imvos aipet To?’ o TaVvT aypev@v 
ovTe Jed akwaTot phves, aynpaws Sé ypove 

, , > 4 , ” 
duvaatas Katéyes "Odvptrov pappapoeccay aiyray. 


— Soph. Ant. 604. 


2 


tov, Eéve twappop’, ed pvrAaEat, 
petactal’, amoBabt. oA- 
Aa KéXevOos EpaTves * 
KAVELS, @ TOAULOXO aATa ; 

, ” > y¥ 
Aoyov el TLV OloELS 
Tpos éuav Nécxav, 4Batav aroBas, 
iva Tact vo“os, 

s , > > / 
gaover* tpocbev & atrepvxov. 

a. BA 

Ovyatep, trot tis hpovTidos EXOor ; 
@® TaTép, aatois loa ypn medeTar, 
eixovtas & Set KaKOvoVTAs. 


1 Codd. kdveuoévrwy. The gender of alyléwy compels emendation. 
2 Jebb’s reading. 


Vol. xxiv. 


Ol. 


XO. 
HA. 


Ol. 
OH. 


Ol. 
Ol. 
Ol. 
OH. 
Ol. 


OH. 


J Deliberatives in Greek. 
mpoabuyé viv wou. — Soph. O. C. 161.4 


avaé”ArrodXov Kal Beol, rot Tis Piryot ; 


— Ar. Plut. 438.1 


kal ti hiros péEays ; yapor wAHOovew avias. 


— Theocr. 27, 24.? 


The Dependent Construction. 


ovk tof bras A€Eaiwt TA Wevdh Kara 
> \ \ co , 
és Tov Todvy Pirovot KapTrodabat ypdvov. 


— Aesch. Ag.620. 


tais cais 8é tUyass, ioO1, cvvadyo@° 
TO Te yap Me, SoKe, suyyeves OUTWS 
é€cavayxatet, 

xepis Te yévous ovK EoTiv OTH 


peilova potpay veiwaim’ 7 coie— Aesch. P. V. 288. 
/ 


e s \ ‘ s ‘ 
Ta: obY Tadala Tapa vewTépas palo ; 
ovK Eotiv baTIs TAN Evos KEipaLTO VLV. 


— Aesch. Cho. 


tis Snr av ein tTHVS Oo Tpocbaxav Edpay ; 
é 7” y e \ > / 
pa kat “Apyos et Tis vpmlv eyyerns 
eo, daT1s av cou TovTO TpoaxpHCoL TUXEtD. 
@ hirtate, oxés odTep el. OH. ti 8 Eats cor; 
pn pou SenOjs. OH. wpdypyartos toiov; Réye. 
»” ee 7 lal > > e , 
G01 axovav TaVd bs éof 0 TpocTaTns. 
\ y ae \ ad B= JA / s 
Kal Tis TOT eaTiV, bv ¥ eyo WéEatpt TL; 
mais ovupds, a@vak, oTuyvos, ov NOyav éyo 
adyict av avdpav éEavacyoiuny Krvov. 
/ >] > > A uv \ \ cal 4 / 
TiS ; ovK axovey Eott, Kal un Spav & pH 
7 ’ age > ‘ \ 4 
xpncers ; Ti aot TOUS eat AvTNpoOV KrUELD ; 


171. 


— Soph. O. C. 1166. 


1] agree with Jebb that the general character of these two passages (the first 
is sufficiently shown here) demands the subjunctive. 
? This example was added by J. D., but is not included in Mr. Sidgwick’s latest 
list, Classical Review, March, 1893. 


- 


184 William Gardner Hale. [1893. 


GAN od6€ vauKAnpiay 
wv vee’ y 
éo@ Strat Tis alias 
oteiras 7 AuKias 
hat Mae \ \ > tA 
elit’ éml Tas avvdpous 
"Aupmovidas &épas 
dvotavou tapadvcat 


uyav:— Eur. Alc. 112.1 
AIl. XaBov 10+ ob yap oiS av ef wetcaipi ce. 
OA. xteivev dv dv ypH; TodTO yap TeTaypcba. 
AII. od«, addr Tots wéAXoveL Oavatov éuBareiv. 
OA. Exw Aoyor 87 Kai rpoOuplav cébev. 
AIl. éo7° ody bras "AXKnotis és yipas morXor ; 
OA. ovK Eote* Tipais Kape TéptrecOas Soxer. 
AI. obras wréov x’ dv } piav Wwoynv AaBors. 
OA. véwv POivovtav peifov dpvupar yépas. — Eur. Alc. 48. 
Tis TOVS Epupvav Swpdtov Eyer KpaTos, 
batis Eévous déEatTo TovTiw cad 
KapvovTas ev YELl“a@VL Kal vavayiats ; 


— Ar. Thesm. 871. 


The following, which has not before appeared in the con- 
troversy, is thus given by Dindorf : 


Eo@ Straws dvev payns Kal Tis Katokeias Bojs 
€s Nayous ENOowpev GAAHAOLCL Kal Siadrrayas ; 
— Ar. Vesp. 471.2 


Mr. Earle’s theory, it seems to me, may be briefly dis- 
missed. Something approaching purpose (though still clearly 
differing from it) might be felt in the case of Alc. 112, but 
in P. V. 288 it would appear a palpable forcing to render by 
any phraseology remotely resembling such a form as “no one 


1 Not given by Mr. Sidgwick. Theogn. 382 is probably intentionally omitted 
by him, together with Homeric clauses with 8s, as too early.— Suppl. 20, Phil. 895, 
Plut. 374, and Av. 172, once read without dy, are now generally emended. 

2 V. alone, according to Dindorf, has €\@ocmev, the rest of the Codd. having 
Z\Owuev. (Liddell and Scott, under dxws, have inadvertently cited the passage 
with both readings.) The subjunctive is impossible, making a sentence unlike 
any that we have seen in Part I. The only question, then, is whether we ought, 
with Hermann, to amend dvev to ap éx. 


Vol. xxiv.] Deliberatives in Greek. 185 


exists for the purpose of, with a view to, my paying him 
greater honor than I pay you.’”’ Mr. Earle’s view seems to 
owe its origin to a desire to find a relationship between the 
optative construction and the subjunctive construction, which 
latter he has treated as final. If that view falls, as in 
Part I. I have tried to show that it must, Mr. Earle’s probably 
strongest reason for his explanation of the optative falls with 
it. It should be added, too, that none of the disputants have 
agreed with him on this point. J. D., who shares his view 
of the subjunctive construction, regards the optative con- 
struction as potential. 

J. D.’s argument upon the optative, as given: in the Decem- 
ber number, 1892, falls into two parts ; vzz. (1) a special argu- 
ment, and (2) a canon with application. 

The argument is in substance as follows: There is no need 
of resorting to “a grammatical novelty,” since the examples 
can all be explained as potentials unaccompanied with ay, 
such as are frequently found in Pindar and Theocritus (to 
say nothing of Homer), and occasionally even in Attic (as 
Aesch. Eum. 265). 

The canon, mingled with the application, is contained in 
the following : . 

“An oblique question is essentially a question repeated in 
word or thought; it is therefore necessarily dependent on 
- some verb or verbal substantive implying perception of mind 
or sense or the outward expression of such perception (verba 
sentiendi et declarandi). In these indirect deliberatives no 
such introductory statement occurs; neither can it be under- 
stood,” 

Briefly put, the canon would read, as Mr. Earle’s canon 
read (in my version) above: If in a given example the full 
apparatus and full force of a given known construction do 
not exist, then the construction found in the example can 
have no relation with that known construction. This canon, 
on which I have already commented in criticising Mr. Earle’s 
theory with regard to the subjunctive idiom, would leave us 
helpless in face of many an idiom in many a language. To 
say nothing of the cwr-question of propriety turned relative, 


186 William Gardner Hale. [1893. 


which I have already brought into court, with what heart 
would one who held such views of method confront a phe- 
nomenon like ti daywowv in Mark 6, 36: amdAuvcov avtods, 
iva atreNOortes eis TOUS KUKAW aypols Kal Kaas ayopacwoww 
éautois TL Paywotv, — which is after all but a step in advance 
of the vod . . . «Alvyn of Luke g, 56: 6 8é vids Tod avOpwrou 
ovK exer Tod THY Keharny Krivy ?} 

This part of J. D.’s argument, therefore, has in my judg- 
ment no weight. If there exists an independent remote 
deliberative in Greek, then it could readily pass over into-a de-- 
pendent relative form, just as the deliberative subjunctive did. 

The other argument of J. D., on the other hand, certainly 
is sound in form. Whether it is sound in substance or not, 
I shall attempt later to show. 

Before proceeding to the minuter examination of the two 
important rival theories, it will be well to consider certain 
fairly evident antecedent objections that have been raised, or 
might be raised, against one of them. - 

Against Mr. Sidgwick’s theory lies the evident and not 
inconsiderable objection that the cause invoked by him to ex- 
plain the phenomena is a cause not known to exist. The only 
certified optative of remoteness is an optative of the past. 
While, then, this state of things is not necessarily fatal to 
the theory, it constitutes a strong objection to it, and a cor- 
respondingly strong presumption in favor of any opposing 
theory which —like that of the potential origin of the idiom — 
should be based upon known causes. 

This objection lies against Mr. Sidgwick’s theory in its 
application both to the independent and to the dependent 
examples of the idiom. A ‘second difficulty? lies in the 


1 See also in Mark 8, 1 and 2, uh éxdvrwy Tl Pdywouw and ovk Exovew Tl Payo- 
ov, and in Luke 12, 17, Ti rolnow, drt obk Exw Tod cuvdsw Tods Kapmrovs mou; 
I owe these examples to the kindness of my colleague, Professor Burton, who gave 
them to me in advance of the publication of the new edition of his Syntax of the 
Moods and Tenses in New Testament Greek (1893). See § 346, and cf. § 319. 

2 Though I should class this among the fairly evident antecedent objections to 
Mr. Sidgwick’s theory, it has not been pointed out in print by any one, nor had - 
it occurred to me before it was suggested by Miss C. E. Millerd, in a discussion in 
my syntactical seminary. 


Vol. xxiv.] Deliberatives in Greek. 187 


numerical relation between these two classes. Accepting 
all Mr. Sidgwick’s examples of the independent sentences, 
together with the one added by J. D., we count but five in 
all, including the rather late one from Theocritus.! Of the 
dependent examples we have found seven or eight. Now 
the construction of which the subjunctive idiom already dis- 
cussed is a special extension, vzz. the independent deliberative, 
was an idiom in constant use. There are in the literature 
undoubtedly many hundreds of cases of the independent con- 
struction to perhaps two dozen of the extended construction. 
It is easy to believe in the occasional extension of an idiom 
repeatedly read in all kinds of literature, and repeatedly heard 
in daily speech. It is correspondingly difficult to believe in 
the extension of an idiom found a few times only in a single 
kind of literature, and never found elsewhere, —and therefore 
probably never heard in speech. And even if this general 
difficulty were to be surmounted, it still remains hard to 
understand why, in the one class of literature in which the 
constructions do occur, the dependent examples should ex- 
ceed the independent in number. In order to meet this 
objection, Mr. Sidgwick will need to shape his theory still 
further, by seeking for some conception of the peculiar nature 
of the idea produced by the combination ov« &o@ or éo with 
87rws, etc. — 

The minuter examination of the general question will fall 
under four heads, the first three having to do with the force 
of the construction, the last with the form: (1) An applica- 
tion to Mr. Sidgwick’s view of certain tests suggested by the 
nature of that view; (2) an application of corresponding tests 
to the theory that the construction is potential; (3) an exami- 
nation of the evidence afforded by the contexts of the examples 


1 As already said, I agree with Professor Jebb that two of these examples 
should be thrown out. And I further agree with Mr. Verrall’s recent note on 
Cho. 593: “As it is possible to supply év from the parallel clause [¢pdoa dv] 
preceding, this cannot be counted with certainty among the examples of the 
simple optative used in this way [for the expression of greater remoteness or 
impossibility] as a variant for the subjunctive.” Cf. the omission under corre- 
sponding circumstances in prose, as in Plat. Rep. 352 E. Wecklein regards the 
explanation as entirely satisfactory. 


« 


188 William Gardner Hale. [1893. 


cited ; (4) the consideration of objections to the solution indi- 
cated by evidence reached in (1), (2), and (3). 

(1) Mr. Sidgwick’s theory that the optative idiom “varies,” 
not from the optative with dy, but from the subjunctive idiom, 
and that its difference from the latter lies in its expressing 
the idea that its contents are “more out of the question,” 
“in the last degree unlikely,” “a wild impossibility,” must 
necessarily rest for its support upon the truth of the fol- 
lowing tenets: (a) This meaning cannot, by the necessary 
nature of examples cited, be proved to be conveyable by the 
potential with dv; (4) this meaning zs proved, by the neces- 
sary nature of examples cited, to inhere in the idiom under 
examination; and (c) the differentiation of this idiom from 
the subjunctive idiom lies only in this one point, the two 
idioms being otherwise identical in force. These three neces- 
sary tenets will be taken up in order. 

(a2) How would the answer: to Mr. Sidgwick’s reav, Zed, 
Suvacw Tis avdpav irepBacia Katacyo. be expressed? Does 
not a form of expression modelled on Homer's téxva gi’, 7 
to. | Znvi Bpotav ovk« av tis épifor (5 78) and Plato's od phy 
éote KaXXiwY O605 OvdS dv yévo.To (Phil. 16 B),—namely, the 
form ovdceyia av . . . Katacyot, suffice? If it does, then Mr. 
Sidgwick’s genealogical tree of “ grammatical affinities,’’—to 
make use of a very good term of his, —is putting very near 
relatives asunder. If it does not, would not ov« éo@ é7ra@s av 
Tis... Katacyo. suffice? . The truth is that the idea of be- 
ing ‘more out of the question,” “in the last degree unlikely,” 
is to be found in an abundance of potential examples, if 
one judges them with the same readiness to find the idea 
with which we approach Mr. Sidgwick’s examples. £.g. his 
ovK Eat batts TAY évos Keipartd vv May be compared with 
TAY TODS av ovdels evdikws péurraito pot, from v. 63 of the 
same play. The idea of being out of the question certainly 
does not more necessarily inhere in xal tis mot éotiv, bv éy@ 
WéEauit te (Soph. O. C. 1172) than in was av 7d yy axov 
Tpaypy av eixoTws Weyos ; in v. 977 of the same play. And 
surely nothing could be more “wildly impossible” than the 
dependent clause in Ar. Nub. 1181: od yap éo@ é7ras | pi’ 


Vol. xxiv.] Deliberatives in Greek. 189 


nuépa yévorr’ dv jpépa Svo, or in Soph. Ant. 911: pntpos & év 
‘AiSov xal matpos KexevOdrow | od Eat adedpos boTis av 
Brdorot Tort. 

The meaning which Mr. Sidgwick conceives to be funda- 
mental in our idiom being, then, perfectly conveyable by the 
potential optative, there must be some grave reason to justify 
a divorcing of the two, and an attaching of the former to the 
subjunctive idiom. 

(4) Since the differentiation of the optative clauses from 
the subjunctive is defined as lying in the “wild impossi- 
bility’ conveyed by the former, the two must of course be 
understood to be in other respects alike. Let us see whether 
they are. 

In Part I. we have found that, with the exception of the 
single example Euripides H. F. 1245, which represents an 
arrested side-growth of the construction, all of the twenty-five 
subjunctive (or corresponding optative) examples from Classi- 
cal Greek express an act demanded by the speaker, or by 
some one else; while the main sentence, if negatived, as it 
is in. all the examples down to Aristophanes, expresses the 
non-existence of a way, a person, etc., for the accomplish- 
ment of this demand. The formula for the sentence of the 
earlier type would then be the following: It is my will (your 
will, his will, etc.) that a certain thing be done, but there 
is no way. If, then, the optative idiom differs from the sub- 
junctive as Mr. Sidgwick believes it to differ, and only thus, — 
the meaning of a negative main sentence with attached opta- 
tive would be expressed by the formula: It is my will (your 
will, his will, etc.) that a certain thing be done, but there 
is no conceivably possible way. Such an interpretation would 
fit Alc. 48, but not the other passages ; for Ag. 620, so inter- 
preted, would mean “I want (or, at any rate, somebody wants 
me!) to make a lie seem fair, but there is no conceivably 
possible way”; P. V. 291 would mean “I want (or somebody 
wants me) to pay greater honor to some one else than to you, 
but there is no one to whom I possibly can” ; Cho. 172 would 


1] waive the fact that, in all the subjunctive examples from Aeschylus and 
Sophocles, the demand is that of the speaker. himself. 


at 


190 William Gardner Hale. [1893. 


mean “it is demanded that some one proceed to cut it, but 
there is no conceivably possible person, save one”’; and O. C. 
1172 would mean “I want (or, at any rate, you want me) to 
censure some one, but there is no one whom I can possibly 
censure,” —all of which renderings are impossible. 

(c) But a still more serious defect in the theory —a fatal 
one, it seems to me—lies in the fact that this meaning of 
“wild impossibility ” is zo¢ fundamental in these examples. 
In two of them, to say nothing of Vesp. 471, the meaning is 
the exact opposite; namely, in Alc. §2 and Thesm. 871, — 
“might she then be spared?” and “who is here of kingly 
power, that mzght receive?’”’ A moment’s examination, too, 
will show that the thought in O. C. 1172 is not “there is 
no one whom I could possibly blame.” These simple facts 
are destructive ; for it is solely upon the idea of ‘“ remote- 
ness from possibility” that Mr. Sidgwick’s explanation of 
the grounds of the optative is based. 

(2) We pass to the theory that the optatives in question 
are potentials. The tests to be applied to this theory 
would be, so far as the force of the idiom goes, (a) the the- 
oretical sufficiency of the potential idea to give the actual 
meanings found in Mr. Sidgwick’s examples, and (4) actual 
correspondence of these examples with others, otherwise .the 
same, in which evidence of the potential meaning is given 
by the presence of av. 

(a2) The potential expresses possibility. When, then, as 
in Alc. 52, Vesp. 471, and Thesm. 871, the main clause in a 
complex sentence is not negatived, the potential yields at 
once the only meaning conceivable in the passages; namely, 
“is there a way by which she might come to old age?” 
“might she possibly in some way come to old age?” “ mzght 
we possibly come to terms?” etc. In the case of these 
examples, accordingly, the theory of a potential origin meets 
the test of consistency with itself which Mr. Sidgwick’s the- 
ory failed to meet. 

When, on the other hand, the main clause is negatived, 
the effect of the sentence as a whole is the denial of the 
existence of any possibility, which is precisely the mean- 


Vol. xxiv. ] Deliberatives in Greek. IgI 


ing attached to the construction by Mr. Sidgwick’s theory, 
though he has mistaken the place where the negation lies. It 
is worth while to say in passing that it is only by this negativ- 
ing of the main sentence that the idea of “wild impossibility” 
can possibly be attached to any dependent optative, itself 
unnegatived. In other words, the question whether a given 
complex sentence shall express a possibility or an impossibil- 
ity turns, not upon the force of the dependent clause, but 
upon the presence of a negative or implied negative (as often 
in an interrogative sentence) in the main member. 

The theory that the construction is a mere potential, then, 
will explain Ag. 620 (“the way does not exist by which I 
could make a false tale seem fair,” etc.), will explain Alc. 112 
(“no place exists by faring forth to which one could set free 
the hapless woman’s life’’), will explain Cho. 172 (‘no one 
else could cut it’’), will explain Alc. 52 (‘does a way exist 
by which she might”), will explain Thesm. 871 (“who is king 
here, that might receive guests weary,” etc.), will explain 
P..V. 291 (“there is no man whom I could place above your- 
self’), and will explain the difficult sentence O. C. 1172 under 
its most probable interpretation (see footnote, p. 202). 

So much for the dependent optatives. As to the interrog- 
ative independent construction of the same mode, it will 
depend wholly upon the nature of the individual case whether 
a given optative be taken as a serious inquiry about the exist- 
ence of a possibility, or.as the virtual expression of an im- 
possibility. ‘‘ How could this be done?” may mean either 
“inform me of the way by which the accomplishment of the 
act would be made possible,” or may mean “ how utterly 
impossible the act is.” 

The potential, then, is in its essential nature entirely com- 
petent to yield the force of complete impossibility which Mr. 
Sidgwick found in certain of his examples. It is also compe- 
tent to account for the resemblance which, though not inhe- 
rent in the idiom, he rightly found to exist between certain 
of his examples and the examples of the subjunctive idiom. 
The idea of daffied impulse which we find in the earlier type 
of the subjunctive examples (the impulse being expressed by 


192 William Gardner Hale. [1893. 


the subjunctive and the baffling by the denying, in the main 
clause, of the existence of a means, etc.) does indeed lie close 
to the idea of the xon-extstence of a possibility (the possibility 
being expressed by the potential, and the non-existence by 
the main sentence). And, finally, it is competent to yield 
the force found in the remainder of Mr. Sidgwick’s examples, 
which express fosszbzlity, and which are widely separated 
from the examples of the subjunctive idiom. 

(6) But we are not confined to a theoretical demonstration. 
For the competency of the essential nature of the potential 
to yield the force actually found in the idiom under examina- 
tion is supported by the exact correspondence of these exam- 
ples with other optative examples the potential character of 
which is made indubitable by the presence of av. A table 
of parallels follows: 


teav, Zed, divacw tis avdpav 
trepBacia Kataoxot ; 
— Soph. Ant. 604. 


70 8 én yav drag weodv Oavaowov 
“mporapo dvdpds péAav alua Tis av 
madw ayKadéoat éxacidwy ; 

— Aesch. Ag. Iolg. 
pever 8¢ pipvovtos év Opdvw Ards 
mabeiy tov épgavta. Oéopuov yap" 
Tis dv yovay apaiov éxBaror dopwv ; 
KexOAAnrat yévos mpos ata. 

— Aesch. Ag. 1563. 


GAN’ tréptoApov avdpos dpdvnua 
tis A€yot ; — Aesch. Cho. 594.1 


Kal ti gidos peau;  yapor 
mAnGovew davias. 

— Theocr. 27, 24. 
ovk é09 dws A€Saye TA Wevdy Kadd. 
és tov modiv ido Kaprovoba 

xpovov. — Aesch. Ag. 620. __ 
gor’ ovv orws “AXxnotis és yypas 
porot ; — Eur. Alc. 52. 


GAAG Ti Kev peta; Geds da wavTa 
teXevTa. —T go. 


ov yap éo8 ows 
pi” uépa yévorr’ av ajuépa Svo. 
— Ar. Nub. 1181. 
xoux €of Sxws duadis av quads Ere 
Aabor. — Ar. Vesp. 212. 
aor «i py Tt Oavpacrov dcov dia- 
pipet TH dperH TOV GAXwY, otk éEoF 
Grws av Tis P¥yot TO KaTayéAacTos 
yeveoOau, pacxwv éxew tTavrnv THV 
émotnunv. — Plat. Lach. 184 C. 


1 In agreement with Jebb, I do not here include Mr. Sidgwick’s two remaining 
examples. There would be no difficulty, however, in giving a long list of par- 


allels if they were to be retained. 


Vol. xxiv.] 


> ” bid ‘ ear ld , 
ovK €oTlv OoTis tANV EVOS KELpatTo 
viv. — Aesch. Cho. 172.1 


xwpis Te yévous ovk Ear OTH 
peilova poipav veipaty 7 ot. 

— Aesch. P. V. 291. 
GAN’ ov6€ vavKAnpiav | 
” a ” 
éoP orot Tis alas 
oreiAas .. - 
dveravov tapadvoat 
Yuxav. — Eur. Alc. 112. 


kal tis mor éotiv, dv y eyo Webarpi 
tt; —Soph. O.C. 1172. 
tis TOvO épuvpvdv Swydrwv exe 
KpaTos, 
datis E€vovs Sé~arro . . .; 
— Ar. Thesm. 871. 


Deliberatives in Greek. 


193 


el yap tis hain THO TOAEE TovTW 
treiotwv ayabayv airias yeyevno Oar 
tois “EAAnot kal peyiorwy KaKav 
peta THY Répfov otpareiay, ovK €oTw 
Grws ovK arnOy ddgeev dv A€yew 
Tots €iddot TL wEpl THY TOTE yeyevy- 
péevwv. — Isocr. 12, 156. 
GAN ovdé hirwv zéXas Ear’ ovdeis, 
dotts dv eiroe rétepov POipevnv 
Bacirieav xph revbeiv, 7) Lao" 
étt Has Aevooe: MeXlov zais. 

— Eur. Alc. 79. 
pytpos 8 & “Atdov kai matpds 
kexevboroww ; 
ovk gor ddeAdds doris dv BAGdoTOL 

motré. — Soph. Ant. git. 
ovk éoF Grou Oiyouy’ dv évdixcsrepov. 
— Eur. El. 224. 


ov yap éoF Sov 

€oOAdv tt Spdcas paptup av Ad Bors 
matpav. — Eur. H. F. 186. 

ovk éo8 Groiov oravt av avOpwrov 
Biov 

our’ aivécay, dv ovre peprpaiunv 
more. — Soph. Ant. 1156. 

ovk got ovdcis Gotis dv péurparro 
oe. — Eur. El. 903. 

6pa kat “Apyos «i tis dpiv éyyevns 

eof’, dotis dv cov TOUTS tporxpy ot 

. Tuxeiv. — Soph. O. C. 1167. 


(3) We come next to a scrutiny of the contexts of the 


examples under examination. 


In the case of two, clear 


evidence seems to be presented that the mode is potential. 


1 Compare the following: 


@s ox @00’, Os ofs ye ktvas Kepadfis 
dmahddkor. — X 348. 


obdé of &AXoe 
elo’, of kev xara SiHuov adddxovev 
kaxérnTa.—o 166. 


194 William Gardner Hale. [1893. 


In the first passage from Aeschylus in the list on p. 182, 
g¢paca is beyond any reasonable doubt an optative, and, be- 
ing an optative, is unquestionably potential, whether an av 
is obtained by emendation or not. But the sentence of which 
it is the verb is exactly parallel with the sentence tis Aéyor, 
differing from that only in being declarative where tis Aéyor is 
interrogative. The passage runs rapidly through some of the 
manifestations of the power of nature, and then continues, 
“‘of the rushing wrath, too, of hurricanes, the fowls of the 
air and the beasts of the field can tell. But who can tell the 
soaring thought of man and the frenzied passion of woman’s 
daring spirit?” .The “grammatical affinities” of the idiom 
in question are thus clearly fixed for us by Aeschylus himself. 

Similar evidence appears in the passage from the Alcestis 
of Euripides. At v. 48, Apollo, giving up the contest so far 
as he is concerned (but all the time knowing, as he after- 
wards prophesies in v. 65, that one is coming to rob Death of 
Alcestis), says, ‘‘Take then thy prey and go; for I know not 
whether I could persuade thee.” Death answers, “ Slay the 
one that should die? for that was our agreement.”’ Apollo, 
seizing the suggested advantage, replies, ‘No, bring death 
on those who soon must die.’ Death retorts, “Now I 
catch your drift and aim.” Apollo, in the familiar word- 
fencing fashion of the stage, again seizes a suggested advan- 
tage, and, as if Death were really yielding, asks, “Is there, 
then, a way by which Alcestis might be spared to reach old 
age?”’ to which Death, ending the matter, makes the cate- 
gorical reply, “ There is no way.” The question in Apollo’s 
mind in v. 48 is thus substantially identical with the question 
in v. 52, the former being “can she be saved through my 
persuasions?” the latter being “can she, then, be saved?” 
and in this parallelism of dv ef wefcaipi oe and émas “AXxn- 
oTls €s ynpas woo the “ grammatical affinities” of the idiom 
in question in the latter are again set forth for us by a Greek 
writer. 

(4) The case in behalf of the potential looks, then, very 
strong. The force which every grammarian assigns to that 
mode would render its just meaning to every example. 


Vol. xxiv.] Deliberatives in Greek. 195 


Moreover, every example is found to have a mate in an unde- 
niable potential sentence exactly or substantially parallel to 
it. And, finally, in two of the passages in which the exam- 
ples under examination occur, the context itself betrays the 
potential force. 

Manifestly, it would require a good deal of evidence to 
overthrow the strong probability thus established. Let us 
see whether the objection brought by Mr. Sidgwick will 
do it. 

This objection will best be stated in Mr. Sidgwick’s words: 
“If these are simple cases of an dy omitted in a relative 
potential clause, why are they all of one kind? Why do we 
never find them in positive clauses? On J. D.’s principle of 
explanation you ought to be able to say eiciv of XNéyorev or 
éotw dmws réEatus, usages of which there is in Attic no 
trace. In the examples the principal verb is always nega- 
tive, or, what comes to the same thing, interrogative. Why 
again do we find this omission of dy with the optative, not 
merely in subordination only to clauses actually or virtually 
negative, but only to one special form of negative phrase, ov« 
éotiv (or the logically identical ris éotiv;)? ... It is not 
therefore an accident (as it would be on J. D.’s theory) that 
they stop with ov« éori draws réEauu, ovK Eotiv boTIS Kel- 
patto, and do not further extend to éc@ das réEarpt, still 
less to méurrw boTIs dyyeiXar or any such usage.” (Classical 
Review for March, 1893.) 

“Tt seems to me that a careful observer of language will 
suspect that he has not got to the bottom of the usage by 
simply saying ‘dy is omitted,’ but that there is probably 
some other instinct at work which restricts the exceptional 
omission of dy to just this class (or these classes) of 
cases.”” (Ibid.) 

The objection as expressed in its first form is not difficult 
to meet. Putting aside Mr. Sidgwick’s statement that “the 
principal verb is always negative, or what comes to the same 
thing,’ which has been disproved above, one may first sug- 
gest that Mr. Sidgwick’s point is easily turned against himself. 
The subjunctive idiom does in time develop a declarative and 


et 


196 William Gardner Hale. [1893. 


positive main sentence, as in éyes dru Néyy. If now the opta- 
tive is only a “remoter” form, why does no corresponding 
declarative example of éyeu dtu Néyou, Extiv STL NEyou, OF Eat 
dmws Aéyou appear in Attic poetry? But a more decisive 
answer than the ad hominem one can be given. The reason 
why Attic does not, by the omission of ay, produce such 
constructions as elow of Aéyotev, Eotiv bras A€Eaum, and 
TéuTre bots ayyeirat, is that Attic has no such constructions 
with av, or, if it has them, they are altogether too rare to be 
a natural field for any variation. It is probably safe to say 
that no case like wéumw botis av ayyeidat Occurs anywhere 
in Attic, though-combinations very like it are occasionally to 
be found in Homer (as in dAN dye 8) Tia pavtiv épeiopev 
) lepia, | Os « eirot..., A 62). I think I may also say with 
safety, from the use of indices, that no cases like eiow of av 
Aéyouev Or Ectiv bras av A€Eauyw declarative exist in Aeschy- 
lus or Sophocles. The second does not exist even in Homer, 
in spite of the greater range of the Homeric grammatical 
apparatus; while the first, so far as my knowledge goes, 
occurs but once, vzz. in K 170, though the negative form 
occurs frequently, ¢.g. in E 192 and 483, & 299, X 348, 5 166 
and 559, andv 125. (X 348, it may be noted, is without ay», 
and has its exact correspondent, even to phraseology, in 
8 166.1) 

Nevertheless, Mr. Sidgwick’s objection, though it has gone 
too far, has, in the second form in which it is put, some 
weight ; for there ave some examples in Attic of the depen- 
dent optative with av, though of a different type from those 
which he suggests. 

In answer to the difficulty thus presented, my colleague, 
Professor Shorey, suggests that though, as he also firmly 
believes, the construction is potential, the omission: of dv 
may be due to the influence of the subjunctive idiom, espe- 
cially in the form which it takes in dependencé upon a past 
tense. This may be the right solution, but my feeling is 
against it. It seems to me that the two idioms would have 


1 The two examples are given in the footnote on p. 193. 


Vol. xxiv.] | Deliberatives in Greek. 197 


to be much nearer neighbors in actual force, —not merely 
in superficial form, —and in the whole range of both, — not, 
as is the case, in a part of their range only, —before the 
one could affect the other. I am more inclined to look for 
a solution in the essential character of the potential construc- 
tion, and to be willing to wait for this solution, even if it be 
not at once forthcoming in a satisfactory and final form. 

In this form it is, indeed, not forthcoming. Some consid- 
erations may, however, be suggested, which may guide some 
one else to better results. | 

As regards the independent construction, the omission of 
av is, in point of fact, not confined to the examples under 
investigation. The true potential occurs without dy in veoy- 
yos avOpmrav pabor (Aesch. Ag. 1163)!; it occurs, according 
to the Codd., in the perfectly unobjectionable icws yap 4 
Khpv& Tis } mpéc Bis poroe (Suppl. 727; Codd. rpeaSrponror) ; 
and it occurs in the parenthetical phrases @o7rep eizrot Tis, 
@s eltrot Tis, Odooov 7 Aێyou Tis, in Eur. Andr. 929 and Hipp. 
1186, and Ar. Av. 180 (see Goodwin, Moods and Tenses, § 242, 
and Kriiger, Gr. Sprachlehre IT, § 54, 3, A. 8).? 

Nor is the objection by any means so strong within the 
province of the dependent construction as might at first 


1 Wecklein holds that emendation is necessary in all these cases. Yet he 
grants the omission of dv in the dependent clauses under discussion. 

2 The optative of a fixed resolve, a derivative, as it seems to me, of the opta- 
tive of ideal certainty,—to which distinction I shall presently return, — occurs 
without dy, according to the usual reading, in a passage cited by J. D., vzz.: 


amd bé cot 
depoluav Bookay raparos Suemérov. — Aesch. Eum. 265. 


The next line—xal {Gvrd o” isxvdvac’ dmrdtouat xdtw— corroborates J. D.’s 
general argument that gepoluay cannot express a wish. A wish is not only in 
itself impossible here, but would be inconsistent with the positive statement in 
' o loxvdvac’ dedtouat,—which statement, on the other hand, is entirely in keep- 
ing with the idea of a fixed resolve. Wordsworth’s emendation to ¢époiw’ dy is, 
however, so slight and easy as to commend itself to my mind. 

In Cho. 854, where the Codd. have ¢péva x\éWecav, the emendation to gpév av 
kdéYecav is easy, and is hinted at, though not of course made necessary, by the 
metre. The passage beginning at Ag. 1374 is too difficult and uncertain of 
interpretation to cite in support of any doctrine. 


198 William Gardner Hale. [1893. 


hearing seem, as an analysis of the actually occurring exam- 
ples of other kinds will show. 

For these examples I am obliged, in order to save a delay 
which is forbidden me, to confine myself to Aeschylus and 
Sophocles, since there are no complete lexicons for Euripides 
and Aristophanes. Neither can I be sure that the collection 
is complete for the first two authors; for I find omissions, 
Ellendt’s Lexicon Sophocleum giving, e¢.g., only two examples 
of the potential with a;=87, while Dindorf’s gives four. 

The actually occurring examples of the dependent potential 
optative and of the dependent optative of ideal certainty, or 
the optative which is used in ordinary conclusions, softened 
assertions, and the like, are, if I have made no mistake, the 
following. I group them under these two heads : 


Dependent “ Potential” Proper. 


, . 25 a he , , 
(1) rev&er wap’ avdpav kal yuvatkelwv oToXor, 
donv Tap ddr ovtroT av cxéOots BpoTav. 


— Aesch. Eum. 856. 
(2) eloit @ Eevot, 
adrws Te Kal pépovtes of av ovTE TIS 
Sopov atracat ovT av jobein NaBov. 
— Soph. El. 1323. 
(3) ov yap exw Tas av 
orépEatme Kaxov T6de NeVeowv. — Soph. Trach. gol. 


4 lal 
(4) Aoyos pév Eat’ apyaios avOporav daveis, 
¢€ > x tol gee Mae / A \ ” 
@S ovK av aid éxudbos BpoTtav, mpiv av 
/ Ui A" 
Odvy Tis, oT’ ei ypnaTos ovT el TY KaKds* 


— Soph. Trach 1. 


Dependent Optative of Ideal Certainty. 


(5) orddyyvev Te NevdTNTA, Kal ypotav Tiva 
Exovt’ dv ein Saimoow pos ndovnv, 
XOATs AoBod Te mrocxirnv evpopdiar - 
— Aesch. P. V. 493. 





Vol. xxiv.] Deliberatives in Greek. 199 


(6) éuol & Eri pév Oparos ayamos 
6 Bios év tratpos Sopors, 
pnd Kpetaoover €“ov 
Gedv adduxtov dupa tpocdpaxoe pe. 
amroréuiatos bde y’ 6 TOAELOS, ATrOpa TrOpLLOS, 
ovd Ex Tis av yevoiwav~ Atos yap ox Opa 
pati bra pvyouw av.— Aesch. P. V. got. 


(7) dp’ ict’, aowdas Kal yoous mpd Tod Oaveiv 
e Hw A 4 > > / / 
ws ovd av eis TavcaT av, ei ypetn éyerv ; 
— Soph. Antig. 883. 
(8) Aris Néyers wev ApTiws ws, ef AdBois 
4 \ 4 lal > / ww 
abévos, TO ToUT@Y picos éxdelEeLas av - 


— Soph. El. 347. 


> a? a a 4 >, ? ‘ 
(9) GAy@ ’re Tols Tapotcw: WaT ay, ei cOEVOS 
AaBoun, Snrwocam’ av of” avdtois dpova. 
— Soph. El. 333. 
(10) Aéy’, exrel cé A€yos Sovpradwtov 
7 b] 4 4 »” 
atépéas avéyet Govpios Aias- 
@oT ovK av adpis UTrelmois. — Soph. Ai. 211. 


(II) GAN ed ye pévtot TOUT errictac’ ws eyo 
col ev vésouw av THode Kal peifm yapiv* 
— Soph. Ai. 1370. 
(12) dor’ ody pavtelias y dv ote THE eyo 
Brera av obven’ ovte THD Av toTepor. 
— Soph. O. T. 857. 
(13) GAN’ Drew ene Tov ixérnv SeEatato - 
@s ovy Edpas yas THAD av E€EAOowp’ Er. 
— Soph. O. C. 44. 


(14) dote Evov y av oddév GVO, BaTrEp av Vir, 
c / \ > 4 
bmektparroiuny un ov cuveca@ ley 


— Soph. O. C. 565. 


Two examples remain which cannot be assigned with cer- 
tainty to the one class rather than to the other, though the 
first would appear more probably to belong to the second: 


200 William Gardner Hale. [1893. 


cal nal , , 

(15) ov« €o@ orotov otavt av avOpwrrov Biov 
aed Oe > x ” / eT 
OUT aivécaip av ovTE pmeprpaiuny Tore. 


— Soph. Antig. 1156. 


(16) @or’ ov« dv évdixws y atiysafolTo cot: 


— Soph. Ai. 1342. 


To some of my readers, this division of the non-wishing 
optative into two classes will have no weight, and the argu- 
ment founded upon it will have no justification. I know, 
indeed, of no grammar that sufficiently insists upon the dis- 
tinction. But the distinction, nevertheless, has long seemed 
to me not only a real, but an important one. There is a vital 
difference between “I can” and “in a certain event I surely 
should,” between “he may perhaps” and “he surely would.” 
If we should not tolerate a translation in which “might” was 
used where the idea was “would,” or vice versa, no more 
should we be indifferent to the same difference in the expo- 
sition of Greek or Latin syntax.! 

Now the examples of the idiom which we have to explain 
are all translated by Mr. Sidgwick by the formula “could ”’ 
or ‘“‘could possibly,” and cannot be translated by “would.” 
We should, therefore, in weighing the force of his last argu- 
ment, set aside all examples of the optative of ideal certainty, 
— of which the meaning is clearly different. 

If this be done, there remain five sure and six possible 
examples of the true potential, three (1, 2, and 15) in a rela- 
tive clause, one (3) in an interrogative, one (16) in a consecu- 
tive with #ore, and one (4) in the Oratio Obliqua after as. 

Of the three in the relative clause, one, namely (16), is 
cast in the same form with the examples of the problematical 
idiom. Similar sentences could also, of course, be framed 
with the other two relative forms ofos and déaos, as ovK éoTt 
TUYN bony av [oiav av] aivécao mpiv av Bary Tis. 


1 There is, of course, a neutral border country between the two, and examples 
are easy to find which you may translate by either “could ” or “ would,” as in 
“quis hoc credat?” But this does not lessen the difference where the difference 
exists. 


Vol. xxiv.] Deliberatives in Greek. 201 


We may, then, describe the general state of affairs thus: In 
potential relative clauses, dv may be either used or omitted. 
In interrogative and consecutive clauses and clauses quoted 
after @s, we have no evidence that it is ever omitted; but, 
on the other hand, we have but a single example of each of 
these to go upon. The number of actual cases that support 
Mr. Sidgwick’s objection is thus so small—so far as we 
can judge from two out of the four authors concerned — as 
to carry little weight. 

Before leaving the subject, however, I wish to suggest a 
possible line of inquiry with regard to the really controlling 
reason for the omission of dv. Our knowledge of the history 
and force of this particle is exceedingly small; but most 
writers agree in describing it as having meant originally 
something like “in a given case” (éevra), and having come 
finally to be a sign of “contingency,” of dependence upon 
some assumption, —not of course necessarily expressed, but 
underlying the thought. If this is so, we should not expect 
to see the particle omitted with any freedom in eases of the 
optative of ideal certainty, since that always has the feeling 
of contingency underlying it. In two of the examples just 
given, vzz. in (8) and (9), an actual condition is expressed ; in 
(6), the questions in the last two verses do not turn upon 
things as they are, but upon an imaginary case previously 
clearly implied ; in (5), éyovra is in effect a condition; in (13), 
the clause of wish in the first verse quoted forms a condition ; 
while in the rest also there is a sense of some assumption 
underlying each statement. The particle of “contingency,” 
the particle suggesting some idea‘like “in that case,” or “in 
an imaginable case,” is therefore much in place. 

It is also much in place in strictly potential sentences where 
the “possibility” is not absolutely existent, but hangs upon 
something, as in « 268: ddda Edy Ttoicdeot Odcaov | devyo- 
fev’ Ett yap Kev [=evyovtes] advEapev Kaxov jap. In the 
examples on which Mr. Sidgwick founded his theory, on the 
other hand, there is no sense of contingency whatever, but 
only the bare idea of “possibility.” Now the use of dy has 
become fixed in ordinary use, by Attic times, in the poten- 


- 


202 William Gardner Fale. [1893. 


tial construction, whether with or without contingency, as 
(substantially) it had already done by Homeric times in the 
construction of ideal certainty.1 But the Attic writers for 
the stage liked to impart an archaic flavor to their style. 
They used vw for the pronoun of the third person. They 
used és for eis. In older Greek they found an occasional 
omission of av. They accordingly omitted it occasionally 
themselves, using this license with comparative freedom in the 
case of present and future conditions in the subjunctive and 
in the case of mpiv-clauses, etc., where the omission could not 
possibly lead to confusion, and, much more sparingly and 
cautiously still, in the case of the potential optative, where 
they never passed outside the limits of a bare unconditioned 
possibility, as seen in phrases like @s eizrou Tis, @oTreEp eEi7roL 
tis, Oaooov % Aéyou Tis, and in the examples of similar feeling 
(a point by no means to be overlooked) which have formed 
the subject of this paper. 

Such a view has also something to commend it in the 
phraseology and historical sequence of the eight dependent 
examples. The oldest of them—the three from Aeschylus 
—are in the negative (ovx éo@ Grrws, ovK Ect GoTis, ovK 
éotiv 6tT@). So is also the example ovdé... &o@ éao from 
Euripides. Two others take the corresponding form of 
inquiry, without the negative (é¢@ des ;). There then 
remain two cases, Thesm. 871 and O.C. 1172. The former, 
which is the latest of all, is put by Aristophanes into the 
mouth of Euripides, and might therefore be expected to be 
somewhat unusual in expression. In simpler form it would 
have been “is there some one here in authority, that might 
receive us ?—can some one here receive us?” (=éo@ éctis ;), 
and so is essentially of the same feeling as the éo@ das of 
Aristophanes or of Euripides himself. The example O. C. 
1172, though not late, is possibly farther removed from the 
type. Seven of the eight examples, however, fall within the 


1 Such an omission as the one in Pindar’s xecvds env (Ol. 3, 45) is very rare 
even in Homer. 

2 This much discussed example (see Madvig, Advers. Crit. I., pp. 189 seq.) is 
the only one of the whole eight the meaning of which is not perfectly clear. It 


Vol. xxiv.] Deliberatives in Greek. 203 


type ov« Eotwv, or éortwv interrogative, plus relative pronoun. 
Now, if we remember the extremely frequent recurrence of 
the idea of the denial of possibility in the combination ov« 
éorev plus relative (seen not only, in Homeric and Attic Greek, 
with the potential plus dy, but also, in Attic, with the subjunc- 
tive, as appears in Part I., and even, indeed, with the indica- 
tive itself, in one of the commonest uses of the combination), it 
would seem that, if dy were anywhere to be omitted, here was 
a place where the mode and the introductory phrase together 
might safely be trusted to convey the potential meaning, — 
just as, in the independent interrogative, it was safe to trust 
to the association of the potential idea with the optative ques- 
tion obviously implying a negative answer. This step having 
been taken with the dependent examples in the denial of 
possibility, as in our three earliest examples, it would then 
be a natural further advance to use the same formula in the 


surely cannot mean “ who is there whom I could possibly censure?” as it would 
have to do if Mr. Sidgwick is right in saying (see p. 195 of this paper) ‘in subor- 
dination . . . only to one special form of negative phrase, od« or, or the logi- 
cally identical ris éorwv;”. The main clause, following ¢0.8’ dxotwy TSvd" bs 20% 
6 mpoordrns, must mean, not “who zs there?” but “who is the man?” The 
force wanted for the sentence as a whole in the context is “ who in the world can 
the man be whom — following your commands —I am to reject?” Donner gives 
it this force (“ und wer denn ist es, den ich so verwerfen soll?”), and so, appar- 
ently, does Jebb in his translation (“and who can he be, against whom I should 
have a grief?”’). To get this exact force, however, we ought to have a subjunc- 
tive clause, similar to the riv’ Zxwv phunv dyabhy nas, €¢ bry xvoGpev dyuids 
of Ar. Eq. 1321. The true optative does not lend itself easily to subordination, 
and so could not express the wish of another than the speaker. The optative 
with &v sometimes has the meaning of propriety, but hardly in such a combina- 
tion. We thus seem driven to the theory of a mixture of two thoughts in the 
main sentence, the relative clause being attached to the one which is not formally 
expressed, so that the sentence means “ who in the world can this man be, —at 
least that could incur my censure?” or, in expanded form, “ who in the world can 
this man be?” and “is there some man that might incur my censure [so that I 
should yield to yourdemand]?” This would seem to be substantially the force of 
Jebb’s rendering in his commentary: “ who is he, to whom I could possibly have 
any objection?” But if the interpretation is right, then the relative clause with 
its suppressed antecedent conception, involving as it does an éorw dv y’ éym 
éEaul te; is parallel to the €¢@ Srws uddor of Euripides and the ris @xee xpdros 
boris 5éEacro of Aristophanes, and so is after all essentially of the same type with 
the other seven examples. 


204 William Gardner Hale. [1893. 


guestion of possibility, especially if, as may well have been 
the case, there had previously been an intermediate use in 
a question implying denial. 

My conception of the present state of the question, to sum 
up, is as follows : 

Mr. Sidgwick’s theory is in any case disproved ; for, even 
if the independent use claimed were not so infrequent as to 
make an extension into the dependent form improbable, and 
even if the fact that it is based upon an unknown cause were 
to be overlooked, yet (1) the fundamental meaning which, in 
keeping with that assumed cause, and as against a potential 
origin, Mr. Sidgwick assigns to his examples, is precisely the 
meaning which, in many of the examples, the potential would 
yield, and (2) this same fundamental meaning is wholly absent 
from others of his examples, and (3) the subjunctive idiom 
and the one in question are essentially unlike, in that the 
former, in all but one of the twenty-five examples thus far 
adduced, expresses an impulse or demand, which meaning is 
absent from several of Mr. Sidgwick’s examples. Whatever 
may be thought of any other solution, then, this theory must, 
it seems to me, be abandoned. 

Only one other theory deserving serious consideration has 
thus far been advocated, the theory that the verbs in question 
are potentials in the strict sense. This theory meets every 
condition except one. It accounts for the force of impossi- 
bility found in a number of the examples. It accounts for 
the force of possibility overlooked by Mr. Sidgwick in the 
others. It receives great strength from the fact that every 
one of the examples can be shown to have an exact corre- 
spondent in an unquestioned potential. It receives great 
strength from the fact that, in the case of two of the exam- 
ples, the Greek poet has, for our better guidance, expressed 
an exactly parallel idea, in the immediate context, by an 
unquestionable potential. It is possibly defective at one 
point, in case, namely, of the failure of my explanation that 
the omission of av in these cases of unconditioned poten- 
tiality with (ov«) éc@ bmws and é7o and the equivalent ovd« 
éotiv Gots andiallied phrases, was, if not first suggested, at 


Vol. xxiv.] Deliberatives in Greek. 205 


least made easier and safer through the association of the 
potential idea with the phrase ovd« éc@ és, in its frequent 
use with optative, subjunctive, and indicative, and that there 
is accordingly nothing remarkable in the fact that the omis- 
sion is not found in the widely removed constructions of 
result, quotation, and indirect question. Yet, even if this 
explanation be not accepted, the defect is not a grave one, 
since’ there is reason to believe that the total number of 
examples with which the comparison is to be made is small 
(possibly only two, certainly not more than three, for two out 
of the four scenic poets), and since our knowledge of the 
force of the particle on which the difficulty turns is con- 
fessedly imperfect. My own conclusion, then, is that there 
is an extremely strong probability that the ultimate solution 
of the problem will justify the view now commonly held, by 
making of the examples in dispute nothing but ordinary 
potentials. 











aa 


TH. 


APPEN DEX: 


PROCEEDINGS OF THE TWENTY-FIFTH ANNUAL SESSION, CHICAGO, 
Tix.,. 1893. 

TREASURER’S REPORT (Pp. V.). 

List OF OFFICERS AND MEMBERS (p. liv.). 

CONSTITUTION OF THE ASSOCIATION (p. lxvi.). 


PUBLICATIONS OF THE ASSOCIATION (Pp. lxviii.). 


: oa 
: i =? ve, 
“ht atl wy ert 


a 


ae 
sae 


oe 


= a - a 
= ae eee a 
. fo 4 ere 





MEMBERS IN ATTENDANCE AT THE TWENTY-FIFTH 
ANNUAL SESSION (CHICAGO). 


F. F. Abbott, University of Chicago, Chicago, Il. 
Louis F. Anderson, Whitman College, Walla Walla, Washington. 
H. L. Baker, Detroit, Mich. 
M. Bloomfield, Johns Hopkins University, Baltimore, Md. 
D. Bonbright, Northwestern University, Evanston, Ill. 
H. C. G. Brandt, Hamilton College, Clinton, N. Y. 
James W. Bright, Johns Hopkins University, Baltimore, Md. ' 
Demarchus C. Brown, Butler University, Irvington, Ind. 
Henry F. Burton, University of Rochester, Rochester, N. Y. 
A. Guyot Cameron, Yale University, New Haven, Conn. 
Edward B. Clapp, Yale University, New Haven, Conn. 
John Pitt Deane, Brooklyn, N. Y. 
M. L. D’Ooge, University of Michigan, Ann Arbor. 
Herbert C. Elmer, Cornell University, Ithaca, N. Y. 
Harold N. Fowler, Adelbert College, Cleveland, O. % 
James M. Garnett, University of Virginia, Charlottesville. 
Basil L. Gildersleeve, Johns Hopkins University, Baltimore, Md. 
Thomas D. Goodell, Yale University, New Haven, Conn. 
Alfred Gudeman, University of Pennsylvania, Philadelphia. 
William Gardner Hale, University of Chicago, Chicago, Ill. 
G. L. Hendrickson, University of Wisconsin, Madison. 
David H. Holmes, Johns Hopkins University, Baltimore, Md. 
Milton W. Humphreys, University of Virginia, Charlottesville. 
George E. Jackson, Washington University, St. Louis, Mo. 
Gustaf Karsten, University of Indiana, Bloomington. 
Martin Kellogg, University of California, Berkeley. 
David A. Kennedy, Orange, N. J. - 
J. H. Kirkland, Vanderbilt University, Nashville, Tenn. 
Abby Leach, Vassar College, Poughkeepsie, N. Y. 
J. H. T. Main, Iowa College, Grinnell. ~ 
F. A. March, Lafayette College, Easton, Pa. 
C. W. E. Miller, Johns Hopkins University, Baltimore, Md. 
Charles A. Mitchell, University School, Cleveland, O. 
W. B. Owen, Lafayette College, Easton, Pa. 
W. H. Parks, City of Creede, Colorado. 
Ernest M. Pease, Leland Stanford Jr. University, Palo Alto, Cal. 
Samuel B. Platner, Adelbert College, Cleveland, O. 
Horatio M. Reynolds, Yale University, New Haven, Conn. 
Louisa H. Richardson, Carleton College, Northfield, Minn. 

i [ovER] 


- 
- 


American Philological Assoctation. 


W. S. Scarborough, Wilberforce University, Wilberforce, O. 
Charles P. G. Scott, Yonkers, N. Y. 

Paul Shorey, University of Chicago, Chicago, IIl. 

Josiah R. Smith, Ohio State University, Columbus. 

Herbert Weir Smyth, Bryn Mawr College, Bryn Mawr, Pa, 
W. O. Sproull, University of Cincinnati, Cincinnati, O. 
Lewis Stuart, Lake Forest University, Lake Forest, Ill. 

G. V. Thompson, Yale University, New Haven, Conn. 
Frank L. Van Cleef, University of Wisconsin, Madison. 
John H. Walden, Harvard University, Cambridge, Mass. 
Benjamin I. Wheeler, Cornell University, Ithaca, N. Y. 
Frank E. Woodruff, Bowdoin College, Brunswick, Me. 

A. C. Zenos, McCormick Theological Seminary, Chicago, Ill. 


[Total, 52.] 


AMERICAN PHILOLOGICAL ASSOCIATION. 





Cuicaco, ILt., Tuesday, July 11, 1893. 


Tue Twenty-fifth Annual Session’ was called to order at 3.30 P.M., 
in the Art Institute, by the President, Professor William Gardner 
Hale, of the University of Chicago. 

The Secretary of the AssociaTIoNn, Professor Herbert Weir Smyth, 
of Bryn Mawr College, presented the following report : — 


1. The Executive Committee had elected as members of the ASSOCIA- 
TION : — 


William F. Abbot, High School, Worcester, Mass. 

F. G. Allinson, Professor of Greek, Williams College, Williamstown, Mass, 

Edward P. Baillol, Professor of Romance Languages, University of Indiana, Bloom- 
ington, Ind. 

Miss Mabel Banta, Bloomington, Ind. 

George K. Bartholomew, English and Classical School, Cincinnati, O. 

William J. Battle, University of Texas, Austin, Texas. 

C. C. Bates, Professor of Latin, Buchtel College, Akron, O. 

E. C. Benson, Professor of Latin, Kenyon College, Gambier, O. 

J. R. Bishop, Hughes High School, Cincinnati, O. 

George F. Bristol, Assistant Professor of Greek, Cornell University, Ithaca, Noakes 

Robert Baird, Northwestern University, Evanston, Ill. 

Samuel Brooks, Professor of Latin, Kalamazoo College, Kalamazoo, Mich. 

Demarchus C. Brown, Butler University, Irvington, Ind. 

F. W. Brown, Professor of Latin, Franklin College, Franklin, Ind. 

A. H. Buck, Professor of Greek, Boston University, Boston, Mass. 

W. I. Burnap, Instructor in Greek, Lake Forest Academy, Lake Forest, Ill. 

R. W. Crowell, High School, Columbus, O. 

H. A. Dearborn, Professor of Latin, Tufts Coltege, College Hill, Mass. 

Joseph H. Drake, University of Michigan, Ann Arbor, Mich. 

William S. Ebersole, Professor of Greek, Cornell College, Mount Vernon, Ia. 

Miss Kate M. Edwards, Cornell University, Ithaca, N. Y. 

F. H. Ellis, Wesleyan Academy, Wilbraham, Mass. 

Miss E. Antoinette Ely, University of Chicago, Chicago, Ill. 

Vernon J. Emery, University of Chicago, Chicago, IIl. 

Frank H. Fowler, University of Chicago, Chicago, Ill. 

1 The regular summer meeting of the AssociaTION was held in Chicago in conjunction with the 
Congress of Philologists, which convened in that city during the week July 11-15, at the invi- 
tation of the World’s Congress Auxiliary of the World’s Columbian Exposition. Two of the 
regular sessions of the AssocIATION were merged into General Sessions, where papers were 
presented by foreign and other scholars in attendance at the Congress. Some of these papers are 
incorporated in the present volume of Transactions. 

iii 


+ 


iv American Philological Association. 


J. B. Garritt, Professor of Greek, Hanover College, Hanover, Ind. 

W. N. Guthrie, Professor of French, Kenyon College, Gambier, O. 

E. L. Hale, Professor of Latin, Hiram College, Hiram, O. 

L. B. Hall, Professor of Latin, Oberlin College, Oberlin, O. 

Charles Harris, Professor of German, Adelbert College, Cleveland, O. 

William A. Heidel, Professor of Greek, Illinois Wesleyan University, Blooming- 
ton, Ill. ‘ 

K. F. R. HochdGrfer, Professor of Modern Languages, Wittenberg College, 
Springfield, O. 

H. A. Hoffman, Professor of Greek, Indiana University, Bloomington, Ind. 

Hon. Samuel E. Hunt, Cincinnati, O. 

Mrs. Julia J. Irvine, Professor of Greek, Wellesley College, Wellesley, Mass. 

Gustaf Karsten, Professor of Germanic Languages, University of Indiana, Bloom- 
ington, Ind. | 

A. P. Keil, Professor of Latin, Hanover College, Hanover, Ind. 

R. A. King, Professor of German and French, Wabash College, Crawfordsville, Ind. 

H. M. Kingery, Professor of Latin, Wabash College, Crawfordsville, Ind. 

E. G. Kinkead, Assistant in Latin, University of Cincinnati, Cincinnati, O. 

A. F. Kuersteiner, Hughes High School, Cincinnati, O. 

Henry B. Longden, Professor of Latin, De Pauw University, Greencastle, Ind. 

C. M. Lowe, Professor of Latin, Heidelberg University, Tiffen, O. 

E. W. Manning, Professor of Modern Languages, De Pauw University, Green- 
castle, Ind. 

W. J. McMurtry, Yankton College, Yankton, South Dakota. 

F. J. Miller, Professor of Latin, University of Chicago, Chicago, Ill. 

Charles A. Mitchell, University School, Cleveland, O. 

W. O. Mussey, Assistant in English, University of Cincinnati, Cincinnati, O. 

Carl Osthaus, Assistant Professor of German, University of Indiana, Bloomington, 
Ind. 

W. H. Pabodre, Woodward High School, Cincinnati, O. 

T. H. Paden, Professor of Latin and Greek, New Concord, O. 

W. F. Palmer, Ph.D., Instructor in Latin, Lake Forest Academy, Lake Forest, Ill. 

Thomas M. Parrott, Ph.D., Dayton, O. 

William Morton Payne, Esq., Chicago, Ill. | 

Miss S. Frances Pellett, University of Chicago, Chicago, IIl. 

John Pickard, Assistant Professor of Latin, University of Missouri, Columbia, Mo. 

A. C. Pierson, Professor of English, Hiram College, Hiram, O. 

Julius Howard Pratt, Jr., Ph.D., Milwaukee Academy, Milwaukee, Wis. 

Benjamin F. Prince, Professor of Greek and History, Wittenberg College, Spring- 
field, O. . 

George Scott, Professor of Latin, Otterbein University, Westville, O. 

Miss Helen W. Searles, Instructor in Greek and German, Ferry Hall Seminary, 
Lake Forest, Ill. 

T. H. Sonnedecker, Professor of Greek, Heidelberg University, Tiffen, O. 

Hiram A. Sober, University of Wisconsin, Madison, Wis. 

Robert B. Steele, Professor of Latin, Illinois Wesleyan University, Bloomington, IIl. 

W. F. Swahlen, Professor of Greek, De Pauw University, Greencastle, Ind. 

A. T. Swift, Master in Modern Languages, Lakeville, Conn. 


Proceedings for Fuly, 1893. v 


F. W. Tilton, Rogers High School, Newport, R. I. 
A. H. Tolman, Professor of English, Beloit College, Beloit, Wis. 
W. H. Wait, Ph.D., Peoria, Ill. 
F. Whitlock, Professor of Latin, Ohio Wesleyan College, Delaware, O. 
W. G. Williams, Professor of Latin, Ohio Wesleyan College, Delaware, O. 
Charles A. Wilson, Assistant in French, Wittenberg College, Springfield, O. 
Theodore L. Wright, Professor of Greek, Beloit College, Beloit, Wis. 
G. H. Young, Professor of Latin, Wittenberg College, Springfield, O. 
A. S. Zerb, Professor of Hebrew, Heidelberg University, Tiffen, O. 
[Total, 85.] 


2. The TRANSACTIONS and PROCEEDINGS for 1892 (Vol. XXIII.) were 
issued together in May of the present year. Separate copies of the Pro- 
CEEDINGS may be obtained of the Secretary or of the Publishers. 


In presenting his report as Treasurer, Professor Smyth alluded, to 
the fact that, despite the heavy outlay for Vol. XXIII., one of the 
largest volumes yet published, the finances of the AssociaTION were 
in a satisfactory condition. The following is the report for the fiscal 
year ending July 8, 1893 : — 


RECEIPTS. 
Balance fromt: 1891-92 .1-.,.0- = cep 44-78 Sy 2 a, Dea oe le 
Heegand Arrears: as 34; Bl 6 as = 9 oP 3 SS ao 
Sales of Transactions . . . at We ee 225.17 
Reprints and Authors’ Coscectiaa. dic: <b ieee 71.20 
Dividends Central N. E.& Western R.R. . . .- . . * 6.007 
RRL CERSE SG 8 or ey wast, PA Blase, apes 33-04 
Total weesigan ioe the yearsa*9 5 SS 1473-41 
$3125.63 
EXPENDITURES. 
Transactions and Proceedings Seo XXIII. a -'. . “Surges 
Salary of Secretary. . ... She Asha HES. 250.00 
RM par ee a eG 8 Fee ae athe YM 47.86 
Sepeeeeee s  aN oe ge BS o te eS es 3.00 
A aly OS ae a 17.25 
Stakiomatger s bteas 2. ae oS eg ee ne 4.50 
Briding Sr cee es ee le 8 ee ame 1.20 
POCIOEIN St Wines ro os bes is. 5) e heute ae 2.84 
Total expenditates foe: theFear «6s ye we taue $1919.39 
Balance, July G-380s > <2 -3. sa ko Cee Gates 1206.24 
$3125.63 


The Chair appointed as Auditors of the Treasurer’s Account, 
Professors Kellogg and Hendrickson. 


- . 
- 


vi American Philological Association. 


As a Committee on Place of Meeting for next year were appointed 
Professors Sproull, Goodell, and Sterrett; on Officers for 1893-94, 
Professors D’Ooge, Fowler, and Abbott. 

The reading of papers was then begun. At this time there were 
about sixty persons present. At subsequent meetings the number 
fluctuated greatly, at times being as large as one hundred. 


1. The Latin Prohibitions,’ by Professor H. C. Elmer, of Cornell 
University, Ithaca, N. Y. 


Part I. — This paper owes its origin to a feeling the writer has long had. that 
certain uses of the Latin perfect subjunctive are inadequately, and, in some par- 
ticulars, inaccurately, treated in Latin grammars. It is customary, for instance, 
in dealing with ze and the second person of the subjunctive in prohibitions, to 
dismiss the subject with the statement that, when the prohibition is addressed 
to no definite person, the present tense is used; otherwise, the perfect. All 
attempts, like Gildersleeve’s,? for instance, to make any further distinction be- 
tween the tenses have been frowned down. Scholars in general have been 
inclined to accept the views of Madvig (Opzusc. acad. altera, p. 105)? and of 
Weissenborn (on Livy 21, 44, 6) as final, viz. that the perfect is used, when a 
definite person is addressed, only because the present cannot be used. The 
reason for this remarkable state of things they do not trouble themselves to seek. 
Even Schmalz in the second edition of his Lat, Synt. § 31, would have it under- 
stood that the perfect tense in this use has no special significance. Such ignoring 
of all distinction between tenses is common also in other constructions, e.g. in the 
so-called potential subjunctive. One of the latest grammars (Allen and Greenough, 
§ 311) says that in aliguis dicat and aliguis dixerit, the two tenses refer without 
distinction to the immediate future. The same grammar, in dealing with modest 
assertion, draws no distinction between pufaverim and putem. It is customary, 
again, to dismiss the perfect subjunctive in prayers with the mere statement that 
it is a reminiscence of archaic formulae, without a hint that the perfect necessarily 
means anything. It has seemed to me that this looseness of interpretation is 
entirely unjustified by the facts of the language, and I have accordingly under- 
taken an investigation of the whole range of those independent constructions of 
the perfect subjunctive in which that tense deals with future time. I have included 
also in my investigation such uses of the future perfect indicative as are frequently 
said to be “ equivalent to the simple future.” 

For the purposes af this paper I have collected and classified all the instances 
of the uses concerned that are to be found in the remains of the Latin language 


1 The paper, of which this is an abstract, will be found in complete form in No. 58 (Vol. XV.) 
of the American Journal of Philology. 

? Latin Grammar, § 266, Rem. 2, which is, as far as it goes, in perfect harmony with the 
results reached in this paper. 

% Madvig is inexcusably careless in some of his statements in this connection. On page ros, 
e.g., he says that we with the present is apud ipsos comicos rarissimum et paene inusitatum. 
As a matter of fact, it is extremely common afud comticos — far more so than any other form of 
prohibition. 


Proceedings for July, 1893. vii 


up to the end of the Augustan period (except the late inscriptions), together 
with important parts of Silver Latin. I ought, perhaps, to say that for four 
volumes of the Teubner text I accepted a collection of instances made by one 
of my students. He is, however, one in whose care and accuracy I have great 
confidence, and I feel sure that his collection is substantially complete. That 
part of my investigation, the results of which I have chosen for the present 
paper, deals chiefly with the second person, present and perfect tenses, of the 
subjunctive in prohibitions. For the purpose of simplifying the discussion, I 
shall, for the present, exclude the few cases (commonly called prohibitions, and 
classified under ze with the subjunctive) introduced by nec, numguam, nihil (e.g. 
nec dixeris, nec putaveris). There are so many serious objections to explaining 
any one of those introduced by mec (megue) in the best prose writers, and some 
of those introduced by xzhil, numquam, as instances of the same construction as 
that found in we feceris, that I shall here merely refer to my full discussion of 
the subject in the American Journal of Philology. Furthermore, the limits 
of this abstract are such that I must omit references and citations, and can 
give hardly more than a few bare results of my investigation. All the state- 
ments, however, that are here made will be found fully substantiated in my 
complete paper. ' 

The impression is very generally given that ze with the perfect subjunctive 
is one of the most common methods of expressing prohibition in the best classical 
prose. Asa matter of fact, it is almost entirely unknown to such prose. It will 
- be understood, of course, that the Letters of Cicero do not represent the usage 
of what is understood by “classical prose.” Tyrrell has clearly shown that the 
diction and constructions in the Letters are the diction and constructions of the 
early comic drama, and not at all those of what is commonly meant by Ciceronian 
Latin. Indeed, Cicero himself (ad. fam. IX. 21, 1) calls especial attention to 
the wide difference in this respect between the Letters and his other productions. 
We must not consider them in determining the usage of the best classical prose any 
more than we should the usage of early comedy —they both reflect the language 
of familiar everyday life. Throwing the Letters aside, we may say that ze with 
the second person perfect subjunctive does not occur in any production, whether 
prose or poetry, of the whole Ciceronian period, except in seven dialogue passages 
of Cicero where the tone distinctly sinks to that of ordinary conversation, or 
unceremonious ordering.! Were it not for four instances in Horace we might 
make the same statement for the entire period between Terence and Livy. It is 
not to the point to say that a prohibition is in its very nature familiar, nor would 
such a statement be true. The orations and the philosophical and rhetorical 
productions of Cicero, as well as the productions of other writers belonging to the 
same period, abound with prohibitions. The orations of Cicero alone contain 
81 prohibitions (or probably twice this number if we count such expressions 
as guacso ne facias, obsecro ne, etc.); and still in his orations no instance can be 
found of ze with the perfect subjunctive except pro Murena 31, where Cicero is 
quoting the supposed words of a teacher to his pupil. 

Again, the grammar-rule which says that the present tense is used when the 


1 There is‘no manuscript authority whatever for ze s7rzs (Catullus 66,91). The manuscript 
reading mon siris is the true one. This will be fully discussed in the American Journal of 
Philology. . 


\ 


viii American Philological Association. 


prohibition is general, i.e. addressed to no one in particular, while the perfect 
is used when it is addressed to some particular person, or persons, is entirely 
misleading and blinding in the form in which it is given. The truth which 
the rule contains is rendered useless by the absence of any hint as to the 
principle involved. Furthermore, exceptions to the rule are not uncommon, 
despite the sweeping assertions of the grammars. Sometimes general prohibitions 
take the perfect tense, e.g. Cato de agr. cult. 4;1 37,13 45,2; 933; 113,23 158, 2; 
161, 2; X// Tabulae, quoted in Serv. in Verg. Ecl. 8, 99; Cic. pro Murena 
31, 65 (quoting general precepts of the vestri praeceptores, who had just been 
mentioned. Notice the singular verb side by side with ves¢ri (instead of ¢u2), 
which seems to show that the prohibition is general); Hor. Sat. 2, 2,16. On 
the other hand, it is probable that prohibitions addressed to definite persons 
occasionally take the present tense at all periods of the literature, and that it 
is not, even in classical times, confined to poetry, as is commonly supposed. At 
any rate there are passages in prose which it requires ingenuity and violence to 
explain in any other way, and which, if found in Plautus or Terence, no one 
would have thought of explaining in any other way. This use is very common in 
early comedy, and I have collected the following instances from Cicero and later 
prose: Cie. zz Verr. II. 4, 23, 52 ne putetis; tb. de republica 6, 12, 12 (where 
the imperative “‘ audz¢e,” instead of a subordinate subjunctive, makes it probable 
that ze excitetis is also independent); 26. ad fam. 1, 9, 23 ne pertimescas,; ib. 
16, 9, 4 (where cautus sis and the form taken by the rest of the sentence show . 
that ze naviges also is probably independent); 74. ad Att. 9, 18, 3 ne agas (a 
proverb applied here to a particular person); 24. ad Quintum fratrem 1, 4, 1 
amabo te, ne .. . adsignes (Cicero never uses @mare in this sense with a 
dependent clause, though its parenthetical use is common in his Letters with 
independent imperative constructions, e.g. ad A¢t. 2,2,1; 76.16,16c; 75. 10, 10, 33 
ad Quint. 2, 8 (10)”); Phil. II. 5, 10 ne putetis (most naturally taken as inde- 
pendent); Livy 44, 22 me alatis (this, or some reading which involves the same 
construction, seems inevitably correct, and would undoubtedly be accepted by 
everybody were it not for the supposed rule); 2d. 22, 39, 2 megue desis neque 
des (Livy and later writers freely use segue for neve); Tac. Dialogus 17 ne 
dividatis. It was formerly customary among editors of the Dialogus to take this 
as a prohibition. Recent editors use only a comma, or a semicolon, before ze 
dividatis, understand an ellipsis, i.e. Haec dico ne, and then apologize for the 
awkwardness of the sentence they have made Tacitus use. Why make this so 
difficult? Why not let it be what it seems to be on the face of it, namely a 
prohibition ? Here are ten probable instances in prose of the present sub- 


1 The attempt of Nitzsch to show that this production of Cato was intended for the manage- 
ment of a particular estate is, on every hand, acknowledged to have been a failure. The evidence 
against such a view is overwhelming. 

2 Even in Plautus and Terence amaéo in this sense is almost invariably thrown in paren- 
thetically. 

3 When this paper was read before the American Philological Association Professor Gudeman 
objected to my distinction between the two tenses of the subjunctive in prohibitions, saying, if I 
understood him correctly, that he had in mind a certain very impassioned prohibition in Tacitus 
in which the present tense was used. As I had not then examined Tacitus with reference to this 
construction, I could not answer his objection. An examination of this author, however, will 
show that Professor Gudeman’s memory was certainly playing him false. The passage above 


Proceedings for July, 1893. ix 


junctive with e addressed to a definite person. The reason why it is not more 
common will appear later in this discussion. But even if none of these examples 
existed (and there have been ingenious attempts to explain away most of them 
in deference to the supposed rule), there would still be no sufficient ground for the 
rule laid down by the grammars. In the whole field of classical prose, from 
the beginning of the Ciceronian period to the end of the Augustan period, and 
even later, there is but a single example of ze with the second person of the 
present subjunctive in an indefinite prohibition. There are a few examples from 
poetry, but these have no bearing upon the point in question, as it is everywhere 
acknowledged that ze with the present is common in poetry even in addressing a 
definite person. The single example just referred to is, of course, the one cited 
under this rule, with suspicious uniformity, by all Latin grammars, viz. Cic. de 
senectute 10, 33, though even here it might be noticed that Cato is speaking to 
definite persons, addressing at one time Scipio individually, again Laelius, and 
still again both together. The truth is that a general prohibition in Latin is 
nearly always expressed by the use of the third person, e.g. nemo putet, etc.,,or 
some circumlocution introduced by cavendum est ne, oportet, or the like. It 
will, I think, be admitted that the above considerations at least cast serious 
doubt upon the validity of the grammar-rules regarding the use of #e in pro- 
hibitions. The question as to the true distinction between the tenses in such 
constructions seems to me to be still an open one, and this paper is intended as a 
contribution to its solution. 

Let us start with certain general principles. All will agree that the perfect 
subjunctive, when dealing with a future act, differs, at least in some uses, from 
the present in representing the act as one finished in the future. For instance, in 
the expression sz venerit, videat the act of coming is conceived of as a finished act 
in the future, about to be completed prior to the beginning of the act of seeing. 
In st veniat, on the other hand, the act is conceived of as in progress in the 
future. Such a distinction between the tenses of me feceris and ne facias would 
not be entirely satisfactory at all points of the parallel. Me feceris cannot mean 
literally “ Do not, prior to a certain point in the future, have done it.” In one 
respect, however, the distinction still holds. In me feceris there is at least no 
thought of the progress of the act. The expression deals with an act in 
its entirety. The beginning, the progress, and the end of the act are brought 
together and focussed in a single conception. The idea of the act is not dwelt 
upon, but merely touched for an instant, and then dismissed. The speaker, as it 
were, makes short work of the thought. There is a certain force about the tense. 
When a man says me facias he is taking a comparatively calm, dispassionate view 
of an act conceived of as one that will possibly de taking place in the future; 
ne feceris, on the other hand, implies that the speaker cannot abide the thought; 
he refers to it only for the purpose ot insisting that it be dismissed absolutely as 
one not to be harbored. As far as the comparative vigor of the two expressions 
is concerned, the difference in feeling between them is similar to that between 
“Go!” and “Be gone!” “Go” dwells upon the progress of the act. A man 
never says “ Be gone!” except when aroused by strong emotion, which does not 
allow him to think of the progress of the act, but only the prompt accomplishment 


cited contains the only instance in Tacitus in which the present tense of the subjunctive is used 
in prohibition, and this, instead of making against my view, is a striking confirmation of it. 


- 


x American Philological Association. 


of it. In a similar way me feceris betrays stronger feeling than me facias — it 
disposes of the thought with the least possible ado. This feature of the tense, if 
my characterization is correct, would lead us to expect it to be used only, or 
chiefly, in animated, emotional, or unusually earnest discourse, and to such 
passages, as we shall presently see, is it almost exclusively confined. I wish to 
insist upon this as the only real distinction between the two tenses with me. We 
shall now, of course, expect that in the majority of cases where a prohibition is a 
general, indefinite one, the present tense will be found. When a man is soberly 
philosophizing and writing precepts for the world at large, he is not often aroused 
by emotions so strong as he is when, actually face to face with a person, and 
perhaps under the influence of anger, alarm, or some other intense feeling, he 
orders that person not to do a certain thing. But even in this sort of writing, 
when he feels that his precept is of prime importance, he may occasionally fall 
into the more vigorous form of expression. For the satisfactory study of such 
expressions, we look for some production abounding in general precepts and still 
not written in the form of dialogue and not addressed to any one in particular. 
Naturally we turn to Cato’s de agricultura. In the seven different passages 
of this work cited above, Cato uses ze with the perfect in a general prohibition. 
In each case the context makes it probable, or, in the light of facts which I shall 
present later, practically certain, that he considers of especial importance the 
particular thing prohibited. : 

By far the best place to study the difference in feeling between the two tenses 
is Plautus and Terence, because in them (and only in them) both tenses are 
very freely used with we in prohibitions. It is there, too, that the tone of the 
prohibition can best be determined, because the dramatic action makes clear 
the feeling of the speaker. I give, in my complete paper, classified lists of all the 
passages in Plautus and Terence containing prohibitions of this sort. There are 
in these two authors 31 instances of we with the perfect subjunctive. In 
nearly all of these the feeling of strong emotion of some sort, e.g. great alarm, 
fear of disaster if the prohibition is not complied with, or the like, is very promi- 
nent. Many of them are accompanied by other expressions which betray the 
speaker’s earnestness, e.g. fer deos atgue homines, opsecro, hercle, etc. And there 
is not one of them in the least inconsistent with my explanation of the meaning 
of the tense. 

The same feeling that prompts the use of the perfect tense with #e explains 
the use of the same tense in prohibitions introduced by cave. Plautus and Terence 
present 33 instances of cave with the perfect, though elsewhere in Latin only 
the present tense is found with cave. 

If now we turn to me and cave with the present subjunctive, we find a very 
different state of things. There are in Plautus and Terence more than 100 instances 
of me and 18 (19 ?) instances of cave, in this form of prohibition, all of which 
are given in my complete paper. 

There are certain remarkable differences between the prohibitions expressed 
by the present tense and those expressed by the perfect, which a casual observer 
might not notice. If my distinction between the two tenses is correct, we should 
expect that a prohibition dealing with mere mental action, e.g. “ Do not suppose,” 
“Do not be surprised,” “Do not be afraid,” would commonly take the present 
tense, because such prohibitions would not commonly be accompanied by strong 





Proceedings for July, 1893. xi 


emotion, and, as far as the interests of the speaker are concerned, it matters little 
whether the prohibition be complied with, or not. Such a condition of things 
is exactly what we find. Among the instances of ze with the perfect tense, not a 
single example of a verb of this class will be found; but among those of me with 
the present there are no less than 31 instances of such verbs, or nearly a 
third of the entire number. Again, such prohibitions as “ Do not ask me,” ‘ Do 
not remind me” (i.e. I know already), would not ordinarily imply any emotion, 
and no such verbs will be found among the instances of ze with the perfect. But 
there are 13 such verbs among the instances of the present. Substantially 
the same holds true for the cave-constructions. Among the 33 instances 
of cave with the perfect there is no instance of a verb belonging to any of 
these classes. There is no avoidance of such verbs with cave used with the 
present tense (in spite of the fact that there are only about half so many 
instances of the present as of the perfect); or with mo/i (though of is com- 
paratively rare in Plautus and Terence); or with #e followed by the imperative, 
a construction which, in Plautus and Terence, occurs 32 times with verbs 
of this sort out of a total of 84 instances. A similar state of things is found 
outside of Plautus and Terence. Scores of such verbs are found in other 
forms of prohibition. But nowhere in this whole period is such a verb to be 
found in the perfect tense in a prohibition. Why this mysterious absence of all 
such verbs from this one sort of prohibition ? Recurring to the instances of the 
present tense in Plautus and Terence, we notice that, in eleven of the passages, 
the prologue or some one else is calmly addressing the audience with “ Do not 
expect me to disclose the plot of the play,” or some prohibition equally calm. 
But there is not one instance in the prologues either of Plautus or Terence of the 
perfect tense in this use. And this again is exactly what we should expect. 
(It matters little, for our present purpose, whether Plautus wrote the prologues 
to his plays or not.) In general the fact may be emphasized that me with the 
present is chiefly confined to prohibitions of the most common-place sort. -Where 
this is not apparent from the nature of the verb itself, a study of the context will 
show that the speaker is not under the influence of any strong emotion. There 
are in all only 5 instances (a small number out of so many) which can fairly be 
said to be accompanied by emotion, and in all of these cases the verb is the 
same; so they should really count for only one instance. 

Whatever differences of opinion may be held regarding individual instances, I 
feel sure that no one who compares carefully the instances of the present and of 
the perfect tenses respectively can resist the general conclusion to which I have 
come. 

If now the distinction I have drawn between the two tenses holds so clearly 
for the only two authors who make frequent use of ze and cave with both tenses 
of the subjunctive in prohibitions, a strong presumption is established in favor of 
a similar distinction in the few instances to be found in later writers, where there 
‘are not always so many indications at hand, as in dramatic productions, to make 
clear the feeling of the writer. And a study of these instances confirms the pre- 
sumption. There are, in classical prose, from the time of Terence up to near 
the end of the Augustan period, only 7 instances of me with the perfect in 
prohibition, and these are all in Cicero. As pointed out above, each of these 
occurs in dialogue where the tone sinks to that of ordinary conversation, in which 


. 


xii American Philological Assoctation. 


some one is delivering himself of an earnest, energetic command. One is naturally — 
more unceremonious in addressing a familiar friend than in addressing a mere 
acquaintance — he falls more readily into energetic forms of expression. Often 
he assumes an off-hand imperious tone in such cases merely as a bit of pleasantry. 
This would be especially natural when one was urging his friend not to do what 
he feared that the friend might do, viz. in prohibitions. One can hardly fail to 
notice this at any talkative gathering of intimate friends. The 7 instances 
mentioned are: de div. 2, 61, 127 (a supposed command of a god to a man); 
de rep. 1, 19, 32 (addressing the adulescentes before him); de leg. 2, 15, 36 
(Atticus replying sharply to Marcus); Ac. 2, 40, 125 (in conversation with 
Lucullus at a familiar gathering of friends); Zusc. Disp. 1, 47, 112 (replying, 
in a deprecatory tone, to a suggestion that had just been made); A/ur. 31, 65 
(quoting a supposed command of a teacher to his pupil); Par. Sto. 5,3, 41 (in a 
vigorous protest). An unusually earnest and energetic tone is to be found in 
each one of these. Notice, for instance, the strongly contrasted pronouns and 
the other indications of strong feeling. The reason why this construction is so 
rare in classical productions is that they are, for the most part, of a very dignified 
character. The prohibitions they contain are therefore commonly expressed by 
noli with the infinitive (a construction that occurs 123 times in Cicero, twice in 
Nepos, 3 times in Sallust, 3 times in Caesar), or by cave with the present sub- 
junctive (30 times in Cicero, once in Nepos, once in Sallust), or by vide me with 
the subjunctive (18 times in Cicero, once in Nepos). Even ze with the present 
subjunctive is less deferential than the constructions just named; it smacks some- 
what of its sister construction, and so is comparatively rare. Where, next to the 
early comedy, do we find the most familiar tone prevailing ? One may answer, 
without hesitation, in the Letters of Cicero. And it is in these letters that most 
of the instances of me with the perfect in classical times are found. It is also a 
significant fact, and one, I think, not hitherto noticed, that all but 2 of the 14 
instances here found are addressed to his bosom friends, or relatives: 8 of them 
to Atticus, 2 to his brother Quintus, and 2 to his intimate legal friend Trebatius, 
upon whom he was always sharpening his wits, and whom he never lost an 
opportunity to abuse good-naturedly to his face. One of the 2 exceptions is in a 
very impassioned passage of a letter written by Brutus (ad Brut. 1, 16, 6); the 
other is in ad fam. 7, 25, 2, where Cicero is enjoining upon Fadius Gallus in 
the most urgent terms possible not, under any circumstances, to reveal a certain 
secret. To his other correspondents he uses only 70/2, or, in 2 instances, cave 
with the present subjunctive. Excepting the passionate remonstrance referred to 
in a letter written by Brutus, the correspondents of Cicero use only mo/f when 
addressing him. In the treatise ad Herennium, I might add, ze never occurs in 
prohibition, though other forms of prohibition are common. 

Most of the instances to be found in the prose of classical times, of ze with the 
second person present subjunctive in prohibitions, have been cited above. The 
following should be added to complete the list: Cie. Cato Maior 10, 33; ad Att. 
2, 24,1. That ze with the present subjunctive is not more common in the best 
prose is due to an increasing fondness for the #o/z construction, which in dignified 
address became the regular usage. In early comedy there was comparatively 
little call for the more calm and dignified forms of expression, and there, accord- 
ingly, we find that #o/7 is comparatively rare. It occurs in Plautus and Terence 


Proceedings for July, 1893. xiii 


only in addressing some one who must be gently handled. It is found only where 
the tone is one of pleading — it never conveys an order in the strict sense of that 
word. It is almost never used by a superior in addressing an inferior. In the 
two or three exceptions to this rule the superior has some motive for adopting 
the mild tone. 

As regards the different forms of prohibition in classical times, nothing can 
show more strikingly the difference in feeling between we with the perfect sub- 
junctive and zo/i with the infinitive than a comparison of the classes of verbs 
found in the two constructions. Of the 123 instances of mo/i in Cicero, 76 
are used with verbs indicating some mental action, or some action which would 
not be likely to be accompanied by emotion on the part of the speaker, e.g. “ Do 
not suppose,” “Do not be afraid,” etc. In the Letters 21, out of 32 instances, 
are verbs of this sort. Of the 29 instances of cave with the present subjunctive, 
16 are of this sort. In the Letters the proportion is 10 out of 17. A glance at 
the instances of ze with the present subjunctive will show that most of the verbs 
in this construction also belong to the same class. We find the same state of 
things also in Plautus and Terence. Now side by side with these facts put the 
fact that, in the whole history of the Latin language, from the earliest times down 
to, and including, Livy, there are to be found in prohibitions expressed by xe 
with the perfect subjunctive only two, or at most three, verbs that denote 
merely mental activity. The only other instances (four in number) of verbs 
dealing with mental action at all, distinctly involve also other sorts of action. 
The almost entire absence, until the beginning of the period of decline, of this 
whole class of verbs in prohibitions expressed by me with the perfect subjunctive 
and its remarkable frequency in other forms of prohibitions can, it seems to me, 
be explained only in one way. Verbs of this class are, from their very nature, 
such as would not often be accompanied with passionate feeling, and so are con- 
fined to the milder forms of expression. And this, it seems to me, goes far to 
establish my contention that #e with the perfect subjunctive is reserved for 
prohibitions that are prompted by uncontrollable emotion, or else that are in- 
tended, generally from some serious motive, but sometimes merely as a bit of 
familiar pleasantry, to be as vigorous as possible in tone. This tone is commonly 
one of commanding. Rarely it is one of earnest entreaty, though in such cases 
the prohibition is commonly introduced by zo/z. Noli with the infinitive is the 
expression best calculated to win the good will of the hearer, as it merely appeals 
to him to exercise his own will (i.e. “Be unwilling”), or to forbear using it, 
while ze with the perfect subjunctive disregards altogether the will of the person 
addressed, and insists that the will of the speaker be obeyed. 


The paper was discussed by Professors Shorey, Gildersleeve, 
Gudeman, and Hale ; and in reply by Professor Elmer. 


2. On the Interpretation of Sa/ura in Livy VII. 2, by Professor 
George L. Hendrickson, of the University of Wisconsin, Madison, Wis. 
1 Such éxpressions as ue vos guidem timueritis (Cic. Tusc. Disp. 1, 41, 98), nunqguam 


putaveris (Sall. Fug. 110, 4), nec putaverts (Cic. Acad. 2, 46, 141) represent very different 
uses, as I shall show in my fuller discussion. 


+ 


Xiv American Philological Association. 


According to the prevailing view, the compositions designated as saturae, in 
this celebrated summary of the origins of the Roman drama, were rude and 
uncouth specimens of rustic banter, for the most part extemporized and quite 
devoid of plot. They are looked upon as representing a third and original form 
of Roman satire, in addition to the Lucilian and the Ennian forms recognized by 
ancient critics. The passage is one of great importance, and scholars from Scali- 
ger and Casaubon down have vied with each other in their efforts to cast light 
by its aid upon the early history of satire and the drama at Rome. No important 
progress, however, was made in its elucidation (with the exception of a sug- 
gestion by Casaubon, which neither he nor subsequent scholars carried to its 
logical conclusion) before O. Jahn (Hermes II. 225) pointed out that it betrayed 
a sharpness of division into periods suggesting philological combination rather 
than authentic history, and that this was due to an effort to explain certain 
existing institutions by the circumstances of their origin, according to the 
well-known aetiological methods of the Roman philologians of the first and 
second centuries B.c. The suspicion thus cast upon the trustworthiness of this 
account received an extraordinary confirmation in the fact observed by F. Leo 
(Varro und die Satire, Hermes XXIV. p. 67 ff.) that the passage reproduces in 
some of its most essential features Aristotle’s description of the origin and devel- 
opment of comedy. But while furnishing the clue to the solution of Livy’s 
description, without which it must ever have remained an enigma, Leo does not 
seem to have given a correct explanation to his own observations, nor to have 
realized the closeness of the parallelism existing between this account and Aris- 
totle. To point this out is the object of this paper, in which the writer will seek 
to show that the sa¢ura of Livy’s account is an assumed parallel to the old Attic 
comedy, the designation of which was chosen with reference to the vehement 
personal invective (évoyacri xwuwdetv) of the saturae of Lucilius. Leo held that 
satura is here used either as the designation of a loose and irregular poetical form 
in the sense introduced by Ennius (fer saturam), or else that it was chosen, 
under the influence of the Greek cdrvpo, to designate an assumed analogue to 
the satyr-drama, maintaining (with most interpreters) that a connection be- 
tween the satyr-drama and the saturae is suggested by Livy’s account. Both 
these views, however, are incorrect, the latter admitting of refutation without 
reference to the true interpretation. That the sa¢urae are made analogous to the 
odtupo is a view which depends upon the assumption quite universally made, 
from the time of Casaubon, that the erodia of Livy’s account (sec. I1) are a 
survival of the sa¢urae, while, as the name implies and as their connection with 
the Ave/lanae reveals, these exodia were afterpieces in the manner of the Greek’ 
satyr-drama. In this account two classes of performers are sharply distinguished, 
voluntary (zventus) and professional (Aistriones). The periods into which it is 
divided are four in number (exclusive of the Etruscan /udiones): (1) The Fescen- 
nine zocularia of the zuventus, (2) the saturae of the native histriones, (3) the 
fabulae argumento sertae of the histriones, inaugurated by Livius Andronicus, and 
(4) the ridicula (exodia) of the izventus. The latter productions are represented 
as a revival of earlier performances, not, however, as has commonly been assumed, 
of the saturae, which were produced by Aéstriones, but of the ancient (antiguo more, 
11) zocularia of the z#ventus, to confirm which a comparison of the almost iden- 
tical descriptions of the performances of the zwventus in sections §, 7, and 11 will 


Proceedings for July, 1893. xv 


suffice. The saturae are not, therefore, put in any relation to the satyr-drama by 
Livy, but are simply described as the step in the development of the drama pre- 
ceding the employment of the general plot by Livius Andronicus. 

According to Aristotle, comedy had its origin in extemporary phallic verses. 
Its early history was obscure, and only late was it given a chorus at public ex- 
pense. The most important event in its development was the introduction of the 
general plot (u#@0), an innovation ascribed to Epicharmus of Sicily; but at 
Athens Kpdrns rpGros hpiev adgpéuevos Tis lauBixjs ldéas kabddovu moretvy Adyous Kal 
uvOous (Poetics 5). With this compare Livy l.c.8: Livius . .. ab saturis ausus 
est primus argumento fabulam serere. It is only necessary to put these passages 
side by side to reveal the fact of their relation, and that in the Latin account 
saturae corresponds to the laufixh ldéa, the latter phrase designating the element 
of personal satire in the old comedy, which Aristotle censured. ‘H layBixh ldéa 
is therefore at once a designation and a description of the old comedy, with which 
Horace (Sat. I. 4, 6) connects Lucilius (Hinc omnis pendet Lucilius). What, 
then, is more natural than that a descriptive designation of the old comedy 
should be interpreted by the name of Lucilius’ compositions ? Our conclusion, 
therefore, is that satura, employed in the sharply defined meaning given to that 
word by the aggressive quality of the poems of Lucilius, is the name of an assumed 
Roman parallel to the old Attic comedy, and that it is chosen as containing the 
most significant suggestion of the aggressive character of the old comedy which 
could be conveyed by a single Latin word. The description of the saturae 
as impletae modis may refer to the moduuerpia (especially of the parabasis, 
cf. Platonius de diff. com. Diibner I. vs. 52), while the two predominant charac- 
teristics of the dpxala xwupéla recognized by the ancient critics, its yéAws and 
its oxdppata (dbea rot cxwarrev), are here reproduced by risus ac solutus 
zocus (11). ; 

Parallel and related to this description of Livy is a well-known passage of 
Horace (Epp. II. 1, 145 ff.), descriptive of the origin and development of the 
drama at Rome. Here, as in Livy, the beginnings are connected with the Fes- 
cennina licentia (the gaddxd of Aristotle), after which follows a description of 
the transition of this playful banter to open abuse (aferta rabies), which had to 
be restrained by law. Now this account is nothing more nor less than a descrip- 
tion of the layBixy lééa of the old Attic comedy, with its unrestrained directness 
and openness of attack (gavep@s oxwwrev, aperta rabies), to which the check of 
legal restriction was applied at Athens. The compositions thus described by Hor- 
ace correspond therefore to the saturae of Livy’s account, and both represent an 
assumed parallel to the old Attic comedy, devised perhaps to afford an aetiologi- 
cal explanation of phenomena of the literary history of Rome (cf. the paper 
referred to below), or perhaps merely for the sake of constructing a literary 
history for Rome on the Greek pattern. 

In the passage from Livy’s account, quoted above, Livius Andronicus is said to 
have been the first to abandon saturae and to compose the play with general plot 
(argumentum), a change which is represented as an advance in artistic form. 
Now it is well known that Aristotle’s estimate of the old comedy, as compared 
with the new (xa.vj = the later uéon), was very similar to this, For the old 
comedy of personal (ra xa0’ &xacrov) satire was the most direct conceivable 
antithesis to his fundamental principle of the universal (xa@é\ov). In Horace’s 


xvi American Philological Assoctation. 


account Aristotle’s conception of this relation is also preserved, though less clearly 
and in a somewhat different way; for it has apparently escaped observation that 
Horace’s description of the transition from the aferta rabies (vs. 149) to a milder 
form of composition (vertere modum ... ad bene dicendum delectandumque 
redacti, VSS. 154, 155) reproduces Aristotle’s definition of the character of true wit 
as illustrated by the history of comedy: mérepov ody rév eb oxdmrovra dpioréov 

. TQ ph duweiv Tov dxovovra 7 Kal Tréprew (Eth. Nic. IV. 8,7). Just as in 
Livy the drama iz artem paulatim verterat under the influence of Livius Androni- 
cus, ‘a captive Greek,’ so in Horace, Graecia capta ferum victorem cepit et artis || 
intulit agresti Latio (vs. 157). 

It appears, therefore, that the descriptions of Livy and Horace reproduce the 
three stages of development presented by the history of comedy, as set forth by 
Aristotle: (1) The gadduxd (Fescennina licentia), (2) the lauBixy idéa (saturae, 
aperta rabies), and (3) the artistic comedy of general plot (ui@o, axgumentunt), 
designed to please (répmrev, delectare) and not to hurt (uy Auzety, denedicere, i.e. 
non maledicere). 


[The detailed arguments in support of the views here advanced and additional 
instances of this assumption of an old comedy in Roman literature will be found 
in the writer’s paper entitled “The Dramatic Sa/urva and the Old Comedy at 
Rome,” in the American Journal of Philology, Vol. XV. pp. I-30. ] 


3. Solution of Some Problems in the Dialogus,' by Dr. Alfred 
Gudeman, of the University of Pennsylvania, Philadelphia. 


The paper deals (1) with the question concerning the relevancy of the intro- 
ductory chapters to the main point at issue. (2) It is shown, on the basis 
of hitherto overlooked material, that the statement in c. 17 centum et viginti 
anni, etc., is neither an erroneous addition of the preceding enumeration nor a 
round number, but the pivotal point upon which the speaker’s entire argument 
rests. (3) Arguments derived from ‘culture-historical’ conditions are adduced, 
which show that the Dialogus could not have been written in the reign of Domi-_ 
tian or later. / 


Remarks were made by Professors Sproull, Hendrickson, and 
the author. 


4. “Hunc Inventum Inveni” (Plautus, Captivi, 442), by Professor 
W. S. Scarborough, of Wilberforce University, Wilberforce, O. ‘ 


Whether Plautus was purposely obscure or whether the obscurity is due to the 
license to which Schlegel refers when, in speaking of Terence, he says, “‘ Even his 
contemporaries reproach him with having falsified or corrupted a number of Greek 
pieces for the purpose of making out of them a few Latin ones,” is in some 
respects an open question. No one will deny that the Latin comic poets assuined 
liberties and licenses in attempting to copy the new Greek comedy and to adapt 


1 
1 Incorporated in the author’s edition of the Dialogus (Ginn & Co.). 


Proceedings for July, 1893. xvii 


it to the Roman vehicle of thought, liberties which were out of harmony with the 
Greek originals. : 

This may not be due either to indifference, to laziness, or to that “ negligentia” 
which Terence praises and which Dr. West in his excellent edition of Terence 
says must not be confounded with slovenliness. The liberties referred to may be © 
attributed to some other causes, —to undue haste prompted by the need of money, 
to the genius and structure of the Roman tongue, possibly to the character of the 
audience for whom the plays were intended. Whether one of these or all of 
these, the fact remains the same that Plautus has succeeded well in weaving into 
his plays here and there an obscure passage that neither context nor the circum- 
stances of the play itself seem to throw much light upon. The passage under 
consideration is one of them. 

In this passage it will be observed the alliterative element appears, to the fre- 
quent use of which Plautus was especially addicted. It occurs, indeed, with almost 
clock-like regularity, and to my mind the indication is that there was method in 
its use, that it was not merely accidental or a fortuitous coincidence. ‘ 

Be this as it may, no one can deny its large presence in his plays. 

Plautus was not alone in its use. Alliteration was a Latin characteristic. To 
quote Professor Peck,} “Those who to-day doubt, as Lachmann doubted, the pres- 
ence of alliteration as characteristic in Latin diction, should in this particular 
compare such contemporary and fairly comparable writers as Lucretius and Catul- 
lus, Cicero and Caesar, Vergil and Horace.” 

But we are forced to the conclusion that in too many instances Plautus sacri- 
ficed clearness to the swing of the verse that alliteration enabled him to give. 
The present passage, in my opinion, is a strong illustration of this. 

Let us turn to the act and scene itself. 

The personae of the 3d scene (Act II.) of the Captivi are Hegio, the rich old 
man of Aetolia, Philocrates, a captive, and his slave, Tyndarus (Hegio’s own son, 
but as yet unknown to him as such). Hegio turns to Philocrates, mistaking him 
for the slave, tells him that his new master desires that he should pay faithful 
obedience to his former owner in whatever he may wish, and further that he is 
desirous of sending him to his father in order that he may secure the return of 
his son. 

Philocrates declares himself ready to do anything that he is commanded to do. 
Tyndarus appeals to Philocrates not to forget him when he has returned to his 
own country. Philocrates assures him that he will be true to the trust reposed in 
him. The language of both is designedly ambiguous, as Hegio is standing in 
hearing distance and it is the purpose of each to deceive the old man. I quote: 


“ Serva tibi in perpetuom amicum me atque hunc zzventum inveni.” 
Some editions have the following: 
“ Serua tibi in perpetuom amicum me atque hunc zzwen/u inueni.” 


Hallidie gives the following note on the passage: 

“Inuentu ‘dy finding his son’; so Sch., who refers to Merc. 847, eorum in- 
uentu. The MSS. reading, inuentum, is taken to mean ‘(on your return) find a 
friend in this man, in whom we have already found one’; in support of it Brix 


1 Transactions American Philological Association, 1884. 


XViil American Philological Association. 


quotes Men. 452, homines occupatos occupat, Cur. 540, subiges redditum ut red- 
dam tibi, Cic. Fam. XIV. 1, uide ne puerum perditum perdamus.” 

Lindsay says: “unc - i.e., Hegio, ‘and do not lose this one you have found.’” 

Harrington gives this as his opinion: 

“ Hunc inventum invent. The meaning of this passage is much disputed. 
Hunc is referred to Hegio and to his son, and to Tyndarus in the character of 
Philocrates. Some translate, ‘And still find Hegio yours, as you have found 
him’; others, ‘ Find this young man, already in part found by the information we 
have given of him’; or, ‘ Find a friend in this young man, discovered and restored 
to his father.’ Brix says, ‘Gain one already gained to be wholly a friend to you.’ 
Insure Hegio’s perpetual friendship by the restoration of his son.” 

It will be seen from the extracts given that editors are not at all agreed as to 
the exact meaning and translation of this line; and no one of them, so far as I 
have observed, ventures to give an interpretation of his own, but each contents 
himself with giving what others say. 

Dr. Proudfit approaches nearest an acceptable interpretation of the passage in 
question of any of the editors and annotators of Plautus that I have consulted. 
He disposes of it as follows: 

“ Hunc inventum invent. This obscure passage has given rise to many con- 
jectures. Some interpret thus: ‘ Find a friend in Hegio, already found, i.e., 
confirm his friendship to you by restoring him his son. Others suppose it to 
refer to the son of Hegio: ‘ Find this young man, already in part found by the 
information we have given of him.’ Both are unsatisfactory. It most probably 
refers to the son of Hegio in a different sense, and the whole line may be inter- 
preted thus: ‘ Make me your friend forever, and fizd (gai) a friend in “#is young 
man, discovered and restored to his father.’ ” 

The meaning of this particular line is determined by the antecedent of Auzc. 
If we make this antecedent Hegio, then it would read, “ Find this person, Hegio, 
a friend still as you have found him.” This is not a common-sense translation, 
taking the material we have to make it out of. What ground have we for declar- 
ing that Hegio was ever a friend to either of these, Tyndarus (though his son) or 
Philocrates, both of whom were as yet unknown to him at the time the play begins? 
Philocrates was a prisoner of war, and was purchased with others by the old gen- 
tleman, who hoped to find his son among the number. Tyndarus was the servant 
of Philocrates at the time of the purchase. Both were strangers, at least so far as 
Hegio’s knowledge went, and were thus regarded till the discovery was made by 
the return of Philopolemus through the agency of Philocrates, and till Tyndarus 
had been sent for to come home from the quarry to which he had been taken. 

Again, no such translation as the following is allowable, neither will the Latin 
permit it, whatever be the suggested relations of the words of the passage: “ Make 
this old man, Hegio, a friend and keep him so by finding his son and returning 
him to his father.” The editors who adopt this view have no ground for it 
whatever, it seems to me, and are doing violence to the verse that they are striv- 
ing so hard to explain by making it mean what it has never meant and cannot 
now mean. I quote the context and a portion of what follows: 


“‘Scito te hinc minis viginti aestumatum mittier. 
Fac fidele sis fidelis, cave fidem fluxam geras. 


Proceedings for July, 1893. xix 


Nam pater, scio, faciet quae illum facere oportet omnia. 

Serva tibi in perpetuom amicum me atque hunc inventum inveni. 
Haec per dexteram tuam te dextera retinens manu 

Opsecro, infidelior mi ne fuas quam ego sum tibi.” 


If we make Azne refer to the son of Hegio, Philopolemus, the meaning is clear 
and the interpretation is simple. The thought in the mind of Tyndarus, doubt- 
less, was the absent Philopolemus for whom Philocrates was now to be sent. He 
is not lost, as Hegio supposes, but found (inventum). The play upon words 
comes in the finding and the already found. To the old man, Hegio, he was lost, 
hence the word find (invenz) could be with propriety used; to Tyndarus he had 
already been found (inventum). The cleverness of Plautus appears in the use 
of the two words zzventum and inveni,—the one referring to one person, the 
other to another; invent (find) from Hegio’s point of ‘view and inxventum 
(found ) from that of Tyndarus. With Zune referring to the son of Hegio, the 
thought suggested by the passage would be, and do you now seek out this person 
whom we have discovered to be in possession of Menarchus. Go fetch him to his 
father, for we know where he is. He is no longer lost, but found (inventum). 
This latter rendering seems to me to be in keeping with the idea intended to be 
expressed by Plautus himself, and therefore to be the only intelligent and rational 
view to take of it with the light we have at hand. 

It will be observed that I have based the remarks of this paper upon the 
reading zzventum, as found in the text of Fleckeisen (Teubner series) and upon 
which Harrington and others based their editions of Plautus, and not upon the 
reading zzventu. (Vide Ausgewahlte Komédien des T. M. P. fiir den Schulge- 
brauch erklart von Julius Brix, II., 2d ed., 1870 (Captivi); T. M. P. comoediae. Ex 
recognitione Alfredi Fleckeiseni, 2 vols.; Fr. Ritschl iiber die Kritik des Plautus, 
eine bibliographische Untersuchung (1836) in his Opuscula philologica, I1., 
1868, 1 sqq.) 

: Hallidie, who substitutes zzventu for znventum, avails himself, as he tells us, 
of the apparatus criticus in Professor Schoell’s edition of Plautus and a collation 
of V included in his preface to the Casina, 1890. 

Inventum is found in all the MSS., BD VEJ. This being true, thé question 
arises, How did izventu creep into the text? Is it an interpolation, a mistake 
of the copyist, or did’ some editor or scholiast insert it to help himself out and to 
make the text read as he thought it ought to read? I have not accepted the sub- 
stitution because it lacks MS. authority, so far as I have been able to observe. 

With an interpretation based upon imven¢u the translation suggested by this - 
paper would necessarily be modified, and the interpretation given of it by the 
majority of editors would stand. 


At 5.15 the Association adjourned. 


EVENING SESSION. 


At 8 o’clock the Association convened to listen to the address 
of the President of the Association. Professor Hale was gracefully 
introduced to the large audience present by William Morton Payne, 
Esq., of Chicago, Chairman of the Committee of Arrangements. 


eh 


XX American Philological Association. 


5. Democracy and Education, by Professor William Gardner 
Hale, of the University of Chicago. 


“Es ist dafiir gesorgt,” says Goethe, “dass die Baume nicht in den Himmel 
wachsen.” “Care is taken that even the tallest trees shall not invade the 
heavens.” 

We Americans cannot, even in the possible presence of visitors from other 
countries, conceal our belief in the essential value of a democratic form of govern- 
ment. But it would be idle for us, whether in their presence or in the completest 
secrecy of our own hearts, to deny that the advantages which our political system 
brings us have their accompanying defects. Care has been taken that even the 
young Republic shall not scale the heavens, 2 

These defects lie in the unequal workings of the parts of our machinery, in 
consequence of the freedom of the individual state, the individual county, and the 
individual town, and of the great power for mischief which, in a country that 
lacks an aristocratic class and a conservative force in legislation, lies in the power 
of individuals of mistaken or low ideals. Hence our Civil War, our fiat-money 
and silver-coinage schemes, our slow gains in civil service reform, our misgoverned 
cities. Yet against this mischief wrought by the energy of individual leaders 
there rise up other leaders who, in the newspaper office, in Legislatures, in 
Congress, in State House or White House, mould and consolidate public opinion 
for good. And so a civil war is successfully carried to its issue, and the country 
purged of a national shame; so popular economic follies are checked; so the 
once received doctrine “to the victor belongs the spoils” begins to be dis- 
reputable; so some, at least, of our Tweeds die in prison; and so, when the 
shock of disillusion, suffered by the young voter nurtured upon Whittier’s poem 
of the freeman and the ballot-box, is once over, there grows up a deepening faith 
that, rough as is the mechanism by which human life in the mass governs itself, 
the masses learn by their mistakes, and the better side in the end triumphs. ° 

If this be true in politics, then one may with a stouter heart face and unre- 
servedly discuss the difficulties and discouragements which our system of indi- 
vidualism brings to the cause of education. 

The largest conglomeration known in education is the state. Yet even this 
has, in general, no serious controlling power. Each little town will shape the 
education it gives its children according ‘to the views of shifting officers, more or 
less controlled by local opinion. Even the state universities cannot rise much 
above the convictions of the average voter; and these convictions will differ 
widely in different states. In this country, the average man is, in the main, the 
court of ultimate appeal in education, and, in a matter so far removed from his 
practical knowledge, the average man is very likely to go astray. The average 
local school-board is consequently in great danger of not knowing what con- 
stitutes an education; and in particular, it is in danger of regarding only those 
subjects as desirable for young men and women the immediate advantage of 
which in earning a living is obvious to a shallow observer. In its ignorance 
of the extent to which specialization has gone, it generally assigns several subjects 
to each teacher; so that, in place of a Latinist, a Germanist, and a historian, 
you may, perhaps, in large schools, find three men each of whom has to teach all 
three subjects. For the same reason, it generally, if it seeks college graduates for 


Proceedings for July, 1893. xxi 


teachers, supposes them capable of teaching anything, without inquiring whether 
they have gone beyond: the minimum demands of the institution in which they 
were trained ; but it much more frequently is content to appoint, even to positions 
in high schools, young men and young women who are themselves graduates of 
such schools only, and who therefore have not advanced a step beyond the point 
to which they may have to carry their own pupils. The same want of apprecia- 
tion of differences also leads communities to pay smal] salaries to school teachers, 
to lay upon them many hours of work, and to grant them no security of tenure, 
and little honor outside of a high-sounding title that once belonged to specialists. 
The result is a school system that puts us far behind Germany, England, or 
France. And a further result is that, while a large quantity of advanced work 
comes out of the German gymnasium, almost none issues from the American high 
school. The idea of a creative scholarship has no home there. 

When it comes to the universities, the same tale has to be told, with some 
additions. The average board of trustees overloads the college teacher with 
hours of stated teaching, and, finding always many competitors for vacant places, 
pays scantily. The result is twofold. The professor’s life, with all its charm when 
at its best, does not attract as large a number of thoroughly able men into its ranks 
as could, in the interests of education, be desired; and, on the other hand, 
in many of our strongest universities, those professors who possess no independent 
means are obliged, instead of devoting their scanty leisure to investigation and 
publication, to devote it to earning enough money from outside sources to make 
up what is necessary to satisfy the claims of the butcher, the grocer, and the coal- 
combine. And the same law of supply and demand has led to the almost entire 
absence of any organized arrangements for the maintenance of the professor, if, 
after a life of ill-paid and _saving-forbidding toil, he has the misfortune to outlive 
his usefulness. 

It is not, however, the outside governing body or the outside community alone 
that will be found to go astray under the freedom of individualism. The views of 
professors themselves, in the things which it is their function to settle, are likely 
to be colored, to an extent by no means inconsiderable, by the local conditions 
under which they have lived, and especially by the local conditions under which 
they have been educated. The result is the greatest diversity of opinion upon 
the most fundamental questions of education. 

But, as I have spoken of these difficulties, two ideas must have repeatedly 
forced themselves upon your minds, namely, first, that in a system in which 
individualism can and must work harm, the remedy for this harm lies in the very 
thing which brings it about, that is, in individual action; and second, that upon 
the individual there rests, in such a system, a heavy obligation. In education, as 
in politics, Americans of convictions have no right to sit idly by. 

Obligation rests upon individual schools, and upon parents who have children 
in schools, since it is only the development of the best schools that can give to the 
rest the evidence of what is possible. Obligation rests upon individual universities 
to lead the community which forms its environment and feeder to higher and juster 
conceptions of education than it possesses. And upon universities founded and sup- 
ported by private means obligation of an especially heavy kind rests, since no other 
institutions can act upon their beliefs so fearlessly. It would be difficult, for exam- 
ple, to bring a body of state regents, themselves largely affected by the opinions of 


- 


xxii American Philological Assoctation. 

the masses which form their constituents, to entertain certain ‘convictions about 
salaries and hours of instruction and character of work which were unanimously 
reached by the trustees of the university of this city. And finally —for to this 
Rome all the roads in our country lead—a grave responsibility lies on every 
individual man or woman, in or out of schools or universities, that has firm 
beliefs. It is individuals that form the nucleus for false opinion; it is individuals 
that must lead the fight for sound opinions. 

The picture I have drawn has been a dark one. But I should do myself an 
injustice if I were to leave an impression of fundamental doubt. I believe in the 
democratic idea. The doctrine is sound which stands written upon the gate of the 
water-front of the White City: “Civil liberty the means of building up personal 
and national characters” And even if for a while the price of widely varying 
and often faulty systems of formal education has to be paid for this education 
of character, the purchase is well made. But there are many signs of hope. 
What has been done by the accumulated work of individuals in a comparative 
length of years may be seen in our oldest university, Harvard, of which an 
American has no reason to speak with anything but pride. What may be done 
by the work of individuals in less than a score of years may be seen in one 
of our youngest universities, Johns Hopkins; and the lesson seen in the record 
of such a university, of the rapid power of creation possessed by individual minds 
working under democratic conditions, is most significant, when one .recalls at 
how many points in these United States colleges and universities, already estab- 
lished, stand ready as vantage-grounds for the battle of educational ideals. In 
the lower education, too, signs are already to be seen of a consolidation of 
opinion similar to that of the consolidation which I have spoken of as carrying 
great movements in politics. In particular, the now-famous work of the Com- 
mittee of Ten — whatever may be the value of this or that opinion advanced — 
constitutes an event of national importance. The time is one of wide-spread 
interest in educational questions, and of busy discussions of them in journals and 
at conventions. Not only, then, as I believe, will education in America finally 
reach a high efficiency, but—a matter of some interest for us who have not 
yet left the stage — we are, unless signs fail, at the threshold of a time of rapid 
advances. 


At the conclusion of the address an informal reception was held. 


GENERAL SESSION. 
Cuicaco, July 12, 1893. 


Professor Hermann Osthoff, of the University of Heidelberg, who 
had been requested to preside over this joint meeting of the AMERICAN 
PHILOLOGICAL ASSOCIATION, the MODERN LANGUAGE ASSOCIATION, and 
the DiaLecr Society, opened the proceedings at 10 a.m. by a brief 
address of welcome. On motion, Professor Smyth was appointed 
Secretary. 


Proceedings for July, 1893. xxiii 


6. The Connection between Indian and Greek Philosophy, by 
Professor Garbe, of the University of Konigsberg. 


Professor Garbe first briefly outlines the fundamental principles of the idealistic 
monism of the Vedanta and the dualism of the SAinkhya philosophy. The ideas 
of both these systems are found in Greece: the monism of the Vedanta in Xeno- 
phanes and Parmenides, and the doctrines of the Samhkhya philosophy in the 
Ionic physiologers, Anaximander, Heraclitus, and the rest. In Empedocles, 
Anaxagoras, Democritus, and Epicurus, indubitable points of agreement with the 
Samkhya philosophy are found, especially with respect to the ideas of metempsy- 
chosis and the eternity and indestructibility of matter. 

But all these coincidences are coincidences of general thought, and not of 
special or arbitrary details. Hence, Professor Garbe will not give an apodictic 
opinion as to the source of the doctrines of these Greek philosophers; for the 
Greek and Hindu doctrines might each have arisen independently, the resem- 
blances being due to the natural sameness of human thought. He inclines, 
however, to the opinion that the Greek systems mentioned weve derived from 
India, and substantially adopts the conclusions, though not all the arguments, 
of Ed. Réth, Aug. Gladisch, and C. B. Schliiter on this question. He regards 
Persia, not India, as the place of mediation of these ideas. 

But if the resemblance between the two philosophies in the case of the above- 
mentioned philosophers is only a general one, in the case of Pythagoras it is very 
clear and striking, even in details. The very word “samkhya” denotes number. 
The two systems further agree in the doctrine of metempsychosis, the prohibition 
of eating beans, the doctrine of the five elements, in the possession of the so-called 
Pythagorean theorem, the irrational number V2, and in the character of their 
religious and philosophical fraternities. But while in Pythagoras all these doc- 
trines are unconnected, and lack an explanatory background, in India they are at 
once rendered intelligible by the intellectual life of the times. On this point 
Professor Garbe accepts the main conclusions of Schroeder. Unquestionably, the — 
system of Pythagoras is derived from Indian sources; but according to Garbe, 
Pythagoras acquired his knowledge of the Samkhya philosophy in Persia, not 
in India. 

One other point is elucidative here. With Pythagoras number is the essence 
of all things. In the Satkhya system, however, number does not play so im- 
portant a rdle, that system being simply zemed after the enumeration of the 
material principles. The Pythagorean form of the doctrine Professor Garbe 
attributes to a misunderstanding on the part of Pythagoras, and disagrees with 
Schroeder in the belief that it is an o/der form of the Samkhya philosophy. 

The next influence of the Hindu philosophy is that on Christian Gnosticism 
and Neo-Platonism. Disagreeing with Lassen as to the share of Buddhism in the 
formation of the Gnostic systems, Professor Garbe is rather inclined to attribute 
the greatest part of Hindu influence exerted on these doctrines to the Samkhya 
philosophy — especially with respect to the Gnostic identification of soul and 
light, which is exactly the idea of the Samkhya that soul and light are the same. 
Also, the Gnostic classification of men agrees with the Samkhya doctrine of the 
three Gunas. There are also many other points of contact. 

The agreement of Neo-Platonism with the Samkhya doctrines is striking and 


- 


Xxiv American Philological Assoctation. 


considerable. Especially is this the case with Plotinus and the Yoga philosophy, 
a branch of the Samkhya doctrine. But perhaps the most significant loan of 
Grecian thought from Hindu philosophy is the Neo-Platonic conception of the 
Aéyos. This is derived from the Hindu doctrine of the VAch (voice, speech, 
word). Garbe accepts Weber’s opinion on this point (/udische Studien, Vol. 1X.), 
only he thinks that the date of the derivation of the \éyos idea from India is to be 
put 500 years earlier than Weber would seem to put it. 

The influence of the Samkhya doctrines and of Hindu philosophy generally on 
Occidental thought does not extend beyond Neo-Platonism. With the exception 
of a tinge of Buddhism in Schopenhauer and Hartmann, no modern influence of 
Hindu ideas is noticeable. This state of affairs will be bettered by new and 
more complete expositions of the Indian philosophy. In this lecture, which will 
appear in The Monist of January, 1894, Professor Garbe only attempts to seek out 
the Azstorical connection between Indian and Greek philosophy, and does not 
discuss the internal character of these relations. 


7- Some Problems in Greek Syntax, by Professor Basil L. 
Gildersleeve, of the Johns Hopkins University. 


The problems of syntax, like all problems of grammar, are problems of method 
and problems of material. At one time method comes to the front, and in the 
dissertations of the forties and the fifties one grew somewhat weary of via atgue 
ratione, or, by way of variety, ratione viague. The logical method was dis- 
carded years and years ago because, as we were told, language is not logical but 
psychological; as if the psychological did not involve the logical; as if there could 
be any orderly presentation of truth without some kind of reasoned arrangement; 
as if the establishment of categories were not important for the discovery of law. 
Then there was a period when the ‘ organic’ method was rife, the organic method, 
which was after all only the logical method in disguise, the method against which 
Kriiger protested so vehemently because it made syntax the theory of the sentence, 
said sentence being a telescopic thing, now shut up into the compass of a word, 
now drawn out to the starry-pointing extreme of the Ciceronian period. But in 
spite of Kriiger’s protest against the identification of syntax with ‘Satzlehre,’ the 
method had and has its advantages, and for parallel grammars will always be 
available, though all manner of crimes have been committed in its name. 
Whether the danger of further crime is averted by the cloud of controversy that 
has gathered about the definition of the sentence is questionable, because under 
these conditions discussion is apt to assume the character of a religious war, 
always the most merciless of wars. The jelly-fish theorists have also had their 
say in the matter, though the doctrine that the sentence preceded the.word and 
the word was differentiated from the sentence has had little influence on syntax. 
The earlier attempts to handle syntax etymologically, that is, to arrange the clauses 
of the sentence after the supposed case-relations of the introductory particles, was, 
as might have been expected, not an unqualified success. The uncertainty of 
the etymology and the perverseness of language in smothering etymology under 
analogy combine to make such a method as a universal method hopeless. Still, 
one cannot say that the attempt has been fruitless, and the neogrammarians have 
been busy in sorting the good from the bad. Among other things, this method 


‘ Proceedings for July, 1893. XXV 


has led to a very conspicuous line of work, has led to what may be called the 
rehabilitation of parataxis. The primitive Aryan, we are told with all the 
solemnity of a revelation, spoke in sentences, and these sentences were simple 
sentences. How these sentences were combined to make compound sentences 
is the problem. The question was primarily one of proximity, as dangerous in 
language as in life. Then came correlation, then all the other complex groups 
that we call hypotactic. The only explanation of a hypotactic sentence, there- 
fore, is the restoration of the original parataxis. Up to a certain point nothing 
has been more fruitful than this ‘ paratactic’ method, But we have to be on our 
guard. Hypotaxis is older than our record, and we cannot argue safely as to 
prehistoric processes with consciousness lost and analogy working its will; we can- 
not insist on the steadiness of the original function. How wild such theorizing may 
become is shown by the fact that one adventurous gentleman wished to make ply 
with the infinitive a paratactic sentence in spite of the dependent nature of the 
infinitive itself, and another has recently considered it a great gain to make the same 
infinitive an imperatival infinitive, as if an imperatival infinitive were any less 
dependent than an infinitive of purpose, of which it is only a form. Not one 
of these methods is to be rejected out and out, none to be applied without caution, 
and, to exemplify the importance of caution, a few words may be said in regard 
to some of the recent studies in the range of historico-statistical syntax, for it is 
hardly fair to speak of historico-statistical method. The historico-statistical school 
deals with the registration of facts on lines that are supposed to be historical, but 
when it comes to the interpretation of the facts there is great divergence, and 
first one method, then another, is applied. As in every sphere of so-called 
intellectual activity, so here human absurdity asserts itself unabashed, and the lies 
of the census bureau and the foolish inferences of the manipulators of figures are 
not lacking in this new order of philological knighthood —the Knights of the 
Decimal Table. Not that the use of figures is illegitimate. Some minds are 
accessible to nothing else, and to so many is an array of figures an end of con- 
troversy that the weary investigator, whose tabulation has proved what his insight 
had divined before, counts all his toil a gain if he can thereby escape the din 
of a fruitless logomachy. Then the statistical way is often the only way. It is 
only thus that the question of proportion is to be determined, and statistic has 
often called attention to significant facts, and so led to truly scientific results. 
But in the majority of cases the conditions of the problem were settled in advance, 
and mountains of statistics are of no avail without a clear appreciation of those 
conditions. Statistic has to be taught what categories are worth watching, 
otherwise the cum pulvisculo exhaurire brings forth nothing but dust. The 
position of clauses makes an enormous difference; the position of words makes 
an enormous difference; positive and negative conceptions have often a con- 
trolling influence, and these are things that not unfrequently slip through the 
meshes of statistic. Of course, no matter how arid the statements, the investigator 
who knows what he is after can make some use of them. So the work, mostly 
mechanical, that has been going on in the range of the prepositions has aided 
scholars in reinforcing lessons that have long been known to those who choose to 
watch. But apart from the distinct relegation of certain prepositions and certain 
uses of certain prepositions to the field of poetry and dialect, the ‘favorite 
preposition’ business seems to have been very much overdone. In his treatise 


- 


XXV1 American Philological Association. 


on the prepositions in the Attic orators, which is one of the most laborious and 
on the whole one of the most valuable works in this line, Lutz has passed over 
some of the most significant phases, and the elaborate tables are nothing more 
than confirmations of a frzorz conclusions. An author’s favorite preposition is in 
many instances nothing but the preposition demanded by his subject, and what 
Lutz has seen to be true of Isaeus’ use of éf is true elsewhere. In many of the 
dissertations that swarm over this field, history and chronology are treated as 
if they were practically identical, as if the emergence of a construction in literature 
were the emergence of the construction in language, and the disappearance of it 
from literature were its death. The individual counts everywhere, even if not so — 
manifestly in Greek as in Latin; and in Greek the department is more potent than 
the individual. Take the familiar example of the articular infinitive. To my 
mind nothing is clearer than that it belonged to the people, and was so slow in 
making headway because it belonged to the people. Hence the exclusion of it 
from the aristocratic epic; hence the occasional use of it in lyric poetry, which 
had to break bounds if it was to be truly lyric. The articular infinitive is an 
indispensable organon of philosophic thought, and hence it is not disdained by 
Parmenides, who forces the plebeian construction into -epic verse, and shows 
thereby that his epos is no true epos. It is the same Parmenides, be it noted, 
that uses od 7), that familiar turn which is strictly conversational, strictly dramatic. 
That od 44 is young is a zon seguitur from its late appearance. It is old even if 
it is not so old as the articular infinitive, which in turn can hardly be so old as the 
historical present, for the historical present is Aryan if anything is Aryan, and yet 
the historical present is absent from epos and higher lyric. Students of the 
Greek language who study Greek as a language and not as a form of literature, 
rebel against the stylistic reason.. But those whose chief interest in language 
lies in language as a form of art are too familiar with similar phenomena in other 
spheres to see anything dangerous in the admission of style as a norm of usage. 
The realm of the artificial as well as the artistic has to be extended into the past. 
Ornament is older than clothing. In raiment and apparel one hardly thinks of 
the covering of nakedness. The English language was at its noblest in the time 
of Elizabeth, the Greek language at its noblest in the time of Pericles, and while 
the comparison is evidently not fair, if urged, and while the complexity of the 
English problem is almost infinitely greater, still Spenser might help us to under- 
stand how the dramatic poets consciously used obsolete words and hyperepic 
syntax, and what is true of the Attic dramatists is true of Pindar, and who knows 
how far back it is safe to push this use of ornament ? These are things that 
statistic will not reach, and yet statistic tries to reach. Still statistic has stirred 
many problems, none more noteworthy than the use of the tenses. After passing 
along a number of formule about the historical tenses, in which conative, 
progressive, panoramic, ingressive, complexive have figured largely, in which 
convenient translation has too often taken the place of convincing argument, 
grammarians have at last been aroused to the serious study of the imperfect and 
the aorist. The machinery of statistics has been brought to bear on the problem, 
and is even now at work — apparently to the detriment of the aorist in Greek. 
But what is an aorist, what is an imperfect ? Are we to call everything an 
imperfect that has the form of an imperfect ? Are we to huddle together first 
aorist and second aorist, so manifestly different in their build ? And what is to 


Proceedings for July, 1893. . xxvii 


become of our accepted school grammars, if we do not? Statistic will not relieve 
us of this important condition of the investigation. Nor can we put aside the 
question of the dialects. So Cobet says that Herodotus uses the imperfect very 
freely, indeed implies that he uses it amiss, ‘after the Ionic fashion.’ But the 
same Cobet says that the imperfect is the tense for actual vision; imperfectorum 
usus, he says, oculatis testibus proprius. Why may not Herodotus haye assumed 
the position of an oculatus testis? The artist and the Ionian would beat one. 
But what of Thucydides, who, when he falls into the descriptive vein, has no 
objection to the imperfect?. As to Xenophon, he seems to have caught Herodotus’ 
trick during his campaign in Asia Minor, and the influence of Herodotus and 
Xenophon was potent in after times. We cannot get out of the abhorred cate- 
gory of stylistic, if we will. Nor will the counting of imperfects and aorists suffice 
even after we settle our definition of imperfect and aorist, because we have to 
reckon with the effect produced by the temporal rélation of the moods and 
verbs. A pageful of evolving present participles cannot fail to temper the sharp 
aorist indicative. Soa comparison of the relative use of aorist and imperfect in 
Greek with the relative use of historical perfect and imperfect in Latin will not 
yield the clear results that are anticipated. There are too many other temporal 
factors, and a practical grammarian but the other day made the portentous 
blunder of léaving out of the calculation, though not out of the statistic, the Latin 
use of the historical present. This omission of important elements is one of the 
serious and at the same time amusing defects of modern ‘methods.’ So in the 
study of the prepositions what may be called the mefastasis of the prepositions 
has often been overlooked. ‘After it became a familiar thing to vilify Xenophon 
for his use of the preposition ovr, one would have thought that otv was avoided 
by model prose as if it were a pestilence, and every school-boy is taught nowa- 
days to model his prose composition on the law of the Attic orators. ovv dead ? 
Yes, somewhat as dvd is dead. Its prepositional life may be over, but the death 
as a preposition is compensated by its enhanced life as an element in compound 
verbs, and that enhanced life is the original life, and breathes the strong person- 
ality of the primal cJv. But such things do not enter into the calculation of 
the mere statistician. In fine, for illustrations might be multiplied without 
number,—in fine, no mechanical scheme can take the place of the loving 
sympathy to which alone language will yield her inmost treasures. The wise 
Centaur, the first philologian on record, was right when he said 


kpurrat kratdes évri cops MeOods tepav pidordrwv. 


The kingdom of syntax is not to be won by the violence of brute numbers. 


8. On the Origin of the so-called Root-Determinatives, by Professor 
Maurice Bloomfield, of the Johns Hopkins University, Baltimore. 


The writer’s article ‘On Adaptation of Suffixes in Congeneric Classes of Sub- 
stantives’ (American Journal of Philology, XII. 1-29) furnished the starting- 
point for his remarks on the ‘root-determinatives.’ He began by pointing out 
additional cases of congeneric attraction. Vedic angusthd (Sk. angustha) = Zd. 
ayguita ‘great toe’; Ved. dstha ‘lip’; Ved. updstha ‘lap’; Ved. késtha ‘ab- 
domen’ (cf. Arm. £u¢ ‘ venter’) exhibit a case of adaptation of the suffix -stha 


+ 


XXViil American Philological Association. 


in four designations of parts of the body removed from another as far as possible. 
So also Ved. dsthi ‘bone’ (cf. Obg. ost ‘ bone’), sékthi ‘thigh,’ and asthi-vdntau 
‘the knees,’ are parts of the body, and that, too, dozy parts. The last case 
exhibits a more narrow congeneric domain than the first, but there is no law 
which dictates the lengths to which language may go in feeling that things are 
congeneric. Vague and half-relevant associations are as much at play in this 
kind of operation as the sober and matter-of-fact. The nom. wovs may have 
been formed after ddovs, the distance being no greater than that between dstha 
and angusthd. : 

Congeneric assimilation and adaptation may take place, first, between words 
which designate things absolutely or nearly identical. Lith. sa/dus, Obg. 
sladuki ‘ sweet,’ are derivatives from the I.-E. stem sa/d ‘salt’ (Schmidt, Indo- 
germanische Neutra, p. 182), but they owe the particular conformation of. their 
suffix and the specialization of their meaning to I.-E. suadus ‘sweet.’ The v of 
the oblique cases of els ‘one’ supplanted I.-E. # [stem sem, (o) ula] owing to the 
influence of stems ofvo- (European ozo-) and *uovfos (uobvos, ubvos). Avestan 
asti ‘minister’ and Sk. a¢(z) tht ‘guest’ (I.-E. *oth-ti) seem to exhibit the effect 
of the congeneric relation with Lat. Aostis, Germ. gast(i)s, Obg. gosti, whose basic 
form may have been I.-E. *zhostis for *zhoth-tis. Greek dvrujyv is formed from the 
twiced reduced form agy- ‘breathe.’ The identity of its suffix with Sk. amdn 
is due to their identity of meaning. The latter may be z/mdn from the dis- 
syllabic root aza1 ‘to breathe,’ and may again represent a radically independent, 
but congeneric formation with Germ. *2f-ma, the basis of Obg. a¢um, Old Saxon 
athma, Ags. athom. Hosts of doublets, occasionally triplets, owe their similarity 
to congeneric influence; words like dgru: ddxpv; oxwp, oxarbs : ¢dkyt caknds 
(xémpos); Zend partou: Sk. sétu ‘bridge’; dlio-: dnya-; ursén: rsen, and many 
others which will be discussed in the fuller presentation of this paper. 

Secondly, words belonging to the same broad class frequently call up for conge- 
neric assimilation those members of that class which stand related to them by more 
special traits. Thus, of parts of the body those of the head are especially prone 
to influence one another; hence augo and auso in German; ¢/nga and daga 
in Sk. (even the Zend has svva in srvo-jan ‘breaking the horns’). Germ. mus 
and /us are not only united by the common bond of ‘designations of animals,’ 
but they are both ‘varmint,’ ungeziefer. Gr. Adpvyf and gdpuvyf evidence a 
special intimacy within the class which designates ‘ hollowness,’ of which cipry& 
‘pipe,’ oddmcvyé ‘trumpet,’ ompdrvyé ‘cave,’ and cfpayé ‘cleft,’ are the more 
broadly related representatives. : 

Thirdly, opposites exercise the same influence upon one another as identical 
congeners, les extremes se touchent. To the hosts of well-known examples may 
be added e.g. Vedic sayz- ‘thief’: Payz- ‘protector.’ Just as udjdé ‘hunch- 
backed’ seems to betray in its final sound its congeneric relation to my-ubja 
‘ crooked-back,’ so also urudbjd ‘wide open’ is the opposite of kubjd. We have 
here a start in the direction of a suffix -wéja2, whose productivity is limited, how- 
ever, by lack of opportunity. 

Fourthly, the broadest categories produce congeneric assimilation and adapta- 
tion. Not only those which readily suggest themselves, as designations of animals, 


1 The mutilated forms, e.g., instr. dna Sk. are clearly formed under the influence of san# 
‘ body, self’: the loss of the 2 cannot be due to phonetic influence. 





Proceedings for July, 1893. Xxix 


colors, but such as are hardly categorized consciously at all, as e.g., the Greek 
words designating hollowness, above. The London public and the London papers 
have created of recent years a suffix -eries (plurale tantum), designating public 
exhibitions. It appears to have started with the fisheries exhibition, which was 
called ‘ the fisheries’ for short; next came an exhibition of flowers, which required 
no violent adaptive process to be turned into ‘ the floweries.’ Again, the hygienic, 
or health exhibition, became ‘the healtheries,’ and finally the Colonial and Indian 
exposition appeared as ‘the colinderies,’ a word which the purist would say ought 
to turn the very printer’s ink vermilion. 

Verbal suffixes are in general preémpted for the expression of indispensable 
relations: voice, tense, and mood. But occasionally a verbal suffix is free to 
adapt itself to some more special function, totally foreign to its original value. 
A considerable number of Greek verbs with the suffix -idw (ddovridw, etc.) 
designate ‘to suffer from a certain disease,’ or ‘to have the symptoms of a certain 
disease’; an equally large number in -id{w designate religious acts and celebra- 
tions (Baxxidfw, dpyidfw, etc.). In Lat. -Zsco the inherently inchoative value of 
verbs like senesco, adolesco, cresco, etc., has been transferred to the suffix, wherever 
it occurs. The I.-E. stems fr(4)-sk- ‘ask,’ i(s)-sk- ‘search, wish,’ un-sk- ‘ wish,’ 
point to a proethnic adaptation of the same suffix (-sko) to the idea of ‘asking, 
searching, wishing.’ The suffix -fo has adapted itself to the acts of ‘ binding, 
twisting, bending, braiding, folding’: Lat. plecto, Obg. fihtu,; — Goth. falpan, 
Obg. pletqg, Sk. puta ( pita); — Lat. pecto, Obg. fiktu ; — Lat. necto; cf. Sk. nadh 
(1.-E. *nedh abstracted from I.-E. *#egk and *negdh for *negh-t), and Obg. nista 
(*nedh-to with suffix -¢o, unconsciously doubled), Goth. ga-vida, Obg. witu ;— 
Sk. vestate, Lith. vfstau. 

Romance ‘/astare (Ital. ‘astare, French ¢éter, Germ. tasten, Engl. taste) is a 
modification of late Lat. sexare ‘to touch vigorously’ in deference to gustare,! 
just as German eischen [I.-E. i(s)-sko-, Obg. etscOn] owes its h to hetissen, and 
dialectic English sguench is guench with the s of congeneric sguelch added. The 
Vedic root dhkyas ‘to be frightened’ is clearly root 442 with as of root tras 
‘tremble’ as an extension. All such cases throw a strong light upon the so-called 

‘ root-determinatives: I.-E. yers ‘flow’ may be ers ‘ flow’ with the ¢ of congeneric 
yor or ueér ‘water’ udd-r, ud-nds ‘water’ prefixed; I.-E. stembh (*sthembh) ‘sup- 
port’ may be I.-E. skembh ‘support,’ formally modified in deference to séa (stha) 
‘stand’; the Aryan root ¢vaés is built up upon the I.-E. root “és, Aryan éaks 
with a dash of the root vaks ‘ grow’ (I.-E. yess) through it. Only a shade less 
certain is the genesis of Ved. ésar ‘to sneak up with malicious intent’ as a con- 
tamination of root sav ‘go’ by root Zar ‘ overcome,’ or the origin of the I.-E, root 
Rleys ‘hear’ from &/ey under the influence of verbs and nouns for ‘ hearing,’ which 
end in s: I.-E. ous ‘ear,’ Goth. hausjan, dxov(c)w, etc. 

In some such way the intrinsically harmless I.-E. alveolar voiced stop d has 
established itself as a ‘root-determinative’ for actions and things pertaining to 
the ‘podex’: Lat. pedo, podex, Biéw, New Slovenic pesdéti from I.-E. pezd; — 
wépderat, Sk. pardate, Obg. firzan from I.-E. perd; — x&w, xddavos, Sk. Addati, 
Zd. zadanh from I.-E. ghed ; — Norse skita, Ags. scttan, Ohg. scizan with Lith. skédz 
again point to a root in d;—note also Sk. d4asdd‘ podex’ and perhaps, as an 
opposite to I.-E. perd, I.-E. skerd‘ vomit,’ Sk. chard; cf. Obg. skargdié* nauseating.’ 


1 For the current view of the origin of tastare see Kluge s.v. tasten. 


XXX American Philological Association. 

One more group may be indicated briefly: it is a group ending in # in roots 
for ‘sound.’ 

1) Sk. dhvdnati, ON. dynja, Ags. dynnan (Engl. din). 

2) Sk. dhranati (Dhatupatha), Goth. drunjus ‘sound,’ Nhg. dréhnen, Gr. 
Opjvos ‘ dirge.’ 

3) Sk. stdnati, Gr. orévw, enbiies: ON. stynja, Nhg. stéhnen. 

4) The variant root ¢ex in Ved. tanya ‘thunder,’ Lat. sonare, tonitru, and 
the German derivatives. 

5) Sk. svanati, Zd. hvangnt, Lat. sonere (arch.) sonare, with nominal derivatives. 

6) Obg. zvinéti ‘to sound,’ svoni% ‘sound,’ is a direct modification of I.-E. 
Pheu ‘call,’ Obg. zove (cf. Ved. Advate) to sound-roots in en. 

7) We may mention also Lat. cazo: Gr. xavd{w ‘sound,’ Erse canaid ‘ canit,’ 
and point out the special congeneric relation between xavax7 ‘clash’ and 
orovaxy ‘groan.’ Perhaps also Sk. 4van1 ‘sound’ may belong to the same 
root, having borrowed its v from adkvan and svan, thus still further accentuating 
its character of congener with these.? 

In the light of such cases the ordinary view of the origin of the ‘ root-determi- 
natives’ “as agglutinative’ elements needs to be strongly modified. The deter- 
minatives are more frequent at the beginning of roots than at the end. The 
phenomena in question are due less to agglutination and more to congeneric 
assimilation and adaptation of certain final elements to certain categories of 
roots grouped as a semasiological unit. The writer will shortly present his views 
more completely and with a greater abundance of illustration. 


g- On the Implicit Ethics and Psychology of Thucydides, by 
Professor Paul Shorey, of the University of Chicago. This paper 
appears in full in the TRANSACTIONS. 

Adjourned about 1 o'clock. 


AFTERNOON SESSION. 


The President called the meeting to order at 3.15 pM. The 
reading of papers was resumed. 


10. The Language of the Law, by Herbert L. eanet, Esq., Detroit, 
Michigan. 


The language of the law, as the meeting point between linguistic and legal 
science, offers interesting material for the student of either science. Its study will 
tend to further the advancement of both sciences, and especially the latter. It is 
worthy of special study for the accomplishment of two purposes amongst others; 

z.: first, improvement in the means of legal expression; secondly, the elucidation 


1 ‘The cerebralization need have no etymological value any more than in vada =nada ‘ reed,’ 
2 In Nhg. drdhnen, stihnen, and ténen (the latter secondarily from the loan-word on, roves) 
the congeneric character of the group is indicated vividly even to-day. 





Proceedings fox July, 1893. XXxxi 


of legal history, by tracing the origin and history of the elementary legal concep- 
tions represented by single words. 

Improvement in Legal Expression.—The need of improved means of legal 
expression in English law is great. It is a want which is very generally felt, 
though perhaps not so generally recognized. The crying need of the hour in the 
field of law is a better arrangement and statement of the law, one which will 
render the law clearer, more systematic, and more accessible, —in a word, more 
scientific. A great difficulty in the way of supplying such scientific statement is 
the want of a scientifically accurate terminology. The present terminology is 
inadequate for scientific uses. It partakes of the unscientific character of the 
legal system which produced it. It has been for the most part produced in the 
course of judicial administration, in which it was moulded by lawyers and judges 
to meet immediate practical needs; and it is therefore adapted to those uses, but 
not to scientific arrangement, or to expression for purposes of codification or other 
legislation. Any statement of the law for purposes of legislative enactment calls 
for the highest powers of accurate and comprehensive expression of which human 
language is capable. If we are ever to have a legal language fitted for such uses, 
it must be formed with those ends in view, and must come from scholars conver- 
sant with both legal and linguistic science. It will not come from the courts or 
from legislatures, or from lawyers engrossed in legal practice. It is to be hoped 
that the subject may erelong engage the attention of persons having both the 
ability and the opportunity to deal with it successfully. 

The most valuable service, however, which philology is capable of rendering to 
the science of law is in the field of legal history. Legal science must be based 
upon legal history, and legal history must be essentially a history of ideas; that 
is to say, a history of the development of legal conceptions and principles, as well 
as of the various rules and enactments which have found a place in positive law. 
Much of this history of ideas is to be read in the history of the words’ by which 
they have been expressed. A study of the terminology of English law discloses 
facts which are of the first importance to a right understanding of that system of 
law. English law is found permeated with, and largely made up of, legal concep- 
tions and principles which are practically identical with those found in the Roman 
civil law. At the same time, English law is held to be an indigenous and inde- 
pendent system, produced, developed, and sanctioned by the customs or immemo- 
rial usages of the English people. If this theory of English law were true, the 
fact that English law contains so much that is common to Roman law would be 
of great significance in comparative jurisprudence. If these elements which the 
later system has in common with the earlier were acquired by independent growth 
and development from within, and not by borrowing from the older system, that 
fact would be remarkable testimony to the permanency and universality of those 
elements. But the philology of English legal language leads to wholly different 
conclusions, In a paper read before this Association in 1889 upon “The Roman 
Element in English Law,” I pointed out the following facts, viz.: (1) That these 
elements which English law has in common with Roman law constitute more than 
four-fifths of the elementary legal thought entering into the composition of English 
law; (2) that the words by which these elementary legal conceptions are ex- 
pressed in English law are of Roman origin, and were directly borrowed from the 
Latin language; and (3) that there are no native English words in existence by 


- 


XXxil American Philological Association. 


which to express these ideas. From these facts the conclusion is that these legal 
ideas in English law were not developed there independently, but were, on the 
contrary, borrowed from Roman law. If this-conclusion is correct, the presence 
of these elements in English law becomes much less significant in comparative 
jurisprudence, but much more significant in determining the essential nature of 
English law. In the light of these philological evidences the accepted theory 
of English law alluded to must be abandoned or profoundly modified. Native 
English custom could not grow up and find its first and only expression in words 
of a foreign language. These words could not in the nature of things have been 
borrowed by the people at large for the purpose of giving expression, to their own 
existing customs. A custom presupposes current ideas held in common, and ideas 
cannot become current until they are clothed in words. Native custom having 
the force of law therefore could not exist without being expressed in native words; 
and if it once existed and was so expressed, it is impossible to conceive how the 
native words could have not only given way to foreign words, but also entirely 
disappeared. 

By whom, then, were these Roman words borrowed? and how did they find 
their way into English law? To this question an examination of the words them- 
selves will furnish an answer; and it will, at the same time, inferentially disclose 
the true nature of English law. Appended hereto are groups of words, belonging 
to the five great branches of law, which exhibit the nature, proportion, and impor- 
tance of the Roman and native elements respectively. From these it will be seen 
that the Roman element greatly predominates everywhere, but especially in the 
law of procedure, which is practically all Roman in its elementary thought. It is 
well known that the law of procedure, or adjective law, precedes substantive law 
in point of time; that substantive law is produced by the workings of adjective 
law in the practical administration of justice. When, therefore, we find the 
organic and the adjective law of the English so completely dominated by Roman 
thought as it is shown to be in the first and second groups, it is evident that 
English national life and its attendant national administration of justice, though 
retaining necessarily many English characteristics, were from the outset developed 
very largely upon lines marked out by the Romans, and with the free and constant 
use of Roman thought. And it is also evident that these Roman words were 
borrowed by those who were in authority during the early stages of English 
national life, those upon whom devolved the task of organizing the nascent 
nation, and supplying it with workable machinery for the orderly administration 
of justice. Those who did this were chiefly the king and his counsellors, the 
judges and lawyers. They borrowed the words and the thoughts because they 
needed them and as they needed them, and they put them to immediate 
practical use. English society was feeling its way to a vigorous and compact 
political unity, and those who were directing its energies learned from the Romans, 
and utilized, the means by which Roman political unity had been accomplished. 

In the three remaining groups of words we find constant confirmation of this, 
and also abundant evidences of its effect upon English law. In the criminal law, 
the law of property, and that of contract, we find that the line of contact between 
English life and English governmental control is coincident with the line of 
contact between English and Roman thought. 

English legal thought is primarily Roman thought super-imposed upon English 


Proceedings for July, 1893. XXXili 


non-legal thought, to meet the necessities of government. And the resulting 
English law is not a mass of customs having the force of law through long con- 
tinuance, but rather a great body of rules developed by English judges and lawyers 
in the course of the actual administration of justice. This body of rules is what 
is known as English Case Law, and its authority rests, not upon custom, but upon 
what may be called the concensus of public opinion as expounded and applied by 
the courts in the administration of justice. It is not directly connected with cus- 
tom, and is generally affected by custom only in so far as existing customs have 
entered into the formation of public opinion. 

The conclusion thus indicated by the philological evidences is important not 
only to the theory of English law, but to its practical application. If it shall come 
to be accepted in lieu of the theory hitherto prevailing, it will help to solve some 
perplexing problems in English legal history, and materially aid in the advance- 
ment of legal science. 

Groups of words exhibiting and contrasting the Roman and native ele- 
ments : — 


‘ 


1. Organic and International Law. 


Advocate, Alien, Allegiance, Ambassador, Amnesty, Attorney, Appropriation, 
Arbitration, Chancellor, Circuit, Citizen, Code, Constitution, Comity, Committee, 
Compact, Congress, Constable, Coroner, County, Court, Democracy, Denizen, 
Deputy, District, Domain, Edict, Enfranchise, Exchequer, Excise, Exequatur, 
Expatriation, Extradition, Forum, Franchise, Function, Funding, Government, 
Heptarchy, Herald, Inauguration, Interdict, International, Intervention, Legation, 
Legislate, Loyal, Magistrate, Majority, Mandatory, Mandate, Manifesto, Mayor, 
Mediation, Minister, Mob, Monarchy, Municipal, Mutiny, Nation, Nobility, Nisi 
Prius, Nominate, Notary, Office, Officer, Oligarchy, Opinion, Ordain, Panel, Par- 
don, Parish, Parliament, Peace, Pension, People, Political, Posse, Precinct, Presi- 
dent, Prize, Proclamation, Province, Public, Quarantine, Quorum, Rebellion, 
Record, Renunciation, Repeal, Republic, Repudiate, Resident, Resignation, 
Respite, Revenue, Revolt, Rogatory, Royal, Sedition, Senate, Sine die, Society, 
Solicitor, Sovereign, Sovereignty, State, Status, Statute, Subsidy, Suffrage, Sumptu- 
ary, Superior, Supremacy, Supreme, Surrogate, Tariff, Tax, Term, Territory, Traitor, 
Treaty, Tribunal, Ultimatum, Unconstitutional, University, Usurpation, Vacancy, 
Vassal, Veto, Vicinage, Viscount, Visne, Vote. Borough, Domboc, Earl, Folk- 
gemote, Gemote, Gerefa, Hundred, King, Lord, Mark, Queen, Reeve, Shire-gemote, 
Sheriff, Thane, Town, Town-ship, Witenagemote, Wood-mote. 


2. The Law of Procedure. 


Pleading. — Abatement, Action, Amendment, Allegation, Assumpsit, Aver- 
ment, Certainty, Covenant, Declaration, Debt, Demurrer, Departure, Dilatory, 
Disclaimer, Ejectment, Inducement, Intendment, Issue, Joinder, Multifarious, 
Non-joinder, Pleading, Rejoinder, Repleader, Replication, Similiter, Surplusage, 
Surrebutter, Surrejoinder, Trespass, Traverse, Variance, Verification, Versus, Vi 
etarmis. Practice.— Appeal, Appearance, Application, Attachment, Case, Cause, 
Certiorari, Challenge, Citation, Client, Compurgator, Continuance, Decree, Defense, 
Defendant, Detinet, Detinue, Discontinuance, Docket, Elegit, Elisor, Engross, 
Enjoin, Enroll, Execution, Garnishee, Habeas corpus, Impanel, Imparlance, In- 


- 


XXXIV American Philological Association. 


junction, Inquest, Inrollment, Interlocutory, Interpleader, Intervention, Judgment, 
Levari, Levy, Lis pendens, Motion, Ne exeat, Non-suit, Order, Oyer, Practice, 
Precept, Pro confesso, Proceeding, Procedure, Process, Profert, Prohibition, Pro- 
visional, Quash, Recoupment, Regular, Remedy, Replevin, Respondent, Retainer, 
Return, Reversal, Review, Revival, Revivor, Scire facias, Sequester, Severance, 
Subpeena, Suggestion, Suit, Summary, Summon, Supersedeas, Supplemental, Sur- 
charge, Surprise, Temporary, Transcript, Transitory, Trial, Triors, Trover, Vendi- 
tioni Exponas, Venire, Venue, Verdict, Voir-dire, Answer, Forswear, Oath, 
Set-off, Speaking, Wager of Battel, Wager of Law. 


3. Criminal Law. 


Accessory, Adultery, Amercement, Arrest, Arson, Assault, Attainder, Battery, 
Burglary, Bigamy, Bribery, Capital, Carnal, Champerty, Conviction, “Crime, 
Defense, Embezzlement, Embracery, Flagrante delicto, Forgery, Fugitive, 
Homicide, Hue and cry, Suicide, Treason, Impeachment, Imprisonment, Indict- 
ment, Infanticide, Information, Innocent, Larceny, Maintenance, Penal, Penalty, 
Penitentiary, Perjury, Pillory, Piracy, Premeditation, Prosecution, Provocation, 
Punishment, Recrimination, Rescue, Reprieve, Reward, Robbery, Sentence, Solic- 
itation, Vagrant. Blackmail, Breaking, Guilt, Manslaughter, Mayhem, Murder, 
Outlaw, Steal, Theft, Thief. 


4. The Law of -Real Property. 


Accession, Adverse, Alienation, Amotion, Appendant, Apportionment, Benefi- 
ciary, Close, Common, Courtesy, Conveyance, Coparcenary, Copyhold, Curtilage, 
Dedication, Defeasance, Deforcement, Demesne, Demise, Descent, Detainer, 
Detention, Devastation, Devastavit, Devise, Devisee, Disinherit, Disseizin, Dis- 
tribution, Disturbance, Dower, Emblements, Eminent domain, Encroach, Entail, 
Entry, Enure, Escheat, Escuage, Estate, Eviction, Extinguishment, Feoffment, 
Feud, Fine, Fixture, Formedon, Habendum, Heir, Hereditaments, Heritage, 
Homage, Immovable, Inalienable, Inchoate, Incorporeal, Incumbrance, In esse, 
In fieri, Intrusion, Irrigation, Lease, Livery, Merger, Messuage, Metes, Mortgage, 
Mortmain, Occupancy, Perpetuity, Possession, Pre-emption, Quit, Real, Recovery, 
Re-entry, Release, Remainder, Rent, Resulting, Reversion, Riparian, Scutage, 
Seizin, Servitude, Severalty, Succession, Tenancy, Tenant, Tenement, Tenure, 
Terre-tenant, Title, Turbary, Vacant, Vadium-vivum. Aackwater, Betterments, 
Bote, Building, Burgage, Deed, Dwelling, Fardel, Farm, Fee, Folkland, Free- 
hold, Gavelkind, Gift, Glebe, Ground, Grant, Haybote, Hidage, Hedge-bote, 
Hothbote, Homestead, House, Hudegeld, Land, Landlord, Socage, Squatter, 
Things, Thainsland, Waste, Warren. 


5. The Law of Contract. 


Acceptance, Accord, Account, Agreement, Bailment, Bargain, Barter, Bill, 
Bilateral, Bona fide, Bonus, Broker, Charter, Cheat, Cognation, Collateral, Com- 
modatum, Commutation, Compact, Concurrent, Condition, Confirmation, Consent, 
Consideration, Consolidation, Construction, Contract, Contribution, Covenant, 
Cy pres, Damages, Debenture, Debt, Deceit, Defalcation, Default, Delivery, De- 
mand, Deposit, Discharge, Due, Express, Execute, Factor, Foreclosure, Guarantee, 
Hypothecation, Implied, Inception, Indebtedness, Indemnity, Inducement, Inter- 


Proceedings for July, 1893. XXXV 


est, Interpretation, Laches, Liquidate, Locatio, Marriage, Maturity, Misrepresen- 
tation, Mutual, Negligence, Negotiable, Note, Novation, Obligation, Obligor, 
Offer, Option, Partner, Party, Payment, Performance, Pignus, Pledge, Policy, 
Precedent, Premium, Principal, Privity, Promise, Protest, Provision, Purchase, 
Quid pro quo, Ratification, Reciprocal, Recognizance, Re-insure, Relation, Re- 
scission, Respondentia, Risk, Salary, Satisfaction, Security, Simple, Solvent, 
Special, Stipulation, Suppressio veri, Supra protest, Surety, Tender, Ultra vires, 
Unilateral, Usury, Valuable, Vendue, Waive, Warranty. Bearer, Bond, Borrow, 
Bottomry, Breach, Drawer, Holder, Loan, Maker, Sale, Seller, Settlement, Sight, 
Warehouse. ' 


11. Professor F. A. March, of Lafayette College, reported as 
Chairman of the Committee on Spelling Reform. 


He began with an outline of the action of the Association on this subject. 
At the annual meeting in 1875, President J. Hammond Trumbull suggested in 
his address that a list of words exhibiting side by side the present and a reformd 
spelling, such as prominent scolars in England and America would recognize as 
allowabl, would go far towards insuring the success of the reform. A committee 
was appointed to consider the subject, and prepare and print such a list if they 
thought best. The Committee wer Professors W. D. Whitney and J. Ham- 
mond Trumbull, of Yale; Professor F. J. Child, of Harvard; Professor F. A. 
March, of Lafayette; and Professor S. S. Haldeman, of the University of Pennsyl-. 
vania. At the next meeting, in 1876, the Committee reported the general prin- 
cipls which should guvern attempts to amend orthografy. In 1877 it made an 
application of these principls to English, and stated the prevailing sounds of the 
letters in English, and so gave an alfabet with constant sounds for each letter 
such as it is hoped may finally be establisht in use. In 1878 a beginning of the 
list was made with eleven words. In 1880 the president of the Philological 
Society (London), Dr. Murray, brought the matter before the scolars of England. 
The Philological Society discust amendments for six meetings, and adopted a 
body of them, January 28, 1881, which wer issued by the Society in a pamflet. 
This was discust in the American Philological Association the same year. In 
1882, at the suggestion of the English Society, communications wer opend be- 
tween the Societies, to effect a complete agreement, so that “a joint scheme _ 
might be put forth under the authority of the two chief filological bodies of the 
English-speaking world.” In 1883 this complete agreement was effected, and a 
scheme of partial reform was jointly approved, and recommended for immediate 
use. It was in the form of comments on the letters, mentioning with each letter 
when it should be dropt or changed, sumtimes mentioning particular words, 
sumtimes giving general rules with limitations and exceptions. It was accom- 
panied by an alfabetic list of sum three hundred and fifty words. In 1886 the 
list, enlarged to sum thirty-five hundred words, was presented to this Association 
and printed in the Transactions of that year. The corrections ar made in the 
interest of etymological and historical truth, and confined to words which the 
changes do not much disguize from the general reader. Other reports hav been 
litl more than reports of progress from year to year. The list was reprinted 
in 1887 by the Spelling Reform Association, in 1891 in the Century Dictionary, 


- 


XXXVi American Philological Association. 


in 1893 by the U. S. Bureau of Education. The Committee has taken no official 
action during the last year. It reports progress. The Modern Language 
Association of America, at their annual meeting last December, adopted a 
resolution uniting with the Philological Societies in “recommending the joint 
rules for amended spelling and the alfabetical list of amended words publisht 
in the Transactions of the American Association and in the Century Dictionary.” 
The Anthropological Society of Washington held a symposium on the question, 
“Is simplified spelling feasible as proposed by the English and American Philo- 
logical Societies?” It was continued for three sessions, and participated in by 
Hon. W. T. Harris, U. S. Commissioner of Education; A. K. Spofford, Librarian 
of Congress; J. W. Powell, Director of the U.S. Geological Survey; Alexander 
Melville Bell, William Dwight Whitney, Professors W. B. Owen, C. P. G. Scott, 
and others. The speeches and papers wer publisht, several of them spelt ac- 
cording to the rules and list, in the American Anthropologist for April, 1893. 
The Hon. William Mutchler, representativ from Pennsylvania, moved as an 
amendment to the House Bill on Printing, that the public printer be instructed, 
whenever variant spellings ar found in the current dictionaries, to use the simplest. 
The amendment was adopted. The Bill has not yet passed. Under the influence 
of the Association Phonetique des Professeurs de Langues Vivantes, and of 
Volapiik, the hed-quarters of both of which ar at Paris, there is now much 
discussion of French reform. M. Paul Passy, Professor G. Paris, Professor A. 
Darmesteter, M. Bréal, and many other prominent teachers and linguists ar 
taking part. Permission has been obtaind to try fonetic teaching. in certain 
scools, and the reformers ar very hopeful and activ. 


The report was accepted, and the Committee continued. It now 
consists of Professors March (Chairman), Child, Lounsbury, Price, 
Trumbull, and Whitney. 

Remarks were made by Professors Sproull, Hale, and March, and 
by several gentlemen in the audience. 


12. English words which hav Gaind or Lost an Initial Consonant 
by Attraction, by Charles P. G. Scott, Yonkers, N. Y. 
This paper is printed in the TRANSACTIONS. 


13. The Hebrew Names in the English Bible, by Professor W. O. 
Sproull, of the University of Cincinnati. 

Professor Sproull reported briefly upon the work done upon this 
paper. 

Adjourned about 5.30 P.M. 


Cuicaco, July 13, 1893. 


At ro a.m. the Association assembled with President Hale in the 
chair. The reading of papers was at once begun. 


Proceedings for July, 1893. XXXVii 


14. Critical Notes on Certain Passages in Sophocles’ Philoctetes 
and Antigone, by M. L. Earle, Ph.D., Barnard College, New York 
City. 

I. PHILOCTETES. 
Vv. 43 sq. GAN 7 al hops voorov éeAHAvOev 


] pbdov ef Te vSdvvov Kdrodé Tov. 


The traditional text of v. 43 is quite out of the question, nor does any one 
of the emendations hitherto proposed (so far as they are known to me) seem 
to restore the manus Sophoclea. This was, I conjecture, d\N 7 él gopBh- 
NECTIN. éeAnr\vOWC (dopBiv éorw éfeXndrAvOds). The ease with which 
this could corrupt into the traditional form needs no comment. 


Vv. 54sqq. NE. rh dr dvwyas; OA. rhv didoxrhrov oe Set 
Yux hy brws Adyourtw éxxrAéVers A€yov. 
bravo’ épwrdi ris re kal wdébev wdpet, 
Aéyerr "AxiAdéws wats * 753° obxl Krerréov: 


The syntax of these verses as they stand is very dubious. But I do not think 
it is to be helped adequately by changing Aéywv to an inf. (e.g., cxowetv). A 
simple solution of the difficulty may, I think, be found, if we stick closely to the 
connection of the speech of Odysseus with that of Neoptolemus, observing also 
the opening of the speech of Odysseus. Neoptolemus says ti 547’ dvwyas; 
*“ What then do you direct?” The natural answer to such a question is an oblique 
form of expression = imv. of oratio recta. This is, of course, in the case in ques- 
tion, the infin. Note now that the preceding speech of Odysseus, which Neop- 
tolemus does not regard as imperative, has the de? construction; and, further, that 
in v. 57 \éyecv would fall in much more naturally as second member of a com- 
pound infin. (= imv. in oratio recta) structure. In fine, I would read (adopting 
Gedike’s S6Aovcw for Adyourw in v. 55, a conjecture that might occur to any one 
as it had to me independently, and Nauck’s suggestion Srav 8’ in v. 56) as 
follows: : 


‘ 


Tl bAr dvwyas ; —Thv PiroxrHrov cKotretv 

Wuxny drws Sdrororw éxxrépers Adyov’ 

brav 8 épwrak ris re kal whbev rape, 

Aévyerr "Axiddéws wais* 745 obx) Kpumréov * 
(xpurréov Nauck for xXerréov). Az. 556 sq., cited by Professor Jebb in support 
of the construction de? 8rws c. fut. ind., admits of easy correction by substituting 
okorretv Straws for o Stws watpds at the close of v. 556. 

V. 567. as radr’ érlorw Spwper’, ob wéddovt’ Ert. 

Though this verse is undoubtedly corrupt, it does not appear that either Nauck 
or Blaydes has been successful in his conjecturing,—the former reading od for 
ws; the latter, Ws Spdyev toh trabra, or rar’ éteriorw Spwpev. The last, how- 
ever, comes nearest to what Sophocles seems to have written, viz., «3 tair’ émi- 
orw Spwpev’: cf. Electr. 616 e} vv érlorw rdvie w aloxvvnv exev. The 


corruption may easily have arisen from contamination of ans C (568) —such 


contamination being a fruitful source of error in the Sophoclean text; or perhaps 
it may be due to v. 563 ws éx Blas xré. 


+ 


XXXViil American Philological Association. 


Vv. 900 sqq- P|. od 5% ce Sve xépera Tod vorhuaros 
trewev wore uh Ww dye vavrny er; 
NE. dravra dvoxéepeca, Thy abrod ptow 
6rav uray Tis Spat TA WH Tpocekdra. 
Pl. adAdX dev Ew rod Gutevoavros cv ye 
Spais ovd5e Gwvreis, EoOddv dvdp erwHeddr. 

A similar case of contamination to that just suggested is to be detected, I think, 
in v. 901, where instead of vabrnv rt we should read weds émt. The source of 
corruption here was AY TOY in v. 902, probably aided by NOCHMATOC 
in v. goo. Again in y. 904 there. has been a somewhat similar degeneration. 
OYAEN owes its origin in part to OYAE in v. 905. But, whatever were the 
details of the process, the original form of the verse I venture to think was this: 


d\n” OYZENONTI (708 durevoarros ov ye 
(= ob gévoy Tt). 
tévor c. gen. = alienum-a c. abl. can be supported by O. T. 219 sq.: a *ya& tévos 
peév Tod Névyou Todd éfepG, | Eévos dé Tod mpaxGévros, It also falls in most aptly as 
a retort to Ta wh mpoceckéra in v. 903. 

V. 917. ofuo, tt elas; Valckenaer’s ri yw’ elwas; is, of course, out of the 
question here. ti 7’ elas of B is, as Jebb says, “weak.” But why may not the 
phrase ré elas ; (common to-day, as always) have taken the place of a less com- 
mon equivalent? I would suggest rf daveis; as in Electr. 1349. 


V. 991. @ pivos, ola xdtaveuploxers éyeuv. 


Aéyev certainly strikes one oddly. Jebb suggests that it should be Aéyw». 
That seems hardly likely, however, with KPATUWN at the close of v. 989 and 
A€lW at that of 990. Perhaps we should rather write ETTH. There are 
several ways in which AE EIN might have come into the text. 


II. ANTIGONE. . 


Vv. 82 sqq. IC. otpor radralyns, ds brepdédouxd cov. 
AN. ui’ nod rpordpBe* tov cdv efdpPov mérpor. 
IC. GX ody rpopnvions ye TodTO undevi 
Toupyov, Kpuph dé xede, odv & atrws éya. 
AN. otpor, xaravéa+ réddov éxbiwv Zone 
ovy&o’, éav wh waor xnpvinus Tdde. 


In y. 86 ofuor has long seemed to me wrong. Jebb slides over it in his com- 
mentary. In his translation he gives “Oh, denounce it!” which may be natural 
English, but does not adequately represent the Greek. The Schneidewin-Nauck 
commentary gives nothing; nor is Blaydes’s note (‘‘ An exclamation here of indig- 
nation. Cf. 320”) of much more service. Professor Humphreys offers us at 
least something more when he annotates ofuo: thus: “Of disapproval or dis- 
satisfaction. So even ofuo xaxodaluwy, of rage, Ar. Av. 1051. In id. 1260 olor 

*rddas may be ironical.” But even this is unsatisfactory; the presence of xaxodal- 
wv and Tddas after ofuo in the Aristophanic citations robs them of all apposite- 
ness, and we are left no better off than before. Let us confess it frankly, this is 
a case for emendation, not explanation. Why should we not restore the vigorous 


Proceedings for July, 1893. XXXiX 


and apt pf pot (perhaps better written here uh ‘uol or uh éuol)? The source of 


th ‘uption is not far to seek, being contained in OIMOI above. A scribe 
a - MHMOY 


was quite capable of jumbling the two neighboring passages in such a way as to 
produce the present state of affairs in the text. In further support of this emenda- 
tion may be compared vv. 544-7: 


IC. ph ro, cacvyvhrn, w adrindoms 7d uh ob 
Oavety re cdv col Tov Oavévra & ayvioa. 

AN. ph wot Odvms od Kowwd, und & wh Ovyes 
tow ceauTAs* apxécw OviirKova’ eyo. 


Here, though there is no ellipsis, there is yet a striking similarity in the tone of 
harsh refusal and repulsion. 


Vv. 404 sq. Tavrnv y i80v O4eroveay by od Tov veKpov 
dretras. 


Every one feels the harshness of the position of rév vexpév. The words may be 
sound, but they look like a gloss. Perhaps they may have taken the place of 
TovTo Spav. 

Vv. 417 sq. kal rér’ étalgyns xOovds 
tupads delpas oxnrréy, oipaviov &xos. 


ovpdviov &xos has given trouble for more reasons than one. I would suggest 
&x 80s odpavod as possibly the original form of the words. If &y@os became dxos, 
transposition and a change from gen. to adj. might follow. - 


Vv. 478 sq. ob yap éxéXer 
gpovety péy’ Satis dodNbs éore THv wédas. 


So dubious a word as éxré\ec is in more than suspicious company when it 
stands over wéAas. There has doubtless been contamination between the ends 
of vv. 478 and 479. Blaydes writes: “Qu. od yap odv mpére (or wéXec).” The 
former is nearer what I believe Sophocles wrote; viz., edmpewés. (Cf. Class. 
Rev, VIL., p. 344.) 


Vv. 1001 sq. déeyvisr’ dxovw Pbbyyov dprvlOwv, kaxde 
KAdfovras olorpux kat BeBapBapwpévar. 


We read smoothly enough through oforpw xai; but after the cal we get a 
mental jolt. We are all ready for another attribute to the épudas implied in 
Pbbyyov épvidwy, when we have an attribute to oforpux suddenly thrust upon us. 
Here again I feel sure there has been contamination, an original BeBapBapwpévous 
= doadgeis having been assimilated to xax& above it. The loci classici for the 
‘barbarism’ of birds (Hdt. 2, 57; Ar. Av. 199; Aesch. Ag. 1050 sq.) are also 
in favour of the reading proposed. (In the very similar passage, Eur. Ac. 
777, we should follow Nauck’s suggestion [Zur. Studd. I1., p. 85], and read 


cuvadpvepévos [Evy-] for cvvadpvepévar. ) 


This. paper, which in the absence of its author was read by 
Professor D’Ooge, was commented on by Professors Shorey, Gilder- 
sleeve, Smyth, D’Ooge, and Humphreys. 


- 


xl American Philological Association. 


15. Some Suggestions Derived from a Comparison of the Histories 
of Thucydides and Procopius, by Dr. W. H. Parks, City of Creede, 
Colorado. 


This is merely a general discussion of the subject, preparatory to a more 
detailed treatment of the grammatical side at a future meeting. 

There is always something fascinating to the human mind in comparing the 
literary work of two men, one of whom has, either consciously or unconsciously, 
been influenced by the other. And if this statement be true in general, how much 
more so in the case which we have before us. Both writers participated, to 
some extent, in the events of which they wrote; but how different the circum- 
stances under which they lived! 

Let us pause for a moment to contemplate the contrast of light and Shadow 
afforded by this picture. Thus we find Thucydides living in the very atmosphere 
of freedom, in the springtime of the world’s life and thought. We behold Proco- 
pius, on the other hand, living at a time when no one dared to call his life or 
even his thoughts his own, and when spies lurked in the innermost recesses of a 
man’s household. With these changes in the political world had also come 
changes in the language, the religion, and the national life of both the Greeks 
and the Romans. 

But in investigating this matter of imitation, a considerable degree of caution 
should be observed, it being almost an axiom in logic that, of two or more phe- 
nomena, each may be derived from a common source, as well as one from another. 
To give one from the infinite number of examples of this principle, the story of 
“Puss in Boots” is found in so many languages, and among such widely separated 
peoples, that to extract its origin and the details of its subsequent progress from 
the evidence which we have at hand is well-nigh impossible. 

Further, we must constantly keep in mind the fact that Procopius’ history is 
sensibly affected by his imitation of Herodotus. 

Before proceeding to a further discussion of the subject, it may be well to 
refresh the memory of our readers by the barest outline of the life and works of 
Procopius. 

Procopius of Caesarea was born in Palestine, about 500 A.D. In early life he 
went to Byzantium, where he was made secretary of the great general Belisarius. 
This event coincided in time with the accession of the famous Justinian, whose 
death in 565 A.D. must have been nearly coincident with that of Procopius. His 
works consist of somewhat full descriptions of the three great wars of Justinian 
(the Persian, Vandalic, and Gothic), besides a work on the Edifices of Justinian 
(Ilepi Kricudrwv) and the so-called *Avéxdora or Secret History, which adminis- 
ters a scathing rebuke to the avarice of the emperor and the early vices of his 
empress, Theodora. The last-mentioned work is generally explained on the 
supposition that Procopius became offended at various slights received from his 
imperial patron. 

We shall consider the resemblances and differences between our two authors 
from five standpoints. 

I. HistoricaL.—The beginning of Procopius’ Persian War is very similar to 
the beginning of Thucydides’ History; but Procopius is more diffuse, and the gen- 
eral plan of his work bears traces of conscious imitation of Thucydides, notably 


Proceedings for July, 1893. xli 


in the chronological arrangement of his campaigns. But he is of course decidedly 
inferior as an historian to his predecessor. In general, he imitated him in form 
rather than in substance, especially in copying more or less faithfully his stock- 
phrases. In matters of accuracy and impartiality, he resembles Herodotus more 
than he does Thucydides. He takes sides, for instance, in the factions of the 
circus and in religious dissensions, particularly in the Historia Arcana. — 

In the matter of gossip, Procopius is far inferior to Thucydides. We may, 
indeed, say that the times in which the former lived were much more favorable 
to the Homeric ¢7uy. But let us consider what a scandal Thucydides might 
have made of the intercourse of Socrates and Pericles with Aspasia, or how he 
could have revelled in the midnight escapades of Acibiades! 

II. LirERARY.— We find perhaps less difference between our authors under 
this head than under the former. As Gibbon has said: “ Procopius at times 
rivals the strength and even the elegance of the Attic historian.” But it is only 
at times. For, while Thucydides generally subordinates manner to matter, we_ 
find- many noble sentiments in his pages (cf. II. 43, dvép@v ... évduarGrar). And 
his noticeable harshness of diction would undoubtedly have disappeared had he 
written at a more finished period of the language, to such an extent are the 
strongest characters influenced by their surroundings. Now Procopius had all 
of Thucydides’ roughness, unredeemed by many of his better qualities. Thus, 
instead of the periodic arrangement, he often jumbles the members of a periodic 
sentence in well-nigh inextricable confusion (cf. Th. I. 143, xal éxl 7@ kuvdivy . 
ovvaywviferbar with Pr. Bell. Vand., p. 103, 21 (Hoeschel’s Fanon, ovx Sov 
movets ... NaBety, K.T.d.). 

Thucydides? transitions, too, are often elegant, resembling those of Pindar (see 
I. 23), while Procopius wearies one with his continual womep pou Aedékerar od © 
WOAAG vorTepoy, K.T.r. 

Procopius often imitates Thucydides quite closely in forms of expression, nota- 
bly at the beginning and end of paragraphs (note dua dé Hpe dpxouévw) passim. 

In conclusion, while Procopius can by no means equal the vivid pictures of 
Thucydides (cf. the plague, Th. II. 47-54), yet some of his narrations are decidedly 
striking, such as his account of the self-abasement of Belisarius (H. A. ch. IV.). 

III. GRAMMATICAL. — As we hope, at some future time, to discuss this part 
of our subject more thoroughly, we shall omit it from this abstract for want of 
space, merely pausing to remark that, in this regard, Procopius imitates Thucydi- 
des more in phrases than in constructions. 

IV. MISCELLANEOUS. —1. Zhe Plague. While Procopius is generally consid- 
ered to have been a lawyer by profession, such eminent authorities as some of the 
French medical dictionaries include him among the physicians, basing their claim 
on the fact that he describes, in B. P. II. 22, 23, the plague which devastated 
Byzantium in 543 A.D. with much minuteness and scientific method. But, in real- 
ity, his description bears a striking resemblance to Thucydides’ account of a 
similar calamity which befell Athens in the spring of 430°B.c. This resemblance 
consists not only in style and forms of expression, but also in the more important 
matters of the origin and character of the disease. 

2. Speeches. These form a prominent part of the works of both historians. 
Procopius undoubtedly copied this feature from Thucydides, as the latter got it 
from Herodotus. Now the Byzantine imitates many of the characteristic features 


- 


xlii American Philological Association. 


of the speeches of the Athenian, such as the opposition of pairs of speeches, the 
use of pithy proverbs, and of similar forms of expression. But there seems to be 
more variety in the speeches of Thucydides, in spite of their general sameness of 
character (cf. the funeral oration of Pericles, II. 35-46, with the speech of the 
Spartan ephor, I. 86). 

V. ConcLusion. — We may say in summing up that, the more one studies 
Thucydides, the more one finds to admire, the more depth one discerns in his 
pages, while the greater part of Procopius’ excellence is apparent on the surface. 
But when we consider the circumstances under which the latter wrote, the artifi- 
cial character of the language, the blighting influence both of an effete civilization 
and a despotic system of government, we are constrained to wonder that he did 
as well as he did. 


The Committee to audit the Treasurer’s Report announced that 
they had compared it with the vouchers and found it to be correct. 
The Committee on Officers for 1893-94 reported as follows : — 


President, Professor James M. Garnett, of the University of Virginia. 
Vice-Presidents, Professor John H. Wright, of Harvard University. 
Professor Bernadotte Perrin, of Yale University. 

Secretary, Professor Herbert Weir Smyth, of Bryn Mawr College. 
Treasurer, Professor Herbert Weir Smyth, of Bryn Mawr College. 
Additional members of the Executive Committee. 

Professor B. L. Gildersleeve, of Johns Hopkins University. 

Professor W. W. Goodwin, of Harvard University. 

Professor Abby Leach, of Vassar College. 

Professor F. A. March, of Lafayette College. 

Professor W. D. Whitney, of Yale University. 


The report was adopted. 

The Committee on Place of Meeting in 1894 reported, through 
Professor Goodell, that the invitation of Williams College be accepted. 
The report was adopted. The next annual session of the ASSOCIATION 
will therefore be held at Williamstown, Mass., beginning July 10, 1894. 

Considerable discussion then ensued as to the feasibility of holding 
a session every other year in conjunction with the AMERICAN ORIENTAL 
Society, the MopEerN LANGUAGE AsSOCIATION, the D1aLecr Society, 
and other kindred organizations." Upon the motion of Professor 
D’Ooge, it was resolved that the Executive Committee be requested 
to ascertain whether it is feasible to hold such a joint meeting. 


16. Vedic Studies, by Professor M. Bloomfield, of the Johns 
Hopkins University. 

This paper will appear in the Journal of the American Oriental 
Society, XVI. 1-42. 


1 See PRocEEDINGS for 1892, p. xi. 


Proceedings for July, 1893. xliii 


17. ‘Extended’ and ‘ Remote’ Deliberatives in Greek, by Professor 
W. G. Hale, of the University of Chicago. 

This paper is printed in full in the Transactions. Remarks were 
nade by Professors Sonnenschein, Gildersleeve, Shorey, D’Ooge, and 
»y Drs. Parks and Miller; and in reply by Professor Hale. 


Professor Sonnenschein contributed a new instance of the subjunctive (Arist. 
<nights 1320: rly exw Phunv dyabhy axes, ep bry kuoGpyev dyuids;) and 
called attention to the ultimate identity of the final and the deliberative subjunc- 
ive, both being forms of “ will-speech” and developments of the subj. of “ com- 
nand,.” The question is therefore not one of choosing between two different 
cinds of subj., but of deciding by what road the unusual subj. of the relative 
slause made its way into Greek, The subj. of Aesch. Prom. 471 ox éxw obgiop’ 
irw | Tis viv wapovcns mnuovas draddayG is ultimately the same subj, as that of 
he Latin 2on habeo artificium quo liberer, though it was probably developed in 
. different manner. He agreed with Sidgwick and Professor Hale that the imme- 
liate source of the Greek subj. of the relat. clause was to be found in the delib- 
srative question (the command-question), whereas the origin of the Latin guz 
vith subj. may be seen in such an instance as Plaut. Rudens 1329 guo nihil invi- 
us addas, talentum magnum, ‘a great talent, to which you are not ¢o (need not) 
idd anything against your will,’ — an instance which. is not final in the ordinary 
ense: cf. the Plautine eas ‘go,’ ne eas ‘go not,’ and oicg@ ody 6 Spacov,—In 
egard to the immediate origin of the optat. in Soph. Phil. 281 (doris dpxéoese) 
ind other instances dependent on a past tense, there is the difficulty that the 
Greek did not use the optat. in zzdependent questions as to what ‘was to be 
lone’: mot Tis pvyor; means not guo fugeres? but guo fugias? Thus @or’ ody 
irws” Adknoris és yhpas wddo.; (Eur. Alc. 52) is easier to account for as of delib- 
erative origin than Soph. Phil. 281. Still it is quite possible that when the use of 
he subj. with the relat. had become established, the use of the optat. in depend- 
ence on a past tense arose by way of an adjustment (cf. héyw tva pdOys, édeyor 
‘va pdOo.s). 


At 1 o’clock the AssocraTIon adjourned. 


AFTERNOON SESSION. 


The reading of papers was taken up at 4 P.M., with President Hale 
n the chair. 


18. Fastgium in Pliny, H. N. 35, 152, by Professor Harold N. 
Fowler, of Western Reserve University (College for Women), Cleve- 
land, O, 


Pliny’s words are “ Butadis inventum est rubricam addere aut ex rubra creta 
ingere, primusque personas tegularum extremis imbricibus imposuit, quae inter 
initia prostypa vocavit, postea idem ectypa fecit. hinc et fastigia templorum 
orta.” 


xliv American Philological Association. 


By comparison of passages in which the word /fastigium is applied to any part 
of a building, the result is reached that when used accurately (i.e. not in the gen- 
eral sense of roof or slope or top), it means the sloping cornice. Nearly all these 
passages are in Vitruvius. Incidentally it appears that corona is used by Vitru- 
vius to designate a cornice exclusive of the sima, and is therefore frequently used 
of the horizontal cornice. ‘ 

When /astigium is used of sculptural adornment it applies to acroteria, not to 
pediment groups. This is most clearly shown by comparison of Sueton. Div. 
Jul. 81, Calpurnia uxor imaginata est conlabi fastigium domus, with Plutarch, v. 
Caes. 63, 738, Av yap Tt TH Kaloapos oixig mpockeluevoy .. . dkpwrhpiov ..., TOTO 
bvap 7 Kadrovupria beacauévn xarappnyvipevoy @d0te mormacbar Kal daxpvev. It 
is therefore evident that the passage of Pliny cannot be used as an argument for 
the origin of pediment sculptures from terra cottas. (This paper has been 
published in full in the American Journal of Archaeology, Vol. VIII. (1893), 


p- 381 ff.) 


19: On Some Greek Inscribed Wax Tablets in the University 
Library at Leyden, by Professor Harold N. Fowler, of Western 
Reserve University (College for Women), Cleveland, O. 


The University Library at Leyden has recently received from Mr. A. D. Van 
Assendelft de Coningh, Burgomaster of Leidendorp, seven Greek wax tablets, : 
bought in Palmyra in 1881 by the late Lieutenant H. Van Assendelft de Coningh. 
These tablets seem to have formed a little book. On the outside of the first tablet 
is the line (Hesiod op. 347) Eumopé ror requi(s) bs 7° Eupope yelrovos alcOdod. 
On the other tablets are fables of Babrios. The photographs shown at the meet- 
ing represented the first and the fourth tablets. On the fourth is Babrios fad. 117 
(Schneidewin, Lewis, Gitlbauer) and the beginning of fad.91. The writer evi- 
dently intended to give the exact wording of Babrios, but numerous variations 
occur, some of which show a marked disregard of metre. This may be because 
the writer was unable to appreciate quantitative metre. The chirography and 
orthography are those of an early period,— possibly not later than the second 
century after Christ, —and the tablets therefore help to fix to some extent the 
date of Babrios. An exhaustive discussion of these tablets—by far the most 
important Greek wax tablets extant—is at present impossible, pending their 
publication by Dr. D, C. Hesseling of Leyden. (This publication has now 
appeared in the Journal of Hellenic Studies, Vol. XIII. (1892-93), p. 293 ff.) 


Remarks were made by Professors Sonnenschein, Smyth, and 
Fowler. 


20. Ileptrérea and Allied Terms in Aristotle’s Poetics, by Professor 
Horatio M. Reynolds, of Yale University, New Haven, Conn. 


Commentators and critics, ancient and modern, have often used the term 
mwepirérea in a general and incorrect sense. Especially in recent times it has 
been employed frequently as synonymous with ‘catastrophe’ or dénofment 
whereas Aristotle, from whom, whether he coined it or not, dramatic criticism 


Proceedings for July, 1893. xiv 


obtained the word, seems to have employed it consistently in a technical and 
carefully defined sense. - In the Encyclopaedic Dictionary under ‘ peripetia’ we 
read, ‘Old drama. The sudden reversal or disclosure of circumstances on 
which the plot in tragedy hinges: the dénofiment of a play.’ In Worcester, 
*A sudden change of fortune.’ In the Century Dictionary, ‘That part of a 
drama in which the plot is unraveled and the whole concludes: the dénof- 
ment.’ These definitions are doubtless fully borne out by modern usage, 
and to that extent justified in dictionaries; that they are not Aristotle’s is also 
clear. According to them, peripetia seems applicable to any play. This may 
find its explanation in the fact that modern drama is generally ‘ complex.’ Greek 
tragedy, Aristotle teaches, is not so; and ‘ peripetia’ is not a universal phenomenon 
even in complex tragedy. In the first definition, the phrase ‘ disclosure of cir- 
cumstances’ would better represent dvayvwpiots. In Worcester, there is not even a 
limitation of the term to the drama. In the Century Dictionary, we seem to have 
a translated echo of Aristotle’s definition of Avots — Avow 5é Thy awd THs apxis 
THs peraBdcews wéxpe TéXovs. So, too, of the general definition, dénofiment. 
Translators also fail to give the strict meaning, rendering dvayvupus éx mepure- 
telas ‘sudden recognition,’ regardless of the paraphrase in the same chapter, 
dvayvaputs 7 € abtdv TOv mpayudtwv THs éxmdhtews yryvouévns Se elxérwv. 
Teachers and editors of tragedy also use the term in a loose way. The foremost of 
the English scholars of to-day in his edition of the Oedipus Tyrannus paraphrases 
the word with ‘reversal of fortune,’ which is inadequate, and yet he is dealing 
with the Aristotelian criticism of the play. And a well-known American scholar, 
in his edition of the Antigone, on 1. 988, speaks of ‘the beginning of the repiréreca 
of the play,’ where, if he preferred a Greek word, dots or werd Bacis would have 
been more exact. To use a word generally understood to be Aristotelian in 
an un-Aristotelian sense is productive of confusion. What, then, is the Aris- — 
totelian meaning? The /ocus classicus is found in the Poetics 1452 a: 


elol 5¢ rGv whOwv of wev aot ol 5¢ wemreyuEévor~ Kal yap ai mpdées, oy pit- 
ces ol ud00l eioy, bmdpxover eVOds odoat ToratTrar. Aéyw Se awhfy pev mpaiu, 
Hs yuwouerns Worep wpiorat cuvexois kal puds dvev wepirerelas } dvayvwpicuod 7 
perdBacts ylverat, rerdeypuévyn 5€é éoriv, €& Fs weTa dvayvwpicpod 7H wepirerelas 7 
dugoty y perdBacis éoriv. raidra dé det ylverOar é& abris Tis sveTdcews TOD 
pibou, wore éx Trav mpoyeyernuévwv cupBalvey 7 €E dvdyKns H Kara Td elxds ylyve- 
0a tradra* diadéper yap word Td yiyverOar Tdde Sid Tae H wera Tae. 

éore 5¢ wepiréreca pév 4 els Td évavTiov TOyv mpaTrTouévwy peraBodH, Kabdrep 
elpnrat, kal rodro 5¢ womrep déyouev kara Td elxds } dvayKaiov: Wowep ev TO 
Oldtrodt €Mav as edppavdy rdov Oldirovy cal dwadd\dtwy tov wpds Thy unrépa 
PbBov, SnrXéoas, ds Hv, robvavrloy érolncev- xal év T@ Avyxe? 6 pev dydbpuevos as 
dro8avotpevos, 6 5€ Aavads dxodovdav ws dmoxrevGv, Tov wey cuvéBn éx TSv 
wempayyuévwv adrobaveiv, roy 6¢ cwOvat. 


In the discussion preceding this passage, after giving his celebrated definition 
of tragedy and deducing therefrom the six qualitative elements of tragedy, —the 
myth, the ethical element, the diction, the sentiment or thought, the scenic deco- 
ration, and the lyric element, — Aristotle proceeds to discuss the myth as first in 
importance, and disposes of certain general questions of dramatic form, — the 
proper extension of the myth, dramatic unity, and poetic truth in contrast with 


xlvi American Philological Association. 


historic truth. He then turns to his main theme — how the action is to be shaped 
in order to be ¢vagic. Things that awaken pity and terror!1—rd édeerva kai Ta 
goBepéd —are in his view the peculiar field of tragedy. Without here defining 
these, he elucidates at some length their nature: they will exert the better their 
peculiar influence if they happen contrary to expectation (rapa rhy dégav), and 
still more if they happen contrary to expectation, one from another (6¢ &\AmAq), 
i.e. in causal sequence. And the @avyaoréy, or éxmdnxrixéy, as he elsewhere? 
terms it, is shown by more than one reference® to be in Aristotle’s view a 
substantial element of tragedy. Such myths as have the qualities of surprise 
and causal sequence are those to which the poet should direct his gaze as 
more worthy of the tragic muse. 

In the passage quoted above, the train of thought is somewhat as follows: 
Since a tragic action without change of situation (serdfSacrs) is unthinkable, 
what change of situation is appropriate to tragedy? An action, and hence its 
imitation in the myth, is simple when the transition from one situation to another 
is without wepiréreta or dvayvepiots; it is complicated, when it is brought about 
by or with weperéreca or dvayvepiois, or both. A transition of some kind is abso- 
lutely necessary, for the action cannot conclude as it begins. mepurérea and 
dvayvwpiots are only special forms or means of this transition, and either or both 
may be dispensed with. If the transition is effected without these, the myth is 
simple; if with one or both, the myth depicts a movement which fails to reach 
the goal at which it aimed, and this failure is brought about as a necessary or 
probable consequent. Aristotle then defines repiréreca, and gives in detail two 
instances of its employment in tragedy. In explanation of xa@dzep elpnra, it is 
well to cite with Vahlen 1450 a, cuuBalver eis edruxtav éx dvoTuxlas ... weraBad- 
Ae (cf. weraBodry in the definition and the example of Lynceus, in whom 
the reversal of fortune centered). By r&v rparrouévwv, Vahlen understands 
not ‘circumstances’ or ‘situation,’ but ‘what a character does for a special 
purpose,’—a purpose, however, which is defeated and the direct opposite 
wrought. This distinction is certainly borne out in the two examples cited by 
Aristotle from the Oedipus and the Lynceus. Each of these plays contains 
a thwarted purpose, and the form of statement gives prominence to this 
element ; but the distinction is more apparent than real. A drama implies 
action, and hence actors who must have purposes. A sudden change of 
situation without a thwarted purpose, therefore, is hard to imagine. The 
element of suddenness and surprise, on which Aristotle elsewhere‘ lays so 
much stress, is here implied in the use of the word peraBody instead of werd Bacrs, 
and still more in the phrase els 7d évayriov. We may now paraphrase Aristotle’s 
definition: mepumréreca is the sudden, striking reversal, in necessary or probable 
sequence, of the situation or action to the directly opposite, i.e. from happiness 
to unhappiness, or the contrary. 

In regard, then, to Aristotle’s use of the term, the following conclusions are 
warranted: First, that weperérea is not synonymous with werdBaors, the latter 


1 Subsequently defined and classified under meperéreta, avayvwprors, and ra8os. 

276. 1454, 3 Cf. Rhet. I. 11. 24 quoted below. 

4 Cf. Poet. 1452 a and Rhet. I. 11. 24 Kai ai wepimérecat kai 7d mapa pixpoy owlerOar ex Tav 
xivdvver (sc. 76)* mavta yap Oavpacra tavta, ‘From the love of wonder arises the pleasure 
we derive from tragic reperérecac and narrow escapes from danger,’ etc. 


~~ 


Proceedings for July, 1893. xIvii 


being applicable to every tragic action, simple or complex, the former merely to 
some complex actions; secondly, it may denote the special manner in which (pera 
xTX.), or the special means by which (ék xrX.), the transition is effected; further, 
it consists of or includes a special act that brings consequences unforeseen by the 
agent, who may or may not be the hero of the drama; and lastly, it is such as to 
excite wonder by its suddenness and completeness. A term liable to con- 
fusion with wepuréreca is defined by Aristotle (cf. Poet. 1455 b), ore 5¢ rdons 
rpaywpolas rd pev Séois TO SE AVows .. . AUow Se Thy dwd Tis apxhs THs peraBa- 
cews wéxpt Tédovs. Etymologically, therefore, Avots is exactly dénodment, which 
is often made the synonym of wepiréreca. But Advors includes the werdBaors and 
all that follows; it is hence far more general, both in its application to a single 
tragedy and in the fact that it is applicable to every tragedy. 

So much for the Aristotelian usage, which is strict and consistent.! In later 
Greek authors, we may not@ occasional conformity to that usage. Cf. Plutarch de 
Socratis genio, I. 596. 29, and perhaps Diodorus Siculus III. 57.8. In Sextus 
Empiricus 310, the word occurs with reference to tragedy, though in the sense of 
iréGecis. But usually in later Greek authors it is employed in the general sense 
of a chance event, favorable, unfavorable, or neutral, as in Polybius. Cf, 
Schweighiuser’s Lexicon Polybianum and Stephanus’ Thesaurus. 


Remarks upon this paper were made by Rramaapre Shorey and 
Reynolds. 


21. Libration in the Periods of Cicero, by Professor W. B. Owen, 
Lafayette College, Easton, Pa. 


The paper treated the methods by which Cicero secured the enrichment of 
style, — especially duplication, the balancing of related parts of sentences, and 
the grouping of ideas and synonyms in pairs. De Oratore, Book L., was closely 
examined with a view to this vibratory movement. 

There is a natural tendency in emfatic speech to reinforce meaning by repeti- 
tion, or double strokes, and it is in the nature of expression, further, to seek 
clearness and emfasis by throwing parts of sentences into the attitude of correla- 
tion. The Latin developt an elaborate machinery for these purposes: First, in 
the particles which throw clauses and frases into correlation and admit them 
in pairs; and secondly, in its ample vocabulary, furnishing the materials for 
duplication. ; 

Cicero’s sentences vibrate with pairs, in which the added members give 
differences of meaning if you look into the shades of meaning, but very ofn 
without distinct intention beyond the dual cadence, where the duplication is one 
of movement and sound rather than of thought. This is conspicuously the case 
in sentences which start out with the/vibratory swing, and the movement is then 
kept up thru a succession of pairs. Such a style gives an excellent opportunity 
for the study of synonyms; but that method applied to a particular text is likely 
to inject more thought into the frases than the writer was conscious of, Of course 
Cicero is accurate and discriminating in the choice of words; but when his 


1 De hist. Animal. 590b. 13 forms no real exception, 


xlviii American Philological Assoctation. 


sentences fall into vibration we must not press too hard our search for an argu- 
ment in every frase. 

A few variations wer noted and illustrated. 

The libration ofn extends to the structure in detail of hole frases and even 
clauses. ; 

The second member is very ofn more fully developt, especialy at the close 
of a period. 

On the other hand, where more exact rhythmic balance is desired it is secured 
by attaching common adjuncts partly to one and partly to the other member. 

The effect of a pair is at times greatly hightend by reversing in the second 
member the terms of the first. 

Pairs ofn occur within pairs. 

In the management of larger groups of particulars when they ar subjects which 
require discussion they ar frequently introduced in pairs. ‘ 

When mention is all that is required, the mere list stil bears evident traces 
of libration, the vibratory grouping being determined by similarity or contrast 
of meaning. 

The'series of particulars at times presents a climactic saa and yet easily 
falls into the vibratory swing. 

In fcrmal libration, the second member is sumtimes Brake: into another pair ; 
in such cases the enlargement of a sing] member frequently balances a lighter pair. 

Even without the forms of libration a difference of connectiv ofn puts the 
items of apparent triplets on a slightly different footing, the result being sum 
variation of the dual cadence. 

Then again, in formal triplets, the third member ofn contains a pair, thus 
giving it the effect of double libration. q 

Groups of four ar almost always thrown into the form of two pairs. ~ 

Words ar ofn displaced to bring them more obviously into pairs. These 
remarks wer abundantly illustrated by passages from De Oratore. 

The bearing of this trait upon certain grammatical figures involving duality, as 
Hendiadys, Hysteron-proteron, Zeugma, etc., was then discust; also certain fixt 
combinations which recur as redy-made pairs. 

Its relations to the criticism of uncertain text wer also noted, and passages 
cited in which considerations of rhythmic symmetry ar important if not decisiv. 

Then followd a comparison of Cicero in this respect with Caesar, Livy, Sallust, 
and Tacitus. 


Professors D’Ooge, Shorey, and Gudeman briefly discussed the 
above paper. 


22. Varro and Chrysippus as Sources of the Dialogus of Tacitus," 
by Dr. Alfred Gudeman, University of Pennsylvania. 

A comparison between Quintilian’s I. bk. and the treatise Ilept raléwyv dywyjs, 
ascribed to Plutarch, shows that both authors are largely indebted to Chrysippus’ 


work Ilepi raléwy dywyfs. It is next shown that Ps. Plutarch and the chapters on 
the education of children in the Dialogus, reveal some remarkable parallelisms in 


1 Incorporated in the author’s edition of the Dialogus (Ginn & Co.). 


Proceedings for July, 1893. xlix 


thought and language, coincidences which can only be explained on the sup- 
position that both authors are indebted to Chrysippus as their common source. 
The Dialogus finally contains a remarkable parallelism with Varro’s de liberis 
educandis, who in his turn was very likely influenced by the Greek treatise. The 
question whether Tacitus borrowed directly from Chrysippus or got his informa- 
tion through the medium of Varro cannot be decided with confidence. 


Remarks were made by Professors Sproull, Hendrickson, and the 
author. 


23. The Indo-European root s/@ ‘stand’ in Italic, by Professor 
Carl D. Buck, University of Chicago. 


The present systems in which the I.-E. root s¢a ‘ stand’ most frequently appears 
may be grouped as follows: 

I. Reduplicating-class: A. Unthematic, e.g. t-orn-u, i-cra-uev; B. Thematic, 
e.g. Skt. ¢2-stha-tt. 

II. Root-class, e.g. Skt. d-stha-t, d-sthi-thas, Gr. @-orn, ¢-ord-Ons. i 

III. zo-class: A. with strong ablaut-form of the root, I.-E. sta!-20, e.g. Lith. 
stéju, O.B. stajg; B. with weak ablaut-form of root, I.-E. s¢a-70, e.g. Skt. sthi-yd-te 
(for *stha-ya-te, cf. Brugmann, Grundriss II. p. 897), O.B. sfojg. Another sub- 
division, based on the ablaut-change of the suffix, is: 1. With ablaut zo-ze (corre- 
sponding to thematic o-e), e.g. Lith. s#éju, 3rd sing. sééja, O.B. stajg, 3rd sing. 
stajetii; 2. With ablaut zo-2, again subdivided a) i0-2, e.g. O.B. stojg, 3rd sing. 
stojité (like veljq-velité), b) io-i, e.g. Lith. stéviu (with root-increment x), Ist 
plur. stévim. Here, as in many other verbs, A. seems to go hand in hand with 1., 
and B. with 2. Yet the divisions do not always coincide. 

As regards the semasiological difference between the present classes of sa, we 
may assume, on the authority of Greek and Latin, that the reduplicated forms had 
transitive force ‘set.’ In Sanskrit, however, ¢ésthami is intransitive, the meaning 
‘set’ being brought out by the causative sthapdyami. Within the o-class, while 
we cannot maintain complete coincidence between the semasiological divisions 
transitive-intransitive and the morphological divisions A-B or I-2, yet it is 
certainly true that the intransitive meaning is especially prominent in groups 
B.and 2. It is from the forms of B. that the Sanskrit passive has been developed, 
and in Balto-Slavic it is the forms of 2. which outside the present show the 
element @ which is seen again in the Greek ‘second aorists passive,’ e.g. O.B. 
monja, monéti ‘think’ = palvouat, éudvnv, cf. Brugmann, Grundriss II. p. 1082. 
In O. Bulgarian the simple intransitive meaning ‘stand’ is expressed by stojq, 
while s¢ajq is rather an iterative, and used only with prepositions, as vustaj¢g 
‘stand up.’ So, too, Lith. s¢ojw is not the word for ‘stand,’ but is used in the 
reflexive form s¢éjd#-s in sense of ‘I place myself, take stand,’ and also in active 
with propositions like the O.B. s¢ajq. 

After this classification of the formations in which the root s¢a appears in 
other languages, the paper discusses the Italic forms with reference to their 
position in the scheme. 

The Latin sés¢o offers no difficulty. It belongs, of course, under I. (redupli- 
cating-class), and in the main subdivision B, though a form like s¢stimus may 
equally well be placed under A (unthematic). 


] American Philological Association. 


The points with which the paper deals are 1) the position of the Latin sf, 
2) that of the various forms of the Italic dialects, such as Umbrian stahu, stahitu, 
staheren ; Oscan stahint, staiet, stait, staieffuf, etc. 

In regard to the inflection of Latin so the paper claims a far more important 
share for the root-class (II.) than is allowed it in the latest treatment (Brugmann, 
Grundriss II. passim). It is doubted whether'even in the first person s/0 we are 
to see an exact equivalent of the Umbrian stahuz, and consequently a representative 
of the zo-class, cf. also Bartholomae, Idg. Stud. II. p. 142. 

The Oscan and Umbrian forms are plainly connected with III. (zo-class), but 
beyond this there is a difference of opinion. The writer considers the sup- 
position of an Oscan-Umbrian present-stem *s¢a-2, as assumed by both Brugmann 
(Grundriss II. p. 1066) and Bronisch (Osk. I und E Vocale, p. 185), but with 
totally different explanation of the same, as unnecessary. Not merely Umbrian 
stahu, but also stahitu, Osc. stahint may be referred to an I.-E. *sta-io-, *sta-i2-, 
that is, according to our scheme, to III, B, 2, a. We say B, not A, because of 
the syncope in Oscan eestin¢ ‘extant’; 2, not 1, there being no trace of 1 in 
Italic; a, not b, on account of the lack of syncope in Umbr. stahkitu and the 
analogy<of other examples of the formation in which the length of vowel is 
indicated in the writing, as Umbr. herezzz ‘ volito.’ 

Umbr. staheren ‘stabunt’ is, as already noted by Biicheler, on a line with 
other Oscan-Umbrian futures. It represents an older *s¢a-ze-s-en¢ in form sub- 
junctive of an s-aorist. That we should have *s¢azes-, not *sta-s, need not surprise 
us. It has already been remarked (Buck, Vocalismus, d. Osk. Sprache, p. 53) 
that wherever a difference between present-stem and verbal-stem exists, it is the 
former which appears in this future formation, cf. Osc. dzdest ‘ dabit’ with present 
reduplication, Osc. Aafiest, Umbr. Aadiest ‘habebit,’ Umbr. eriest ‘volet.’ The 
form staheren is the proper plural to a singular *s¢ahest formed exactly like heriest. 
Biicheler’s conjecture of [Aereset] in line 27 of the Cippus Abellanus is altogether 
probable, in spite of the objection of Bronisch, l.c. p. 100, note. 


GENERAL SESSION. 
Cuicaco, July 14, 1893. 


This general meeting of the various philological organizations was 
presided over by Professor Gildersleeve, of the Johns Hopkins 
University. Professor bias was elected Secretary. 


24. Dunkles und helles 7 im Lateinischen, by Professor Hermann 
Osthoff, of the University of Heidelberg. 

This paper, which is printed in the TRANSACTIONS, was briefly dis- 
cussed by Professor B. I. Wheeler. 


25. The Scientific Emendation of Classical Texts, by Professor 
E. A. Sonnenschein, of Mason College, Birmingham. 
This paper appears in the TRANSACTIONS. 


Proceedings for July, 1893. li 


26. The Greek Nouns in -is, -idos, by Professor Benjamin Ide 
Wheeler, Cornell University, Ithaca, N.Y. 


This important class of nouns is yet without historical explanation. Recent 
writers seem convinced that it is connected in some way with the Sanskrit nouns 
in -s; cf. Joh. Schmidt, Pluralbildungen der indog. Neutra p. 55; Johansson, K.Z. 
XXX. 401; Kretschmer, K.Z. XXXI. 346; but no account is attempted of the 
stem consonant -6-. Brugmann, Grundriss II. § 128, discusses the nouns in 
question under the heading of “the suffix -d-”’ in connection, on the one hand, 
with the scanty group of Latin nouns like cassis, cuspis, capis, lapis, and on the 
other with Greek nouns in -ds, -déos; see also Gr. Gramm. p. 2110. Schmidt 
speaks of the genit. -idos as “heteroklitisch,” and Gustav Meyer, Gr. Gramm. 
2§ 321, speaks of the confusion of z-stems and the dental stems. An I.E. a 

‘suffix seems to be therefore the common presumption. That the stem-ending 
-.6- has its origin in a Indo-European prototype -zd- is, however, apart from any 
other positive explanation, highly improbable, and for the following reasons: — . 

(1) There was no I.E. suffix -2¢d-. The Indo-Iranian branch shows none. 
There is no trace of such a suffix except in Greek and Latin. The one possible 
comparison, xovis, -(dos : O. Eng. hnite : O.H.G. niz - Bohem. Anida, does not 
affect the case, as the -d- is apparently part of the root (xvifw). The Teutonic 
and Sclavic words have vowel-stems. The Latin words mentioned above show 
no connection in meaning with the Greek nouns, and are insufficient by them- 
selves to establish an I.E. stem-class. If the suffix is not Indo-European, it must 
be a special Greek development. z 

(2) The existence in I.E. of any form of a vital suffix ending in -d is to be 
doubted. There were probably stems ending in-d; cf. the probable etymology 
drsdd-: Gr. dep46-, and the Teutonic verbs in -a4an. The Greek nouns in -ds, 
-450s were a numerous class, but were in part at least derived from the weak 
form of the stem in -ovr-; cf. Kretschmer, K.Z. XXXI. 347 f. The older 
Sanskrit has but three certain examples of the suffix: dysdd-, bhasdd-, ¢ardd-. 

(3) Such words as afford the opportunity of direct etymological comparison 
show the -d- to be peculiar to the Greek: «Ants, kAnidos : Lat. clavis; medXls, 
-lSos : Lat. pelvis, peluis > Skr. palavi; éuals, -ldos : O.H.G. imdi (?). 

-The I.E. inflectional type with which our class of nouns shows clearest marks 
of relationship is that in -2s -zjos, represented in the older Sanskrit by aféis, 
naptiyam, naptiya, etc.; pl. naptiyas, naphibhis. This type in Skr. is to be 
distinguished on the one hand from the monosyllabic nouns like dhzs, dhiyam, 
dhiyds, with varying accent, and on the other from those with the feminine 
differentiating suffix I.E. i@-i(i2) ; devi, devim, devyas. From the latter it was 
distinguished in the following points: (1) Suffix; -2i- without ablaut vs. ia +2(7). 
(2) Constant accent upon the -2- except in compounds. The other accented 
the root in the nomin., the endings in genit., etc., cf. ula, weds; Spyura, dpyuas, 
Joh. Schmidt, K.Z. XXV. 36. In Sanskrit the most of these nouns follow in 
accent the masculines from which they are derived: devi - devd-, whereas those 
which are independent of their primitives in accent are all barytone; /dvisz 
(¢avisd-). The importance of this fact has thus far been concealed under a 
descriptive form of statement; cf. Lanman, Noun Inflec. p. 368. (3) Nom. sing. 
ends in -s, vrdis. (4) Accus. sing. in -iyam vs, im. 


- 


lii American Philological Association. 


The Skr. group in question is evidently made up of diverse materials. Most 
of the nouns may be explained as originally adjective derivatives. I regard 
the inflection as representing a development upon the basis of the nominative 
singular, and this nominative singular as representing the type, Goth. 4airdeis, 
Lith. gaidjs, Lat. a/is, ie. a nomin. ending in I.E. -és for earlier -iios (Streitberg, 
Paul Br. Beitr. XIV. 165 ff.). The appearance of 2 in the form -zz- is due to the 
drawled (sweigiffliger) accent; cf. Ved. -aam. The connection with forms in 
-iios is betrayed, e.g. by rathi-s, “ belonging to a wagon,” “ driver,” etc.: rathiya- 
“belonging to a wagon”; duét-s - dutiya-; samudri-s: samudriya-; purusi-s : 
purusiya-. The evenness or “ regularity” of the paradigm is evidence of youth. 
It is a paradigm indeed which is ever tending to reconstruct itself anew. Its basis 
is the nomin. sing., its materials the commonest analogies of noun inflection. 
Under this point of view are explained the following features: (1) Absence of exact 
etymological correspondences. (2) Use of this stem-formation parallel to z-stems 
and ya-stems without apparent difference of value. (3) The parallelism of 
u-stems: fans: tanivam, cf. ix6ds, lyOGos. (4) The appearance occasionally 
of the same method in z-stems, Skr. dvi-s, dvy-as, possibly I.E.; cf. Gr. dts, oids 
(<éfz0s). 

The Gr, nouns in -fs, -(d0s, correspond to the Skr. -Zs, -zyas in the following points : 
(1) Prevailing oxytonesis of nom. sing. Paroxytones like ps admit an accus. in 
-v, oxytones never. (2) The continuance of the accent upon the z. (3) The 
nom. sg. ending -s. There are abundant traces of a long z, though the short 
vowel of the majority of the cases has predominated ; thus xvnuidas, Wndidas, 
dior, etc., also kAnls, BNoovpGmis, wdis (X. 499, Hes. Theog. 178); cf. Lesb. 
mdecs, Collitz, Sammlung 299, BoGri (voc,) Hartel, Hom. Stud.? I. 105 f. Cf. also 
rodirns, 6pirys, etc., whose paroxytonesis is I.E., Kretschmer, K.Z. XXXI. 344. 

It may be that neither the Skt. nor the Gr. type is I.E., but the basis and the 
plan appear to be the same. Whence is the -6-? The nouns in -ds, -ddos and 
-ls, -l60s doubtless assisted each other. If either class is old, it is the former. 
The nouns in -és, -(S0s are rare in the earliest monuments. The dialects have few 
examples. There are none in the Gortynian inscription. It is a possibility that, 
after all, the -5- is a special development of -z-, — not, however, as loosely stated by 
Curtius (cf. Gr. Etymol.5 636 ff.; Stokes, B.B. IX. 87), but under the special 
conditions of an accented -z- preceded by a labial, or perhaps only by an ¢. The 
old word Hom. dis, waidés (for *rarls, mwapléos) is here of great value, as its 
contraction and partial heteroclisis give it the value of an isolated form. The 
contraction in Homer of a + «>a belongs to the trisyllabic forms; olds is a pos- 
sible parallel. Wackernagel’s “ wagjés” is impossible (K.Z. XXVII. 277). The 
-5- is here removed from the possibility of explanation by suffix extension. The 
syllable -.6- appears here under like conditions with that in a large class. of words 
like dwdotdas, Satéas, Bacidntdos, Antdos, Kiconts, Nnpytdes, "Axattdos, Xpvenis, 
"AiSos (’Atdos?), xAnts, etc. To be noted is also the parallelism of *Epex@nfs : 
*"EpexOetdai, “Aides : *Atdns, in which the ending -ns (as) serves the same indi- 
vidualizing purpose as the -wv of Ovpaviwy : obpdmos; cf. ’Arpelwv : *Arpeldns. 
Other labials precede -i- in, e.g., édwls, dowls, yAugls, rparldes, cxagls, ynpls. 
The development of a dental explosive from z after a labial has its parallel in 
mT < pi, XadérrTw, dorpdrrw, rrvw; Grassmann K.Z. XI. 13; Osthoff M.U. IV. 
13 ff.; Brugmann, Gr. Gramm. § 40; Froehde, B.B. VI. 179. As ~i>-77, so 


Proceedings for July, 1893. liti 


bi> Bd; pdBdos : Lith. viFbas; potBdos : potfos; FAdouar < *éABSouar < édm:-, 
éd\mls; érlBdat (repotia) << *émimiBia to Skr. pibami, Bury, B.B. XVIII. 292. 
The -d- of fdcos is unexplained. The meaning of the word points to ,/syi-; cf. 
Skr. svaydm, self, Cretan fly. Is it for syzios with re-added -tos? pntdios in its 
relation to p¢-@upos still awaits explanation. Osthoff’s connection with Lat. varus 
is unsatisfactory on the side of meaning; is not yraui- as reasonable as yraso-? 
Whatever the origin of the -6-, the striking parallelism of the type éArls, édrldos 
and madis, nadiyas cannot be overlooked, and the stem-ending -d- must be 
explained as a Greek product. 


27. On the Canons of Etymological Investigation, by Professor 
M. Bréal, Collége de France. 

This paper was read by Professor Wheeler, and discussed by 
Professor Osthoff. It is to be found in the TRANSACTIONS. 

Adjourned at 1 P.M. ' 


AFTERNOON (GENERAL) SESSION. 


At 2.30 the General Session of the morning was resumed. 


28. Ein Ablautproblem der Ursprache, by Professor Wilhelm 
Streitberg, University of Freiburg (Switzerland). 

Remarks were made by Professors Bloomfield and Osthoff, and 
in reply by Professor Streitberg. This paper is printed in the 
_ ‘TRANSACTIONS. 


29. The Importance of Uniformity in the Transliteration of non- 
Roman Alphabets was presented as a subject for general discussion 
by Professor M. Bloomfield, of Johns Hopkins University. 

Remarks were made by Professors Bloomfield, Wheeler, Osthoff, 
Streitberg, and by Dr. Parks. 

A vote of thanks, especially to the Committee of Arrangements 
and President Hale, was then carried. In seconding this motion 
Professor Sonnenschein gave expression to the pleasure experienced 
by all the European visitors in participating in the sessions of the 
Philological Congress. 

Adjourned at 3.45 P.M. 


~ 


OFFICERS OF-THE ASSQCIATION. 


1893-04. 





PRESIDENT. 
JAMES M. GARNETT. 


VICE-PRESIDENTS. 
JOHN H. WRIGHT. 
BERNADOTTE PERRIN. 


SECRETARY. 
HERBERT WEIR SMYTH. 


TREASURER. 
HERBERT WEIR SMYTH. 


EXECUTIVE COMMITTEE. 
The above-named Officers, and — 
BASIL L. GILDERSLEEVE. 
WILLIAM W. GOODWIN. 
ABBY LEACH. 
FRANCIS A. MARCH. 
WILLIAM D. WHITNEY. 


MEMBERS OF THE AMERICAN PHILOLOGICAL 
ASSOCIATION|! 


1893-’94. 


William F. Abbot, High School, Worcester, Mass. (26 William St.). 
F. F. Abbott, University of Chicago, Chicago, Ill. 
J. W. Abernethy, Adelphi Academy, Brooklyn, N. Y. 
Charles D. Adams, Dartmouth College, Hanover, N. H. 
Cyrus Adler, Smithsonian Institution, Washington, D. C. 
Frederic D. Allen, Harvard University, Cambridge, Mass. 
George Gillespie Allen, Boston, Mass. 
Francis G, Allinson, Williams College, Williamstown, Mass. 
Joseph Anderson, Waterbury, Conn, - 
Louis F. Anderson, Whitman College, Walla Walla, Washington, 
Alfred Williams Anthony, Cobb Divinity School, Lewiston, Me. 
W. Muss-Arnolt, University of Chicago, Chicago, Ili. 
Sidney G. Ashmore, Union University, Schenectady, N. Y. 
Robert S. Avann, Albion College, Albion, Mich. 
Charles W. Bain, Portsmouth, Va. 
Robert Baird, Northwestern University, Evanston, Ill, 
H. L. Baker, Detroit, Mich. (70 Griswold St.). 
-C, H. Balg, Milwaukee, Wis. (623 Fifth Street). 
Cecil F. P. Bancroft, Phillips Academy, Andover, Mass. 
Mabel Banta, Bloomington, Ind. 
George K. Bartholomew, English and Classical School, Cincinnati, O. 
Isbon T. Beckwith, Trinity College, Hartford, Conn. 
A. J. Bell, Victoria College, Cobourg, Ont. 
George Bendelari, Harvard University, Cambridge, Mass. 
Charles E. Bennett, Cornell University, Ithaca, N. Y. 
Charles W. Benton, University of Minnesota, Minneapolis, Minn. 
R. C. Berkeley, West Virginia University, Morgantown, W. Va. 
Thomas S. Bettens, “ The Kensington,” cor. Fifty-seventh St. and Fourth Ave., 
New York, N. Y. 
Louis Bevier, Rutgers College, New Brunswick, N. J. 


1 This list has been corrected up to March, 1894; permanent addresses are given, as far as 
may be. Where the residence is left blank, the members in question are in Europe. The Secre- 
tary and the Publishers beg to be kept informed of all changes of address. 

: ie : 


a 


lvi American Philological Assoctation. 


Hiram H. Bice, St. Joseph, Mo. 

George H. Bingham, Pinkerton Academy, Derry, N. H. 
Charles Edward Bishop, College of William and Mary, Williamsburg, Va. 
M. Bloomfield, Johns Hopkins University, Baltimore, Md. 
Willis H. Bocock, University of Georgia, Athens, Ga. 

C. W. E. Body, Trinity University, Toronto, Ont. 

D. Bonbright, Northwestern University, Evanston, IIl. 

A. L. Bondurant, 7 Holyoke House, Cambridge, Mass. 

Hugh Boyd, Cornell College, Mt. Vernon, Ia. 

Charles F. Bradley, Garrett Biblical Institution, Evanston, Ill. 

J. Everett Brady, Smith College, Northampton, Mass. 

H. C. G. Brandt, Hamilton College, Clinton, N. Y. 

Josiah Bridge, Foster’s, Warren County, O. 

Walter R. Bridgeman, Lake Forest University, Lake Forest, Ill. 
James W. Bright, Johns Hopkins University, Baltimore, Md. 
John A. Broadus, Southern Baptist Theological Seminary, Louisville, Ky. 
Jabez Brooks, University of Minnesota, Minneapolis, Minn. (1708 Laurel Ave.). 
Demarchus C. Brown, Butler University, Irvington, Ind. 

Edward Miles Brown, University of Cincinnati, Cincinnati, O. 

F. W. Brown, Franklin College, Franklin, Ind. 

Mariana Brown, Earlham College, Richmond, Ind. 

Carleton L. Brownson, Yale University, New Haven, Conn. 

Carl D. Buck, University of Chicago, Chicago, IIl. 

Walter H. Buell, Scranton, Pa. (243 Jefferson Ave.). 

George Woodbury Bunnell, University of California, Berkeley, Cal. 
Isaac B. Burgess, Morgan Park Academy, Morgan Park, Ill. 

Sylvester Burnham, Colgate University, Hamilton, N. Y. 

Henry F. Burton, University of Rochester, Rochester, N. Y. (63 East Ave.). 
Henry A. Buttz, Drew Theological Seminary, Madison, N. J. 

Leona Call, State University of lowa, Iowa City, Ia. 

A. Guyot Cameron, Yale University, New Haven, Conn. 

Edward Capps, University of Chicago, Chicago, Il. 

William H. Carpenter, Columbia College, New York, N. Y. 

Franklin Carter, Williams College, Williamstown, Mass. 

Clarence F. Castle, University of Chicago, Chicago, Ill. 

Eva Channing, Forest Hill St., Jamaica Plain, Mass. 

A.C. Chapin, Wellesley College, Wellesley, Mass. 

Henry Leland Chapman, Bowdoin College, Brunswick, Me. 

S. R. Cheek, Centre College of Kentucky, Danville, Ky. 

Bradbury L. Cilley, Phillips Academy, Exeter, N. H. 

Edward B. Clapp, University of California, Berkeley, Cal. 

J. S. Clark, University of Minnesota, Minneapolis, Minn. 

F. W. Colegrove, Ottawa University, Ottawa, Kan. 

Hermann Collitz, Bryn Mawr College, Bryn Mawr, Pa. 

William T. Colville, Carbondale, Pa. 

D. Y. Comstock, Lakeville, Conn. 

Albert S. Cook, Yale University, New Haven, Conn. 

J. Randolph Coolidge, Jr., Chestnut Hill, Mass. 


. 


Proceedings for July, 1893. lvii 


Oscar H. Cooper, University of Texas, Austin, Tex. 

William L. Cowles, Amherst College, Amherst, Mass. 

Edward G. Coy, Lakeville, Conn. 

E. W. Coy, Hughes High School, Cincinnati, O. 

John M. Cross, Kingston, N. Y. 

George O. Curme, Cornell College, Mt. Vernon, Ia. 

William L. Cushing, Dobbs Ferry, N. Y. 

Charles H. S. Davis, Meriden, Conn, 

Francis B. Denio, Bangor Theological Seminary, Bangor, Me. 
Otto Dietrich, Milwaukee, Wis. (255 Fifteenth St.). 

Daniel Kilham Dodge, University of Illinois, Champaign, IIl. 
Martin L. D’Ooge, University of Michigan, Ann Arbor, Mich. 
Manuel J. Drennan, Vassar College, Poughkeepsie, N. Y. 

M. E. Dunham, University of Colorado, Boulder, Col. 
Mortimer Lamson Earle, Barnard College, New York City. 
Morton W. Easton, University of Pennsylvania, Philadelphia, Pa. ‘ 
William Wells Eaton, Middlebury College, Middlebury, Vt. 
Herman L, Ebeling, Miami University, Oxford, O, 

William S. Ebersole, Cornell College, Mt. Vernon, Ia. 

Thomas H. Eckfeldt, Friends’ Academy, New Bedford, Mass. 
Katherine L. Edwards, Cornell University, Ithaca, N. Y. 
James C. Egbert, Jr., Columbia College, New York. 

A. Marshall Elliott, Johns Hopkins University, Baltimore, Md. 
Herbert C. Elmer, Cornell University, Ithaca, N. Y. 

L. H. Elwell, Amherst College, Amherst, Mass. 

E. Antoinette Ely, University of Chicago, Chicago, Ill, 

C. W. Ernst, Back Bay, Boston, Mass. 

Margaret J. Evans, Carleton College, Northfield, Minn. 

Arthur Fairbanks, Yale University, New Haven, Conn. 
Charles E. Fay, Tufts College, College Hill, Mass. 

Edwin W. Fay, Washington and Lee College, Lexington, Va. 
Thomas Fell, St. John’s College, Annapolis, Md. 

O. M. Fernald, Williams College, Williamstown, Mass. 

F, J. Fessenden, Berkeley School, 20 West Forty-fourth St., New York City. 
Joseph T. Fischer, Syracuse University, Syracuse, N. Y. (618 Crouse Ave.). 
Edward Fitch, Hamilton College, Clinton, N. ¥. 

Thomas Fitz-Hugh, University of Texas, Austin, Tex. 

William Winston Fontaine, Austin, Tex. 

Frank H. Fowler, University of Chicago, Chicago, III. 

Harold N. Fowler, Adelbert College for Women, Cleveland, Ohio. 
Susan B. Franklin, Vassar College, Poughkeepsie, N. Y. 

A. L. Fuller, Adelbert College, Cleveland, O. 

Charles Kelsey Gaines, St. Lawrence University, Canton, N. Y. 
William Gallagher, Williston Seminary, East Hampton, Mass. 
Paul C. Gandolfo, St. Louis, Mo. (2608 Park Ave.). 
James M. Garnett, University of Virginia, Charlottesville, Va. 
George P. Garrison, University of Texas, Austin, Tex. 

Henry Gibbons, Edgewood Park, Allegheny Co., Pa. 


. 


Ivili American Philological Association. 


Seth K. Gifford, Haverford College, Pa. 

Basil L. Gildersleeve, Johns Hopkins University, Baltimore, Md. 

Farley B. Goddard, Malden, Mass. 

Julius Goebel, Leland Stanford Jr. University, Palo Alto, Cal 

Thomas D. Goodell, Yale University, New Haven, Conn. (284 Orange St.). 

Ralph L. Goodrich, U. S. Courts, Little Rock, Ark. 

Charles J. Goodwin, Wesleyan University, Middletown, Conn. 

William A. Goodwin, Portland, Me. 

William W. Goodwin, Harvard University, Cambridge, Mass. 

Herbert Eveleth Greene, Johns Hopkins University, Baltimore, Md. 

James B. Greenough, Harvard University, Cambridge, Mass. 

Wilber J. Greer, Miami University, Oxford, O. 

Gustav Gruener, Yale University, New Haven, Conn. 

Alfred Gudeman, University of Pennsylvania, Philadelphia, Pa. 

William Gardner Hale, University of Chicago, Chicago, IIl. 

Arthur P. Hall, Drury College, Springfield, Mo. 

Isaac H. Hall, Metropolitan Museum, Central Park, New York, N. Y. 

J. Leslie Hall, College of William and Mary, Williamsburg, Va. 

Randall C. Hall, General Theological Seminary, New York, N. Y. (245 West 
Forty-eighth St.). 

Charles S. Halsey, Union Classical Institute, Schenectady, N. Y. 

William McD. Halsey, New York, N. Y. (34 West Fortieth St.). 

B. F. Harding, Belmont School, Cambridge, Mass. 

Albert Harkness, Brown University, Providence, R. I. 

William R. Harper, University of Chicago, Chicago, IIl. 

Karl P. Harrington, University of North Carolina, Chapel Hill, N. C. 

Samuel Hart, Trinity College, Hartford, Conn. 

John J. Harvey, West Virginia University, Morgantown, W. Va. 

Paul Haupt, Johns Hopkins University, Baltimore, Md. 

Edward Southworth Hawes, Polytechnic Institute, Brooklyn, N. Y. 

G. L. Hendrickson, University of Wisconsin, Madison, Wis. 

John H. Hewitt, Williams College, Williamstown, Mass. 

Addison Hogue, Washington and Lee University, Lexington, Va. 

E. Washburn Hopkins, Bryn Mawr College, Bryn Mawr, Pa. 

William A. Houghton, Bowdoin College, Brunswick, Me. 

William Houston, Toronto, Can. 

Albert A. Howard, 1717 Cambridge St., Cambridge, Mass. 

Ray Greene Huling, tor Trowbridge St., Cambridge, Mass. 

L. C. Hull, Lawrenceville, N. J. 

Milton W. Humphreys, University of Virginia, Charlottesville, Va. 

A. J. Huntington, Columbian University, Washington, D.C. (1010 N St., N. W.). 

George B. Hussey, University of Nebraska, Lincoln, Neb. 

Edmund Morris Hyde, Lehigh University, So. Bethlehem, Pa. 

Andrew Ingraham, Swain Free School, New Bedford, Mass. 

A. V. W. Jackson, Columbia College, New York, N. Y. (Highland Ave., Yonk- 
ers, Nz°¥.): 

George E. Jackson, Washington University, St. Louis, Mo. 

M. W. Jacobus, 149 High St., Hartford, Conn. 


Proceedings for July, 1893. _lix 


Hans C, G. von Jagemann, Harvard University, Cambridge, Mass. ’ 
J. Haywode Jennings, Martin, Tenn, 

Henry Johnson, Bowdoin College, Brunswick, Me. 

Henry C. Johnson, Cortland, N. Y. 

J. C. Jones, University of Missouri, Columbia, Mo. 

Gustaf Karsten, University of Indiana, Bloomington, Ind. 

Robert P. Keep, Free Academy, Norwich, Conn. 

Martin Kellogg, University of California, Berkeley, Cal. 

Francis W. Kelsey, University of Michigan, Ann Arbor, Mich. 
David A. Kennedy, Orange, N. J. 

Charles W. Kent, University of Tennessee, Knoxville, Tenn. 

H. W. Kent, Norwich, Conn. 

Alexander Kerr, University of Wisconsin, Madison, Wis. 

John B. Kieffer, Franklin and Marshall College, Lancaster, Pa. 
Robert A. King, Wabash College, Crawfordsville, Ind. 

J. H. Kirkland, Vanderbilt University, Nashville, Tenn. 2 
George Lyman Kittredge, Harvard University, Cambridge,. Mass. 
Charles Knapp, Barnard College, New York City (304 W. 16th St.). 
Charles S. Knox, St. Paul’s School, Concord, N. H. 

A. G. Laird, Cornell University, Ithaca, N. Y. 

William A. Lamberton, University of Pennsylvania, Philadelphia, Pa. 
Charles R. Lanman, Harvard University, Cambridge, Mass. 

Lewis H. Lapham, 28 Ferry St., New York, N. Y. 

C. W. Larned, U. S. Military Academy, West Point, N. Y. 

William Cranston Lawton, Bryn Mawr College, Bryn Mawr, Pa. 
Abby Leach, Vassar College, Poughkeepsie, N. Y. 

J. T. Lees, University of Nebraska, Lincoln, Neb. 

Walter Lefevre, University of Texas, Austin, Tex. 

Thomas B. Lindsay, Boston University, Boston, Mass. 

Alonzo Linn, Washington and Jefferson College, Washington, Pa. 
Gonzalez Lodge, Bryn Mawr College, Bryn Mawr, Pa. 

Lee Davis Lodge, Columbian University, Washington, D. C. 
Frances E. Lord, Wellesley College, Wellesley, Mass. 

George D. Lord, Dartmouth College, Hanover, N. H. 

Frederick Lutz, Albion College, Albion, Mich. 

George Edwin MacLean, 328 Tenth Ave., S. E., Minneapolis, Minn. 
H. W. Magoun, Oberlin College, Oberlin, O. 

J. H. T. Main, lowa College, Grinnell, Ia. 

E. W. Manning, De Pauw University, Greencastle, Ind. 

F. A. March, Lafayette College, Easton, Pa. 

F. A. March, Jr., Lafayette College, Easton, Pa. 

Winfred R. Martin, Trinity College, Hartford, Conn. 

W. W. Martin, Vanderbilt University, Nashville, Tenn. 

Allan Marquand, College of New Jersey, Princeton, N, J. 

Ellen F, Mason, 1 Walnut St., Boston, Mass. 

W. Gordon McCabe, University School, Petersburg, Va. 

John J. McCook, Trinity College, Hartford, Conn. 

Nelson G. McCrea, Columbia College, New York City. 


« 


Ix American Philological Association. 


J. H. McDaniels, Hobart College, Geneva, N. Y. 

George F. McKibben, Denison University, Granville, O. 

Harriet E. McKinstry, Lake Erie Female Seminary, Painesville, O. 

H. Z. McLain, Wabash College, Crawfordsville, Ind. 

W. J. McMurtry, Yankton College, Yankton, South Dakota. 

George F. Mellen, University of Tennessee, Knoxville, Tenn. 

Augustus C. Merriam, Columbia College, New York, N. Y. 

Elmer T. Merrill, Wesleyan University, Middletown, Conn. e 
William A. Merrill, of Indiana University, Bloomington, Ind. 

J. C. Metcalf, Soule College, Murfreesboro, Tenn. 

Arthur B. Milford, Wabash College, Crawfordsville, Ind. 

C. W.-E. Miller, Johns Hopkins University, Baltimore, Md. 

Charles A. Mitchell, University School, Cincinnati, O. 

A. P. Montague, Columbian University, Washington, D. C. (1514 Corcoran St.). 
Clifford H. Moore, Phillips Academy, Andover, Mass. 

Frank G. Moore, Dartmouth College, Hanover, N. H. 

George.F. Moore, Theological Seminary, Andover, Mass. 

J. Leverett Moore, Vassar College, Poughkeepsie, N. Y. 

Morris H. Morgan, Harvard University, Cambridge, Mass. 

Edward P. Morris, Yale University, New Haven, Conn. 

Frederick S. Morrison, High School, Hartford, Conn. 

Augustus T. Murray, Leland Stanford Jr. University, Palo Alto, Cal. 

Wilfred P. Mustard, Haverford College, Haverford, Pa. 

Francis Philip Nash, Hobart College, Geneva, N. Y. 

Barker Newhall, Brown University, Providence, R. I. 

Frank W. Nicolson, Wesleyan University, Middletown, Conn. 

Edward North, Hamilton College, Clinton, N. Y. 

W. B. Owen, Lafayette College, Easton, Pa. 

William A. Packard, College of New Jersey, Princeton, N. J. 

Arthur H. Palmer, Yale University, New Haven, Conn. (42 Mansfield St.). 
Charles P. Parker, Harvard University, Cambridge, Mass. 

W. H. Parks, City of Creede, Amethyst P. O., Colorado. . 
James M. Paton. 

Endicott Peabody, Groton School, Groton, Mass. 

Calvin W. Pearson, Beloit College, Beloit, Wis. ' 
Ernest M. Pease, Leland Stanford Jr. University, Palo Alto, Cal. . 
Henry T. Peck, Columbia College, New York, N. Y. ‘ 
Tracy Peck, Yale University, New Haven, Conn. (124 High St.). . 
S. Frances Pellett, University of.Chicago, Chicago, III. 
Emma M. Perkins, College for Women, Cleveland, O. 
Bernadotte Perrin, Yale University, New Haven, Conn. ; 
Edward D. Perry, Columbia College, New York, N. Y. 

William E. Peters, University of Virginia, Charlottesville, Va. 

Edward E. Phillips, Marietta College, Marietta, O. 

William Henry P. Phyfe, 12 East Forty-third St., New York, N. Y. 
William T, Piper, Cambridge, Mass. (179 Brattle St.). 

Samuel B. Platner, Adelbert College, Cleveland, O. 

William Carey Poland, Brown University, Providence, R. I. (7 Cooke St.). 


_—s 


i 





Proceedings for July, 2 Ixi 


John Pollard, Richmond College, Biintond, Va. 
Samuel Porter, National Deaf-Mute College, Washington, D. C, 
William Porter, Beloit College, Beloit, Wis. 
Edwin Post, De Pauw University, Greencastle, Ind. 
L. S. Potwin, Adelbert College, Cleveland, O. (2108 Euclid Ave.). 
John W. Powell, Washington, D. C. 
Henry Preble, 42 Stuyvesant Place, New Brighton, Staten Island, N. Y. 
Thomas R. Price, Columbia College, New York, N. Y. (263 West 45th St.). 
Sylvester Primer, University of Texas, Austin, Tex. 
Benjamin F, Prince, Wittenberg College, Springfield, O. 
Louis Dwight Ray, 2125 Fifth Ave., New York, N. Y. 
John W. Redd, Centre College, Danville, Ky. 
Horatio M. Reynolds, Yale University, New Haven, Conn. 
George M. Richardson, University of California, Berkeley, Cal. 
Louisa H. Richardson, Carleton College, Northfield, Minn. 
Arthur W. Roberts, William Penn Charter School, Philadelphia, Pa. 
Harley F. Roberts, Yale University, New Haven, Conn. 
David H. Robinson, University of Kansas, Lawrence, Kan. 
W. A. Robinson, Lehigh University, So. Bethlehem, Pa. 
F, E. Rockwood, Bucknell University, Lewisburg, Pa. 
George Rodemann, Bedford Heights Institute, 63 New York Ave., Brooklyn, N. Y. 
John C. Rolfe, University of Michigan, Ann Arbor, Mich. 
Julius Sachs, Classical School, 38 West Fifty-ninth St., New York, N. Y.. 
Benjamin H. Sanborn, Wellesley, Mass. 
Thomas F. Sanford, 513 Winthrop Ave., New Haven Conn. 
W. S. Scarborough, Wilberforce University, Wilberforce, O. 
Charles P. G. Scott, 57 High Street, Yonkers, N. Y. 
Henry S. Scribner, Western University of Pennsylvania, Allegheny City, Pa. 
Helen M. Searles, Ferry Hall, Lake Forest, Ill. 
Charles D. Seely, State Normal School, Brockport, N. Y. 
William J. Seelye, Wooster University, Wooster, O. 
J. B. Sewall, Thayer Academy, South Braintree, Mass. 
T. D. Seymour, Yale University, New Haven, Conn. (34 Hillhouse Ave.). 
J. A. Shaw, Highland Military Academy, Worcester, Mass. 
Edward S. Sheldon, Harvard University, Cambridge, Mass. 
William D, Shipman, Buchtel College, Akron, O. 
Paul Shorey, University of Chicago, Chicago, Ill. 
‘Henry A. Short, Columbia College, New York, N. Y. 
Helen W. Shute, Bryn Mawr College, Bryn Mawr, Pa. 
Mary A. Shute, 591 Orange Street, New Haven, Conn. 
E. G. Sihler, University of the City of New York, N. Y. 
M. S. Slaughter. 
Charles Forster Smith, Vanderbilt University, Nashville, Tenn. 
Clement L. Smith, Harvard University, Cambridge, Mass. 
Frank W. Smith, State Normal School, Westfield, Mass. 
Josiah R. Smith, Ohio State University, Columbus, O. 
Richard M. Smith, Randolph-Macon College, Ashland, Va. 
Albert H. Smyth, 126 North Twenty-second St., Philadelphia, Pa. 


- 


Ixii American Philological Association. 


Herbert Weir Smyth, Bryn Mawr College, Bryn Mawr, Pa. 

H. A. Sober, University of Wisconsin, Madison, Wis. (27 Mendots Court). 

George C. S. Southworth, Salem, Col. Co., O. 

Edward H. Spieker, Johns Hopkins University, Baltimore, Md. _ 

W. O. Sproull, University of Cincinnati, Cincinnati, O. 

Jonathan Y. Stanton, Bates College, Lewiston, Me. : \ 

J. R. S. Sterrett, Amherst College, Amherst, Mass. 

Edward F. Stewart, Easton, Pa. 

Austin Stickney, 35 West Seventeenth St., New York, N. Y. 

F. H. Stoddard, University of the City of New York, N. Y. 

Lewis Stuart, Lake Forest University, Lake Forest, Ill. 

Charles W. Super, Ohio University, Athens, O. 

Marguerite Sweet, Vassar College, Poughkeepsie, N. Y. 

Frank B. Tarbell, University of Chicago, Chicago, Ill. 

Julian D. Taylor, Colby University, Waterville, Me. 

John Tetlow, Girls’ High and Latin Schools, Boston, Mass, 

J. Henry Thayer, Harvard University, Cambridge, Mass. (67 Sparks St.). 

G. V. Thompson, Yale University, New Haven, Conn. 

William E. Thompson, Hamline University, Hamline, Minn. 

Fitz Gerald Tisdall, College of the City of New York (Buckingham Hotel, Fifth 
Avenue and Fiftieth St.), N. Y. ‘ 

Henry A. Todd, Columbia College, New York City. 

H. C. Tolman, University of North Carolina, Chapel Hill, N. C. 

Edward M. Tomlinson, Alfred University, Alfred Centre, N. Y. 

James A. Towle, Westminster School, Dobbs Ferry, N. Y. 

J. Hammond Trumbull, Hartford, Conn. 

Frank L. Van Cleef, University of Wisconsin, Madison, Wis. 

Addison Van Name, Yale University, New Haven, Conn. 

George M. Wahl, Williams College, Williamstown, Mass. 

W. H. Wait. 

John H. Walden, Harvard University, Cambridge, Mass. (13 Mt. Auburn St.). 

Benjamin B. Warfield, Princeton Theological Seminary, Princeton, N. J. 

Andrew McCorrie Warren, Brown, Shipley & Co., Founders Court, London. 

' Frederic M. Warren, Adelbert College, Cleveland, O. 

Henry C. Warren, 28 Quincy St., Cambridge, Mass. 

Minton Warren, Johns Hopkins University, Baltimore, Md. 

William E. Waters, University of Cincinnati, Cincinnati, O. . 

Helen L. Webster, Wellesley College, Wellesley, Mass. 

Andrew F. West, College of New Jersey, Princeton, N. J. 

J. H. Westcott, College of New Jersey, Princeton, N. J. 

J. B. Weston, Christian Biblical Institute, Stanfordville, N. Y. 

L. B. Wharton, College of William and Mary, Williamsburg, Va. q 

Albert S. Wheeler, Sheffield Scientific School, New Haven, Conn. 

Benjamin I. Wheeler, Cornell University, Ithaca, N. Y. 

James R. Wheeler, University of Vermont, Burlington, Vt. 

G. M. Whicher, Packer Collegiate Institute, Brooklyn, N. Y. 

Andrew C, White, Cornell University, Ithaca, N. Y. 

John Williams White, American School of Classical Studies, Athens, Greece. 





_——- 





Proceedings for July, 1893. Ixili 


Henry Whitehorn, Union University, Schenectady, N. Y. 

William Dwight Whitney, Yale University, New Haven, Conn. 

B. Lawton Wiggins, University of the South, Sewanee, Tenn. 

John R. Wightman, Iowa College, Grinnell, Ia. 

Alexander M. Wilcox, University of Kansas, Lawrence, Kans. 

Charles R. Williams, Indianapolis, Ind. 

Charles Tudor Williams, 871 Case Ave., Cleveland, O. 

George A. Williams, Vermont Academy, Saxton’s River, Vt. 

George T. Winston, University of North Carolina, Chapel Hill, N. C. 
E. L. Wood, Amherst College, Amherst, Mass. 

Henry Wood, Johns Hopkins University, Baltimore, Md. 

Frank E. Woodruff, Bowdoin College, Brunswick, Me. 

B. D. Woodward, Columbia College, New York City. 

Henry P. Wright, Yale University, New Haven, Conn. (128 York St.). 
John Henry Wright, Harvard University, Cambridge, Mass, 

Clarence H. Young, 308 West Fifty-eighth St., New York City. ' 
A. C, Zenos, McCormick Theological Seminary, Chicago, IIl. 


[Number of Members, 378.] 





THE FOLLOWING LIBRARIES AND INSTITUTIONS (ALPHABETIZED BY TOWNS) 
SUBSCRIBE FOR THE ANNUAL PUBLICATIONS OF THE ASSOCIATION. 


Akron, O.: Buchtel College Library. 

Amherst, Mass.: Amherst College Library. 
Andover, Mass.: Theological Seminary Library. 
Ann Arbor, Mich.: Michigan University Library. 
Baltimore, Md.: Johns Hopkins University Library. 
Baltimore, Md.: Peabody Institute. 

Berea, Madison Co., Ky.: Berea College Library. 
Berkeley, Cal.: University of California Library. 
Boston, Mass.: Boston Athenzeum. 

Boston, Mass.: Boston Public Library. 

Brooklyn, N. Y.: The Brooklyn Library. 
Brunswick, Me.: Bowdoin College Library. 

Bryn Mawr, Pa.: Bryn Mawr College Library.. 
Buffalo, N. Y.: The Buffalo Library. 

Burlington, Vt.: Library of the University of Vermont. 
Cambridge, Mass.: Harvard College Library. 
Champaign, Ill.: University of Illinois Library. 
Chicago, Ill.: The Newberry Library. 

Chicago, Ill.: Public Library. 

Cincinnati, O.: Public Library. 

Cleveland, O.: Library of Adelbert College of Western Reserve University. 
College Hill, Mass.: Tufts College Library. 
Columbus, O.: Ohio State University Library. 
Crawfordsville, Ind.: Wabash College Library. 


- 


lxiv American Philological Association. 


Detroit, Mich.: Public Library. 

Easton, Pa.: Lafayette College Library. 

Evanston, Ill.: Northwestern University Library. 

Gambier, O.: Kenyon College Library. 

Geneva, N. ¥.: Hobart College Library. 

Greencastle, Ind.: De Pauw University Library. 

Hanover, N. H.: Dartmouth College Library. 

lowa City, Ia.: Library of State University. 

Ithaca, N. Y.: Cornell University Library. 

Lincoln, Neb.: Library of State University of Nebraska. 

Madison, Wis.: State Historical Society. 

Marietta, O.: Marietta College Library. 

Middletown, Conn.: Wesleyan University Library. : 

Milwaukee, Wis.: Public Library. 

Minneapolis, Minn.: Athenzeum Library. 

Nashville, Tenn.: Vanderbilt University Library. ° 

Newton Centre, Mass.: Library of Newton Theological Institution. 

New York, N. Y.: Astor Library. 

New York, N. Y.: Library of Columbia College. 

New York, N. Y.: Library of the College of the City of New York (Lexington 
Ave. and Twenty-third St). 

New York, N. Y.: Union Theological Seminary Library (1200 Park Ave.). 

Olivet, Eaton Co., Mich.: Olivet College Library. 

Philadelphia, Pa.: American Philosophical Society. 

Philadelphia, Pa.: The Library Company of Philadelphia. 

Philadelphia, Pa.: The Mercantile Library. 

Philadelphia, Pa.: University of Pennsylvania Library. 

Poughkeepsie, N. Y.: Vassar College Library. 

Providence, R. I.: Brown University Library. 

Rochester, N. Y.: Rochester University Library. 

Springfield, Mass.: City Library. 

Tokio, Japan: Library of Imperial University. 

University of Virginia, Albemarle Co., Va.: University Library. 

Washington, D. C.: Library of Congress. 

Washington, D. C.: United States Bureau of Education. 

Waterbury, Conn.: Silas Bronson Library. 

Waterville, Me.: Colby University Library. 

Wellesley, Mass.: Wellesley College Library. 

Worcester, Mass.: Free Public Library. 


[Number of subscribing institutions, 62. ] 


Proceedings for July, 1893. Ixv 


To THE FOLLOWING LIBRARIES AND INSTITUTIONS HAVE BEEN SENT COMPLETE 
SETS OF THE TRANSACTIONS, GRATIS, 


American School of Classical Studies, Athens, Greece. 
British Museum, London, England. 

Royal Asiatic Society, London. 

Philological Society, London, 

Society.of Biblical Archzology, London. 

Indian Office Library, London. 

Bodleian Library, Oxford. 

University Library, Cambridge, England. 

Advocates’ Library, Edinburgh, Scotland. 

Trinity College Library, Dublin, Ireland. 

Asiatic Society of Bengal, Calcutta. 

Bombay Branch of the Royal Asiatic Society. =, 
North-China Branch of the Royal Asiatic Society, Shanghai. 
Japan Asiatic Society, Yokohama. 

Public Library of Victoria, Melbourne, Australia. 

Sir George Grey’s Library, Cape Town, Africa. 
Reykjavik College Library, Iceland. 

University of Christiania, Norway. 

University of Upsala, Sweden. 

Russian Imperial Academy, St. Petersburg. 

Austrian Imperial Academy, Vienna. 

Anthropologische Gesellschaft, Vienna. 

Biblioteca Nazionale, Florence, Italy. 

Reale Accademia delle Scienze, Turin. 

Société Asiatique, Paris, France. 

Athénée Oriental, Louvain, Belgium. 

Curatorium of the University, Leyden, Holland. 
Bataviaasch Genootschap van Kunsten en Wetenschappen, Batavia, Java. 
Royal Prussian Academy of Sciences, Berlin, Germany. 
Royal Saxon Academy of Sciences, Leipsic. 

Royal Bavarian Academy of Sciences, Munich. 
Deutsche Morgenlindische Gesellschaft, Halle. 

Library of the University of Bonn. 

Library of the University of Jena. 

Library of the University of KGénigsberg. 

Library of the University of Leipsic. 

Library of the University of Tiibingen. 


Smithsonian Institution, Washington, D. C. 


{Number of foreign institutions, 37.] 
[Total (379 + 61 + 37 +1 =), 478.] 


CONSTITUTION 
OF THE 


AMERICAN PHILOLOGICAL ASSOCIATION. 


ARTICLE I.— NAME AND OBJECT. 


1. This Society shall be known as “The American Philological Associa- 
tion.” 

2. Its object shall be the advancement and diffusion of philological knowl- 
edge. , 


ARTICLE II. — OFFICERS. 


1. The officers shall be a President, two Vice-Presidents, a Secretary and 
Curator, and a Treasurer. ; 

2. There shall be an Executive Committee of ten, composed of the above 
officers and five other members of the Association. 

3. All the above officers shall be elected at the last session of each annual 
meeting. 


ARTICLE III. — MEETINGS. 


1. There shall be an annual meeting of the Association in the city of New 
York, or at such other place as at a preceding annual meeting shall.be deter- 
mined upon. 

2. At the annual meeting, the Executive Committee shall present an annual 
report of the progress of the Association. 

3. The general arrangements of the proceedings of the annual meeting shall 
be directed by the Executive Committee. 

4. Special meetings may be held at the call of the Executive Committee, when 
and where they may decide. 


Ixvi 


_—S ~ 


Proceedings for July, 1893. Ixvii 


ARTICLE IV.— MEMBERS. 


1. Any lover of philological studies may become a member of the Association 
by a vote of the Executive Committee and the payment of five dollars as initiation 
fee, which initiation fee shall be considered the first regular annual fee. 

2. There shall be an annual fee of three dollars from each member, failure in 
payment of which for two years shall ipso facto cause the membership to cease. 

3. Any person may become a life member of the Association by the payment 
of fifty dollars to its treasury, and by vote of the Executive Committee, 


ARTICLE V.— SUNDRIES. 


1. All papers intended to be read before the Association must be submitted 
to the Executive Committee before reading, and their decision regarding such 
papers shall be final. : 

2. Publications of the Association, of whatever kind, shall be made only under 
the authorization of the Executive Committee. : 


ARTICLE VI.— AMENDMENTS. 


- Amendments to this Constitution may be made by a vote of two-thirds of 
those present at any regular meeting subsequent to that in which they have been 
proposed, 





PUBLICATIONS OF THE ASSOCIATION. 





THE annually published “ Proceedings” of the American Philo- 
logical Association contain an account of the doings at the annual 
meeting, brief abstracts of the papers read, reports upon the progress 
of the Association, and lists of its officers and members. 


The annually published “Transactions” give the full text of such 
“articles as the Executive Committee decides to publish. The Pro- 
ceedings are bound with them as an Appendix. 


The following tables show the authors and contents of the volumes 
of Transactions thus far published : — 


1869-1870. — Volume I. 


Hadley, J.: On the nature and theory of the Greek accent. 

Whitney, W. D.: On the nature and designation of the accent in Sanskrit. 

Goodwin, W. W.: On the aorist subjunctive and future indicative with drws and 
ov ph. 

Trumbull, J. Hammond: On the best method of studying the North American 
languages. 

Haldeman, S. S.: On the German vernacular of Pennsylvania. 

Whitney, W. D.: On the present condition of the question as to the origin of 
language. 

Lounsbury, T. R.: On certain forms of the English verb which were used in the 
sixteenth and seventeenth centuries. 

Trumbull, J. Hammond: On some mistaken notions of Algonkin grammar, and 
on mistranslations of words from Eliot’s Bible, etc. 

Van Name, A.: Contributions to Creole Grammar. 


Proceedings of the preliminary meeting (New York, 1868), of the first annual 
session (Poughkeepsie, 1869), and of the second annual session (Rochester, 
1870). 

1871.— Volume IL. 


Evans, E. W.: Studies in Cymric philology. 

Allen, F. D.: On the so-called Attic second declension. 

Whitney, W. D.: Strictures on the views of August Schleicher respecting the 
nature of language and kindred subjects. 

Hadley, J.: On English vowel quantity in the thirteenth century and in the nine- 
teenth. 

March, F. A. : Anglo-Saxon and Early English pronunciation. 

Bristed, C. A.: Some notes on Ellis’s Early English Pronunciation. 


lxviii 





i; 2% “ ft aoe ee ae 


— . le a 


Proceedings for July, 1893. Ixix 


Trumbull, J. Hammond: On Algonkin names for man. 
Greenough, J. B.: On some forms of:conditional sentences in Latin, Greek, and 
Sanskrit. 


Proceedings of the third annual session, New Haven, 1871, 


1872. — Volume III. 


Evans, E. W.: Studies in Cymric philology. 

Trumbull, J. Hammond: Words derived from Indian languages of North 
America, , 

Hadley, J.: On the Byzantine Greek pronunciation of the tenth century, as illus 
trated by a manuscript in the Bodleian Library. 

Stevens, W. A.: On the substantive use of the Greek participle. 

Bristed, C. A.: Erroneous and doubtful uses of the word szch. 

Hartt, C. F.: Notes on the Lingoa Geral, or Modern Tupi of the Amazonas. 

Whitney, W. D.: On material and form in language. 

March, F. A.: Is there an Anglo-Saxon language? 

March, F. A.: On some irregular verbs in Anglo-Saxon. 

Trumbull, J. Hammond: Notes on forty versions of the Lord’s Prayer in Algon- 
kin languages. 


‘ 


Proceedings of the fourth annual session, Providence, 1872. 


1873. — Volume IV. 


Allen, F. D.: The Epic forms of verbs in dw. 

Evans, E. W.: Studies in Cymric philology. 

Hadley, J.: On Koch’s treatment of the Celtic element in English. 

Haldeman, S. S.: On the pronunciation of Latin, as presented in several recent 
grammars. 

Packard, L. R.: On some points in the life of Thucydides. 

Goodwin, W. W.: On the classification of conditional sentences in Greek syntax. 

March, F. A.: Recent discussions of Grimm’s law. 

Lull, E. P.: Vocabulary of the language of the Indians of San Blas and Cale- 
donia Bay, Darien. 


Proceedings of the fifth annual session, Easton, 1873. 


1874.— Volume V. 


Tyler, W. S.: On the prepositions in the Homeric poems. 
Harkness, A.: On the formation of the tenses for completed action in the Latin 
_ finite verb. 

Haldeman, S. S.: On an English vowel-mutation, present in cag, keg. 

Packard, L. R.: On a passage in Homer’s Odyssey (A 81-86). 

Trumbull, J. Hammond: On numerals in American Indian languages, and the 
Indian mode of counting. 

Sewall, J. B.: On the distinction between the subjunctive and optatives modes in 
Greek conditional sentences. 

Morris, C. D.: On the age of Xenophon at the time of the Anabasis. 

Whitney, W. D.: édce: or 0éce: — natural or conventional? 


Proceedings of the sixth annual session, Hartford, 1874. 





Ixx American Philological Association. 


1875. — Volume VL . 


Harkness, A.: On the formation of the tenses for completed action in the Latin 
finite verb. 

Haldeman, S. S.: On an English consonant-mutation, present in proof, prove. 

Carter, F.: On Begemann’s views as to the weak preterit of the Germanic verbs. 

Morris, C. D.: On some forms of Greek conditional sentences. 

Williams, A.: On verb-reduplication as a means of expressing completed action. 

Sherman, L. A.: A grammatical analysis of the Old English poem “The Owl 
and the Nightingale.” 


Proceedings of the seventh annual session, Newport, 1875. 


1876. — Volume VII. 


Gildersleeve, B. L.: On ei with the future indicative and édy with the subjunctive 
in the tragic poets. 

Packard, L. R.: On Grote’s theory of the structure of the Iliad. 

Humphreys, M. W.: On negative commands in Greek. 

Toy, C. H.: On Hebrew verb-etymology. 

Whitney, W. D.: A botanico-philological problem. 

Goodwin, W. W.: On shad/ and shoudd in protasis, and their Greek equivalents. 

Humphreys, M. W.: On certain influences of accent in Latin iambic trimeters. 

Trumbull, J. Hammond: On the Algonkin verb. 

Haldeman, S. S.: On a supposed mutation between / and z. 


Proceedings of the eighth annual session, New York, 1876. 


1877.— Volume VIII. 


Packard, L. R.: Notes on certain passages in the Phaedo and the Gorgias of 
Plato. 

Toy, C. H.: On the nominal basis on the Hebrew verb. 

Allen, F. D.: On a certain apparently pleonastic use of ‘és. 

Whitney, W. D.: On the relation of surd and sonant. 

Holden, E. S.: On the vocabularies of children under two years of age. 


Goodwin, W. W.: On the text and interpretation of certain passages in the 


Agamemnon of Aeschylus. 
Stickney, A.: On the single case-form in Italian. 
Carter, F.: On Willmann’s theory of the authorship of the Nibelungenlied. 
Sihler, E. G.: On Herodotus’s and Aeschylus’s accounts of the battle of Salamis. 
Whitney, W. D.: On the principle of economy as a phonetic force. 
Carter, F.: On the Kiirenberg hypothesis. 
March, F. A.: On dissimilated gemination. 


Proceedings of the ninth annual session, Baltimore, 1877. 


1878.— Volume IX. 


Gildersleeve, B. L.: Contributions to the history of the articular infinitive 
Toy, C. H.: The Yoruban language. * 

Humphreys, M. W.: Influence of accent in Latin dactylic hexameters. 
Sachs, J.: Observations on Piato’s Cratylus. 


~~. . +.” s 


EEE — a 





Proceedings for July, 1893. Ixxi 


Seymour, T. D.: On the composition of the Cynegeticus of Xenophon. 
Humphreys, M. W.: Elision, especially in Greek. 
Proceedings of the tenth annual session, Saratoga, 1878, 


1879.— Volume X. 


Toy, C. H.: Modal development of the Semitic verb. 

Humphreys, M. W.: On the nature of caesura. 

Humphreys, M. W.: On certain effects of elision. 

Cook, A. S.: Studies in Heliand. 

Harkness, A.: On the development of the Latin subjunctive in principal clauses. 
D’Ooge, M. L.: The original recension of the De Corona. 

Peck, T.: The authorship of the Dialogus de Oratoribus. 

Seymour, T. D.: On the date of the Prometheus of Aeschylus. 


Proceedings of the eleventh annual session, Newport, 1879. 


1880. — Volume XI. 


Humphreys, M. W.: A contribution to infantile linguistic. 

Toy, C. H.: The Hebrew verb-termination zz. 

Packard, L. R.: The beginning of a written literature in Greece. 

Hall, I. H.: The declension of the definite article in the Cypriote inscriptions. 

Sachs, J.: Observations on Lucian. : 

Sihler, E. G.: Virgil and Plato. 

Allen, W. F.: The battle of Mons Graupius. 

Whitney, W. D.: On inconsistency in views of language. 

Edgren, A. H.: The kindred Germanic words of German and English, exhibited 
with reference to their consonant relations. 


Proceedings of the twelfth annual session, Philadelphia, 1880. 


1881.— Volume XII. 


Whitney, W. D.: On Mixture in Language. 

Toy, C. H.: The home of the primitive Semitic race. 

March, F. A.: Report of the committee on the reform of English spelling. 
Wells, B. W.: History of the a-vowel, from Old Germanic to Modern English. 
Seymour, T. D.: The use of the aorist participle in Greek. 

Sihler, E. G.: The use of abstract verbal nouns in -o1s in Thucydides. 


Proceedings of the thirteenth annual session, Cleveland, 1881. 


1882. — Volume XIII. 


Hall, I. H.: The Greek New Testament as published in America, 

Merriam, A. C.: Alien intrusion between article and noun in Greek. 

Peck, T.: Notes on Latin quantity. 

Owen, W. B.: Influence of the Latin syntax in the Anglo-Saxon Goagele 
Wells, B. W.: The Ablaut in English. 

Whitney, W. D.: General considerations on the Indo-European case-system. 


Proceedings of the fourteenth annual session, Cambridge, 1882. 


xxii American Philological Association. 


1883.— Volume XIV. 


Merriam, A. C.: The Caesareum and the worship of Augustus at Alexandria. © 
Whitney, W. D.: The varieties of predication. 

Smith, C. F.: On Southernisms. 

Wells, B. W.: The development of the Ablaut in Germanic. 


Proceedings of the fifteenth annual session, Middletown, 1883. 


1884.— Volume XV. 


Goodell, T. D.: On the use of the Genitive in Sophokles. 

Tarbell, F. B.: Greek ideas as to the effect of burial on the future life of the soul, 

Perrin, B.: The Crastinus episode at Palaepharsalus. 

Peck, T.:. Alliteration in Latin. < 

Von Jagemann, H. C. G.: Norman words in English. 

Wells, B. W.: The Ablaut in High German. 

Whitney, W. D.: Primary and Secondary Suffixes of Derivation and their ex- 
changes. 

Warren, M.: On Latin Glossaries. Codex Sangallensis, No. 912. 


Proceedings of the sixteenth annual session, Hanover, 1884. 


1885.— Volume XVI. 


Easton, M. W.: The genealogy of words. 

Goodell, T. D.: Quantity in English verse. 

Goodwin, W. W.: Value of the Attic talent in modern money. 

Goodwin, W. W.: Relation of the Mpéedpa to the Mpurdve:s in the Attic BovA 
Perrin, B.: Equestrianism in the Doloneia. 

Richardson, R. B.: The appeal to sight in Greek tragedy. ; 

Seymour, T. D.: The feminine caesura in Homer. 

Sihler, E. G.: A study of Dinarchus. 

Wells, B. W.: The vowels e and z in English. 

Whitney, W. D.: The roots of the Sanskrit langnage. 


Proceedings of the seventeenth annual session, New Haven, 1885. 


1886.— Volume XVII. 


Tarbell, F. B.: Phonetic law. 

Sachs, J.: Notes on Homeric Zodlogy. 

Fowler, H. N.: The sources of Seneca de Beneficiis. 

Smith, C. F.: On Southernisms. 

Wells, B. W.: The sounds o and # in English. 

Fairbanks, A.: The Dative case in Sophokles. 

The Philological Society, of England, and The American Philological Associa 
tion: Joint List of Amended Spellings. 


Proceedings of the eighteenth annual session, Ithaca, 1886. 





Proceedings for July, 1893. ~ Ixxiii 


1887.— Volume XVIII. : 


Allen, W. F.: The monetary crisis in Rome, A.D. 33. 

Sihler, E. G.: The tradition of Czesar’s Gallic Wars, from Cicero to Orosius, 
Clapp, E. B.: Conditional sentences in Aischylos, 

Pease, E. M.: On the relative value of the manuscripts of Terence. 

Smyth, H. W.: The Arcado-Cyprian dialect. 

Wells, B. W.: The sounds 0 and z# in English. 

Smyth, H. W.: The Arcado-Cyprian dialect. — Addenda. 


Proceedings of the nineteenth annual session, Burlington, 1887. 


1888.— Volume XIX. 


Allen, W. F.: The Zex Curiata de Imperio. 

Goebel, J.: On the impersonal verbs. 

Bridge, J.: On the authorship of the Cynicus of Lucian. 

Whitney, J. E.: The “Continued Allegory” in the first book of the Fairy Queené, 
March, F. A.: Standard English: its pronunciation, how learned. 

Brewer, F. P.: Register of new words. 


Proceedings of the twentieth annual session, Amherst, 1888. 


1889.— Volume XX. 


Smyth, H. W.: The vowel system of the Ionic diaiect. 

Gudeman, A.: A new source in Plutarch’s Life of Cicero. 

Gatschet, A. S.: Sex-denoting nouns in American languages. " 

Cook, A. S.: > Metrical observations on a Northumbrianized version of the Old 
English Judith. 

Cook, A. S.: Stressed vowels in Aflfric’s Homilies. 

Proceedings of the twenty-first annual session, Easton, 1889. 

Index of authors, and index of subjects, Vols. I—XX. 


1890.— Volume XXI. 


Goodell, T. D.: The order of words in Greek. 

Hunt, W. I.: Homeric wit and humor. 

Leighton, R. F.: The Medicean Mss. of Cicer6’s letters. 
Whitney, W. D.: Translation of the Katha Upanishad. 


Proceedings of the twenty-second annual session, Norwich, 1890. 


1891.— Volume XXII. 


Capps, Edw.: The Greek Stage according to the Extant Dramas 

Clapp, Edw. B.: Conditional Sentences in the Greek Tragedians. 

West, A. F.: Lexicographical Gleanings from the P%z/odib/on of Richard de Bury. 
Hale, W. G.: The Mode in the phrases guod sciam, etc. 


Proceedings of the twenty-third annual session, Princeton, 1891. 





Ixxiv American Philological Assoctation. 


1892.— Volume XXIII. 


Whitney, W. D.: On the narrative use of imperfect and perfect in the Brah- 
manas. 

Muss-Arnolt, W.: On Semitic words in Greek and Latin. 

Humphreys, M. W.: On the equivalence of rhythmical bars and metrical feet. 

Scott, Charles P. G.: English words which hav gaind or lost an initial con- 
sonant by attraction. 3 

Proceedings of the twenty-fourth session, Charlottesville, Va. 


1893. — Volume XXIV. 


Sonnenschein, E. A.: The scientific emendation of classical texts. = 

Bréal, M.: The canons of etymological investigation. 

Streitberg, W.: Ein Ablautproblem der Ursprache. 

Osthoff, H.: Dunkles und helles 7/im Lateinischen. 

Shorey, Paul: The implicit ethics and psychology of Thucydides. 

Scott, C. P. G.: English words which hav gaind or lost an initial consonant 
by-attraction (second paper). 

Hale, W. G.: “ Extended” and “remote” deliberatives in Greek. 

Proceedings of the twenty-fifth session, Chicago, IIl. 


The Proceedings of the American Philological Association are dis- 
tributed gratis upon application to the publishers until they are out 
of print. 

Separate copies of articles printed in the Transactions are given to 
the authors for distribution. 

The “ Transactions for” any given year are not always published 
in that year. To avoid mistakes in ordering back volumes, please 
state — not the year of publication, but rather—the year for which 
the Transactions are desired, adding also the volume-number, accord- 
ing to the following table : — 


The Transactions for 1869 and The Trans. for 1882 form Vol. XIII. 
1870 form Vol. I. ss ‘ike aiameg 0 Pee $6) EVE 
The Trans. for 1871 “ .  , iL. = = 1884 “ te Ng 
“ “ 1872 “ “ IRI “ “é 1885 “ “ XVI. 
“ “ 1873 “ “ec IV. “ “ 1886 “ “ XVII. 
“ ‘“ 1874 ‘“ cy RAF ge “ “ 1887 “ « XVIIL 
“ “ a 875 “ ee Mil “ “ 1888 “ “ XxX 
“ “ I 876 “ “ VIL. “ “ I 889 “ “ xx 
“ “ 1877 “ “ VIII. ’ ‘“ “ 1890 “ c XX 
“ “ 1878 “ SES “ “ 1891 “ “ XXTE 
“ “ 1879 “ ee, “ “ 1892 “ “ XXTiE 
ae (::.”-  .« & a «= 4893 4 > -«. Xxae 
Pg ts) esa Sa 





Proceedings for July, 1893. xxv 


The price of these volumes is $2.00 apiece, except Volumes XV., 
XX., and XXIII, for which $2.50 is charged. ‘The first two volumes 
will not be sold separately. A charge of fifty cents is made for the 
Index of Authors and Index of Subjects to Vols. I.—XX. 


REDUCTION IN THE PRICE OF COMPLETE SETS. 


Single COMPLETE SETS of the Transactions will be sold, until 
further notice, at forty-two dollars a set. 


It is especially appropriate that American Libraries should exert themselves to 
procure this series while it may be had. It is the work of American scholars, 
and contains many valuable articles not elsewhere accessible; and, apart from 
these facts, as the first collection of essays in general philology made in this 
country, it is sure to be permanently valuable for the history of American scholar- 
ship. 














; 
4 


‘BINDING SECT. NOV2 5 1970 











<< S i: 
Fig 
thea oh) 

” 


ae 

a 

‘ Vest haat 
re WE 


fi 


PT tae) 
$e 
ee 





Manat 

plateaus 

; 2 Sees 
LA Pe 

J pines ary Wh 


sue 
Vee 


& 
+ 

Hy 
% 






































ry i 
aN > 
) 


ipa 
sons 


Ri et oe 
ee eek 
abst .8 3 


vat 


Hate!