This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project to make the world’s books discoverable online.
It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books are our gateways to the past, representing a wealth of history, culture and knowledge that’s often difficult to discover.
Marks, notations and other marginalia present in the original volume will appear in this file - a reminder of this book’s long journey from the publisher to a library and finally to you.
Usage guidelines Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing this resource, we have taken steps to prevent abuse by commercial parties, including placing technical restrictions on automated querying.
We also ask that you:
+ Make non-commercial use of the files We designed Google Book Search for use by individual personal, non-commercial purposes.
and we request that you use these files for
+ Refrain from automated querying Do not send automated queries of any sort to Google’s system: If you are conducting research on machine translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the use of public domain materials for these purposes and may be able to help.
+ Maintain attribution The Google “watermark” you see on each file is essential for informing people about this project and helping them find additional materials through Google Book Search. Please do not remove it.
+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other countries. Whether a book is still in copyright varies from country to country, and we can’t offer guidance on whether any specific use of any specific book is allowed. Please do not assume that a book’s appearance in Google Book Search means it can be used in any manner anywhere in the world. Copyright infringement liability can be quite severe.
About Google Book Search
Google’s mission is to organize the world’s information and to make it universally accessible and useful. Google Book Search helps readers discover the world’s books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web ai[http: //books . google. com/|
NS
AN
ASSOCIATION,
869-70.
PUBLISHED BY THE ASSOCIATION. _TED BY CASE, LOCKWOOD & BRAINARD, HARTFORD.
1871.
Pi AS tN Set yl-2 CONTHN TS.
I. On the Nature and Theory of the Greek Accent. By Pro-
fessor JAMES IITADLEY. . . ; 1 II. On the Nature and Designation of the Accent in Sanskrit.
By Professor WiLt1amM D. WulITNEY. . . . . 20 Ill. On the Aorist Subjunctive and Future Indicative with "Owe
and Ov ph. By Professor WiLtL1aAM W. Goopwin. 46 IV. On the best Method of Studying the North American Lan-
guages. By J. LTAMMOND TRUMBULL. . ; . 55 V. On the German Vernacular of Pennsylvania, By Professor
S.S. HALDEMAN. . . . . . . ° 80 VI. On the Present Condition of the Question as to the Origin of
Lanvuage. By Professor Wittiam D. WHITNEY. . 84
VII. On Certain Forms of the English,Verb which were used in the Sixteenth and Seventeenth Centuries. By THomas R. LOUNSBURY. . . . . . . . . . 95 Vill. On some Mistaken Notions of Algonkin Grammar, and on Mistranslations of Words from Eliot’s ibe &e. BY J. HAMMOND TRUMBULL. . 105
IX. Contributions to Creole Grammar. By Appison Van Nantz. 128
PROCEEDINGS :— | Preliminary meeting, November, 1868. First annual session, Poushkeepsie, 1869. Second annual session, Rochester, 1870.
9820
TRANSACTIONS
OF THE
AMERICAN PutLoLocicaL ASsoctaTion, 1869-70.
I.— On the Nature and Theory of the Greek Accent. Br JAMES HADLEY, PROYVESSOR OF GREEK IN YALE COLLEGE.
Every Greek word of two or more syllables had one syl- lable which was sounded on a higher key than the rest of the word: thus Av in AeAvKouust, Ke IN AeAvKEvat, Koc IM AeAvKdc. For a long time the Greeks in writing their language made no attempt to distinguish the syllable which was thus sounded on a higher key: they aimed to represent the substance of their sounds, the different articulations, but not their relative pitch. It was not until the development of grammatical study in the Alexandrian period, that the grammarian Aristophanes of Byzantium, about 200 years before our era, invented a sign for this purpose. Over the vowel which was sounded ona higher key he placed a wedge-like mark, sloping downward to the left, which was called 4 d&eia mpocydta, ‘sharp accent,’ ‘acute accent.’ But it often happened, in the utterance of a long vowel or diphthong, that the higher key with which the word began was not maintained to the end; that, after pro- nouncing the first part on a higher key, the voice dropped down to a lower, and on this pronounced the last part of the longsound. For such cases Aristophanes introduced a com- pound sign: representing the higher key as before, he added, to represent the lower key, a similar mark, but sloping down-
1
e 6 ) - o e474
2 J. Hadley,
ward to the right. The roof formed by the joining of the two
marks was rounded off in writing, and the whole sign was” called % wepomwpévn mpocpdia, ‘twisted accent,’ ‘ circumflex accent.’ The ordinary lower key was not generally repre- sented by any distinctive mark: if the vowel of a syllable had no mark of higher pitch written above it, this was a suf- ficient indication of its lower pitch. And indeed, there was nothing in this lower pitch that called for designation. The essential fact to be recognized and made evident in the writing was not that some syllables were lower than the rest: it was that some were higher than the rest; or, rather, that one syllable in each word was made conspicuous and important above all others by the higher key on which it was sounded. Yet there were two cases in which the lower pitch was repre- sented in writing. One of these has just been noticed, — when the lower pitch followed the higher in the same syllable, in the same long vowel-sound. The circumflex accent used for such a syllable consisted of two marks, a first representing the higher, and a second representing the lower pitch. The other case relates to oxytone words, where the higher pitch comes at the end of the word and belongs to the last syllable: aya0cc, orparnycce- If such a word is followed by other words in immediate grammatical connection, the higher pitch of its last syllable changes to a lower one, as in dyaGdc orparnyoc éyévero. Here now, on the last syllable of an oxytone word, when in the connection of discourse its higher pitch changes to a lower, the lower pitch is represented in writing; and represented in the same way as in the latter part of the circumflex accent, that is, by a mark sloping downward to the right, and called j; Bapeia xpocwdia, ‘heavy accent,’ ‘grave accent.’ Aside then from these two cases, the ordinary lower pitch is always left without designation.
In this description of the Greek accent, it has been taken for granted that there was an actual regular difference of pitch between different syllables of a word, and that the proper use of the written accent was to represent that difference. The correctness of this assumption is implied in the very names of the accents. The words cfc, ‘sharp,’ and apis, ‘heavy,’
On the Nature and Theory of the Greek Accent. 3
are the ordinary words used in Greek music for what we in our music call ‘high’ and ‘low.’ They are not used to denote difference in stress, or strength of utterance. We might find it natural that dféc, ‘sharp,’ should be applied to a syllable which was pronounced with marked stress; but it would be strange if Bapic, ‘heavy,’ was used of syllables pro- nounced without stress, the weaker or lighter syllables of a word. The term zpocydia itself, as well as the Latin accentus, which is used to translate it, comes from a root which means ‘to sing’; and in explaining the name, the ancient grammari- ans, both Greek and Latin, tell us that it is a singing of the syllable. There is a remarkable passage in a work of Diony- sius of Halicarnassus (de Comp. Verb., 11), in which he compares the melody of speech with that of song. The melody of spoken language, he tells us, is measured by a single inter- val, the so-called fifth, or, as he explains it immediately after, the interval of thrgc tones and a semitone. He says that when the voice rises to the acute, it does not go higher than this interval; and when it falls off again to the grave, it does not sink lower than this interval. He adds that the two into- nations, the acute and the grave, may be combined in the same syllable, and that such syllables are said to be circumflexed. It is perfectly evident that he is here speaking of the accent, that he describes it as a difference of pitch, and that he makes this difference about the same as the musical interval of a fifth, that is (as he himself says), three tones and a semitone. All this is pretty clear ; but it is made still clearer by the contrasted description of music, which comes directly after it. Music, whether vocal or instrumental, uses a number of in- tervals, and does not confine itself to the fifth alone (these are his own words), but employs, for purposes of melody, first the octave (this being first named, as the most important in- terval), and the fifth, and the fourth, and the full tone, and the semitone, and, as some think, even the quarter-tone (the chromatic diesis) so as to be distinctly perceived. Further, he says, music claims the right to subordinate the words to the tune, instead of having the tune subordinate to the words ; by which he means that it is the right of the musician to sing
4 J. Hadley,
the words to any tune he pleases, without reference to the natural tune (if we may call it so) which by the accent they have in spoken utterance. That this is his meaning becomes abundantly evident from a subjoined illustration. He quotes two or three lines from a chorus of Euripides, and points out at some length how the natural tune or accent of the words was wholly disregarded in the music to which they were sung. Thus, the first word, otya, is sung upon a monotone, both syl- lables on the same pitch, though the spoken word had two tones, the acute and the grave, and indeed both of them eom- bined in the circumflex accent of its first syllable. In another word, apB%Anc, which had the acute on the second syllable, the music gave both second and third syllables the same pitch, though in a spoken word the higher pitch of the acute accent was never maintained through two successive syllables. In riOere, the natural tune of the word, a high sound followed by two low ones, was completely reversed by the music, the first syllable being sung to a lower note and the last two syllables toa higher note. In «rvzeire, the circumflex of the second sylla- ble, with its combination of acute and grave, was lost in the music, that syllable being sung upon a single note. And in axonrpdfare, &® word of five syllables with acute accent on the middle one, the higher intonation, which belonged to the zpo, was in the music transferred to the Ga. These are the illus- trations and explanations given by Dionysius himself, whose authority on such a subject must be very high from his intelli- gence and learning, as also from his date, in the first century before our era. This passage alone, if there were nothing else to confirm it, would leave no doubt as to the melodic character of the Greek accent.
But is it not possible that this elevation of pitch, which characterized the accented syllable, was accompanied by an increased effort of the vocal organs, by a greater stress of pronunciation, such as marks the accented syllable in our own language and in other languages of modern Europe? It is possible, certainly; but there is scarcely any evidence to prove it true. In all that the ancient grammarians and other ancient writers have left us on the Greek accent—and the aggre-
On the Nature and Theory of the Greek Accent. §
gate is far from inconsiderable—there seems to be no state-. ment or expression which implies that the accented syllable was pronounced with more force than the rest of the word. No such implication can be found in the remarkable passage just referred to, where Dionysius speaks of the melody of spoken language. We might perhaps expect to find it rather in a following section, where his subject is the rhythm of spoken language. But nothing of the kind appears there. In speaking of rhythm, he refers to quantity of syllables, to the succession of longs and shorts; and he remarks that com- mon speech, the utterance of prose, does no violence to these, but keeps the syllables long and short according to their re- ceived or natural quantity. But music, which disregards the natural tune of words, disregards likewise their natural rhythm ; it changes (so he says) the length of the syllables, increasing or diminishing their quantity, and occasionally even reversing the natural proportions ; for (he adds), instead of making its own musical times subordinate to the natural quantity of the syllables, it makes the quantity of the sylla- bles subordinate to the times of the music. In all this, there is no hint that any one syllable of a word was regularly dis- tinguished from the others by its more forcible enunciation. Géttling, who believes that the accented syllable was actually pronounced with greater stress, can only refer in support of his belief to the modern Phavorinus. The silence of the an- cient authorities may not prove that there was absolutely no difference in stress between accented and unaccented syllables; . but it certainly warrants the conclusion that the difference, if there was any, cannot have been great or striking: it must have been far slighter than in English or in modern Greek. The grand fact about the accented syllable, to the mind of the ancients, was its higher pitch; its greater stress, if it had any, was either not noticed by them, or was felt to be com- paratively unimportant.
The same conclusion — that the stress of voice on the ac- eented syllable was little, if at all, greater than on other syl- lables — may be supported by probabilitics resting on other grounds. It is the natural effect of a decided stress-acceut
6 J. Hadley,
to weaken the following syllables of the word, and especially the one which immediately follows the accent, so that the vowel of that syllable is apt to be shortened or to be omitted altogether. In our own language, this tendency may be seen in the short e of ‘mystery’ as compared with the long sound in ‘mysterious,’ and in the suppressed e of ‘every,’ ‘ wond(e)r- ous.’ In Greek such changes are confined to a few words, as rixre in Homer for ri wore, 4AOoy for earlier #AvOor. They are perhaps hardly more numerous than the cases where an accented vowel has disappeared ; cases like 0vyarpec in Homer for Ouvyarépec, Biv for EBnr, etc. So far from being disposed to shorten the vowel which follows the accented syllable, the Greek shows rather a predilection for such forms as avOpwroc, riOnp, AvOfjooc80r. Latin proper names like Dentatus, Modes- tus, Salernum, the Greeks were perfectly able to pronounce with their Latin accent: there was nothing in their own sys- tem which forbade it; yet we often find such words accented on the first syllable, Aévraroc, Médearoc, Ladepvor, showing that an accented antepenult followed by a long penult was a com- bination agreeable to the Greek ear and regarded with a kind of preference. |
Another consideration which goes to show that the Greek accent was not accompanied by any very decided stress of voice, is found in the structure of Greek verse. In this, the word-accent is wholly disregarded, the ictus of the verse being quite as likely to fall on an unaccented as on an accented syllable. In the heroic hexameter, for example, we know that there was an ictus, or verse-accent, that is, a special stress of voice, on the first syllable of each dactyl or spondee. But if we look at the first seven lines of the Iliad, out of the forty-two cases of ictus which they present, only sixteen are found on syllables which have the written accent. Now it would seem to us unnatural in our own language to take the words, “ regarded with admiration and uncommon esteem,”’ -and read them as a hexameter, “régardéd with admfrati6n and uncummon ésteem;”’ or to take the line, “ ’tis as moon- light unto sunlight or as water unto wine,” and read it as an iambic tetrameter catalcctic, “ ’tis 4s moonlight unté sunlight
On the Nature and Theory of the Greek Accent. T
or 4s watér unté wine.” But we must suppose that the Greeks did something very much like this, if "we assume that with them, as with us, there was a decided stress of voice laid on the accented syllable. It is true that there is something hazardous in such reasoning. The Greeks in the construction of their verse may have treated their language with more freedom than we allow in, the treatment of ours. Different languages, or, rather, the people who use them, differ widely in this respect. Thus the German poet has greater liberty than the French in departing from the established forms and idioms of prose speech. We have seen that in Greek music the natural tune of the word, the differences of pitch depend- ing on the accent, were not observed, but were freely super- seded by other combinations of tones at the pleasure of the composer. If, then, the Greeks in their music were willing to substitute other tones for those given by the accent of the spoken language, it is conceivable that in their verse also they may have been willing to substitute other stresses for those of ordinary speech, to lay the emphasis on other syllables accord- ing to the rhythmical arrangement of the poet. Yet we can- not but regard it as highly improbable that a stress-accent, if it were as decided as ours, should be wholly neglected and superseded in the composition of verse. And in this view we are confirmed by the modern Greek, which, having a decided stress-accent, makes it, as we do, the basis or determining ele- ment of its verse-system.
Taking all these considerations together, we hold it all but certain that the Greeks did not lay any marked or forcible stress on the accented syllable. It is even doubtful whether they laid any stress upon it, more than on other syllables of the word. It is certain, however, that in the history of their language they have either adopted a stress-accent, or have given strength to a weak one existing from the first. In the modern accent, the leading element is stress, and differ- ence of pitch, if it is not wanting, has at any rate ceased to be prominent and uniform. The distinction between acute and circumflex has been wholly obliterated. The forms éy\éca, ‘to make manifest,’ and dy\dea:, ‘might make manifest,’ are
8 J. Hadley,
undistinguishable in the modern pronunciation. With this change in the cltaracter of the accent, was connected, as we just saw, the adoption of accent as the basis of versification. When verses began to be constructed on this basis, they were called oriyor xoderexoi, ‘ political’ or *‘ popular verses,’ in con- trast with the old quantitative verses, which continued to be written, as a kind of literary exercise, long after the pronun- ciation on which they were founded had ceased to be heard. Now these political verses were composed as far back as the eleventh century, and probably much earlier: Whenever they began to be made, we may be sure that the Greek accent had already changed its character, and had come to exhibit a de- cided stress. But the change, we may presume, was very slow, and may have been going on for centuries before the stress element was strong enough to express itself in verse © composition. It is quite supposable, therefore, that a weak stress may have been heard on the accented syllable, as a regular accompaniment, even in the time of Herodian, the principal authority on accent, if not yet earlier in the time of Aristophanes of Byzantium, the inventor of the accentual signs. If this were so, we might find in it an explanation of the fact already noticed, that the last syllable of an oxytone word has an accent written upon it, even when its tone changes from high to low. In aoueug éyévero, the grave accent on Aeve shows (so we are told) that it was pronounced on a low tone. But why then should it have any mark over it, more than the other syllables of the same word BaoAeve? Or why should BacAevc, which in this case has no high tone, which is all pronounced on a low tone, have any mark of ac- cent over it, more than an enclitic word, more than éor/» (for instance) in xaAerdv éorevy, where the verb appears without ac- cent? To such questions it might be replied, not wholly without plausibility, that though Saocdeuc in the case supposed — had no elevated pitch, no accent properly so called, its last - syllable was yet distinguished from the rest by a slightly su- perior stress, and was therefore allowed to have a distinctive mark over it; while the enclitic gor» was written without any such mark, because it had neither elevated pitch nor superior
On the Nature and Theory of the Greek Accent. 9
stress on either of its two syllables. This, I say, would bea possible solution of the difficulty. But theré is another solu- tion, of which I am now to speak, and which brings up a question of much interest for the theory of the Greek accent. Did the Greek accent distinguish only two tones, a high and alow? Or was there some middle tone, having a regu- lar place in the system, some intermediate between the two extremes? It must be confessed that the evidence given on this point by the ancient writers is not so distifct as could be wished. In general, they speak only of two tones, high and low, or (as they term them) acute and grave. Still they do not explicitly assert that all grave tones were equally low. It may be that they thought it important only to distinguish the high tone, as dominating the whole word, and that differ- ences between the lower tones seemed to them of too little practical consequence to require mention. Yet we do find in ancient writings indications of a middle tone. Thus Aristotle (Rhet., 8, 1, 4) speaks of three “tones, acute, grave, and middle ;” though it is possible that by ‘ middle” here he in- tends the circumflex, which, combining as it does both acute and grave, might be regarded as having an intermediate char- acter. But the Greek grammarian Tyrannio of Amisus, as quoted by Varro, enumerates four accents, grave, middle, acute, and circumflex. Tyrannio, indeed, may have been speaking of the Latin accent; but Varro refers to other wri- ters as recognizing a middle accent, Glaucus of Samos, Her- mocrates of Iasos, and the Peripatetics Theophrastus and Athenodorns, some, if not all, of whom must have referred to ' the Greek. And the grammarian Servius says: “It must be _ understood that this doctrine of a middle accent is no inven- tion of recent times, but belongs to all who before the time of Varro and Tyrannio have left any thing on accent; since the majority of these, and the most distinguished writers, have made mention of this middle accent, all of whom Varro refers to as his authorities.” This language of Servius, doubtless, - overstates the case. The number of writers who expressly recognized a middie accent cannot have been so great as here represented. But a reason may be found for this in a remark 2
10 | J. Hadley,
of Servius himself, that ‘‘ the middle accent, which is a sort of border between the two others, resembles the grave more than it does the acute, and is therefore reckoned with the lower rather than with the higher tone.”
The evidence of a middle tone, which has come to us from ancient writers, deficient as it is in definiteness and con- sistency, seems on the whole sufficient to warrant the inquiry, whether the phenomena themselves of the Greek accent fur- nish any indidations of such a tone. G. Hermann, in his essay De emendanda ratione Grammatice Greece (p. 66), - suggested that the grave accent, where it was written for the acute on the last syllable of an oxytone word, was the sign of a middle tone, intermediate between the acute and the un- marked grave. On this point Buttmann also, in his Ausfihr- liche Grammatik, takes substantially the same ground. More recently, G. Curtius, reviewing Bopp’s Accentuationssystem in Jahn’s Jahrbiicher (1855, vol. 71), expresses the opinion that the grave accent, where it forms the second part of the circumflex, represents not the ordinary low tone of the word, but an intermediate tone. This view seems to have betn sug- gested to his mind by the Sanskrit system of accentuation : here, the uddétta (or elevated tone), corresponding to the Greek acute, is regularly followed in the next syllable by the svarita, which certainly differed from the anuddtta (unele- vated, depressed tone) of the other syllables. The Indian grammarians describe the svarita as a combination of the uddilta and anuddita, similar to the Greek circumflex; and this must clearly have been its character when used as an in- dependent accent. But where it is the mere follower of an uddita, and especially where it belongs to a short syllable, the statement that it was a compound accent, a circumflex, seems far from probable: we could much more easily believe it to have been an intermediate tone. Curtius intimates that he would not confine his recognition of a middle tone in Greek to the last part of the circumflex accent; yet he has not actually given further development to the theory. But a recent writer, Franz Misteli, in an able article contributed to the 17th volume of Kuhn’s Zeitschrift, has taken it up and
On the Nature and Theory of the Greek Accent. 11
carried it to a much greater extent. He holds that the acute accent in Greek, as in Sanskrit (?), was regularly followed by a middle tone, this middle tone being either written as tho last part of a circumflex, or being merely understood on the syl- lable which comes after the acute. Ifthe acute stands on the last syllable of a word, where there is no room for a middle tone after it, the acutc itself loses its high pitch and becomes a middle tone, represented by the so-called grave accent. Only at the end of a sentence, or before an enclitic, does the acute under such circumstances retain its high pitch, and the word appear as an oxytone. This theory of a middle tone Misteli applies with much ingenuity to account for the general laws of Greek accentuation. In showing how it may be made to answer this purpose, I shall not confine myself to his state- ments, but shall take the liberty to depart from them in vari- ous particulars, and shall introduce some views (especially those on Latin accent) which do not appear in his exhibition of the subject. |
The general laws here referred to are the four following :— 1. The acute cannot stand on any syllable before the ante- penult.—2. The antepenult, if accented at all, must have the acute; but it cannot be accented at all, if the ultima is long: thus avOpwzoc, but a»fpaxov.— 3. The penult, if accented, must _ have the acute, when the ultima is long (has a long vowel- sound): thus é»Opwzrov, roatrn.—4. A long penult (one which has a long vowel-sound), if accented at all, must lave the circumflex when the ultima is short: thus rocotroe.
The Greek accent is confined to the last three syllables of a word. But in Sanskrit there is no such restriction. The ac- cent may go back to any syllable, however far removed from the end of the word. In dbubodhishémahi, ‘we wished to know,’ it stands at the beginning of a seven-syllable word. The same freedom, we may presume, belonged to the primi- tive Indo-European language. There must have been some time, therefore, in the history of the Greek language, whether before or after it became distinctively Greek, when a change took place in this respect, —some time when all accents stand- ing before an antepenult were carried forward and thrown
12 J. Hadley,
upon one of the fast three syllables. If we ask for the cause of such a change, none could be imagined more natural or probable than a fondness for some particular succession of tones at the end of a word. If the earlier accentuation had a threefold distinction of tones, a high tone, middle tone, and low tone, we can easily conceive that this succession, these three tones in the order of their height, should have been found an agreeable cadence for the close of a word. It is a cadence which appears in Sanskrit in very many words; it may have been common in the Indo-European language prior to the separation of its branches. What we should have to suppose in regard to the ancestors of the Greeks would be that at some time a special taste or fondness for this cadence developed itself among them, with a dislike for any cadence in which the high and middle tones were followed by more than one low tone, so that, in order to secure what they liked - and avoid what they did not like, they came at length to change the accent of words, to shift it from an earlier to a later syllable. Changes in an opposite direction, from a later syllable to an earlier, were not made, even to secure this favorite cadence ; or if such a change took place in particular cases, it did not become the general law. One thing further we must suppose to account for the Greek accent: that in this cadence the Greeks preferred that the final low tone of the word should be a short one: they did not like to have it main- tained through a long syllable. Our hypothesis, then, may be stated in a single sentence :—that the early Greeks changed the older accent of words so as to secure this cadence, ‘high tone, middle tone, short low tone,’ wherever it could be secured without throwing back the accent. This single hy- pothesis will be found sufficient to account for the four general laws already given. Thus :—
1. “The accent cannot stand on any syllable before the antepenult.”’ In éAcirero, ‘ was left,’ the accent (it can hardly be doubted) was originally placed on the augment, as it regu- larly is in Sanskrit: it fell, therefore, on the syllable before -the antepenult. The middle tone would then fall on the antepenult Ae, and the remaining two syllables xe and ro
On the Nature and Theory of the Greek Accent. 18
would have the low tone. But the preferred cadence allowed only one syllable, and that a short one, for the low tone at the end of the word. It was neccssary, therefore, to place the high tone or accent on the antepenult Aa, leaving xe for the middle tone and ro for the low tone.
2. ‘* The antepenult,if accented at all, must have the acute.”’ Of course, if accented at all, it must have either the acute or the circumflex. Suppose, then, that in the word é\cixero the cir- cumflex accent was placed on the antepenult Ae: that syllable, from the nature of the circumflex, would have the high tone on the first part and the middle tone on the last; and thus again as before there would be two syllables, re and ro, for the low tone. The circumflex on the antepenult would, therefore, be incompatible with the cadence required. But the rule asserts also “that the antepenult cannot have any accent, even the acute, if the ultima is long.” For suppose that the first person é\seréuny could have the acute, the high tone, on the antepenult Ax: the middle tone would fall on zo, and the low tone on pny, a long syllable. But the cadence required was ‘high tone, middle tone, short low tone.’ Hence the high tone or acute must be placed on the penult xo, and the final long syllable un» must be divided between the other two tones, its first half being sounded with the middle tone, and its last half (which, of ccurse, has the quantity of a short syllable) being sounded with the low tone.
3. ‘* The penult, if accented, must have the acute when the ultima has along vowel.”’ For in the feminine roaérn, ‘ such,’ suppose that the circumflex could stand on the penult av, This syllable, then, by the nature of the circumflex, would have both high tone and middle tone, and the low tone would fall on 7, & long syllable, which is inconsistent with the required cadence. To secure this, the high tone, or acute accent, must be placed on the penult av, and the final long rn must be divided between the other two tones, the first short time contained in it being sounded with the middle tone, and the last short time with the low.
4. “A penult with long vowel-sound, if accented at all - must have the circumflex when the ultima has a short
14 J. Hadley,
vowel.”” For ‘suppose that the masculine rowitroc, ‘such,’ had the acute or high tone on the penult ov; the middle tone would then fall on the last syllable roc, and the final low tone would be excluded. In many words, as in Adyog, ‘speech,’ this was something unavoidable: the high tone falling on do, the short penult, must fill that syllable, the middle tone must - fill the last short syllable yoc, and there is no room left for the closing low tone. But with a penult long by nature, as in rowvroc, there was no such necessity. It was enough to divide _ the long ov between the high tone and middle tone, in other words, to sound it with the circumflex accent: the ultima roc was then left for the short low tone, and the desired cadence was thus obtained.
We see, then, that these four general rules of Greek accent, which have the appearance of being unconnected, arbitrary and capricious, are all of them direct corollaries from a single hypothesis, all of them necessary results from the extension of a single cadence, ‘high tone, middle tone, short low tone,’ at the end of words. But what shall we say of polysyllables like yaXerde, ‘harsh,’ with acute on the last syllable, or like xarerac, ‘harshly,’ with circumflex on the last syllable, or like AeAvpévoc, ‘having been loosed,’ with acute on the penult be- fore a short ultima? In these words there is room for the favorite cadence, but they do not have it: they close either with the high tone itself (as in yadexde), or with the high tone followed by the middle (as in yxaderéc, AcAupévoc). How can this be accounted for? By the last clause in our hypothesis: “The Greeks changed the older accent of their words, so as to secure the favorite cadence, wherever this could be secured without throwing back the accent.”’? Wehold it to be true, as a general fact, that words like those just given were thus ac- cented in the primitive period, —i.e., either on the last sylla- ble, or on the penult with short ultima following ;— and that they did not assume the threefold cadence, because the ten- dency to this was not strong enough to produce a retraction of the accent from a later to an earlier syllable. By this we do not mean that such a retraction never occurred. It may have taken place in numerous instances, but it never became
On the Nature and Theory of the Greek Accent. 10
the general law. And thus the Greek has thany oxytones, which of course end with a high tone; many perispomena, which end with a middle tone; and not a few paroxytones with short ultima, which likewise end with a middle tone; and these, not in short words only, where the full cadence was impossible, but in very many longer words, where there was room enough and to spare for the succession of ‘high tone, middle tone, and short low tone.’
But there was one branch of the Greek people —the Aeoli- ans of Asia Minor—which went farther than the rest in the fondness for this cadence. The Aeolians did not hesitate to retract the accent for the purpose of securing it. Thus Aehvpévoc was changed by them to AcAdpevoc, xarerdt tO yarerros, xarerac to xarérwe,— yadrerwe they could nat say, for that would make a long low tone. If there was not room for the complete succession of three tones, they secured as much of it as they could, saying (for instance) sédgoc with high tone and middle tone, instead of the common Greek cogdc, ‘wise,’ an oxytone word. Hence in the Aeolic dialect the only oxytones were monosyliables, and even these were oxytone only when the vowel was short, as the articles rdév, rd, ra’. If the vowel was long, there was room for a middle tone after the high tone: the monosyllable was then pronounced with circumflex accent, as rijv, rouc, instead of the common Greek rfy, rove. It was only in prepositions and conjunctions that the Aeolic, in agreement with the other Greek dialects, admitted oxytone words of greater extent than one short syllable. |
We observe now that a similar hypothesis may be used to explain the peculiarities of Latin accentuation. The Latin, beside the acute accent, or hith tone, has a circumflex, which it uses just where the Aeolic dialect would use the circumflex, that is, on all monosyllables with-long vowel (except only ne with the imperative), and on all penults with.long vowel fol- lowed by short ultima. On all other accented syllables the acute was used, viz. on antepenults followed by short penult, — on long penults followed by long ultima,— on all penults long only by position,—on short penults of dissyllables, —and on all monosyllables with short vowel, enclitics of course excepted.
16 | J. Hadley,
It is evident that in changing the primitive accent the Latin has not confined itself to one direction, from earlier to later syllables : like the Aeolic Greek, it has freely moved the accent backward, from later syllables to earlier, in order to secure the desired cadence. But the cadence required to account for the Latin accent is in one respect different from that which served to explain the Greek. The Greek would not allow the middle tone to be followed by a long low tone: the Latin would not allow the low tone to be preceded by a long middle tone, a middle tone extending over the whole of a long syllable, whether long by nature or by position. Hence for the Latin the cadence becomes, ‘high tone, short middle tone, low tone.’ For example, in legere, legeres, legeret, the low tone fell on the last syllable re, res, ret, without reference to its quantity, whether long or short; the middle tone on the short penult ge, and the high tone on the antepenult le. This example is enough to show that, by the necessity of such a cadence, the accent could never go farther back than the antepenult; that the antepenult, if accented at all, must take the acute; and that the antepenult could only be accented when the penult was short. In such forms as gaudere or gauderet, the acute must stand on the long penult de; but if the desired cadence is to be obtained, it must admit the middle tone to the latter half of the long vowel, for the short ultima has only room enough for the low tone: the word, therefore, could only have the circumflex (the combination of high and middle tone) on the penult. But in forms like gauderes, where both penult and ultima have long vowels, the Latin preferred to divide the long ultima res between the middle and low tones, leaving only the high tone for the penult de: the word, then, has the acute on the penult. So too have forms like legendus, where the penult has a short vowel, and is only long by position. Here the circumflex is impossible: the short vowel in gen cannot be divided between two tones, the high and middle: gen must have the high tone, dus the middle, and the low tone is excluded by the necessity of the case. In legendi, the ca- dence could be made complete by dividing the long dt between the middle and low tones; but in legendus, it is necessarily
On the Nature and Theory of the Greek Accent. 17
incomplete, as much as in legit or legunt, where both vowels are short, or in the Greek Adyoc, having only the high and middle tones. .
It is not necessary to go further into details to show that all features of the Latin accentuation may be accounted for by the assumed tendency to close all words with the succession of ‘high tone, middle tone, low tone,’ or so much of it as pos- sible, consistently with the one restriction that the low tone must never have before it a middle tone which occupied the whole of a long syllable. But if we compare the Greek ac- centuation with the Latin, and both with that freer system of primitive Indo-European speech which is best represented to us by the Sanskrit, we may naturally conclude that the first step in the series of changes which gave the accentual systems of the Greeks and Romans their peculiar character, was caused by a simple distaste for a succession of low-tone syllables at the end of aword. I repeat, the beginning of a special Greek and Roman accentuation would seem to have sprung from the mere unwillingness to hear more than one low-tone syllable at the end of a word. This unwillingness, carried into prac- tical effect, would confine the accent (that is, the high tone) to the last three syllables of the word. But it would not cause any retraction of the accent: it would allow such forms a8 xadendc, yaXeric, AcAvpévoc; for these have not even one low tone, still less have they a succession of low tones at the end. And further, this simple distaste or unwillingness, as it im- plies no restriction on the long or short quantity of the single tones, would allow such forms as éAcirouny, drOpwrov, in Greek, and such forms as gaéderet, légendus,in Latin. This first step we may naturally suppose to have been made in the Grzco-Latin or Greco-Italican period, that is, while the common ancestors of these peoples spoke a language, differing indeed from the original Indo-European, but not yet divided into branches having a distinct character as Greek, and Itali- can or Latin. But the next step must have been taken after ‘this division, as it is different in the two branches. In the Greek, it springs from a distaste for a long low tone, a low tone stretching through a whole long syllable, following the
3
18 J. Hadley,
high and middle tones at the close of a word. This would re- quire éAcrduny instead of éAcirouny, avOpwrov instead of avOpwrov. The corresponding step in the Latin springs from a distaste for a long middle tone, a middle tone stretching through a whole long syllable, between the high and low tones at the close of a word. This would require gaudéres instead of gat- deres, legéndus instead of légendus. In Greek the effect of this second step was to make the cadence, ‘high tone, middle tone, short low tone,’ the prevailing one for words which were long enough to admit of it; though there still remained a large number of words, represented by xaAerde, vader, Aedupevoc, which did not have it. Now one section of the Greek race, the Aeolic of Asia, went further than this: they took a third step, probably much later than the second: they threw back the accent of these words, so as to make the already prevail- ing cadence universal, so far at least as the length of the word would allow. Whether the whole Italican race, in all its branches, Umbrian, Oscan, Sabine, etc., took a similar third step, we are unable to say. It is certain that one branch did so, the Latins: they threw back the accent, so that the ca- dence already prevailing should be made universal, as far as the length of the word would allow.
Perhaps I should leave a false impression, if I were to close without calling attention in one word to the hypothetical char- acter of what has been said here about a middle tone. The existence of a middle tone in Greek and Latin has a good deal of positive ancient testimony in its favor. But that a high tone, when it did not come at the end of a word, was regularly followed by a middle tone, is a proposition, which, however supported by Sanskrit analogies, has no direct evidence in the statements of the ancient writers. And of course, if there were no doubt of its truth, still the use here made of it to ac- count for the ante-historic changes and the earliest historical appearances of Greek and Latin accentuation, would be purely hypothetical. At the same time, it may be said with justice, that the hypothesis is so natural in itself, it is so readily sug- gested by known facts, and it offers so simple and perfect an explanation for a variety of seemingly unconnected and capri-
On the Nature and Theory of the Greek Accent. 19
cious phenomena, that one can hardly help believing that it has a foundation in truth.
To some persons it may seem hard to believe that the ordi- nary utterance of discourse and conversation should have had so much of musical intonation: that this threefold distinction of tone should have found place in it as a recognized and con- stant element. But in the Chinese and the languages cog- nate to it, as spoken at the present day, we find the musical element playing a much larger and more important part. In some of the popular dialects of China, a large proportion of the syllables which make up the language are pronounced with seven or eight intonations: thus, as a short abrupt monotone (compare the English preposition ¢o, in its ordinary short pro- nunciation); or as a prolonged monotone (compare the Eng- lish numeral éwo) ; or with mixed falling tone (like the Greek circumflexed rov) ; or with mixed rising tone (like the English two at the end of a question: “two?’’); or with similar in- tonations duplicated on a lower key. Thus the same syllable may be pronounced in seven or eight different ways, having each their special and widely diverse meanings. Compared with such complexity of musical intonation, that which we have hypothetically ascribed to the early speakers of Latin, Greek, and Sanskrit, and the yet earlier speakers of the undi- vided Indo-European speech, is a very simple and easy matter.
Il. — On the Nature and Designation of the Accent in Sanskrit. Bry WILLIAM D. WHITNEY,
PROFESSOR OF SANSKRIT AND COMPARATIVE PHILOLOGY IN YALE COLLEGE.
This paper was originally intended only to give such an ex- position of the nature of the Sanskrit accent as should illus- trate and support the views presented by Professor Hadley in the preceding paper respecting the Greek accent. It has seemed best, however, to treat the subject somewhat more com- prehensively than that intent would demand, because full and correct knowledge of Sanskrit accentuation is less accessible to scholars in America than it ought to be. The Sanskrit grammars in English which are most used among us (Monier Williams’s and Max Miiller’s) hardly touch accent at all;* and the older works ignore it altogether; while to the treat- ment of the matter by Oppert and Bopp (and even, in a less degree, by Benfey) much exception is to be taken. To the student of classical Sanskrit merely, accent is a matter of sub- ordinate consequence; no one pretends to give the Sanskrit words their proper tone, and it is the general custom among Western scholars to pronounce them according to the rules of Latin accentuation. This is because the accented syllable is, in the majority of words, unknown ; the written texts are not marked for accent, and the notices of the native grammarians are not sufficient to supply the lacking knowledge. With the Vedic student, the case is otherwise; the Vedic texts of the first class (namely, the original Sanhitis of the Rig, Sama, Yajur, and Atharva Vedas) are completely accented, and also several works of the second or brdéhmana class (name- ly, the two principal brdhmanas of the Yajur-Veda, the Cata- patha and Taittiriya Bréhmanas, and the Taittirfya-Aranya- ka) ;f and the accent is a highly important element in aiding
* Miiller’s second edition, which has appeared since the above was written, ac- cents its paradigms, and gives an Appendix on the general subject.
t All the works here mentioned are either published or publishing, and will be soon accessible in printed form, complete.
On the Accent in Sanskrit. 21
both the grammatical and the lexical comprehension of the texts. And, of course, every student of the history of the Sanskrit language, or of the history of Indo-European language by the aid of the Sanskrit, requires to understand thoroughly a part of the phonetic structure of the latter which is so funda- mental in its character, and has exercised so powerful an in- fluence in the shaping of words and forms.
In investigating the nature of the Sanskrit accent, we are not limited to the drawing of inferences from the facts of accentu- ation laid befure us in the texts; our chief sources of knowl- edge are the Hindu grammarians, who have treated the sub- ject, as they have most other departments of grammatical theory, with great fullness and acuteness. The great gram- marian Panini, whose work has become the acknowledged authority for all after time, is clear and intelligible in his statements as to accent; and upon the foundation of his work and its commentators alone, withont access to any accented texts, Bolitlingk gave in 1843* an acute, intelligent, and very correct account both of the theory and of the main facts of Sanskrit accent,.one that in many respects has not been sur- passed or superseded by anything that has since appeared. But the brevity of Panini is most acceptably supplemented by the more detailed statements of the Pratigikhyas. These are treatises which attach themselves each to a single Vedic text, as phonetic manual of the school to which that text belongs. They deal with all the elements of articulate utterance — with the mode ‘of production and the classification of articulate sounds, with accent and quantity, rules of euphonic combina- tion, and the like; and they prescribe how the various forms of text in which their Veda is preserved are to be constructed, cataloguing its slightest irregularities of form, and endeavor- ing to fix its readings beyond the reach of question or change. Four such treatises have come to light: one belonging to the Rig-Veda, one to the White Yajus or Vajasaneyi-Sanhita, one to the Black Yajus or Taittirfya-Sanhit&, and one to the Athar- va-Veda; for the SAma-Veda alone none has yet been found.
* Ein erster Versuch itber den Accent tm Sanskrit, von Otto Bohtlingk, in the Me. moirs of the St. Petersburg Academy, vol. vii., series 6th, 4to.
22 W. D. Whitney,
Prior to the publication of any of them, the teachings of the first three with regard to accent were summarily presented by Roth (who was the first to call the attention of scholars to this class of works) in the Introduction to his edition of Yaska’s Mirukta (Gottingen, 1852, pp. lvii.—lxxii.). All have been since edited in full; the Rik-Pratigakhya by Regnier (in the Journal Asiatique, Paris, 1857-9) aud Miiller (Leipzig, 1856-69), the Vajasaneyi-Pratigakhya by Webcr (in his Indische Studien, vol. iv., Berlin, 1858), the Atharva-Praticikhya and the Tait- _ tirtya-Praticikhya by myself (in the Journal of the American Oriental Society, vols. vii., 1862, and ix., 1871). I have also discussed— with more fullness, I believe, than any one else has found occasion to do—some of the general questions rela- — ting to the subject, in a critique on Bopp’s Accentuationssystem, in the Journ. Am. Or. Soc. (vol. v., 1856, pp. 195-218), and in the note to Ath. Prat. iii. 65 (Journ. etc. vii. 494 ff.) ; and have set forth the rules respecting the accent of the verb in the sentence, with illustrations from the Atharva-Veda, in the same Journal (v. 387 ff.).
The information which we derive from all these enumerated sources is full enough, not only to let us see clearly the views held by the ancient Hindu grammarians (their age, unfortu- nately, is not ascertained, but is generally believed to precede by some centuries the Christian era) respecting the accent of their learned and sacred language, but also to enable us in some measure to trace the development of their accentual the- ory, and to criticize it in its details. For, though’ we cannot help admiring and respecting, and in a very high degree, the acuteness and sagacity of those oldest known students of pho- netics, we cannot accept for truth what they give us without first carefully questioning it, and testing it by fact and by the- ory; and we shall be likely to find in their treatment of more than one point their characteristic and national weaknesses, and to be led to modify and amend their doctrines.
The general name given to accent is svara, which means literally ‘tone;’ and, in virtue of this meaning, is applied also to designate other things beside accent. Thus, it is the name of a ‘ vowel,’ as being a ¢one-sound, an utterance in which the
On the Accent in Sanskrit. 23
element of tone predominates over that of articulation ; and it is in the Praticikhyas used a dozen times in this sense to once in any or all others. It designates, moreover, the ‘tones’ or musical notes which compose the scale. There is, then, noth- ing about the word that necessarily implies a particular theory respecting the nature of accent; although the absence of any reference in it to stress or force of utterance, and its connec- tion with musical pitch, are distinctly suggestive of a musical theory. |
This suggestion is made a certainty by the names and defi- nitions of the separate accents or ‘“‘tones.’’ Of these, three are recognized, and they are called respectively uddtla, anu- détta, and svarita.
_ Udétta (passive participle of the root dé, with the preposi- tions wt and @) means literally ‘taken up, raised, elevated.’ And the description of the accent by all the authorities corres- ponds with this title ; it is defined everywhere, in nearly iden- tical terms, as being utterance uccdis, ‘in a high tone.’ (So Ath. Prat. i. 14; Vaj. Prat. i. 108; Taitt. Prat. i. 88; Pan. i. 2.29; the Rik Prat. alone, iii. 1, is less explicit, but its whole doctrine accords with that of the rest.)
Anuddtta is the same word with the negative prefix, and so means ‘not elevated.’ The authorities define it as signifying utterance nicdis,‘in a low tone.’ (See Ath. Prat. i. 15; Vj. Prat. i. 109; Taitt. Prat. i. 89; Pan. i. 2. 30; the Rik Prat. defines uddétta and anuddtta as characterized respectively by “tension”’ and ‘ relaxation’? —i. e. of the organs of articula- tion.) This lowness of tone docs not imply, of course, a fall from the ordinary pitch of voice; the tone is low as compared with uddétta; it is the simple negation of that uplifting of pitch which‘ marks the positively accented syllable.
The name of the third tone is svarita, and it is uniformly explained as consisting in a combination of the other two. Thus, Panini and the Taittirfya-Praticakhya say, in identical phrase, samdhdrah svaritah, ‘the combination [of the two oth- ers, just defined] is svarita,’ the Vajasaneyi-Praticgékhya has ubhayavdn svaritah, ‘one [i. e. a vowel] possessing both [the two tones as already defined] is svarita ;’ the Rik-Pratigaékhya
24 W. D. Whitney,
is more elaborate in its statement: “in ease of the occurrence together of the two preceding [tones] in one syllable, the ac- cent is svarita;” the Atharva-Pratigakhya alone is here am- biguous, using a term (dékshiplam) which is not explanatory, but itself requires to be explained in accordance with what we know by other means of the theory involved. The further specifications added in nearly all the treatises make this defi- nition still more unequivocal. Thus, Panini (i. 2. 82) says, “half [the quantity of] a short vowel at its beginning is udétta;” the Ath. Prat. G.17), “half the quantity of a svarita, at its beginning, is uddétia,’’ the Vaj. Prat. (i. 126), “at its begin- ning, half the quantity of the vowel is uddétta.”” The other two Pratigikhyas complicate the definition with a further de- velopment of the accentual theory, to be explained hereafter ; but in regard to the essential characteristic of the accent, that it is a union of higher and lower tone in the same syllable, they accord entirely with those already quoted. The Rik Prat. (iii. 2, 3) says “half a mora, or half the quantity, of this [svarita| is higher than uddtta; the following remainder is anudétta, but is heard as udétta’’— that is to say, the descent of tone, instead of being from uddtta to anudétta, is from a higher pitch than uddtta down to something sensibly equiva- lent to the latter. And the Taitt. Prat. Cone of whose pecu- liarities it is everywhere to quote conflicting opinions on con- troverted points) says, yet more at large (i. 41-7): “of this svartia, when it immediately follows an uddtta, the beginning, to the extent of half a short vowel, is uttered in a higher tone than uddétta ; the remainder is equivalent to uddtta; or, this following part is in a lower tone; or it is equivalent to anw- ddtta; the teachers say that the first part is equivalent to uddtta and the rest to anuddtta ; some hold that the whole is a downward slide.”
We thus see that there is no discordance whatever among the ancient Hindu grammarians with regard to the nature of the svarda accent, as being the union of a higher and lower tone upon the same vowel, or within the same syllable. No hint of a different explanation is given us, nor do we discover any traces of the former prevalence of another view, crowded
On the Accent in Sanskrit. 25
out and replaced by this. Of the name svarita, however, by which it is called, it is exceedingly difficult to find a satisfac- tory explanation. The word is most likely a quasi-participial formation from svara itself, and means ‘toned ;’ possibly it was applied to the syllable because this exhibited all the tones which have to be taken account of in accentuation ; or because the element of tone was most conspicuous in it; its change of pitch gave it a sing-song or cadenced effect. That it signifies ‘accented,’ in our sense of that term, there is no reason what- ever for supposing. Other conjectural explanations, of which not a few have been ventured, may be left here unnoticed. What is unmistakably clear with regard to it is the view which - the Hindus unanimously held as to the nature of the accent which it designated; and any interpretation which we may try to put upon it must be subordinated to this; we have no right to make an etymology of the name, and then force it into a definition of the thing.
Such being the defined character of the three tones, their accordance with the Greek acute, grave, and circumflex, as defined above, in Prof. Hadley’s paper, is clear and undenia- ble; and the accentual system either of the Greek or the San- skrit, if it needed any support from without, would find it abundantly in this close parallelism with the other. There being no possibility of a copying of the accentual theory on either side, both must be regarded as independently founded upon the facts of the two languages, and as faithfully and fairly representing them. And we are justified in setting aside, when speaking of the Sanskrit accents, the outlandish Sanskrit terms, and employing instead of them the familiar designa- tions “acute,” “ grave,” “ circumflex.’’*
Unless, indeed, we shall find sufficient evidence somewhere in the phenomena of accentuation of either Greek or Sanskrit to convict the native grammarians of having blundered in their observations and deductions. And that this is not the
—
* These were used by Bohtlingk, in the essay which first opened to Europe the knowledge of the Sanskrit accent; and the more the latter is understood, the more generally will they be adopted.
4
26 W. D. Whitney,
case as regards the Sanskrit, I shall endeavor to show by a concise exhibition of the rules for the occurrence of the svaria or circumflex in that language.
The circumflex in Sanskrit is a rare and inconspicuous phe- nomenon as compared with the Greek. Only a very small class of words have it as their proper accent, and it arises chiefly in the course of the combination of words into phrases, by the peculiar euphonic system of the Sauskrit, which, as is well known, does not leave its words to stand independently side by side, but adapts their final and initial elements to one another, avoiding the hiatus, or any collision of incompatible ‘consonants. |
The first class of circumflexes arises when an acute ¢ or 7, or u, is converted into y or v before an unaccented dissimilar vowel. Thus, vi and evé are combined into vyévd* ; nadi’ and asya into nadydsya; apet and agne into apsvdgne. That is to say, the single syllable into which the higher and lower tone are combined still retains the double pitch belonging to its constituent parts. This kind of circumflex is in all the Pr&tigékhyasf styled kshdipra, literally ‘ hasty, quick,’ as being accompanied with an abbreviation of quantity, in the contrac- tion of two syllables into one. —
One of the most peculiar and problematical processes in the whole Sanskrit euphonic system is that by which a final é or 6 absorbs or elides an initial d of the word that follows. This, which has become the regular and necessary mode of combina- tion in the later or classical Sanskrit, was only an occasional license in the older Vedic language ; in the Atharva-Veda, for instance, the @ was elided in only a little more than one- sixth of the cases.in which it followed e oro. Wherever, now, the e or o was acute and the a grave, the accent of the former after the absorption of the latter was made circumflex. Thus, té abravan became tébravan ; 86 abravit became sébravit.
* For want of means to do better, I signify the circumflex accent by what we ordinarily call the sign of grave; the sound represented (in accordance with usual custom) by v is that of our w.
t The distinction, namely, of the kinds of circumflex, with the corresponding nomenclature, is unknown to Panini.
On the Accent in Sanskrit. 27
Here, again, the acute and grave tones of the constituent elements are both represented in the circumflex given to the syllable that results from their combination. This second circumflex is styled abhinthita (or, by the Taitt. Prat., abhini- hata): the literal meaning and ground of application of the term are not very clear.
In case, however, the accent of the two elements is other than has been defined, no circumflex arises; if the second element is acute, the combination is acute; if the first as well as the second is grave, the combination, of course, is grave also.
If, moreover, two vowels are fused together into a single vowel or diphthong, then, if either was acute, the resulting combination, as a general rule, is also acute: that is to say, the acute element is powerful enough to assimilate the other, raising the whole syllable to the higher tone. Thus, sd’ and asti become sd'sti, sé' and eshd’ become sd’ishd’, sd’ and «& be- come 8ét; nayati and {ndrah become nayati'ndrah ; stha d'rjam becomes sfhérjam; and soon. This is the rule laid down in all the Praticakhyas; and all the Vedic texts of the first rank, the Sanhitas, conform to it. Panini, however (at viii. 2. 6), also permits the result of combination of a final acute with an initial grave to be circumflex ; that is, he allows sd’ asti to be- come either sd’sli or s@‘sti, and soon. And there is a single Vedic text of the second order (the Gatapatha-Brahmana, be- longing to the White Yajur-Veda) which makes its combina- tions in this manner circumflex. Of course, this is just the exception which we might expect to see made, considering the nature of the circumflex ; and that it is not universally recog- nized in Sanskrit usage indicates the very different position which the circumflex takes in the Sanskrit system, as com- pared with the Greek: the latter language has a predilection for it, and lets it appear in innumerable cases where it has no etymological justification ; the former has a prejudice against it, and exhibits it only where compelled, as it were, to do so.
But there is one exceptional and infrequent case where the Pratigékhyas require, and Vedic usage shows, a circumflex as result of the fusion of an acute vowel with a grave into one.
98 W. D. Whitney,
It is where two short 2’s are combined, forming long ¢. Thus, divt wa becomes divi'va, abhi thi becomes abhi'hi, and so on. This appears very arbitrary, as we can see no reason why ¢ under such circumstances should receive a circumflex accent any more than @, or @, or the diphthongs e, d7,0, du. And the impression of arbitrariness is increased when we come to no- tice that one Pratigékhya (the Taitt. Prat., at x. 17) denies the circumflex to 7, but gives it to @, combining médsé ut into mdéstt,and soon. This is equivalent to restricting still fur- ther the occurrence of the circumflex in question: there are no more than sixteen cases of ¢ in the whole ample text of the Atharva-Veda ; but there are only five of ¢@ in the yet ampler Taittirfya-Sanhité (which contains about thirty passages in which ¢@ would appear, if the usage of the text permitted). This rare and anomalous circumflex is styled praglishta (or prdglishta), ‘resulting from fusion.’
Such, then, are the cases in which a circumflex arises by the combination into one syllable of a preceding acute and a following grave element. Besides these, there is a limited class of words which show a circumflex as their proper and sole accent: for example, the nouns svar, ‘ heaven,’ and kanya’, ‘girl,’ the particle kva@, ‘where ?’, the adjective budhnyd, ‘fun- damental,’ and so on. But every word of this class contains a y or v before the vowel of its accented syllable; that is to say, the syllable is of the same kind with those which in combina- tion receive the kshdéipra circumflex ; and it is obvious that the circumflex is indeed essentially a kshdipra, its origin lying merely a step further back. By this I mean, that kanya and kvd, for example, stand for more original kanf-¢ and ka-a, so that the circumflex of kanyd' is precisely analogous with that of nadydsya from nadi' asya. And the Vedic verse clearly shows that the fusion of the two syllables into one, with con- sequent circumflex, is a fact not yet accomplished in Vedic times: such syllables are more often to be read as two than as one, kvd becoming dissyllabic, kanyd’ trisyllabic, and so on. Indeed, the Taittirfya-Sanhita (which has a peculiar ortho- graphic usage with regard to a part of these words) regularly writes sévar instead of sudr, budhntya instead of budhnyd, and
On the Accent in Sanskrit. 99
s0 on — while, on the other hand, it resolves indrégnyés into indrdgniyds, bédhvds into bdéhuvés, and so in other like cases.
To the circumflex accent which thus appears as the original accent of a word, and does not arise in the course of the eu- phonic combination of words into phrases, most of the Hindu grammarians give the name of jdélya, ‘native, natural;’ the Taitt. Prat. alone calls it nitya, ‘constant, unchanging.’ Wo may class the cases of its occurrence under four heads:
a. Words like svdr and ked, of which the derivation is ob- scure or without extended analogies.
6. Words having the circumflexed suffix y@ (representing f-a or t-a), like kanya’, viryd, dhanvanyd, samsrdvyd : this class is a pretty large one.
e. Forms of declension made by adding an unaccented case- ending to an accented final 7’ or @': thus, svddhyds from svddhi' ; lalémydm from laldémi'; tanvds, tanvdm, tanve from lant’ (these last are especially frequent ; the Taittiriya-San- hita writes them tantivas, tantive, etc.), prddkvds from prdakd’. These are mostly nominatives singular dual or plural, or accusatives singular or dual, since the endings of all the cases except these five tend to draw the accent forward upon them- selves; and so we have, for example, from nadi’, the nom. dual nadyéi and nom. pl. nadyds, but in other oblique cases nadyd@'i, nadyd's,etc., just as we have from nd‘us the nomina- tives nd'vdu and nd'vas, but the oblique cases ndvé, ndvds, nivt, etc.
d. Accented vocatives, like dyd‘us, jy@ke. It is the rule, namely, in Sanskrit, that the vocative, if accented at all, is accented on its first syllable. If, now, this syllable, as written, is one of which a semivowel preceding the vowel has in the metre the value of a syllable by itself, that element alone takes the acute tone, and the written syllable is circumflexed. The words given above, therefore, are equivalent to df-dus, 7{-dke, and have to be so read in scanning the verses in which they occur. The case is quite a rare one; it occurs, I believe, only twice in the Atharva-Veda (at i. 2.2 and vi. 4.8: but in the latter passage the edited text reads incorrectly dyd’us),
30 W. D. Whitney,
and I presume there are hardly more than that number of cases in the Rig-Veda.
To illustrate the degree of frequency of the circumflex ac- cent in actual use, in its different kinds, it may be stated that in the first twenty hymns of the Rig-Veda (about two hundred lines, mostly of twenty-four syllables each; a part, of thirty- two), there are but ten cases of it, namely six kshdipra, one abhinthia, and three jdtya; of which last, there is one case each falling under a, b,and c. In the first pragna or chapter of the Taittiriya-Sanhita, again (a somewhat less body of text, chiefly prose), there are six kshdipra’s, eleven abhinihita’s, and three jdtya’s (all falling under 4). In the first book of the Atharva-Veda (having about the same extent as the passage from the Rig-Veda already reported), once more, there are forty-six cases; eleven kshdipra’s, three abhinthita’s, and thirty- two jdtya’s ; of which last, ten fall under 4, twonty-one under c,and one under d. .
The diversity thus appearing between the different passages compared is in part accidental, in part characteristic of the texts from which they are respectively taken: in general, it will probably be found that the Ashkdipra circumflex appears oftenest, and the jdlya, its nearest kindred, next; the praglishia, as already explained, is by far the rarest of the four.
So far, there is nothing difficult or questionable either in the theory or in the practice of Sanskrit accentuation, and all the phenomena are of a nature to favor and establish the truth of that description of the nature of the svarita or circumflex accent which is given by the grammarians. But we have next to consider an addition to the theory which is of a very prob- lematical character. The native authorities, namely, teach unanimously that a syllable naturally grave, if it follows an acute, in any part of q word, or in a succeeding word in the same phrase, takes the circumflex tone. Thus, the Rik- Pritigékhya, which makes the fullest statement, says (iii. 9) ‘a grave syllable preceded by an acute is circumflexed , ‘whether separated from it by a hiatus or by a consonant;”’ and the others teach virtually the same thing.* That is to say (as
* Sec Vaj. Prat. iv. 134, Taitt. Prat. xiv. 29, 30, Ath. Prat. iii. 67, Pan. viii. 4. 66.
On the Accent in Sanskrit. 31
we may state to ourselves the virtual meaning of the doctrine), the voice, when once raised to the higher pitch of acute, is not able to descend to the general level of utterance in the inter- val between the acute syllable-and its successor, but slides down in the course of the latter; it occupies a syllable in its descent. Thus, ff is not grave, but circumflex, on ita final ; sé’ and asya become sé‘sya with circumflex on the last a; the bru of the toneless bruvanti is circumflexed in the combination té bruvanti, and so on. This circumflex, by European gram- marians, is conveniently distinguished from those which have been described above by being called the “dependent” or the ‘“enclitic’” circumflex: the term, however, has no correspond- ent in Sanskrit, nor do the Hindu grammarians, by descrip- tion, classification, or designation, intimate a recognition of any difference in character between the enclitic and the inde- pendent varieties of this accent. The Pratigékhyas divide the former, as they do the latter, into sub-varieties, calling tdiro- vyaijana, ‘having consonants interposed,’ a circumflex vowel between which and the occasioning acute consonants inter- vene (as in the examples above cited), and pdédavrtta, ‘with word-hiatus’ (or, in the Rik Prat., simply véiurtta, ‘with hia- tus’), one that is separated from the acute by the hiatus that sometimes intervenes between two words (as in the u of yd’ u ea, from d'hk u ca, and of td ucyante, from té ucyante*). No attempt is made to describe these two varieties as exhibiting
* There are unimportant differences among the different Praticakhyas in re- gard to the classification of the enclitic cireumflexes. The Vj. Prat. (i. 118) makes a special class, ¢dirovirdma, ‘with intervening pause,’ of those which are separated from the occasioning acute by the pause which in pada-text divides the two parts of a compound word: thas, 96 pa‘d« or yajfid patim has this kind of cir- cumflex on pa; and itis probable that the TAitt. Prat. (xx. 3) intends to desig- nate the same as prdtihata, though the native commentary understands the case otherwise, and makes prdtihata signify a circumflex standing in another word than the acute and separated from it by consopants, like the d of yds ird, or the u of té bruvanti, as given above. The Taitt. Prat. further confines the name pddavrtia to cases involving a hiatus between two independent words, and in the excet dingly rare cases where there is a hiatus within the word itsclf (as in prdiigam), classes the circumflex as t4‘rovyaajana, though there is no ‘intervention of consonants ’ at all here. And the Vj. Prat. (i. 120) has a peculiar name, téthdbhdcya, for the ‘very uncommon. and somewhat anomalous case of an enclitic circumflex between two acates.
82 W. D. Whitney,
any difference of character, and their distinction is evidently only an example of that tendency to over-refinement in classi- fication which is characteristic of Hindu systematists.
It is not possible to accept the teachings of the Sanskrit grammarians as to the enclitic circumflex with the same trust as those respecting the independent circumflex. Even the Greek, which shows a marked predilection for the circumflexed tone, never admits it save in an actually long vowel, not find- ing room elsewhere for the exhibition of the double tone or downward slide. And the Sanskrit, as we have seen, has so much less inclination to this tone, or capacity for it, as to ad- mit it only in very rare cases even upon a long vowel or a proper diphthong: to bring it out, there is needed an improper diphthong, as we may fairly call it—a vowel-group in which the first element is an ¢ or w sound, written as a y or v, but still evidently retaining in good measure its vowel value and its capacity of tone. In a language of this habit, then, it is next to incredible that we should find a circumflex tone de- veloped with the utmost freedom even in short vowels (as the finals of ft and mddhu) by the simple neighborhood of an acute. We cannot, so it seems to me, avoid suspecting tho accuracy of the observations which underlie the whole theory of the enclitic circumflex. The Taittirfya-Pratigékhya is in- genuons enough to inform us (xiv. 33) that some authorities denied this circumficx in toto. If we do not carry our own skepticism so far as that, we shall be likely to take refuge in the theory of a “middle tone,” like that assumed by Misteli and Hadley (see the preceding article, p. 11.) in explaining the peculiarities of Greek and Latin accent. This would im- ply that the enclitic tone which was perceived to lead down from acute pitch to grave was in reality a step intermediate between the two, and was hastily and inaccurately apprehended by the Hindu grammarians as a combination of the two, or a slide, and so identified with the independent circumflex, of which the origin and character were too clear to admit of any doubt or question. Through this modification of the Hindu theory, we may win from the Sanskrit enclitic circumflex a degree of support for the ‘“‘ middle tone;”’ but it is necessary
On the Accent in Sanskrit. 33
that we understand and confess the fact of the modification. Quietly to assume, as Misteli does, that the whole Sanskrit circumflex, in both its independent and its enclitic varieties, is only a middle tone, is wholly unallowable, being opposed to the plain and unanimous statements of the Hindu gramma- rians, and, hot less, to the teachings of a sound accentual theory.*
A single exception is | made by all the Hindu text-books to the rule that a grave syllable following an acute takes the enclitic circumflex: if, namely, an acute or a circumflex (of course, independent) immediately follows, the grave retains ita proper quality. Thus, in yé ca, ca is circumflexed ; but in yé ca té or yé ca svar it is grave (Rik Prat. iii. 9, Vaj. Prat. iv. 185, Taitt. Prat. xiv. 81, Ath. Prat. iii. 70, Pan. viii. 4. 67). The Taittiriya-Pratigdkhya (xiv. 82) reports a single gramma- arian (Agnivecyfyana) as disallowing this exception, and maintaining the circumflex even before a syllable having or beginning with high tone; and Panini (viii. 4. 67) mentions three authorities (Gargya, Kacyapa, and Gélava) as holding the same doctrine. I do not see that theoretical considera- tions teach us anything with definiteness upon this point; we should quite as naturally, however, in my opinion, expect the enclitic tone, whether slide or middle tone, to maintain itself before an acute or circumflex as well as before a grave.
These are the principal features of the system of Sanskrit accentuation. We need, however, to follow it out into one or two further details, important as regards the native theory ; and, before we can do this, we must take note of the Hindu
* It would almost seem that Misteli’s view was to be looked upon as the Italian one, since Ascoli also, in his lately published lectures on comparative philology (Corsi di Glottologia etc., first part, Comparative Phonology of the Sanskrit, Greek, and Latin, p. 15), expresses himself upon the subject as follows: “ The syllable, finally, that follows the acute, becomes svarita, ‘ tonic,’ or, in other terms, assumes the scarita accent—which some European grammarians (infelicitously, as it ap- pears to me) have called ‘circumflex’ —; that is to say, it has a tonality higher than the ordinary, but not so high as is that of the syllable with acute.” A scholar of Ascoli’s rank and claims to respect should not allow himself thus summarily to set aside the carefully deduced results of his predecessors, without bringing up a single consideration to support the view he takes. |
2)
34 W. D. Whitney,
mode of designating accent. This is in the main the same in all the accentuated texts: some minor differences will appear as we go on; others may be passed over as of no im- portance to our purpose.*
By a striking —and as IJ think, an ill-judged and unfortu- nate — peculiarity of usage, the Hindu does not designate directly the really accented or acute syllable, but, instead, its surroundings, or the preceding grave and the following cir- cumflex. Or, as we may express it, the acute syllable is left unmarked ; but the grave syllable is marked, by a horizontal stroke beneath the syllable ; and the circumflex (whether en- clitic or independent), by a perpendicular stroke above. In the manuscripts (so far as I know, without exception), these. strokes are added in red ink; in printed text, of course, they have to be represented by black lines.
Thus, examples of single words (with transliteration, ac- cented according to our method, written below each word) are
| of 1] ~ TAT Rote | TPT: a-gni-nd ka-ri-shyd-si co-da-yi-tri'-bhih An independent circumflex is generally to be distinguished
from an enclitic (of which latter kind are those in the above examples) by having no unmarked acute before it: thus
l ] 2 Ta | BAT TPIT sudh ka-nyd yd-tu-dha-nyah A word that is unaccented has the mark of grave under every syllable; an acute monosyllable has no mark at all; and an initial acute has, of course, no grave mark before it, nor a final acute any circumflex mark after it: thus
De Peay a TTT TL | LTH
ca va-ksha-ti sdk pd-sham a-gnim su-ri-pa-kr-tnim We may conjecture plausibly enough that these marks are symbolical: the horizontal line below intimates the lower tone
* The designation of the independent circumflex in the White Yajur Veda is in sundry points peculiar ; see Weber's edition, Prefice, p. x.
On the Accent in Sanskrit. 85
at which the voice moves on; the other directs the voice downward from a higher key. :
This mode of accentuation wonld seem to involve an exces- sive repetition of the sign of grave tone; but the difficulty is avoided (as we may provisionally say) by omitting all signs of grave after a circumflex except under a syjlable that pre- cedes another acute or an independent circumflex. In this way, there is no repetition of the grave sign under successive syllables-except at the beginning of a word or phrase, before the first accented syllable is reached. Thus,
~ | 1 | , «BTA | TATA t PST a ws -_ a, -—_ hé-td-ram __ra-tna-dhd'-ta-mam __ ci-trd-cra-va-sta-mah
The same rule is followed, of course, where words are com- bined into’ phrases; and
RTT ETT saCT Fra aE aa UIRCEROGEED GRICU CMMI ETE d-gne ydm ya-jadm a-dhva-rdm vi-cud-tah pa-ri-bhivdh —dast
becomes, as a sentence,
si tamara: Ue RIGe
In this phrase, it is to be noted, the e of dgne, which was enclitically circumflex when the word stood alone, becomes grave again (by the rule stated above, p. 83) when the acute syllable y¢@m comes to stand after it: and, on the other hand, the grave initial a of adhvardm takes the enclitic circumflex when conjoined with the acute final of yajtdm. The penulti- mate and antepenultimate syllables of the line, as both alike standing between grave and circumflex signs, are both acute, the voice when once sent up by the former sign remaining raised until sent down again by the latter. In the same man- ner, when conducted down by the circumflex sign, it contin- ues at the lower pitch till the other directs it to go up: an example illustrating more fully this effect is
~ | ~ is ! .arreadanyat er sar ar sta émdsag _carshanidhrto vicve devisa a’ gata where the first four words are vocatives, and the last a
36 _ W. D. Whitney,
finite verb in an independent clause —both of which classes of words are toneless in Sanskrit, except when, like the first and third words, they stand at the head of a sentence or of a primary division of a verse.
I will note here but one other peculiarity of the mode of designation: namely, that if an independent circumficx follows an acute, it is in no way distinguished from an enclitic cir- cumflex ; so that, for example, in
. at waa:
it is impossible to tell whether the latter word is a toneless vocative, svarvatih, or a nominative or accusative with its regular accent, svdrvatih.
_ It thus appears that two classes of syllables in accentuated text are left unmarked: first, those which are properly acute or uddita (as 6-, vig-, and d’, in the last example but one); and second, those which are properly grave or anuddtia, but which neither receive the enclitic circumflex as following an acute, nor are marked by the grave sign, as preceding an acute or circumflex. But there is a strange and. perplexing addition to the Hindu theory which Virtually identifies these two classes with one another in respect to tone: and this we have next to examine.
The point is thus stated in the Taittirfya-Praticaékhya (xxi. 10, 11): “in continuous text, after a circumflex, a series (pracaya) of grave syllables has the tone of acute (literally, ‘sounds, or is heard, as acute’), except the syllable that is followed by an acute or circumflex.’’ The other Pratigikhyas have equivalent rules (Rik Prat. iii. 5, 11,12; Vaj. Prat. iv. 188-40; Ath. Prat. iii. 71, 74).*
-*The Vaj. Prat., in describing the character of the accent, uses the term uddttamaya, ‘made of, or of the same material as, acute,’ instead of (like the rest) ucddttagruti, ‘having the audible quulity of acute.’ The Rik Prat. s:ates the rule twice over, calling the tone once (in verse 11) uddttagruti, and once (in verse 5) uddtia, ‘acute,’ outright ; it also adds one or two peculiar items, the first to the effect that ‘“‘some, however,. depress (to grave tone) one syllable, or more than one, at the end of the series; or even all but one or two;” the other (verse 13) is too obscure to be made intelligible without more exposition und discussion than we have room for here.
On the Accent in Sanskrit. 37
To this accent is given the name pracaya, ‘accumulation, | indefinite series,’ or pracita, ‘accumulated, indefinitely con- tinued,’ from the way in which it is liable to run on, to the extent, sometimes, of ten or a dozen syllables, or more.*
According to the Pratigaékhyas, then, the treatment in con- tinuous text of syllables naturally grave is as follows: a single such syllable following an acute before a pause becomes an enclitic circumflex, but before another acute or circumflex retains its grave character (compare dgne, above, illustrations 5 and 6); of two such syllables, the first always has the cir- cumflex, the second is either grave before another accented syllable (as -dhva- in illustration 6), or pracaya, with acute tone, before a pause (as the finals in the lagt two words in 4 and the last in 7); of more than two, either all but the first and last, or all but the first, are pracaya (as -sa¢ carshanidhr- and devd- in 7, or -vastamah in 4) ; — after a circumflex, the treatment is the same, with the exception that the first syllable is not circumflexed, but remains grave or becomes pracaya, according to the character of what follows it; — finally, any number of grave syliables coming before the first acute or circumflex in a phrase retain their grave character. ,
The striking result of this is, that there comes to be a com- plete accordarice between the theoretical tone of each syllable and the way it is marked. Every syllable that has the per- pendicular stroke above it is circumflex, and of the same tone, whether the circumflex is independent or enclitic; every syllable that has a horizontal mark beneath is grave, or of low tone; every syllable that has no mark at all is of high tone, whether it be properly acute, or originally grave and con- verted to pracaya.
An interesting and important question now arises: What is the ground of this complete accordance? Is the mode of designation posterior to the theory? Did the Hindus leave the pracaya grave syllables unmarked, like acute, because
“For example, in the scries of toneless vocatives dgne dudhra gahya kimstla vanya.
38 W. D. Whitney,
they gave them the tone of acute, so that their identity of treatment is due to a perceived identity of character? or, on the other hand, was the omission of a part of the marks indi- cating grave tone made for the sake of convenience, of brevity (as I have, provisionally, assumed above to be the case), and is the theory an afterthought, suggested by the identity of designation, and aiming to establish a corresponding identity of character ?
To many scholars, perhaps even to nearly all, the question I have raised will seem strange, and the latter of the two sug- gested explanations one altogether to be rejected, as implaus- ible and incredible. Nevertheless, I am convinced that the matter merits a serious discussion. And, in the first place, we shall not, I think, be doing the ancient Hindu grammarians an utter injustice, In supposing them capable of fabricating a theory like that of the pracaya-accent, under the induce- ment stated above. The “schools” of Vedic study in which the Praticakhyas originated were far removed from the period in which their sacred texts had grown up, or even had been brought into their present shape; and these texts had begun to be treated in more than one respect in a somewhat formal, arbitrary, and unintelligent manner. The acuteness of ob- servation, and the skill in combination and systematization, displayed by the grammarians, as by the Hindu workers in other departments of science, are worthy of high admiration: but we cannot equally commend their moderation, nor is their soundness to he trusted to the end. They never knew where to stop; and their systems, as has been said elsewhere, always tended to take on a prescriptive character where they were intended to be descriptive only — putting violence upon the facts which they set out simply to examine and classify. I should not be restrained from regarding the pracaga-accent with acute tone as a later and artificial addition to the origi- nal accentual theory by any exaggerated faith in the infalli- bility of the authorities who report it to us.
A single scholar, of high rank, and especiaMy conversant with the Pratigaékhyas,* has, it is, true, maintained that the
* M. Miller, in his History of Ancient Sanskrit Literature, p. 497ff.
On the Accent in Sanskrit. | 39
Pratigaékhyas antedate the use of writing in India, and do not presupposo any recorded text at all. But, if I am not mis- taken, this view of his is generally rejected as paradoxical. The Pratigékhyas do, indeed, like most of the sacred litera- ture of India, and some of its literature in other departments, studiously and successfully ignore everything written, and acknowledge only the oral tradition of the schools as the law- ful channel for the conveyance of knowledge from generation to generation; but, in the view of most men, such phonetic analysis as they make, and such fixation of the nicest details of reading as they imply, would be more th:n wonderful, it would be absolutely miraculous, without the aid of writing. A race that had made advanced and recondite studies in every department of grammar without even possessing an alphabet would be a greater prodigy than “the Anthropophagi, and men whose heads do grow beneath their shoulders.”’
The obstacle in the way of our accepting the doctrine of the pracaya-accent is the seeming impossibility of working it in as a part of the general accentual system, such as this has been described above. The acute is a syllable which is ac- cented or rendered conspicuous by being lifted above the lower or grave tone of utterance: but what does this distince- tion amount to, if the grave syllables also are to be raised, even in masses, to the same level? What a strange grave tone is that which can only maintain itself by the help of an immediately following acute! What are we to think of the independent circumflex, made by the combination of an acute and a grave element, when the grave syllable, even without the encouragement of an acute joined with it, can hardly be held down at grave pitch, but is constantly rising to the higher plane of utterance? Yet worse, what sort of an enclitic ac- cent is that which leads down the voice from acute pitch no- whither, since, the moment the transition is past, the voice leaps back again to acute ?
The last of these difficulties, and the most insurmountable of them, seems to have been felt by the Hindu theorizers themselves, who have made a very curious attempt to avoid it: namely, by shifting the circumflex itself to a higher plane.
40 W. D. Whitney,
Thus, the Rik Praticaékhya says (iii. 2,3) “of this [circum- flex], a half-mora, or the moiety, is more acute than acute; the following remainder is grave, but sounds like acute” literally, ‘is uddttagrutz,’ which is the same term that is used in describing the pracaya). That is to say, the circumflex begins higher than acute, and descends only to acute pitch, thus conducting the Voice to the level at which it then runs on in pracaya. And the Taittiriya-Pratigékhya (i. 41, 42, 44, 45) declares “of this [circumflex], when it follows an acute, the beginning, to the extent of half a short vowel, is higher than acute; the remainder is the same with acute; or the re- maining part is still lower; or it is the samo with grave.” Of these rules, the first two state the theory squarely; the others appear to express the scruples of those who are struck by its inconsistency with the fundamental principle, that circumflex is the combination of acute and grave within the same syllable. The remaining two Pratigakhyas take no notice of any such modification of the nature of the circumflex. And, we may say, with very good reason; for nothing is really gained by it. On the contrary, it would be no less strange that the enclitic accent, the transition step from an acute to a following grave, should rise a grade above acute, and come down to acute it- self, than that the grave to which it led should spring to the height of its first element. And, according to the Rik form of the theory, we should have the independent circumflex, arising out of the union of an acute and a grave element, and having no existence except as the result of such union, lifted up in all or nearly all its substance above acute pitch; while, in the Taittiriya form, whether it were so lifted or not would depend upon the accidental circumstance whether it were pre- ceded by an acute ora grave syllable. There would be no sense in our assuming that even an independent circumflex after an acute might be raised in pitch for the sake of clearer distinction from that acute; for it is sufficiently distinguished by its sliding tone; and, if it had any right to be further dis- tinguished, an acute following an acute would have much more right; while, nevertheless, any number of acutes are
On the Accent in Sanskrit. 41.
allowed to succeed one another, without modification of their natural character. |
To my own mind, I must acknowledge, the difficulties that encompass the Hindu theory of a pracaya accent, giving to grave syllables the tone of acute, are more numerous and formidable than those involved in its rejection. If there be found any one, skillful enough to smooth these difficulties away, or to devise an explanation of the setting up of the theory other than that which I have suggested, no one will rejoice in his success more than I shall. But for the present, I shall discredit the existence of a fourth or pracaya accent, and shall hold it to be at least probable that this was fabri- cated merely in order to establish an identity of character in those syllables which, according to the current method of ac- centuation, agreed in their mode of designation —or rather, in being alike undesignated.
In what has been said of the pracaya, regard has been had only to the teachings of the Pratigadkhyas, since they alone distinctly recognize and name this accent. Butin Panini also appears a somewhat kindred doctrine. The great gram- marian, namely, has the rules Ci. 2. 39, 40) that “grave syllables, following in continuous text a circumflex, are ut- tered in monotone (ekagruti); but one that is followed by an acute or circumflex is more depressed (sannatara, which the commentators explain by anuddttatara, ‘more grave,’ 1. ¢. ‘lower than grave’).” Precisely what is intended, now, by ekagrutt, ‘monotone,’ is less clear than were to be desired. It is by the commentators defined to mean, ‘ without distinc- tion of acute, grave, and ‘circumflex’; but, if it signified only this, there would seem to be no good reason for declaring it to belong to a series of grave syllables anywhere; these would of course all be uttered in the same tone, unless there should be given express direction to the contrary. The first mention of ekagrutz is a little above Cin rule 33) where it is declared to be the tone used “in calling to a person from a distance ;” and it is further prescribed in the rules that fol- low, for certain conjunctures of sacrificial ceremony from which we can infer nothing definite respecting it. But
6
42 W. D. Whitney,
certainly, in shouting to a distance a raising of the voice would seem to us 9 natural and almost necessary accompaniment of the obliteration of ordinary distinctions of tone. We have, then, both a negative and a positive reason for suspecting in Panini’s ekagruti a peculiar and a higher tone, instead of a mere negation of varied tone; and so for recognizing a nearer relationship between his teaching and the pracaya-doctrine of the Pratigakhyas than appears at first sight. An unequivocal sign of the same relationship is the prescription of a lower tone for the syllable, naturally grave, that precedes an acute or circumflex. Whether this actually means, as is generally assumed, a yet lower tone than grave, or only a lower tone than the heightened monotone of ekagrutz, it seems to me an equally unacceptable part of a sound accentual theory. I cannot recognize a positive sinking of the voice as a necessary or natural preparation for its rise to the pitch of acute.
I am not aware that any one has ever undertaken an inves- tigation of the modes of accent usual among the modern lan- guages of India for the purpose of casting light upon the an- cient systems, as laid down in the grammatical text-books and exemplified in the Vedic texts. The mode of recitation of the Vedic texts themselves, however, as practiced by the na- tive scholars of the present day, has been carefully observed by an eminent European Sanskritist, Dr. Haug (now of Munich), and reported in the Journal of the German Oriental Society (vol. xvii., 1863, pp. 799-802). I quote the most important paragraph : |
“ The fundamental law of the Vedic accent is a triplicity of tone. In by far the greatest number of cases the voice begins with a strong tone, the low tone, called anud@tta, rises in uddita —which, however, has absolutely no perceptible tone [i. e. stress ]— and only reaches its full height and force in the svarita, or high tone. The real chief accents are the anuddlta and svarita alone, which are also constantly interchangeable, as is sufficiently shown by a comparison of the pada or word- recitation with the continuous or sanhttd-recitation. The uddtta is only a kind of auxiliary accent, and I could never, with the closest observation of the mode of recitation, perccive
On the Accent in Sanskrit. 43
that the syllable having udétta is really an accented syllable..... The uddtta is only imperceptibly distinguished from the praca- ya or entirely accentless syllable.”
A very cursory comparison of this description with the teachings of the Pratigdakhyas, reported above, will show how far the rules of utterance laid down and follawed in the ancient schools of Vedic study are maintained in their modern successors. The identification of the unmarked acute with the unmarked grave, in point of tone and character, is the same now as of old. The distortion of the circumflex, by pushing it up to a higher position in the scale of tones, which had begun to appear in the Praticéikhya period, has now be- come complete. And, partly under the influence of-these two causes, there has come about an absolute inversion of the original accentual relations. The effort has been to mark and — distinguish .by the voice the two tones that were marked and distinguished by writing. And so the old accented syllable, the acute — the central and determining point. of the whole system, out of which and in strict subordination to which the others have their being —since it has no written sign, has sunk into insignificance, and become a mere “auxiliary” to the other two, which were fortunate enough to be plainly desig- nated. The influence of the written sign on the theory and practice of the spoken accent here exhibited is an important support to the explanation suggested above of the origin of the so-called pracaya accent.
There is one further point in the theory and designation of the Sanskrit accent which it is desirable to consider, in order to make our view of the subject more complete. It is the kampa (literally ‘trembling, shake, trill’), or the peculiar modification undergone by an independent circumflex when followed by another circumflex or by an acute. This modifi- cation consists, according to the Pratigakhyas, simply in a depression of the last portion of the circumflexed syllable to a lower than the ordinary pitch. Its explanation makes a very small figure in the rules of the different treatises. In the Vaj. Prat. Civ. 187) it is so obscurely stated, and in such am- biguous connection, that it would hardly be intelligible, but
44 — W.'D. Whitney,
for the comparison of the other kindred works, and the light that the written form of the texts throws upon it. In the Rik Prat., also, it appears only as an exception to the definition of the ordinary character of the circumflex: the latter part of the circumflex is grave, but has the tone of acute— “except when an acute or circumflex syllable is spoken immediately after it’? (iii. 3). The Ath. Prat., however, says distinctly that, of the varieties of independent circumflex, ‘“‘ when fol- lowed by an acute or a circumflex, a quarter-mora is de- pressed” (iii. 65). And the Taitt. Prat. (xix. 3) mentions it as “the opinion of some authorities that, in a circumflex that is followed by a circumflex [not by an acute, as well], a quarter-mora is depressed.”’ The mode, however, of designa- ting this peculiarity is in nearly all the texts a very conspicu- ous one, and involves an element of prolongation of the vowel suffering kampa, of which the Pratigakhyas give no hint, and which is intimated by writing a figure, 3 or 1, after the vowel. Thus, the Rig-Veda writes tishyd ydthé and yo hyo vartanth as follows:
aoa anit) ar acta:
If the affected vowel is short, the Rig-Veda writes a 1 in- stead of a 8 after it: thus,
I. 0 HIT TAT | ny d-nydim sd-ha-sd
The Sama-Veda (to which there is no known Pratigakhya), however, always lengthens the vowel in such a case— thus, ny d-nydm——and so writes a 3 after it, without exception. What is the proper usage in the Atharva-Veda is somewhat doubtful, since the known MSS. are very irregular and incon- sistent in this whole class of cases; but the editors of the pub- lished text have adopted and carried out the method of the. Rig-Veda. Inthe Taittirfya-Sanhita, the modification is, in accordance with the doctrine referred to in its Praticakhya, restricted to the case of a circumflex followed by a circumflex ; the figure used is always 1, and the affected vowel, if short, is lengthened: thus,
On the Accent in Sanskrit. 45
» SAAT BTL VAT SAY ——
“de-va-tyait hy ¢-tdt 80 ‘po ‘bhy a-mri-ya-ta devatyd'i in the first example being from devaty@m. The Vajasaneyi-Sanhité alone leaves the quantity of the vowel unchanged, and uses no figure, but only a modification of the usual circumflex stroke bencath the vowel.
Whatever may be the origin of this peculiar doctrine and designation, it answers in the majority of texts (Rik, Siman, Atharvan) a useful purpose by distinguishing m many cases an acute syllable from a pracaya grave after a circumflex. Thus, in the last examples quoted, the accentuation does not show us that the final -ta¢ in the one case is acute, and that the final -mriyata in the other is pracaya, while, as written by the Rig-Veda- — namely,
CITY WATT Ate oThy sone
the difference is brought clearly to light. But that this prac- tical advantage had any thing to do with the development of the theory, is not lightly to be assumed. The designation seems to signify that the circumflex vowel, in order to the re- duction of its latter portion to anuddtta tone, or toa yet lower pitch, requires to be somewhat prolonged in quantity, either by a single added mora or by extension to.three moras. But the whole subject is quite obscure, and I do not venture to enter here into the discussion of it, for fear of occupying much space without sufficient result.
It does not belong to what I have undertaken that I should consider at all the correspondences and differences of the Sanskrit and Greek with reference to the actual phenomena of accentuation—the clear evidences which they exhibit of an originally identical system, and of the abandonment of this system in part by the Greek, under the influence of the new law of cadence set forth in the preceding paper. These cor- respondences are most fully and clearly stated in Bopp’s “ Accentuations-system,” a work which, though often wrong in matters of theory, is to be highly commended as a clear and comprehensive exhibition of the facts of which it treats.
TII.— On the Aorist Subjunctive and Future Indicative with "Owe and Ov py. - By WILLIAM W. GOODWIN, ELIOT PROFESSOR OF GREEK LITERATURE IN HARVARD UNIVERSITY.
The rule was first laid down by Dawes, an English scholar of the last century, (Miscellanea Critica, pp. 222, 228,) that érwe py and the double negative od pi can take the future in- dicative of all voices, but the aorist subjunctive active and middle only in the second form. This declares the first aorist subjunctive active and middle.a solecism after these particles. Later writers have extended the rule to érwc (without p73) and indeed there seems no good reason for restricting such a rule to negative clauses. A large number of examples are given by Dawes and his followers, showing that there was a strong tendency on the part of the Greek writers to observe some such principle as the one above stated. This arises from the fact that the future indicative and most of the prohibited forms of the subjunctive differ only by the. quantity of a single vowel; so that, whenever a writer wished to use a subjunctive after a particle which allowed both the future and the subjunct- ive, he naturally avoided, if possible, an expression which was nearly identical in sound, and (before long and short £ and O were distinguished in writing) generally identical in form with the future. For example, the regular use of érwe with the future indicative is found in object clauses after verbs of striving, effect- ing, &c., where the dependent clause is the real object of the leading verb or in apposition with an accusative like rouro; as oxoret Srwe Touro py) Torhoe, See to it that he does not do this. Now when a subjunctive was required in a pure final clause, an au- thor would avoid such an expression as érwc (or drwe pi) rorjon, and would prefer some other final particle (as iva); or he would use some synonymous verb in which the second aorist subjunctive occurs, or perhaps a present subjunctive. This, however, was a mere dendency; and it should not be raised to the dignity of a grammatical principle.
The construction just mentioned, that of object clauses with
On the Forms allowed after *“Orwe and O& py. 4T
éxwe, supplies nearly all the standard examples of érwc and the future. Here the subjunctive is allowed only as an excep tional form, apparently as a less vivid form than the future indicative; and it was especially natural here that a writer should choose the second aorist, when he wished to use a sub- junctive, rather than a form which was almost identical with the ordinary future. Still, unless we determine to carry out the general principle against the highest manuscript authority,. we must admit a few cases of the first aorist subjunctive even here; asin Demosth. Phil. III. p. 128, 25: dxwe pnéeic avarpéyn, rouro oxoneiadat (as all MSS. have it), where modern editions have avarpéfec by emendation. Again, in Demosth. Olynth. I. p- 9,17: rapacxevicacba thy rayiorny drug EvOévee BonOhanre xai po) raOnre ravrov (where modern editions have /onBhcere and xanre, although all the MSS. except one have onfhonre), there seems to be no good reason for objecting tothe first aorist sub- junctive, when all admit the other subjunctive 7a6yre. The mix- ture of the two moods is certainly more objectionable than that of the two forms of the aorist, especially when the former is the result of conjectural] emendation. Whatever we may think of doubtful ‘cases like the first example, where certainly the au- thority of usage is decidedly in favor of the future and against the subjunctive, there is no ground foran extension of the prin- ciple to a construction in which option between the subjunctive and the future was hardly allowed in AtticGreek. I mean that of pure final clauses, in which the subjunctive is the only good Attic prose construction, the exceptional cases of the future being too few to be considered. But even here the emendation on Dawes’s principle has been extended, so that many sentences have been made nearly or quite ungrammatical by unneces- sary emendation. Thusin Thuc. VII. 39, érwe dpororajowrra. kal... émtxeipoory is supported by many of the best MSS.; but the editions have dprororajaovrat, which is entirely opposed to Attic usage. So in Thue. II. 60, durfyayov, érwe tropvijaw Kai pépbopa is very often emended by reading pépyoua, a reading which ought not to be accepted without the highest manuscript authority, and then only as a rare exception. Again in Xen. Cyr. VII. 5, 82, drwe.... aroAaiowpev cat drwg ... yerwpueda is
48 W. W. Gouvdwin,
unobjectionable, although it violates the rule of which we are speaking; if it conformed to that rule it would be irregular, not to say ungrammatical. These cases of the first aorist are rare, for the simple reason given above; but the general statement of the rule, and its application without distinction to both constructions of ézwc, ignore one of the most impor- tant principles of Greek syntax.*
A large number of cases in which the first aorist subjunct- ive with érwe pt) has been emended to the future indicative be- long to the elliptical construction in prohibitions, where a verb like oxoweire, see to it, is understood (as in érwe pi pe Oicere, see that you do not sacrifice me, i. e. do not sacrifice me, Aristoph. Nub. 257). For example, In drwe cé rouro poy éicatne pncéva, but see that you teach this to nobody, Nub. 824, the modern editions read ¢é:daéere against the authority of all the MSS. Now, even if we do not admit the full construction axoret Gru pr) dcdaénc to be classic, it is a new and (as it seems to me) an unwarranted assumption that, when oxdree was omit- ted and the expression became a mere prolhibition,—in which, probably, no common Athenian ever thought of oxdre, or ever knew that he was cominitting an ellipsis of that verb,—every
* In Soph. Elec. 956, saws pi; caroxvfoses, the emendation made on Dawes’s prin- ciple, supported by a few late MSS. and now universally accepted, can be justified ouly on the ground that és ci 8\érw in the leading sentence implies an exhorta- tion or entreaty, so that the clause with Srws may be considered a sort of object clause. (Sce Plat. Rep. VILL. p. 549 Ei: dcaxeAcboyrar dxws ripwphoerat mdvras rovs rotovrovs ) Even then the subjunctive would merely belong to the same class with the doubtful instance above quoted from Demosth. Phil. HI. If, however, it were pure final clause, the future could be justified only by the highest manuscript authority, like any other rare exception.
In Plat. Gorg. p. 513 A, érws pi reccépe0a has generally been interpreted as a final clause, that we may not suffer. It has seemed to me, on the ground above stated, that this must be an object clause depending on dpa; and I am glad to be confirmed by the high authority of Dr. Thompson, who, in his admirable com- mentary.on the Goryias, translates the passage, see that we do nol suffer, &c. I cannot help venturing to dissent, however, from the same learned scholar, when (in his note on p. 510 A) he approves the common emendation déixfcopey as “indispensable” in the passage: xai éxt rotro dpa, ws foie, mapacxevacréioy éori divapty cai rixvny, Ores pi ddixjowpev. Here the clause with érws is not the object Of rapackevacréoy, OF in apposition with a rotro which is the object of that word ; but it explains é#? rodro, fo this end, and is therefore a jinal clause requiring tho subjunctive. oe
On the Forms allowed after "Orwe and Ot pH. 49
one was bound on theory to say érwe mu) deddbecc rather than drwe py Cdaénc. In fact, we may perhaps doubt whether the more common form of prohibition, p) ééaéyc, did not make orwe pr Cdaéne & More natural form here than the other: cer- tainly the analogy will help to justify this form here, when it is supported, as it is in many familiar examples, by the best manuscript authority.
In regard to ob uj, the doctrine of Dawes as modified by Elmsley’s theory (explained below) has long prevailed; and even scholars who reject both Dawes’s Canon and Elmsley’s theory, still follow the former in their practice. As to Dawes’s rule, what has been said of érwe in object clauses holds true of ov py, that two constructions (the future indicative and the sub- junctive) being allowed by usage, the form of the aorist subjunc- tive which was not liable to be mistaken for a future would naturally be chosen, when that tense was to be used at all. But by Elmsley’s theory, two entirely distinct constructions with ov ph are to be recognized; one in ordinary negative asser- tions, where all agree that od yu} with either the subjunctive or the future indicative (but commonly the subjunctive) is equiv- alent to a strong future (ob pa) yévnrae or ob pi) yernoerar = ov yevhoerac); and another confined to the second person in pro- hibitions, whese only the future indicative is to be tolerated, and where, for example, od 1} rovro rojoec; 18 a question, meaning will you NOT NoT do this ?—1.e. do not do this. If this view is accepted, of course it seems very natural that there should have been a distinction in the form allowed in such entirely distinct constructions ; and accordingly the rule is stated absolutely, that only the future can be used in the interrogative construction and that no case of the subjunctive iz to be found in it., Surely enough, if we examine the Greek poets in modern editions, we shall find no example of the sub- junctive, but invariably the future indicative, in the prohibi- tive construction with the second person; although many editors now abandon the interrogative theory, and apparently explain both constructions on the same general principle. A very different result, however, will be obtained if we refer to older editions, or to the manuscripts themselves, as will be
a
50 W. W. Goodwin,
shown below. In the other construction (the mere negative assertion) both moods are found in all our editions, except that the first aorist subjunctive in the proscribed forms is gen- erally changed to the future indicative.
The theory of Elmsley is a mere theory; and it is fast be- coming obsolete, except so far as its influence still keeps up a distinction between the two constructions with od py. Its chief support is found in a few sentences in which a command in the form of a question is followed by a prohibition with 4, as in Soph. Oed. Tyr. 637 : otx ef od 7 otwovc, ob re Kpéwv kara oréyac, kal ji) TO pndév Gdyog eic pey oicere. If this is to be explained as entirely interrogative, itseems more natural to consider the first part a common question with ob, — Will you not go ?—and the last part a question with «7, implying a negative answer, — You will not raise, will you? As it is more than probable, therefore, that there is no combination of od and p»# at all in these expressions, they are entitled to little weight in the dis- cussion; and there are probably few scholars who would be willing to accept Elmsley’s theory for the other clauses with- out this support. Apart from the exceeding awkwardness of the expression, Will you not not do this? for Do not do this, Elmsley’s theory is open to the fatal objection that it attempts to explain expressions so obviously similar as ob i) rovro roujoetc and ob ju) rovro wowjoee On entirely different and opposing prin- ciples. If the former means Will you not not do this? it is incredible that the latter should mean He will not do this, or even There is no fear that he will do this. The theory cannot stand for a moment, if it is shown that the subjunctive is as authentic a form in the prohibitive construction as the future, so that no distinction in form between the two constructions can be maintained. .
The analogy of ordinary prohibitions gives a strong pre- sumption in favor of the aorist subjunctive in the prohibitive construction with ob uj. What was said above of the ellipti- cal prohibitions introduced by érwe 4 applies with still greater force here. The familiarity of such expressions as ji} rovro mona n¢ or pa} Anpneans would seem to make ov py TOUTO TOLHONC or ov ph Anphone a perfectly natural form for expressing the same
On the Forms allowed after "Onwe and Ov ph. 51
idea with greater emphasis. But the question must be tested by an appeal to the only real evidence that we have as to the usage of the Attic writers. It required no new collation of the manuscripts to establish the fact that there is good author- ity in them for expressions like ot j2) Anofenc. But the answer is always ready, that « and 7 are so constantly confused in Greek manuscripts that no argument can be based on the authority of the copyists. I have therefore examined the most important passages of the Clouds and Frogs of Aristo- phanes in which ov #7 occurs, in the two Venetian MSS. of the Library of St. Mark (Nos. 474 and 475),* and in the seven Paris MSS. which contain both comedies (in- cluding No. 2712), and also those of the Clouds in three other Paris MSS. At the same time, in order to ascertain whether the copyists of these MSS. were careless in writing e. and 9, | examined several passages in the same MSS. con- taining futures in -e¢ or subjunctives in -yc, which (on metrical or grammatical grounds) could not be changed to the other forms. Passages of this last class are Nub. 1852, ravrwe d¢ Touro cpaserc (where no MS. has épdonc) 5 1035, ei oprAjoecc (where one MS. has -nc); 1479, pndé p’ exerpidyc (all the MSS.). So we find «xarepeic and oicec, Ran. 298, 524, without variation ; although in 298 most, if not all, MSS. have ob yy Kade Cwhich is evidently intended for a present subjunctive, and might per- haps be allowed on the authority of ob yj dévnrat, ob pu} olde ns &¢.). *AxodovOjeec in Nub. 505 is instructive here ; for al- though one Venetian and six Paris MSS. have Aadjenc (or -nc), only three very poor Paris MSS. have dxoAovbjene Cone of these has -onc by correction from -cec). This shows that there was something in the future in the second case which prevented copyists from writing it with final -yc or -nc, even after they had just written AaAjonc OF AaAjonc (see below). I give these
* The MS. designated Ven. 475 is the elegant folio said to have been copied from the older Venetian MS. 474 in the fifteenth century. As it often differs from the older MS., even in the few lines here quoted, I have thought its readings worth noting. The readings of a Jate MS. are valuable as far as they show the care of the copyist in copying certain forms (as -cs and -ns) or indicate the readings of the older MS. which he copied. Among the MSS. here included, only Ven. 474 and Par. 2712 aro old enough to be of independent authority for the text.
o2 W. W. Goodwin,
results merely to show that, whatever may be the authority on other grounds against od pf and the aorist subjunctive in prohibitions, and however we may choose to despise the copy- ists who have transmitted this form to us, it cannot be said that these particular copyists of the MSS. of Aristophanes were so careless of the distinction between -y¢ and -ee as to write one for the other indiscriminately.
I give the readings of the above mentioned MSS. in Nub. 296, 867, 505, 824, and Ran. 524 retaining the peculiarities of the copyists in writing o% p4 and pyéé in the first and third passages :
I. Nub. 296, 297. O8 pa oxwwne: pncé rotjone &rep of rpvyo- daipoveg ovror* GAN’ ebphper. Ven. 474, Par. 2712. — Ob ph cxawne’ ovse wotjonc, &e. Ven. 475. — Ov py oxwnc pn ce momonc, &C. Par. 2821, 2824, 2902. — Od pu} oxinc pp SE rothanc, &e. Par. 2822.— Od put oxarnc* py 8 wochonc, &c. Par. 2598, 2716, 2820. —Obd> ph) cxwnc* poly C6 nochanc, &c. Par. 28238 (with schol. épa iva &Spionc).— Ob" py oxdnbyc. pu) S€ mothonc, &e. Par. 2717T.—Par. 2716 has gdvapfhone Over oxide.
II. Nub. 867. O8- pa} Anpfonc. Ven. 474, Par. 2828 (-nc). 2824. — O08 pi} Anphonc. Ven. 475, Par. 2712. — All others have ov pr) (ov pu), OF ob pu) Anphonc, Except Par. 2821, which has ov Mm) Anpjoec (With gp’ iva OVer ov), and Par. 2822, which has od p} Anpiic.
III. Nub. 505. Ob: pa) AadAjoeec* GAN’ axodrovdAaere Epoi. Ven. 474. Same, with ov uj, Par. 2716, 2824. O8 pu) AaAhonc’ GAN’ axodovOhoerc éuoi. Ven. 475, Par. 2712, 2821, 2902. 0% pan AaAnonc’® adr’ akorovOnonc Epoi. Par. 2717, 28238; also, Par. 2598 has -en¢ by correction from -cecc. Par. 2820 has od pu) NadF- gotc’ aAdX’ akorovBjcore Epo, (with -cov written over -core in both verbs.) Par. 2822 has ov (corr. to ov) pj, AadAHoetc. adN’ axodovOhaec époi.
IV. In Nub. 824, all manuscripts have drwe dé rovro pi) dddEnc pnoéva, except that Par. 2823 has diddinc. Par. 2716 omits the verse, and has in the margin épa éru¢ éé rov pn deiEne pncéva.
V. Ran. 524, 525. o8 pi) @Avagijone Exwv . . . GAN apdpevog oisere TaXLY Ta orpwuara. Ven. 474, 475; Par. 2712, 2716, 2821, 2820 (with -cov over -cerc). Par. 2717, 2822, 2824 have grvaphoecc and oicerc.
On the Forms allowed after "O7w and Ob 11}. i)
According to the best authorities accessible in this country, the Ravenna MS. has the aorist subjunctive in. Nub. 296° Cin both verbs), 367, and 824; in 505 it has Auadyene cor- rected to -cec, and axodrovtycec. I can find no trustworthy statement of its reading in Ran. 524.
It will be seen that, although there is excellent manuscript authority for the future indicative in some examples, especially those in which this tense follows in a second clause (as in Nub. 505), there is also the very best authority in other cases for the subjunctive. The strongest case, and indeed the case which makes the real nature of these examples clearer than any other, is that of Nub. 296, 297. Here we have the unani- mous authority of the MSS. for exiuyc and zoujenc. Now there is no future oxé~w known in classic Greek; and the change to cxope, which was made by Elmsley and is generally adopted in modern editions, is certainly a bold emendation to make in support of a mere theory, and a still bolder one to maintain after that theory is abandoned. This passage alone is sufficient to show that Elmsley’s theory is unsound, and it refuses to conform to the required punctuation. In Nub. 505, for example, the interrogation-mark (on Emsley’s theory) would follow dxodovOice éuoi. But where can it be placed here? Surely not after ddd’ cigijpe: and if it is to precede this, why should it not precede add’ dxodovOjcee in v. 505? The use of the imperative here in the second clause shows plainly that the preceding words are a mere prohibition and not interrogative. And if such expressions are mere prohibi- tions, it is much simpler and more satisfactory to explain this and the more common construction of ov py on the same prin- ciple. I cannot pretend to account for the doubling of the negative in either construction of od w), any more than in that of ,) ob with the infinitive and participle. The common explanation, that there is an cllipsis of a word like ¢¢jo0c¢ or éevdéy, i8 unsatisfactory, especially in explaining the future indicative with ob py. The future is a well established con-
* While this paper has been printing, I have learned from a distinguished Eng- lish scholar, who has collated the Ravenna MS. of Aristophanes, that this MS. reads ov pi) cxcynts in Nub. 296.
54 W. W. Goodwin,
struction in the first and third persons with ot pf, and yet it is a very rare form with pf after expressions of fear or danger. Whatever may have been its origin, 1 cannot doubt that in an expression like otéérore ovddéy ypiv ob py yévnrat ray deovrwy (Dem. Phil. I. p. 53, 4) the speaker merely used the strong- est negative which the language afforded for what would com- monly be expressed by otdey rev Sedyrwv hiv yervnoerac. Tosup- ply a verb of fearing here would be clearly absurd. The second person is often used in this way, as in aAX’ ob mor’ é& épov ye ph waOne réce (Soph. Elect. 1029), where the position of er’ shows that od belongs to ra8yc quite as much as pi does. In these cases the subjunctive seems to be a relic of the common Homeric subjunctive, which is nearly or quite equivalent to a simple future, as in od rofouc idov obdé tdwpar. The use of the future indicative in prohibitions is amply justified by its use in imprecations, 08 dmoAciobe, oipwtecbe, and still more by examples like rdvrwe é€ rovro épdaecc CNub. 1852), but by all means do this, where the future has nearly the force of an imperative. The same use of the future is seen in dxodov@jcetc (N ub, 505) and oisee (Ran. 525), quoted above, which ex- press mere command. Ov yp») edpn rovro, therefore, means simply he will not find thts; ob yu) etpnc Means you will not find ; SO ob py Tothoee TobTo OF ov pi) Toujon¢g rovro Means you wil not (or shall not) do this, which may be a prohibition equivalent to do not do this.
In Soph. Phil. 881 we find cat ratra . . . . ob py mor’ Eg rv Exipoy éxmAedonc éxwy. Here we could not read éxrdeicecc, as there is no such future; nor even éxrdetce, a8 the metre. for- bids. The meaning clearly is you shall not sail away, &c., and yet the clause cannot be made interrogative. Is not this passage a striking illustration of the inseparable unity of the so called “‘two constructions”? with ob yA, and does it not, with the aid of the imperative in Aristoph. Nub. 297, clearly show that both subjunctive and future are simply emphatic futures in meaning, which may in the second person — like all futures — have the force of commands ?
In conclusion, I would maintain as the result of this inves- tigation, that ot pi can take cither the subjunctive (commonly
On the Forms allowed after “Orwe and Ov pn. Do
the aorist) or the future indicative in every person, and that the common exclusion of the subjunctive of the second per- son in prohibitions is not authorized by the Greek manu- scripts, or by any known principle of the language. As to érwe, the construction always determines whether it regularly takes the future indicative or the subjunctive; and whatever view we may take of exceptional suljunctives in object clauses, no arbitrary rule can ever justify a future indicative in a pure final clause, in Attic Greek, against manuscript authority.
IV.— On the best Method of Studying the North American Languages.
BY J. HAMMOND TRUMBULL, OF HARTFORD, CONN.
THE collection of materials for the study of -American aboriginal languages is already large. Indian vocabularies, grammars and grammatical notices may be reckoned by hundreds, and every year adds to their number. Among these are to be found many works of permanent value, indispensa- ble to students of the languages of which they treat, a few of distinguished excellence, widely known and highly appreciated as contributions to comparative philology, and many others which, without imparting thorough or exact grammatical or lexical knowledge, have been very useful to explorers, mis- sionaries and others, by facilitating communications with savage tribes. But if we look carefully through the entire collection (excluding, however, from present consideration all additions which have been made to it in the last ten years) we shall be obliged to confess—not without some mortifica- tion to the pride of American scholarship—that a great part of it is absolutely worthless to critical students of language, and what is worse, that the real value of original materials has in many instances been lost or much depreciated by the method of their exhibition. We shall find many rash gener-
ov J. H. Trumbull,
alizations from insufficient data, much ingenious speculation from questionable facts and too frequent resort to the com- parative method by writers who had, apparently, very little knowledge of either the vocabulary or the grammatical struc- ture of the languages compared. The materials for the study of a new language or dialect have too often been collected and arranged with constant reference to its supposed likeness to some other languages previously known. This last-mentioned source of error has perhaps been more mischievous than any other. <A great deal of ingenuity has been wasted in tracing analogies and resemblances between Indian languages and the Hebrew, Greek, and Latin. Even John Eliot, to whose rare linguistic genius his version of the Bible in the Massa- chusetts dialect testifies, was not exempt from this tendency to try the new by the old. He would have succeeded better in his “Indian Grammar Begun” if he could have forgotten all the Greek and Latin he knew. He was hampered by his ‘accidence,’ and sometimes overlooked the true characteristics of the Massachusetts language while he was striving to cast it in a classic mould,— scarching for gerunds and supines which were not to be found and puzzling himself about reduplication, augment, and the optative mood.
One writer* after having, as he assured his readers, ‘ pur- sued the comparative examination of the Haytian language and dialects, upon all the languages of the earth” (the italics are his own), pronounces the Haytians, “of Pelagic origin,” but traces the northern Algonkins to central Asia. Another presents a list of words to exhibit the likeness of the Aficmac to classic Greek. A third searches twenty American lan- guages for Egyptian roots, and finds them in abundance. A fourth goes straight to the origin of language, and gleans from the vocabularies of different tribes those words only which he conjectures to have been formed by imitation of natural sounds,—a conjecture which better acquaintance with the structure of these languages would have convinced him was unfounded.
A time may come when the relation of the languages of
#C. S. Rafinesque, in The American Nations, vol. i. pp. 217, 219.
On the Method of Studying the American Languages. 57
America to those of Asia and Europe can be intelligently dis- cussed —if not established: but a great deal of work must be accomplished first in a field in which American scholars have as yet hardly broken ground. Why waste time groping in the dark+after genetic correspondences with Asiatic and «African languages when we have not advanced far enough in our knowledge of the American to discover their relations one toanother? Take, for example, two groups for the study of which our materials are most copious and on the whole most trustworthy,—the Algonkin and the Iroquois: who has demonstrated or is prepared to demonstrate the genetic con- nection of the Mohawk with the Massachusetts, Abnaki or Chippeway ?
The science of language is yet in its infancy. That it has made little progress in dealing with the problems which the almost innumerable languages of America present, need not surprise or discourage us. But true progress ought not to be impeded, at its starting point, by obstacles placed in its path, nor should time which might have been devoted to the inves- tigation of new truths be so heavily taxed for the elimination of old errors.
While a few scholars have labored unprofitably to extract Semitic and Turanian rvots from American words whose structure they had taken no pains to analyze, another and a larger class of collectors and editors have impaired the value of their materials by endeavoring to fashion them to an Eng- lish pattern. I allude now to the very general adoption of some standard English vocabulary as the framework of collec- tions and for the exhibition of results. An erroneous impres- sion appears to have been very generally received, that real progress in the knowledge of a language has been made when one or two hundred words taken from its vocabulary have been set over against certain words of our own language which have meanings not very dissimilar.
Forty or fifty years ago, when Mr. Gallatin undertook his great work of classifying the North American languages, the advantages to be secured by the adaption of a standard vocab- ulary were obvious. Twenty years afterwards, there was stil]
8
58 J. H. Trumbull,
good reason for employing the same vocabulary (with some unimportant changes introduced by Mr. Hale) in the arrange- ment of collections made by the U.S. Exploring Expedition on the North-West Coast. The value of Mr. Gallatin’s “ Synop- sis of the Indian Tribes of North America,” with a ‘“‘ Compar- ative Vocabulary of fifty-three Nations” (in the Transactions of the American Antiquarian Society, vol. ii., 1886), supple- mented by his Introduction to Hale’s “ Indians of North West America” (in Transactions of the American Ethnological So- ciety, vol. ii., 1848), can hardly be over-estimated. These works opened a way to the intelligent study and discussion of what had previously been a chaotic mass of materials. All that Mr. Gallatin proposed to do he accomplished admira- bly, considering the disadvantages under which he labored. His method was well adapted to the end he had in view,— to determine the more obvious groupings of American lan guages and dialects. Its value is not yet exhausted. The standard vocabulary continues to be useful to inexperienced collectors and as a guide in provisional classification. Next to the satisfaction of learning a new language is that of learning something about it—of ascertaining by means of a comparative vocabulary that it is or is not like some other language which we know, at least by name, and that the two belong or do not belong to the same ‘stock,’ ‘family,’ ‘class’ or ‘group,’ — terms which are used with very uncertain ap- prehension of their meaning, when applied to North American tongues.
Duly recognizing the past and present usefulness of these vocabularies as stepping-stones to knowledge, we must at the same time be careful not to estimate their value too highly, — remembering that the real work of the linguistic scholar be- gins where the provisional labors of the word-collector end. Such lists of words give no insight to grammatical structure, contribute little or nothing to analysis, and even with respect to the relationships of languages they enable us to determine only the nearest and most obvious. Professor Whitney (Language and the Study of Language, pp. 246, 247) has shown us “ upon how narrow and imperfect a basis those com-
On the Method of Studying the American Languages. 59
parative philologists build who are content with a facile setting side by side of words; whose materials are simple vocabula- ries, longer or shorter, of terms representing common ideas,” and that “surface collation without genetic analysis, as far- reaching as the attainable evidence allows, is but a travesty of the methods of comparative philology.”
Now while this is true universally, it has special force when applied to the study of languages of the polysynthetic type. The American languages differ from the Indo-European both in grammatical structure and in their plan of thought. To inquire in what, precisely, these differences consist would lead me too far from my present purpose, but the inquiry may at least be so far prosecuted as to show how little value attaches to any list of Indian words set down as the equiva- lents of or as exactly translated by the English, French, or Spanish words of a previously-constructed vocabulary,—to show, in short, how nearly impossible is the translation of any Indian name or verb by an English name or verb, and the converse.
In the English language the analytical tendency has attained its highest results. By employing independent words to ex- press grammatical relations, it has reduced a gre1t part of its vocabulary to monosyllables. The very essence or the Indian languages on the contrary is synthesis, and their capacity for synthetical development is apparently unlimited. Their high- est aim is to express in a single word, “‘ not only all that modi- fies or relates to the same object, or action, but both the action and the object; thus concentrating in a single expression a complex idea, or several ideas among which there is a natural connection.’”’* There is hardly any modification of which the action of a verb is susceptible which may not be effected by means of inseparable particles having the character of adverbs: “thus, the action may be intended, or be about to be done; it may be done well, better, ill, in a different manner, quickly, attentively, jointly, probably, rarely, repeatedly, habitually :”’+ it may be affirmed, doubted, questioned, denied, prohibited.
* Gallatin, in Trans. Am. Antiquarian Society, vol. ii. p. 165. + Gallatin, in Trans. Am. Ethnological Society, vol. ii. p. cxlii. -
60 J. H. Trumbull,
A single example will illustrate this, and I select one which Mr. Bancroft (History of the United States, vol. iii. p. 259) has used for a similar purpose, in his observations on ‘the synthetic character of the American languages.”
‘The Indian never kneels; so, when Eliot translated kneeling [Mark i. 40], the word which he was compelled te form fills a line, and numbers eleven syllables.”’
As an instance of extreme synthesis this word —wut-ap- pe'-sit-tuk-qus'-sun-no-weht-unk'-quoh* -— is well taken, but its significance is by no means limited, as Mr. Bancroft supposed it to be, by that of the English participle ‘kneeling.’ In the verse cited it stands as the translation of the words “ kneeling - down to him” of the English text or, more exactly, for “he kneeled down to him’’— Eliot having substituted the indica- tive mood for the participle, as Indian syntax requires. We have thus five English words represented hy the Indian syn- thesis. But the denotation of the latter is not yet exhausted. Eliot might have found, in the Massachusetts or any other Algonkin dialect, an equivalent for the verb ‘to kneel’, in its literal and primary signification—‘to rest on the bended knees’ or (active-intransitive) ‘toassume the position of kneeling.’ In 2 Chron. vi. 13, Daniel vi. 10, Acts xx. 36, he translated “he kneeled down”’ by ap-pe’-sit-tuk-qus’-sin; but in the verse first cited, something more than the mere act of bending the knees or resting on them is implied. The verb here connotes supplication, submission, and worship, and all this is expressed in the eighth and ninth syllables (-nw-weht-) of the Indian synthesis, the whole of which may be translated, literally: “he, falling down upon his knees, worshipped [or, made supplication to} him.’ Thus the one Indian word of eleven syllables requires for its accurate interpretation erght or ten English words and at least eleven syllables.
* Duponceau pointed out this word as the loogest he had met with in any Indian language, except the Chippeway (of Schoolcraft) in which ‘there were some verbal forms of thirteen and fourteen syllables.” (Afémotre sur le Systéme Gram- matical etc., p. 143). A more remarkable illustration of “the Indian way of com- pounding words” was given by the Rev. Experience Mayhew, preacher to the Indians on Martha’s Vineyard,—in a synthesis of twenty-two syllables, signify- ing, “ our well-skilled looking-glass makers” — Nup-pahk-nuh-té-pe-pe-nau-wut-chut- chuh-qué-ka-neh-cha-e-nin-nu-mun-né-nok ! (MS, Letter, 1722.)
On the Method of Studying the American Languages. 61
This tendency to synthesis is not manifested only in the grammatical structure. It may be traced far back to the roots of the language, and characterizes the primary verbs as truly as it does the many-syllabled cluster-words of later growth. Father Le Jeune, a Jesuit missionary in Canada in 1634, men- tions as a peculiarity of the language of the Montagnars, ‘the infinite number of words which signify many things together” and which yet had no etymological affinity with any of the words which signify those things severally: and he gave as an example the Montagnais verb piovan, meaning “the wind drives the snow,” but in which no trace appears of the words for ‘wind,’ ‘snow,’ or ‘to drive.”* This synthesis which pre- cedes grammar and concentrates complex ideas—thought- clusters—in a single word or syllable, is found in all the American languages of which we have any knowledge. The primary verb affirms conditioned or modified existence, spe- cific and restricted action. There is—I speak now only of that group of languages to which my studies have been chiefly directed, the Algonkin,—there is no independent substan- tive verb; but there are verbs of being under every conceiva- ble condition of time, place, and circumstance. “He is” can- not be exactly translated by any Algonkin verb, but every dialect has verbs signifying ‘he is well,—or ill, ‘he lives,’ ‘he was (and is not),’ he was (and cuntinues to be),’ ‘he has himself,’ ‘he abides,’ ‘he remains,’ ‘he is the same as,’ ‘he is of the kind of,’ ‘ il y a’, etc.
Every standard vocabulary includes the verb ‘to eat,’ yet this verb has not, so far as I can discover, its equivalent in any American language. The Algonkin has: four or five pri- mary and a great many composite verbs of eating, but none of these expresses the simple act of taking food, without reference to the manner, mode, subject or object. One verb, for exam- ple, signifies ‘to eat animal food’ (or that which has or has had life); another, ‘to eat vegetable food;’ another, ‘to eat soft food’ (that which may be dipped up, spoon-victuals, such as samp, succotash, and the like); others, ‘to eat ravenously, to devour like beasts of prey,’ —‘ to graze,’ or take food from the
aoe ee Oe re we eae eee CS eee
* Relation de la Nouvelle France en l’année 1634 (repr. Quebec, 1858, p. 50. )
V4
62 J. HA. Trumbull,
ground as cattle do,—and so on. Others again, by the inser tion of a particle or by receiving a characteristic affix are made to express the act of eating in company with others, of eating enough or satisfying one’s self with food, of eating ad/ that is provided, of feasting, etc.
No Indian language, probably, has any verb which exactly corresponds to the English verb ‘to go’, yet the Indian verbs of motion are almost numberless. There are verbs of going by land,—by water,—by paddle,—by sail; of going from the speaker, —from the place of the action narrated, — and from a place other than that of the speaker or the action ; of going to a person, — place, — inanimate object; of going by running, —jumping,— flying, swimming, etc., (and these are not to be confounded with the verbs which express the acts of running, jumping, flying and swimming); of going fast, slow, before, after, aslant, in a straight course, by a devious path; and scores of others. A special vocabulary of the verbs of motion in any Indian language, giving an analysis of each and its precise signification, would be of some real value to philologists. But what is to be gained by entering against the English infinitive ‘to go,’ in a standard vocabulary, some one or another of these Indian verbs of going,—the entry carry- ing its own evidence of inaccuracy ?
The defects of the vocabulary method are still more obvious when we consider the nature of Indian names. A peculiar strength of the English language lies in its concrete general names, and in the facility with which these names are made - to pass from the concrete to the abstract. The peculiar ex- cellence of the Indian languages is in the nice machinery by which definitions or descriptions of individual objects are made to stand for names, and by means of which, names which in English are general or abstract become individual or concrete. The English abounds with predicates of a class or genus: but the Indian noun— verbum nominale — itself predicates a dtffer- entia or au accidens, occasionally a genus or a species. I say, the Indian noun predicates,—for I can find no less objectiona- ble form of expression, though this conveys only half the truth. Strictly regarded the Indian noun is not separable, as a part of
On the Method of Studying the American Lanyuages. 68
speech, from the verb. Every name is not merely descriptive but predicative,—not as in Indo-European languages by im- plication or suggestion, or by reason of remote derivation from a predicative root, but it retains the verb-form unchanged ; is varied by conjugation, not by declension ; has tenses, not cascs ; may become active, passive, reciprocal, frequentative, like other verbs. In short, every Indian name is in fact a verb, — is formed as a parlicipial immediately from a verb, — or con- lains within itself a verb.
Without pursuing this branch of the subject further at pres- ent or multiplying examples, I repeat, that in view of the fun- damental differences in grammatical structure and in plan of thought between the American and the Indo-European lan- guages, it is nearly impossible to find an Indian name or verb which admits of exact translation by an English name or verb. But the standard vocabularies which have been most largely used in the collection and exhibition of materials are framed on the hypothesis that such translation 7s generally possible. They assume that equivalents of English generic names may be found among Indian specific and individual names, — that English analysis may be adequately represented, word for word, by Indian synthesis. Such vocabularies, as has been remarked, have their uses, but to linguistic science or to com- parative philology they contribute nothing which is worth the cost of obtaining. When a collector or an editor has acquired a thorough knowledge of the grammatical structure of a lan- guage and has learned how to resolve synthesis by analysis, he may undertake the arrangement of his materials in the form of a vocabulary with some probability of imparting to the re- sultreal and permanent value. Without such preparation for his work—no matter how cautiously or with what ability he prosecutes it—he must not hope for great success.
It is easier to discover the defects of the old method than to point out a new and a better one. The details of such a method could not be discussed without excceding the limits of this paper. Nor is such discussion called for. The way to a more thorough and exact knowledge of the Indian languages
64 J. H. Trumbull,
is not unknown or untried. There are laborers already in the field who have not only proved that higher results than tlie com- pilation of brief vocabularies are attainable, but have shown how to attain them: and for the study of a considerable num- ber of languages and dialects of the north, the south, the val- leys of the Mississippi and Missouri and the far west, scholars are no longer restricted in matcrials to quasi-translations of lists of untranslatable English words.
The suggestions I shall offer have to some extent been anticipated by the drift of the foregoing remarks. The first is —
That a constant aim of the student of any of the American languages should be the resolution of synthesis by analysis. What the Indian has so skillfully put together—‘ agglutina- ted’ or ‘incorporated’ — must be carefully taken to pieces, and the materials of the structure be examined separately. Every Indian cluster-word is a sentence, —a description, definition, or affirmation. Mere translation will not exhibit its construc- tion or afford a trustworthy basis of comparison with word- groups in other languages. Something is gained, it is true, by ezact translation, but this cannot be had if the translation must he shaped to the requirements of an English vocabu- lary. A single chapter of the Bible or a dozen scutences of familiar conversation accurately translated into any Indian language or a few selected words and phrases translated from it to English will give a better insight to its structure and do more to determine its relationship to other American lan- guages than long lists of concrete names or verb-forms com- piled on the usual plan. But something more than transla- tion, however accurate, is wanted. These languages must be studied in their roots, — for these are the elements of synthe- sis. The possible forms of synthesis are infinite, but the radi- cals or primaries are, in any language, few. The forms, both inflectional and syntactic, are subject to change from year to year and in passing from tribe to tribe; and these clianges, it is said, have in some instances been surprisingly rapid and ex- tensive. We are told ofa vocabulary compiled by missionaries toa Central-American tribe in 1823 which had become useless
On the Method of Studying the American Languages. 65
in 1833, so greatly had the language changed in the ten years which intervened.* With better knowledge of the structure of these languages such changes would probably have been found to be for the most part only superficial, — the synthesis being differently constructed while its elements, the predicative and demonstrative roots, remained the same. Of such changes some further notice will be taken, in another part of this paper.
To single out and fix the primary meanings of the verbal rools should be the ultimate aim in the study of every Indian language. What excessive synthesis has done, searching analysis must undo. The task is not so difficult as at first sight it may scem to be. As I have before remarked, the roots or primaries are few and constant, or nearly so, in all
*S. F. Waldeck, Lettre A M. Jomard des environs de Palenqué,’—cited by Max Miiller, Lecures on the Science of Language, Ist Series, p. 62 (Am. ed.}. I confess that, without other explanation than appears, find this statement hardly credible, and suspect that the worthles-ness of the vocabulary should not have been attributed solely to the inconstancy of the language. Professor Miiller (1. c.) refers also to Sagard’s Grand Voyage du Pays des Hurons (Patis, 1632), for tho statement “that among these North American tribes hardly one village speaks the same language es another; nay, that two families of the same village do not apeak exactly the same lu guage. And he adds what is important, that their lan- gege is changing every day, and is already so much changed that the ancient Huron language i+ almost entirely different from the present.” But Sagard’s statement must not be received without the qualification he himself gave it. He lid not intimate that the differences of dialect were greater or the tendency to change more apparent in the Huron language than in the French. What ho says
in the introduction to the Dictionnaire de la langue Huronne printed with his Grand Voyage —is in substance this: that there was the same diversity of ac- cent, pronunciation and in the use of words, in provinces, towns and villages in the Huron country os in France; that tho sane words might be differently pronounced or the same object called by different names even by inmates of the same cabin; one person would say ‘ etseignon, and another ‘ etcheignon,’ one ‘ochahenna,’ another tochahenda,’ &e.; and that, as in France (comme par deg) new words were in- vented or brought in fashion and the pronunciation of the court had almost super- seded (presque ensevely) the ancient Gallic, so “our Hurons and, generally, all other nations, have the same instability of language, and change their words 60 that in process of time the old Huron becomes almost entirely different from the modern.” ‘The chungo, as he conjectured, was still going on;—and yet, Sa- gard’s very iraperfect dictionary of this unstablo language, two hundred years or more after it was compiled, cnabled Dupouccau to make himself understood with- out apparent difficulty by the Wyandots, a remnant of the lost nation of the Horons. (Duponceau's Mémoire, p. 110.)
66 J. A. Trumbull.
dialects and languages of the same family, — allowance being made for recognized differences of pronunciation and accent. They preserve their independent signification, however com- bined. They enter into composition without undergoing change of form, while their affixes and formatives obey laws of harmonious sequence of vowels as nicely adjusted as in Turkish. The five, ten, or more syllables of a verbal-synthesis do not grow out of or coalesce with one another, but each is built on: so that when the key is once found the word-puzzle may be taken in pieces as easily as it was put together. In- deed, it is a requirement of the Indian languages that every word shall be so framed as to admit of immediate resolution to its significant elements by the hearer. It must be thoroughly self-defining, for (as Max Miiller has expressed it) “it requires tradition, society, and literature, to maintain words which can no longer be analyzed at once.” ..... In the ever-shifting state of a nomadic society no debased coin can be tolerated in language, no obscure legend accepted on trust. The metal must be pure and the legend distinct.* The more cum- brous and unwieldy the structure, the greater is the necessity for exact adjustment of its parts: and the laws of verbal com- position are well-established, admitting no exceptions.
How far such an analysis as I have suggested can be suc” cessfully carried need not now be inquired. Every step taken in that direction will be something gained, will lead to more exact knowledge and to positive results. To determine and classify the primary verbs in any one language would be to bring a larger contribution to linguistic science than has often been made by students of the American tongues. Back of these verbs and of the primary demonstratives are the ulti- mate roots. These may not now be— possibly, they never will be attainable: yet I do not hesitate to express my belief that through the study of the American languages scholars may as nearly arrive at a solution of the great problem of the genesis of speech, in determining the character and office of its germs, as by any other avenue of approach. All attempts to establish relationship between the several great linguistic families by
* Lectures on the Science of Language, Ist Scries, pp. 292, 293.
On the Method of Studying the American Languages. 67
the identification of roots, may indeed be regarded as hope- less; for few will be disposed to question Professor Whitney’s conclusion (Language and the Study of Language, p. 392) that “the difficulties in the way of a fruitful comparison of roots are altogether overwhelming”: and probably no one is yet “so sanguine as to expect to discover, amid the blind confusion of the American languages, where there are scores of groups which scem to be totally diverse in constituent material, the radical elements which have lain at the basis of their common development.” But if order is ever to be brought out of this blind confusion, —if any satisfactory classi- fication of the hundreds of languages and dialects now so loosely grouped is to be established, — if the genetic relation of one of these to another is to be demonstrated even in those cases where, on grounds independent of language, the proba- bility of such relation is greatest, — analysis must first do its work, until, at least, it shall have determined and classified the earliest traceable constituents of speech, though compelled to stop short of the discovery of ultimate roots.
If the method I have indicated is the true one, the collection of materials for the critical study of an American language should begin, not with the translation into it of a given num- ber of English names, but by looking out its simplest i. e. least composite words, and fixing their meanings, — by detaching from the constant roots or themes terminations and formatives which are merely grammatical, —and by translating from the Indian to the English, provisionally and subject to correction by more rigid analysis, the syntheses which discharge the office of concrete names, by conveying concise definitions or specific descriptions of the objects to which they are severally appro- priated.
Among the words and elements of words which claim earli- est attention, may be mentioned —
1. The Pronouns, separable and inseparable, and pronom- inal suffixes: with which may be included the demonstratives.
2. Particles, which serve as prepositions and post-positions, conjunctions and, occasionally, adverbs. Nearly all of these appear to be remnants of verbs and for the most part are
68 J. H. Trumbull,
susceptible of conjugation as verbs. Their verbal origin may be matter of subsequent investigation, but a careful study of them in their present forms is essential, at the very outset, to thorough knowledge of a lanvuage: for they have much to do with the construction of syntheses and exert great influence in the modification of verbal roots.
38. The Numerals, cardinal, ordinal, and distributive. For the collection and analysis of tlicsc, some suggestions are given in “ Instructions for research relative to the Ethnology and Philology of America,” prepared for the Smithsonian In- stitution by Col. George Gibbs.* As the numerals are always significant, it should be a special aim of the collector to ascer- tain the precise meaning of each. Does the word used for une signify ‘a small thing,’ ‘a beginning,’ ‘ the little one’ (i. e. fin- ger), ‘ undivided,’ or ‘ that which is left behind, or passed by’ ? Does three mean ‘the middle finger? Is five ‘the hand,’ ‘the closed fist,’ or ‘all’ the fingers? Is siz ‘ five-one,’ ‘one more,’ or ‘one held up’ (i. e. one of the fingers which had been doub- led down)? Is nzne‘ one left,’ or ‘ one less than,’ or ‘one want- ing’? Is eleven ‘one again’ or ‘ten more one’? Is dwenty, as in the Eskimo, ‘one man’ (i. e. all the fingers and toes ) ? Every such question that is answered throws some light on the structure and method of synthesis and may help establish the relationship of the language.
4. Primary Verbs. Of these and of the tendency to the concentration of complex ideas in a single word, which is char- acteristic of the American languages, I have already spoken. Recollect that the Indian verb is almost always holuphrastic. It affirms—not action or existence generally, but— some special and limited act or conditioned existence: consequently, it can seldom, if ever, be adequately translated by an English verb without adverbial qualification.
5. Concrete Nouns. We have seen that these are not, as in the inflectional languages so many names have come to be, mere unmeaning marks. ‘They are descriptive and definitive ; specific, not general; and each retains the verb form or em- bodies a verb. Every synthesis is so framed as to differentiate
* Smithsonian Miscellaneous Collections, 160 (vol. vii. art. xi.).
On the Method of Studying the American Languages. 69
the object it serves to name from every other object known to ° the speaker, and this so explicitly as to be intelligible to every hearer. The English word horse tells us nothing about the animal it names. Etymologists who can establish its connec- tion with the Sanskrit Arésh may find a reason for its appro- priation to ‘ the neigher,’ but we use it without having a con- sciousness of any such intrinsic significance, recognizing it, only because we have been taught to do so, as the distinguish- ing mark which has been set upon a species, just as — regard- lees of etymological suggestions — we recognize * Charles’ or ‘William’ as the distinguishing mark of an individual. The American languages permit the use of no such names without meaning. The native of Massachusetts who saw a horse for the first time distinguished it from all animals he had _ previ- ously known, as “the beast that carrics on his back a living burden,” and this name once heard enabled every Indian of the tribe, or who understood the language, to identify the ani- mal whenever it came in his way. So the Chippeway could recognize by its name alone the creature “ whose hoofs are all solid,” and so the Dakota knew at sight the ‘“ wonderful do- mestic animal’’ introduced by the white man.
With this understanding of the nature of Indian names, we sec how tribes speaking dialects of the same language and not widely separated may come to have different names for the same object, —as many names, possibly, as there can be framed definitions or descriptions sufficiently exact for its differentia- tion. One Algonkin tribe calls the beaver a “‘ feller of trees” ; another describes him as “ putting his head out of the water,” 1. e. air-breathing water-animal. The Chippeways and some other tribes of the same family name the humming-bird by the cumbrous synthesis xo no no"k’ aus eé ; the Shyennes, a western off-shoot of the same Algonkin stock, call it ma kd i tat’ wi kis. The two names have no apparent affinity.’ Standing side by side in & comparative vocabulary, their testimony would go to show the unlikeness of the languages to which they respect- ively belong. Yet both names would, probably, be alike in- telligible to a Chippeway and a Shyenne. When we have learned that the one means “an exceedingly slight (or, deli-
70 | J. H. Trumbull,
‘ cate) little creature,”’ and the other, “the iron bird,”’ we shall be less likely to draw a wrong inference from their external non-resemblance.
Where such latitude is allowed in name-giving and where a name is necessarily discarded when the description it gives of an object is no longer sufficient to distinguish it from every other, we must not expect to find the same constancy in the vocabulary as in languages like our own, in which names hold their places not by virtue of their inherent significance but by prescription. And here we have the reason of some of the changes which have been remarked in the languages of certain tribes, of which something was said in another place (p. 65). Such changes are likely to be most considerable and most rapid soon after the opening of intercourse with a civilized race. The significance of old names is lost, in the changed condition of the tribe. One synthesis displaces another which has no longer any distinguishing force: one object after another is divested of the characteristic quality which had given it aname. When Kuropeans first came to New England, the Algonkin name of a pot or kettle (ausuk) described it as ‘made of earth’; but this name — still in use among the western Algonkins — could not long maintain its place in the language of Indians of the Atlantic coast after vessels of copper and iron were generally substituted for pots of clay or steatite. The introduction of fire-arms, — of dogs and horses, — of trading cloth and blank- - ets, — not only called for the invention of a dozen new names but made nearly as many old ones useless.
6. Characteristic particles found in composition with verbs, designating specific modifications of the action or special rela- tions of the action to the subject or object of the verb. These are prefixed, added as terminations, or inserted between the root and the inflection proper.
7. Generic formatives which, in grammatical synthesis, dis- charge the office of appellatives or general names.
These two classes— characteristic particles and generic formatives — present the most formidable obstacles which are to be encountered in acquiring thorough knowledge of any American language. One or the other or both have place in
On the Method of Studying the American Languages. 71
nearly every synthesis. Both must be climinated by analysis, before the primary signification of the verbs with which they are associated can be ascertained. Biliteral or uniliteral — syllables or mere fragments of syllables— they probably all represent, as many of them are known to do, independent words, some of which still maintain their places in the vocab- ulary while others have yielded to phonetic decay. The crit- ical investigation of these particles will compensate the studeut for all the pains it may cost him, for in it he will be brought very near the ultimate roots of the language.
To the former class — characteristic particles — belong all the grammatical machinery for energizing and individualizing the activity of the verb, making it intensive, frequentative, causative, possessive, reciprocal, dubitative, simulative, repre- sentative, ete.,— for designating the manner of acting or of being, and sometimes the instrument or agency by which the act is performed.
The nature and office of these characteristics may he shown by a few examples, from the Massachusctts-Algonkin, the Sioux-Dakota and the Choctaw:* but of their number and variety in any language no adequate conception can be had without study of the language itself.
In the Massachusetts (as written by Eliot), -whk or -ohk interposed between the root and the formative, denotes con- tinued and progressive action —‘to go on’ doing: pet-aii, ‘he puts (or, is put) into,’ petuuK-aii, ‘he goes into;’ assa-maii, ‘he gives food to,’ assavHK-amaii (contracted to sbuKamaii, El.), ‘he keeps on giving food to,’ continues to feed (e. g., a domestic animal); amd-eu, ‘he absents himself, departs’ ; amavuHk-au ‘he drives away’ (goes-after him-going) ; wék-eau “he houses himself, provides a dwelling place’, wékuuK-au ‘he builds or constructs a dwelling place,’ goes on housing him- self.
When the action is performed with the hand, the characteris- tic is -nn before the formative: kenunnum ‘he carries it in his
©The Massachusetts form taken from Eliot’s version of the Bible, the Dakota from the Rey. S, R. Riggs’s excellent Grammar and Dictionary of that Innguage (Washington, 1852), and the Choctaw from the Rev. C. Byington’s Choctaw Grammar (edited by Dr. D. G. Brinton, Philadelphia, 1870).
12 J. H. Trumbull,
han@ ; tohqunnum ‘he holds it fast with his hand’ [comp. Cree tadkwdaénum, ‘he holds it with his hand,’ (@kwdtum ‘he holds it in his mouth’]. If the action is performed by culting or with a knife, -ss takes the place of -un: sohqu-t ‘it is in small pieces’, ‘broken fine’; sokguNNum ‘he breaks or pulls it to pieces, with his hand,’ sohqussum ‘he cuts it in small pieces.’ The act of tying or making fast by a cord or thong is dendted by -z or -pin after the root; AvshPINNum ‘he ties it firmly, with his hand’ (the characteristic is double, here); assepInNum ‘he ties them together’ ; togkupinnau ‘he holds him fast by bonds,’ &c. Sudden, violent, or disastrous action is denoted by the insertion of -8h; petaii ‘he puts (or goes) into’. becomes © petsHaii ‘he falls into’ (e. g. a pit, or a snare); pohqui ‘it parts asunder,’ poksHeau ‘ it breaks, by violence, or suddenly’ ; tugkun, ‘it strikes,’ togkusnin ‘it strikes with violence,’ etc.
In the Dakota group, the instrumentive or modal character- istic is prefized to the verb: b6a- shows that the action is done by cutting or sawing ; bo- that it is done by shooting Clit. by blowing). or by some missile ; ya-that it is performed with the mouth; pa- that itis done by pushing, drawing, pressing, or rubbing with the hand c. g.: Baksa to cut off,’ BAmda ‘to cut in slices’, BApla ‘to cut off a piece’, BApako ‘to cut or saw crooked’ (from pako ‘ crooked’) ; bohdého ‘to loosen by shoot- ing’ (from hohkdé ‘loose’), BOi’yowaza ‘to make an echo by shooting’ (from yaiwowaza ‘to mako an cchio’) ; yAchocho ‘to chew fine’ (from chocho ‘ soft’); Yahdho ‘to make loose, with the mouth (from Aého ‘ loose’) ; Padopa ‘to push into the mud’ (from dépa ‘ to mire’), Pabu ‘ to make a noise with drumming with the fingers’ (from bu ‘to make a noisc’), Pahmfyan ‘to make round like a ball, with the hands’ (from hmt-yan' ‘round’), ete.
In the Choctaw, Mr. Byington (Grammar, p. 86) gives some of these forms for the verb takchi ‘to tie’: ta’kchi ‘to he tying’, fatyakchi ‘to tic firmly’, tana’kchi ‘to keep tying’, taukchi ‘ to tie instantly’ or suddenly, takcutchi ‘to cause to tie’, etc.
In some of the Algoukin languages there is a special form of the verb for denoting a pretence of doing or being, ‘ feigning
On the Method of Stadying the American Languages. 3
to do’. In the Gree, this form has the characteristic -ka's : from nipp'ow ‘he sleeps’ comes nippaxa'soo ‘he pretends to sleep’; muskowissu ‘he is strong’, muskowisseKa Soo ‘ he pre- tends to be strong’, etc. (Howse’s Cree Grammar, pp. 20, $4.) .
What I have called generic formatives have been regarded by some writers on the American languages, especially by Mr. Schoolcraft, as “ primitive nouns never disjunctively used.” All, however, which are found in the Algonkin languages, may be shown to belong to one of two classes: verbals and par- ticipials regularly formed from primary verbs some of which still retain their independent places in the language, —and inflections, with a characteristic particle prefixed to each. They may be described, generally, as terminations which de- note the class or kind to which the object designated by the synthesis belongs. Examples of these formatives may be ob- served in many geographical and local names. In the parts of the country where Algonkin dialects were spoken, -paug or -pdg final (or followed by the locative sign -wt, -it-, -ing,) denotes ‘ water at rest,’ ‘standing water,’ and is the substan- tival component of many names of lakes and ponds: -hanne or ~han ‘flowing’ distinguishes a ‘rapid stream’ or ‘current’ ; ~tuk (Abnaki -tegaé, Delaware -tttuk) ‘driven in waves’, from a root signifying ‘ to strike,’ is found in names of tidal rivers and estuaries and of broad deep streams: -ompsk, contracted to -psk or -msk, (Abnaki -peskoo, Cree -pisk, Chippeway -bik) denotes ‘ hard or flint-like rock.’*
-Minne or its contraction -min is the generic affix of names of berries, nuts and other fruits which may be eaten. It is never used independently, though a nearly-related word meen, pl. meenun, is found, in the Chippeway and some other dialects, specially appropriated to a single species (the blue berry), and in the Cree the diminutives ments and meniests are used for ‘berry’ generally. The cranberry was called by the Narra- gansetts sasé-min ‘very sour berry, by the Chippeways
* Since the above was written a more extended notice of this class of generic formatives, has been given in a paper “ On the Composition of Indian Geograph- ical Names,” printed in the second volume of the Collections of the Connecticut Historical Society.
74 | J. A. Trumbull,
muskegé-min ‘swamp berry’; the strawberry is (Chip.) odé- min ‘heart berry’; Indian corn, in Massachusetts, ewdchi-min or weatchi-min, but among the western Algonkins, monda-min ‘manito (i. e. supernatural or wonderful) fruit.’
-Pin denotes an esculent tuber, or tuberous root; as in Chip.) opin, potato, witti-pin, wild potato, suskode-pin ‘ prairie root,’ wawbeze-pin ‘swan root’ (a species of Sagitta- ria), etc. |
-Asq in the Massachusetts and Narragansett dialects was the generic formative of the names of fruits which might be eaten ‘raw’ or when ‘green,’ — particularly, of melons and edible gourds. In the plural -2sg makes -asgnash, — whence our name ‘ squash’ for several varieties of Cucurbitacezx.*
In the Chippeway language, -gan and yigan (-gun and yeegun, Schoolcraft; Cree, -gun, -chéggun, Howse; Delaware, -can, -schican, Zeisberger ;) are the formations of many names of instruments. Mr. Schoolcraft regarded these names as ‘based upon the word Jeegun, one of the primitive nouns, which, although never disjunctively used, denotes, in its mod- ified forms, the various senses implied by our words instru- ment, contrivance, machine, &c.”’ Sometimes, he says, it is shortened to-gun.f These generics, however, are not primitive words, but the formatives of participles, and ~égan is never shortened to -gan, but is formed by the insertion of the char- acteristic of energetic action, 7, between -gan and the verbal root. Participials in -gan (or -gun) serve as names of what may be distinguished as passtve instruments, — things ‘ used for’ some purpose by an animate agent: e. g. niba-gan ‘a bed’
* The primary meaning of asg or ast seems to have been ‘ before-time’, ‘imma- - ture’, ‘unfinished,’ or the like. As an adverbial prefix to verbs it denotes that the action is net yet performed. Hence, esk-i and ask-m ‘it is raw’, i. e. not yet prepared to be eaten, or ‘it is green’ i. ©. not yet matured. Eskimo is the Algon- kin name of one who ‘eats fish or fiesh raw’, — Abnakj ’ski-moohoo, Mass. aski- moowhau. The Dakota sak corresponds to the Algomkin asq: sa‘ka ‘raw’ ; dimin. sa'ka-da* ‘green, immature’; sa’ka-yutapi ‘something eaten raw’, melons, cucum- bers, &c.
t Lectures on the Odjibwa Substantive. — Gallatin in Trans. Am. Antiq. Soc., vol. ii. p. 228, adopts from Schoolcraft the statement that “a numerous class of compounds is derived from jeegun or gun, meaning ‘instrument’, words never ased alone.”
On the Method of Studying the American Languages. T5
(‘used for sleeping’), opwd-gan ‘a pipe (‘used for smoking’), wassdilshie-gan ‘a window’ (‘used for lighting’), &c. Parti- cipials in -jigan (jeegun) or -chéggun denote inanimate agente, instruments ‘for doing’ something and which are regarded as exerting a degree of energy of their own. Of this class are all labor-saving machines and contrivances for helping the In- dian do what he cannot do without them: e. g. Chippeway kishkiboyjigan ‘a hand saw’, i.e. used for cutting crosswise; tdshkibozigan ‘a saw-mill, or pit saw’, used for cutting length- wise ; bissiboyigan, ‘a corn- mill, or coffee mill’, used for making fine, reducing to powder. Delaware, kinhan-schican (Zeisb.), ‘a grindstone’, used for sharpening.*
The preceding examples have been taken from languages of the Algonkin family, in which the generic annex follows the qualificative. In other groups the order of synthesis is re- versed and the generic is prefixed. The Dakota cha" (ch as in chin) meaning ‘ tree’ or ‘ wood’, corresponds to the Algon- kin -takh, for the designation of articles ‘made of wood’ or ‘belonging to a tree’, e. g. cha"-ha ‘ tree skin’, bark; cha’ha"pt ‘ tree sap’, sugar, chaopiye § wood to put into’, a box or wooden vessel ; cha"shet” ‘tree fat’, gum or resin; cha’shu'shka ‘ good for nothing wood’, the box-elder, &c. Za is a generic prefix of names of ruminating animals, but when used independently denotes the moose, par excellence. Wa limits certain names to the ‘ bear’ species. Ho refers others to the class’ ‘ fish’, — as in ho-a'pe ‘a fin’ (from a'pe ‘ leaf’); ho-wa'sapa ‘ all-black fish’, the cat-fish ; ho-ta"ka ‘great fish’, the sturgeon, &c.
The number and variety of these characteristics — often ° represented by single consonants,—the presence of one or another of them in almost every synthesis, the fact that the modifications they effect in the action of the verb are some-
* This characteristic -ji is itself a compound or derivative, as we find by going back to simpler forms of the verb. In the Cree and Chippeway, ¢ or d (Massa- chusetts, ¢¢ or dt) is the characteristic of verbs of action performed on inanimate objects; but if the object is not expressed, the verb takes a different inflection and its characteristic becomes che or ji (i. e. t-she, d-zhe). From this form of the verb comes the participial in -jigan or chéggun, which by its formative, -an or -un ascribes action to an inanimate subject employed to do an act, generally, or of which the object is not specified ; “it cuts (something, or any thing) crosswise”, “it makes something sharp”, &c.
76 J. H. Trumbull,
times so slight as to be lost sight of in translation to English, and the difficulty of separating them in all cases from the roots and inflections proper and of assigning to each its specific value — have contributed to give false notions of Indian word- making. One of these must be mentioned here,—for the error it involves is fandamental. |
More than fifty years ago, Mr. Duponceau, relying on a state- ment of Hans Egede that “‘ words are formed in the Greenland language by taking and joining together a part of each of the radical words the ideas of which are to be joined together in one compound location,” and believing that he found something like this in the Delaware language, questioned his correspond- ent Mr. Heckewelder on the subject. Mr. Heckewelder replied, that in the Delaware and other languages that he was ac- - quainted with, ‘‘ parts or parcels of different words, sometimes a single sound or letter, are compounded together, in an arti- ficial manner, so as to avoid the meeting of harsh or disagree- able sounds, and make the whole word fall in a pleasant man- ner on the ear”; and he gave two or three illustrations — or _ what he supposed to be such — of this process. It was further explained by Mr. Duponceau, in the Preface to his translation of Zeisberger’s Delaware Grammar, in 1827, as “ consisting in putting together portions of different words, so as to awaken at the same time in the mind of the hearer the various ideas which they separately express.”” Dr. Pickering lent his high authority to this statement by incorporating it with his excel- lent paper on Indian Languages in the Encyclopedia Ameri- “cana (vol. vi., appendix), and Mr. Gallatin adopted it in his Synopsis (Trans. Am. Antiq. Society, ii. p. 201), where “ this manner of compounding words, by uniting in a single one the abbreviations, sometimes a single syllable, or even letter, of five, siz or more words” ,is said to “belong equally to the Eskimau and to the Algonkin,” &c. It attracted the attention of W. von Humboldt, whose observations on the Delaware Language in the introduction to his great work (Ueber die Kawi-Sprache auf der Insel Java, vol. i. pp. ccexxxii.) were founded on Duponceau’s Correspondence with Heckewelder and transla- tion of Zeisberger’s Grammar. From these authorities he
On the Method of Studying the American Languages. (7
inferred, “das es von dem Redenden abhangt, solche [neue] Worter oder vielmehr ganze zu Wortern gestempelte Phrasen gleichsam aus Bruchstiicken einfacher Worter zusammenzu- fiigen”’: but he came nearer than had Mr. Duponceau to the true explanation of the process, in the suggestion that ‘“‘ man miisste aber eine tiefere Kenntniss der Delaware-Sprache und der Verwandtschaft ihrer Wérter besitzen, um zu entscheiden, ob wirklich in den abgekiirzten Wortern die Stammsylben vernichtet, oder nicht vielmehr gerade erhalten werden’ (p. eccxxxiii.).
The publication in Paris of Duponceau’s Mémoire, in 1838, brought his theory of Indian synthesis, or ‘agglutination’ as he preferred to call it, to the notice of European scholars, and since then it has been the common property of writers on language. Mr. Schoolcraft had already adopted it. In his Lectures on the Chippeway Substantive, first printed in 1834, he had contrasted the language ‘viewed in its original, elementary state — in a vocabulary, for instance, of its prim- itive words,” with its “oral, amalgamated form”, whose ‘transpositions may be likened to a picture, in which the opal, the carmine and the white lead are no longer recognized as distinct substances,’ — while ‘it is the painter only who possesses the principle by which one element has been curtailed, another augmented, and all, however seemingly discordant, made to coalesce”: and ‘so completely transpositive do the words appear, that like chessmen on a board thetr elementary syllables can be changed at the will of the player, to form new combinations to meet new contingencies, so long as they are changed in accordance with certain general principles and conventional rules; in the application of which, however, much depends upon the will or skill of the player.”
Now it is certainly true that the elements of every new word in an Indian language may be found also in older words of the same language ; but in regarding these elements in their new combination as the representatives or as “abbreviations” of the old words, selected not because of their inherent signifi- cance but “ at the will of the player” and merely as mnemonic symbols, — Mr. Duponceau’s theory is not only unsupported
78 J. H. Trumbull,
by facts, but it involves a doctrine of Indian synthesis which is directly opposed to the true one. If it be accepted, neither the analytical method I have suggested nor any other could lead to positive results, and mere guess-work must be substi- tuted for critical investigation. The actual derivation of the components of a phrase-word could never be known with cer- tainty, and we could do no more than assign possible values to the several unknown quantities.
As I have said before, the Indian aimed at extreme precision. His words were so constructed -as to be thoroughly self- defining and immediately intelligible to the hearer. In the construction of his synthesis, he was controlled by established and universally recognized laws; in the selection and arrange- ment of its elements, he admitted nothing ambiguous, left nothing to conjecture. What may be called the frame-work of an Algonkin polysynthesis, or phrase-word, is composed of two or more monosyllabic roots, each having its own substantial meaning. These are not to be regarded as “ abbreviations” or fractions of words because they do not bring with them all the pronominal or other affixes, characteristics and inflectional forms they have received in other relations: and they “ com- bine in one” not “all the various ideas contained in the several older words” in which they are found, but only the ideas which are inherent in or inseparable from themselves.
Mr. Duponceau’s theory was adopted on slight examination and mistaken evidence. It appears to have been suggested to him by the statement of Hans Egede respecting the composi- tion of words in the Eskimo language. In his correspondence with Heckewelder, before mentioned (p. 76), and again in- his Mémoire (p. 92), he gives analyses of Eskimo words taken from Egede’s Description of Greenland (Det gamle Gronlands nye ° Perlustration, Kidbenhavn, 1741), perhaps in its German or English translation. But Paul Egede (in his Grammatica Gronlandica, Havnie, 1760,) and Otho Fabricius (Forsdg til en forbedret Grénlandsk Grammatica, 1791 and 1801,) sup- plied more full and clear explanations of the method of com- pounding Eskimo words, and showed how the meaning of the verb is modified by characteristic particles (see before, p. 71)
On the Method of Studying the American Languages. 19
which are to be construed as auxiliary verbs but which have no independent meaning.* Fabricius (pp. 214-343) enumer- ates 287 of the inseparable characteristics (“‘ Verbalske Endel- ser,’ and gives examples of their use. Neither of these writers affords the least countenance to Duponceau’s assumption that such particles are abbreviations or reminders of words which retain independent places in the language, — whatever may have been their remote origin.
In Mr. Schoolcraft’s glowing description of Indian word- making as a process the secret of which is known only to the word-maker, in which the laws of composition are subordi- nated, and vocal pawns are moved about “at the will of the player,’ — this theory receives its reductio ad absurdum.
Dr. R. G. Latham, in his Elements of Comparative Philolo- gy, published in 1862, (p. 520), gave a concise exposition of the true doctrine of grammatical synthesis in the American languages. He dissented explicitly from the view which, since the appearance of Duponceau’s Mémoire in 1838, had been taken by American and European philologists. Admitting “that there are in these languages certain very long words expressive of what in Europe is expressed by short ones, and that out of these long words compounds may be made which are no longer than either of the single elements,” and that ‘this looks as if each were picked to pieces, and a part only taken,’’ he remarks, that “in respect to the phenomenon of a composition with a decomposition to precede it— it would be important if proven. The fact, however, of the decomposition is more than doubtful. Jt i not out of the full-formed pair of primary compounds that the secondary compound is made, but out of the original parts which existed while they — the apparent primary compounds — were merely compounds in posse.”
*“«Certee cum Verbis simplicibus componuntur Particule, loco Auxiliorum usurpande, qu nullius, extra illam compositionem, sunt significationis” (Egede, Gram. Gronl., p. 145); and again, of adverbs (p. 199): “ partim e separatis voci- bus aut particulis, partim e particulis encliticis, nulla, extra compositionem cum aliis vocibus, significatione gaudentibus, constant.”
80 S. S. Haldeman,
V.—On the German Vernacular of Pennsylvania. BY 8. S. HALDEMAN,
PROFESSOR OF COMPAR. PHILOLOGY IN THE UNIVERSITY OF PENNSYLVANIA.
Pennsylvania German—#improperly called ‘“ Pennsylvania Dutch” is not normal German corrupted after its arrival in America, but a fusion of the South German dialects brought from the region of the upper Rhine, including Switzerland. To these localities we must look for its characteristics, except the infusion of English, and in adopting more or less of this foreign element, it follows the natural law of all languages, in taking up new or convenient terms for new objects, so that a word like clearing (land cleared by the destruction of a forest) comes into American German as naturally as prairie came into English.
Nor is this English element confined to Pennsylvania Ger- man, for the advertisements of High German newspapers in the cities of New York, Buffalo, Chicago, Little Rock, Des- moines, Philadelphia, Galveston, and others, cannot be read by a German unacquainted with English. Here we find —
Cin pradtvolleS 3ftid. und Bafement Bridhaus in 10. Ward, volles Lot .-..Gin groped Country Hotel... . Cine fcpdne Farm... . Barn und allen Ge- tathidaften....Mortgages werden al8 Zablung angenommen....BWird ver- tauft gegen Brooklyn Property. — BVerlangt: DMtddden, die Oftrid und Vultures ndhen fonnen....Topfloor, Room 16.... Operators und Baifters .... Gin Front. und Rearhaus....durdhaus gepaintet....ein tid. und Bajement Framehaus....gute Gegend fiir eine Grocery-Store — Cin Gro- cerystore mit Stod und Firtured....Lagerbier-Ealon mit Boardinghaus — Gin Drygoods-Store....an einem Corner — Gin Butdhershop — Cin KaffeeSalon verbunden mit Candy, Soda-Wafser, Rootbier, Cifeream, ete— (New- Yorker Staats-Zeitung, 8 Aug., 1870.)
In Pennsylvania German the a is sounded as in what, fall, the nominative form ‘der’ replaces the accusative den, the imperfect tense is avoided, both as in: Swiss, and the perfect is used instead, so that for id) fagte (I said), id hatte (I had), id fain (I came), we have ‘ich hab ksaat, ich hab kat, ich bin kummo.’ The nominative neuter article da becomes ‘ des,’
On the German Vernacular of Pennsylvania. _ 81
infinitive -n is dropped, and many other finals replace it with nasality — which are features common to Suabian, as well as the almost exclusive use of -li as a diminutival. To the Ger- man strong verbs the weak inflection is given, so that, instead of er ift he eats, er fallt he falls, er frigt he asks, we have (# in fat) ‘er esst, er fallt, er frookt.’
The three following detached stanzas are from Harbaugh’s Harfe (Poems in this dialect, 1870), to which I add transla-
tions. ‘Der Mond is{iss] uf—er is juscht voll — he moon is up — so full and fair, ’R piept zum Fenschter rei[ herein] —guck mol —* t peeps into the window there, Un scheint uf’s Bett un Floor. And shines on bed and floor : Was regt sich an der Wand ? ich glaab, What plays upon the wall? I b’lieve ’S is Schattenschpiel von Babblelaab : . The shadow of a poplar leaf, —
So hab ich’s g’seh’ zuvor. p- 83. As I have seen before.
Ei guck, wie schnell der Mond nuf[binauf] eilt; But see! the moon has upward gone; Wie lang hab ich mich doch verweilt! How long indeed I’ve linger’d on — Ich muss mich widder zeh’. Again I must now go ; Gut Nacht, klee’[f[cin] Stiibche’! halt, ich meen Good night my bedroom! stay — I feel ’S wir Ebbes in mei'm Aag—’n Dhreen! A something in my eye —a tear! Der Abschied, der dhut weh ! p. 35. Departure — this is woe!
’S is Alles schtill! Die Kinner schtecke All is still! the children nestle Schnock im Bett un draame schee’ ; Snug in bed, with pleasant dreams ; Santa Claus werd sie net wecke, Santaclaus will not awake them, ° Er dhut all sei’ Sach allee’ ; He does all his work alone; Hengt d’r Baam mit Scheene Sache, Decks the tree with pretty trifles, - Schleicht herum im ganse Haus, Slips about th’ entire house, Legt sei’ Gabe ’raus mit Lache, Scatters gifts around with laughing, Un dann — Ho!—zum Schornschtes’ naus !” p. 40. And then up the fiue departs. 11
s2 8. 8. Haldeman,
The next example, by Tobias Witmer of Buffalo, has certain ‘Sanday Concerts’ in view. I add a translation and notes, and retain the not quite uniform spelling of the original:
‘““DIE FRESCHLIN. THE FROGS. Die freschlin hen! e? grosze jacht,® The frogs have always quite a din Si peiffe* schiir die ganze nacht, For half the night, and toads strike in Die krotte dun ’ne® helfe ; To aid (perhaps outvie) them ; Doch wann ebmol der reife® kammt, Tow’rd midnight, if the frost appears So is’ ir’ liedle glei® verstummt, . Their voices cease to reach our ears, Si geen ins bet am zwelfe. And then to bed they hie them.
Wan’s nord? ¢ bissel wermer werd, If t were a little warmer now Es grosse dreckloch ausgederrt, To desiccate that muddy slough So mitte in dem jauli, As usual in mid Jaly, Nord?® fange d’ groeze bullfraags an?!° | Then giant bullfrogs would begin Un bloose, das mer!!’s heere kaa, To let us hear their vocal din So weit wie unser muuli.!? As far as our mooley. Wan ich jetz!® weer e musikand,’ Could I instruct in tune and dance Un keem juscht vom Franzooseland, And were I just arrived from France Dart woo sich d’ kaiser mache ; Where kings are self-elected, Ich ht!* e musik’ schuul in gan ewe A school I d soon get under way Di fresch zu lerne, ’s nehm net lang, To teach the frogs to sing and play — Ja gwisz, du brauchscht net lache. Do n’t laugh, ye disaffected ! E dutzet bullfraags miszte nei,!§ A dozen bullfrogs I would get, E hundert gleene!® fresch derbei, A handred little ones I d pet — Die krotte far der tenor ; The toads should sing the minor, E’ groszes loch voll laager biir, A puddle full of lager beer Des miich! de fresch di hels recht kliir® Would make their raucous throats quite clear —
Du weescht,!® — was weer jetz schener ? What could indeed be finer.
On the German Vernacular of Pennsylvania. 83
Nord? kriigt!* ich miir en groosse saal Then would I get a stately hall
Un adverteis? en sunntaags baal And advertise a Sunday Ball, Mit meine musikante ;7 With music-frogs a million ; Dii veegel miszte all derzau, I d have as guests the various fowl —
Dr Condor un dr Uhu-huu The condor and the graceful owl
Dii tantze mit enander.” Would start a grave cotillion.
Des weer der human familii ’T would be to man’s proud family © En pattern of gentilitii ‘A pattern of gentility Si kente sich impruafe, With which I could delight them ;
Si hen! e fact schun ardlich?! glernt, They have in fact begun to learn — - Es gebt e mol™ e groose erndt — The ball would yield a fair return —
Der Guckuk werd si ruufe.” The cuckoo shal! invite them.
1 For haben have, as in South German ; — han of the Minnesingers. A dative form of the pronoun is used in ‘m’r hen’ (Swiss mer hend) we have; ‘d’rhent’ (Swiss der hend) you have.
2 The German article cin, eine reduced, like English a, and pronounced indis- tinctly.
® The common Penn. Germ. word for noise, Garin. * Sie pfeiffen they pipe.
5 For the dative i}nen; or possibly, a regular plural of jn.
© German der Reif rimefrost. About midnight the cold causes them to cease.
7 The s is not used as English z. 8 oleic) presently, directly.
®° Then, a metathesis of Darnad.
19 This word is the vowel of fall nasalised, and the word below rhymes with it.
11 The German plural dative inir used impersonally in the singular, for man, one, as in Swiss, &c. See note}.
12 English mooley a horniess cow. From the Irish and Gaelic.
13¥For jept, but nau (Eng. now) is more used. In a funeral sermon lately heard (1870) in the dialect, although the speaker endeavored to use jets, he often fell into nau, and he used but one other English word, in saying that ‘salvation depends upon certain conditions’ — pronouncing the word as in English, but with a pure hissing final.
14 Greecht, also griicht and kriigt J would get; Ger. friegen to seize, get.
8 miibten binein would have to bein. 16 Plural of flein little.
13 Des miich this would make. Des miich alles recht this would make all right.
18 The English word clear. 9 Di weift thou knowest.
2 Enanner is usually without the d.. .
31 Tolerably ; from yrdentlid) orderly. “ce mil, ein mal once.
VI.— On the Present Condition of the Question as to the Origin of Language. BY WILLIAM D. WHITNEY, PROFESSOR OF SANSKRIT AND COMPARATIVE PHILOLOGY IN YALE COLLEGE.
Ir is far from being my purpose in this paper to enter seri- ously into the discussion of the origin of language, in the way either of putting forward a theory of my own, or of contro- verting those which have been put forward by others. I do not wish to open at this time before the Association so vast and uncontrollable a subject. No theme in linguistic science is more often and more voluminously treated than this, and by scholars of every grade and tendency ; nor any, it may be added, with less profitable result in proportion to the labor expended ; the greater part of what is said and written upon it is mere windy talk, the assertion of subjective views which commend themselves to no mind save the one that produces them, and which are apt to be offered with a confidence, and defended with a tenacity, that are in inverse ratio to their ac- ceptableness. This has given the whole question a bad repute among sober-minded philologists—insomuch that, for exam- ple, the recently established French association of kindred object with our own (the Société de Linguistique) forbids by its fundamental law any introduction of the origin of language into its transactions and debates. The prohibition, however, has not worked unexceptionably well; for there is no similar society a larger part of whose members have rushed into print upon the subject before the general public; so that one may conjecture that if they had been permitted to fight the fight out more among themselves, the community outside would have been the gainer: and hence, we need not feel bound by their example.
The reason of this irreconcilable discordance and regreta- ble waste of activity appears to be that no common basis of discussion is yet established. The question of the origin of language is not one of facts, to be settled by direct evi- dence, like the question of relationship of a dialect, or of the
On the Question as to the Origin of Language. 85
genesis of a form; it does not belong to comparative phi- lology, but to linguistic philosophy, all whose fundamental doctrines are involved in its solution. And it will be readily settled (so far as it is capable of being settled at all) when the grand principles of linguistic philosophy are placed upon a firm basis, when it is no longer the case that even scholars of the highest rank are disagreed as to such points as the na- ture of language and its relations to the mind and to thought (the old dispute as to give or Gécer), and the relation of human expression to that of the lower animals.
My intention here, then, is merely to review briefly the present aspects of the discussion, and to endeavor to straiten its field a little, by directing attention to points that deserve to be regarded as settled, and pointing out directions in which further effort will be likely to lead soonest to valuable result.
And, in the first place, it may be premised that the question of the origin of language is a purely scientific question, and a legitimate one, aud that its investigation is to be carried on by strictly scientific means and methods. There ought to be no need of putting forth this claim, still less of insisting upon it ; yet, as things are, it requires to be made and urged. A sci- entific treatment implies that the known and recorded facts of human language, in combination with known and observable characteristics of human nature, be made the sole basis of the inquiry, and be examined with thoroughness and without pre- judice, till they have been forced to yield the utmost result that they are capable of furnishing. This, on the one hand, excludes the admission as co-ordinate evidence of all opinion, by whomsoever and at whatsoever time expressed ; of all state- ment, traditional or other, and on whatsoever authority re- ported. Nothing but harm and confusion can come from attempting to combine the hints of the Genesis, for example, with the deductions of science, in order to yield a joint con- clusion; or from suffering the one to govern or regulate the other. The student of language should not ask whether the course of inference and deduction which he is pursuing is go- ing to bring him to conclusions in accordance with views heretofore held by any, or not. His business is solely to see
86 W. D. Whitney,
what language itself has to say of its own origin, and how plainly and unequivocally; whether it gives him a solution of the problem that is certain and must be maintained against all attack, or only furnishes probabilities and limits the range of possible hypotheses. When the scientific work is done, then is the time for comparison with views derived from any other quarter, balancing their respective merit and claim to credence, abandoning the one for the other, or trying how they may be reconciled.
The scientific method requires, on the other hand, that no assumption of a different human nature from that which we see and know be made a factor in the inquiry — that no special faculty, or instinct, either in particular individuals, or races, or generations, be postulated, and charged with the beginnings of intelligent and intelligible utterance. To make such an assumption is equivalent to abandoning the scientific ground entirely, and is no better than the admission of a miraculous or superhuman agency. If human capacities as they actually are be found, in the last analysis, unequal to the task of pro- ducing the germs of a method of communication like ours, then that will be the scientific result of the investigation, and the field of conjecture will be thrown open to whoever may desire to enter it; but he is no scientific inquirer who uses such materials in his investigation itself.*
*I need hardly explain that I allude here to the assumption which F. M. Miiller has made a part of his so-called ‘‘ ding-dong theory” of the origin of lan- guage — his assumption of ‘an instinct, an instinct of the mind as irresistible as any other instinct,” which, after it had given ‘to each conception, as it thrilled for the first time through the brain, a phonetic expression, became extinct when its purpose was fulfilled.” It is, indeed, possible to put this doctrine in such a form as shall give it a scientific status. If the claim were made that a faculty and disposition to direct expression of thought and the production of “ phonetic types” forms a part of universal human nature, and would show itself and work its legit- imate results in every individual if its action were not anticipated by the learning from others of already formed and developed speech — that, indeed, would be worth discussing and testing by careful inductive processes, by examination of the facts of human history and the history of speech. But Miiller, with his followers (if in this particular doctrine he has any followers), does not explain himself thus, or show any indications of meaning thus; in his view, that this faculty was “ pe- culiar to man in his primitive state” “‘ must be accepted as an ultimate fact”; no other reason is alleged than that “that faculty must have existed in man, because its effects continue to exist” — which is a palpable begging of the qucstion, a
On the Question as to the Origin of Language. 87
In the second place, if we would make our contributions to this discussion tell upon its result, we need to draw the line distinctly between the historical and the theoretical sides of the question — or, what is nearly the same thing, between what has already been done and what yet remains todo. Histori- cal investigation takes us from the present condition of lan- guage a long way back toward the beginning, but it does not and never can take us the whole way. In the very nature of things, it cannot show us why the first speakers used this and that sign for this and that idea; and practically, it cannot show us that they did use this and that sign at all. There is no prospect that we shall ever be able to say ‘ these are the very first utterances of speaking men; now let us see how they originated.’” We come nearest to such a result, doubtless, in the Indo-European family ; yet, even there, we can only assert the use of certain elements in certain senses before the break- up of the family into its independent branches; of the abso- lute primitiveness of any of these elements we have and can have no assurance. In most or all of the other families we cannot even go so far as this— whence the worthlessness of the attempted comparisons of roots between the different families. The grand conclusion, however, at which historical study has surely and incontrovertibly arrived, is that all the grammatical apparatus of languages is of secondary growth; that the end- ings of declension and conjugation, the prefixes and suffixes of derivation, were originally independent elements, words, which were first collocated with other words, and then entered into combination and were more or less thoroughly fused with the latter, losing their primitive form and meaning, and becoming mere signs of modification and relation; hence, that the his- torically traceable beginnings of speech were simple roots; not parts of speech, even, and still less forms. That these roots, moreover, signified external, sensible, physical acts and qualities ; precisely what ones, we cannot yet tell, and shall perhaps never be able to tell; but this, in its bearing on the question of origin, is of no great consequence. All that there
taking for granted, without argument, that language is its effect and could have been the effect of nothing else : that is to say, it must have existed because it must have existed.
88 W. D. Whitney,
is left to explain, then, is, how such roots as these should have come into being and use. And this amounts to a won- derful simplification of the question of origin; did we not see that primitive speech was thus widely different from the devel- oped discourse of historical epochs, we should give up our in- quiries in despair, and acknowledge that only miraculous power could have been equal to the origination of language.
It would be unfair to claim that the accordance of students of language in this doctrine is absolute. There is here and there an ultra-conservative, who will believe only so far as he is forced by unequivocal testimony, and, while he confesses the later formative elements of speech to be wrought out of inde- pendent words, refuses to infer that the older are of the same character, preferring to hold that there was some mysterious and inscrutable difference between the ancient and modern tongues as regards their principle of growth: and we even meet occasionally with a man who has done good service and won repute in some department of philology, and who yet commits the anachronism of believing that endings and suffixes sprouted out of roots by an internal force. But these are men with whom it is vain to reason; they must be left to their idiosyncrasies, and not counted in as bearing a share in the progress of modern linguistic science. There are also, of course, many whose studies in language have not gone far enough to show them the logical necessity of the views we have described ; but they, too, arc to be reckoned as in the rear of the present movement. He who sets himself seriously to ex- amine or to demonstrate the theory of roots as the historical germs of speech will be accounted as one who threshes straw ; he who does not make that theory the basis of his farther in- quiries into the origin of language must not expect even to obtain a hearing from scholars.
Upon this basis of historically determined fact whatever farther truths are raised must be won by processes of another sort. Strict induction from determinate items of knowledge is no longer applicable ; its place is taken by inference from general views and theoretical conditions—these views and conditions being themselves, of course, not arbitrarily assumed,
On the Question as to the Origin of Language. 89
but derived by inductive process from the known facts of lan- guage and human history. It is here, accordingly, that there begins to he wide discordance among even the best scholars and deepest reasoners; a discordance that is sometimes im- plicit and unacknowledged, sometimes clear and outspoken. And it is highly desirable that the efforts of those who would advance the science of language be brought to bear directly upon some of these points of discordance, whose settlement ought to be already within reach, and would be of decisive influence upon that of the ultimate question which we are considering.
Thus, what difference can reach deeper, or be of wider bear- ing, than that which prevails with reference to the nature of the relation between language and thought? One party con- tends, either impliedly or formally, that there is an actual identity between speech, on the one hand, and thought, mind, reason, on the other; that language is not only a sign of rea- son, but its very substance; that thought without expression is an impossibility ; that the formation of an abstract idea de- pends entirely upon its name—and soon. This doctrine finds probably its extreme expression in the assertion that a deaf mute is destitute of reason, and does not become possessed of it until he learns a mode of expression from the reasonable beings about him. The other side maintains that language is only the assistant of reason and the instrument of thought ; that reason is the indefeasible endowment of humanity, and thought the action of human minds; that they need, in order to their full development and proper working, an auxiliary like speech, and have proved able to provide themselves with it; that, even had men been deprived of voice, they would have made available some other instrumentality for the same purpose; that he whose want of hearing cuts him off from this particular mode of expression is still a man, with all the essential characteristics of humanity, which merely require to be developed and educated by a less usual and less convenient instrumentality ; that, were a generation of infants to grow up untaught to speak, they would from the beginning be possessed of reason no less than ourselves, and that their reason would
12
90 W. D. Whitney,
at once begin a course of training analogous with that through which the liuman race has already passed, one of the essential steps in this course being the production and use of speech. When we come down to smaller details, the one side hold that the idea without the word is an impossibility, and that no con- ception can exist till there is provided a name for it; the other, that the idea or conception always precedes in time, and must precede, the name; that signs are made in order to be applied to ideas which the mind has formed and seeks to ex- press;:that the whole process of language-making, from the beginning of time, has been only a process of names-giving which has followed close upon the growth of knowledge and conscious thought, mastering and making manageable and communicable whatever bit of valuable mental wealth has been wrought out by experience.
Men who hold these two so diverse sets of opinions cannot be expected to agree with one another in their views of the origin of speech. And he who should address himself suc- cessfully to this one subject, should point out the errors and misapprehensions involved in the one or the other theory, or in both, in a convincing manner, so as to lead the way to a mutual understanding and agreement, would, in my opinion, render the very greatest service that can be rendered to the question of origin. Most of those who undertake the latter directly do not treat the other with fullness, or at all; they simply let their discordant views upon it appear, as if the matter were too plain and elementary to call for discussion ; or they dispose of the opposing opinion with an absurd mis- representation or unfair fling. Thus, nothing is more com- mon than that those who hold the former of the two sets of opinions described above should make easy work of vanquish- ing their opponents by simply assuming the latter to maintain that men work out a whole series of new ideas, and then, by an afterthought, set themselves at work to devise and apply appellations for them; or, they attempt to confute the idea of the “conventionality” of spoken signs for thought by showing the laughable absurdity of a gathering or ‘“‘ convention’ of speechless men, discussing and voting the adoption of spoken
On the Question as to the Origin of Language. 91
designations —as if the term “conventional”, in any of its uses, ever implied any such convention! Indeed, so custom- ary is this sort of unfairness, that I may truly say I have never seen the controversy conducted otherwise by the party re- ferred to, or the opposing views squarely met and argued against, in the form in which their present supporters would put them.
Another point of first-rate importance, whose solution is to a great extent bound up in the result of the controversy which we just have been considering, is this: should the first im- pulse to speech have come from within, or from without? were words pushed out by a longing after expression, for the sake of the benefit and relief afforded thereby to the individual’s owl mind, or were they drawn forth by the desire to commu- nicate, to make known to another what lay in the utterer’s thought ? were they framed as the means of expression pure and simple, or of communication? This also is @ point which is apt either to be overlooked altogether by inquirers into the origin of language, or to be carelessly and insufficiently treated by them. Yet its decisive bearing upon the question is evi- dent. Its settlement one way or the other involves a com- plete diversity in the essential character of the first utterances, the germs of after development. On the one hand, we should have to seck in these some internal and necessary tie between the conception and its sign, naturally inherent in the latter, and determining its assignment to its office. On the other, no such tie would be implied, any more than between idea and sign in the later stages of language, and the only adaptedness in the sign would be its adaptedness to be readily understood by the being to whom it was addressed. The first framers of speech would be regarded as standing toward one another in a position essentially the same with that of two persons of wholly different language who should meet at the present time and desire to hold communication together: all the resources of imitative expression would be laid under requisition by them — grimace, gesture, posture, imitative utterance, whether Onomatopeeic or exclamatory, symbolical utterance, so far as in this there was power of suggesting an intended meaning.
92 W. D. Whitney,
The process of mutual understanding would be a tentative “one, every imagined expedient being tried, and adopted if it proved successful; and ere long a foundation would be laid which would admit of rapid and indefinite expansion. We must not overlook, of course, the great differences between this imagined case and that of the primitive language-makers : where two beings with developed powers of thought and ex- pression, and with formed habits of speech, came together, their progress would be indefinitely greater, and the process would soon become one of learning one another’s speech, and framing a common dialect out of the mixture of the two (doubtless with great preponderance of the one over the other); but where the two were before speechless, and that command of the mental powers and dexterity in wielding them which language gives had to be acquired step by step along with and by the production of language, the process would be laboriously slow, and generations instead of days or weeks would be needed to mark the stages of its advance. And yet, in both cases the initial steps would be parallel and essentially alike. That is to say, the recognition of commu- nication as the primary and ultimate object of speech involves as its necessary consequence an acceptance of the “imitative” theory of the origin of speech;* nor, on the other hand, can this theory be established independently of such recognition ; the two doctrines must stand or fall together. Into any de- tailed discussion of their truth, it is not the purpose of this paper to enter. I would only point out one or two difficult implications which seem to be made in the opposing view.
One of these is, that the solitary man would have the same inducement to produce a language as the member of a family or a society. If words are made because the individual feels or knows that the possession and use of such signs will help his consciousness to gain command of the processes of mental action, will render orderly and. consecutive thought possible to him, will be to his reason what the tools he invents are to
* The theory, namely, which some have unauthorizedly divided into an onoma- topetic and an interjectional theory, but which in fact includes both these, and more beside. See Wedgwood’s Origin of Language, and the writer’s Language and Study of Language, eleventh lecture.
On the Question as to the Origin of Language. 93
his hands, then the whole efficient force and its occasion of action are within the individual, and society adds only a means of perpetuating what he originates for his private benefit. And it needs to be inquired whether what we know of solitary hu- man beings, or of those who by some special local deficiency are cut off from the usual avenues of communication with their fellows, and whether what we see of the relation of society to language during the recorded history of its growth and em- ployment, are in accordance with this view.
Again, a questionable degree of forecast, of comprehension of what would make for his advantage in the development of his capacities, is thus attributed to primitive man. That human: beings at even the lowest stage of existence are accessible to in- ducements founded in their social nature, no one will think of denying; but that they are capable of anticipative pleasure in the ‘* projection of their thought outside of themselves,”’ in set- ting it forth as an object of contemplation by themselves, is vast- ly more doubtful. Experience, and only experience, it would seem, is capable of making the individual realize the advantage and take pleasure in the exercise — if, indeed, the realization comes at all until a considerable degree of culture is reached, and if, in all the early.stages of development, men do not, so far as they themselves know and are conscious, talk solely for the sake of intercourse with others. Thus it was with the his- tory of writing, an art that stands only second to that of speak- ing in its bearing upon the culture of the individual and the advancement of the race; it was not devised as a means of culture and advancement, but as a means of communication merely; and all the advantages which it has in the former respect also have been attained unconsciously, without being anticipated or aimed at.* Thus it has been, too, with the in- vention of instruments. The instinct to devise and use such aids to his physical powers is not a whit less distinctively char- acteristic of man than the instinct of speech; but the earliest human beings did not sit down to satisfy that instinct by ex-
* The analogy in this respect between speaking and writing, an analogy preg- nant with meaning aod instruction, has been more fully sct forth in the author’s Language and the Study of Language, lecture twelfth.
94 W. D. Whitney.
ercising their inventive capacity; they provided for each special practical exigency that arose, by such means as were readiest at hand, and could best be made available; and so they have advanced from clubs and stones to power-looms, steam-engines, and telegraphs, as in language from the rudest sigus of thought to such intricate and perfected instrumentali- ties as Sanskrit, Greek, and English. This is the usual and normal way in which the latent and unsuspected capacities of human nature are drawn out by the pressure of external cir- cumstances and trained by experience; and if the history of language has been different, the burden of proving it so de- volves upon those who hold the doctrine.
These are, if I am not mistaken, the most effective tests by which the work of every investigator of the origin of language may be tried. If ho mingles authoritative statements, from whatever quarter, with his inductive reasonings, or fails to recognize the results of historical linguistics in the establish- ment of the initial radical stage of language, he is out of har- mony with the whole present condition and spirit of linguistic science, and cannot expect to command the attention of schol- ars. If respecting the relation of language to thought, the order of genesis of the conception and its sign, and the nature of the primary impulse to utterance, he does not hold definite opinions and defend them by solid arguments, or if he passes lightly over these questions as of subordinate consequence, he will add little or nothing that is valuable to the enormous and constantly increasing mass of disquisition and discussion of the subject — and, in conclusion, it may be claimed that, if he takes the wrong side of these questions, he will never reach a sound and defensible theory. A theory, what we hold respect- ing the origin of language must always remain, since (as has been already pointed out) direct inductive reasoning cannot reach so far back in the history of language; but the elements of uncertainty in it may, with right views and a sound method, be reduced within very narrow limits.
VII. — On Certain Forms of the English Verb which were used in the Sizteenth and Seventeenth Centuries. BY THOMAS R. LOUNSBURY, INSTRUCTOR IN ENGLISH IN YALE COLLEGE.
It is a well known fact that Early English has three meth- ods of forming the present indicative of the verb: that of the South of England in which the Anglo-Saxon declension was closely followed ; that of the North in which the original ter- mination -th was changed into -s, and that of the Midland counties from which the modern forms have been derived. For a long period all these may be said to have existed side by side, though after the time of Chaucer at least, the suprem- acy of the last cannot be disputed. Still, the occasional use in the sixteenth century of forms belonging to the two first has been observed by several writers, and in particular has been pointed out in detail by Maetzner, who (in his Englische Grammatik, vol. 1. p. 220. ff.) shows that third persons plu- ralsin -th and -s, as they loveth and they loves, and second per- sons singular in -s, as chow loves, can be found even late in the seventeenth century. Maetzner calls attention to the fact that these plural forms seem to have escaped the notice of most commentators upon Shakespeare, who either silently change them into the singular or else explain them upon high scientific grounds. Thus in the edition of King Richard II. (Oxford, 1869), prepared: by W. G. Clark and W. A. Wright for the Clarendon Press series of text-books, the plurals draws and makes contained in the following lines found in Act ii. Sc. 3,—
“‘ These high wild hills and rough uneven ways Draws out our miles and makes them wearisome,”’—
are explained in the following manner: This is, say the edit- ors, ‘a construction which the Greeks called a cyijpa mpoc ro onpavepevov. These wild hills and rough ways blend, as it were, into one idea in the speaker’s mind, and he proceeds as if he had said ‘journeying over these hills and ways’ &c.”
96 " TR. Lounsbury,
The bearing of this whole question on several disputed pas- sages of Shakespeare is so direct, that it becomes a matter of some importance to know how general was the use of these forms during the period in which he flourished ; to ascertain to how great an extent they were employed, not only by that writer himself but by other authors of his time and of the time immediately preceding. Investigations of this kind are attended in this country with considerable embarrassment and trouble, from the perpetually recurring difficulty, not to say impossibility, of obtaining either copies of editions which ap- peared in the life time of the author, or exact reprints of the same, preserving every peculiarity of the original orthography, etymology, and syntax. Occasionally, indeed, some of these ancient forms are retained in editions, published in our own time, where the language has been modernized. Especially is this the case in poetry, as there the demands of rhyme or metre are frequently imperative. But necessarily no conelu- sion can safely be drawn from such occurrences save the fact that such forms were used by the authors in whose works they are found: the extent or peculiarity of use can only be inferred and even that with little certainty.
In the preparation of this paper I have examined the fol- lowing works or parts of works: the poems of Wyatt and Surrey; Ascham’s Toxophilus (1545) and Scholemaster (1570) ; Bishop Latimer’s Seven Sermons before Edward VI. (1549) ; Bishop Bale’s Play of Kynge Johan (about 1550) ; Udall’s Comedy of Roister Doister (before 1553); More’s Utopia, translation of Robinson (2d edition, 1556); Sack- ville’s Induction to Mirror for Magistrates, and Complaint of Henry, Duke of Buckingham; Gascoigne’s Steele Glas (1576) and Complaynt of Philomene (1576); Whetstone’s Remem- brance of Gascoigne (1577); Gosson’s School of Abuse (1579), and a short poem entitled Speculum Humanum (1576) ; Sidney’s Apologie for Poetrie (1595) ; Correspondence of Robert Dudley, Earl of Leicester, during his government of the Low Countries in the years 1585-6 ; Spenser’s Shepherd’s Calendar (1580), Fairy Queen (1590-6), and Virgil’s Gnat (1591); Bacon’s Advancement of Learning (1605); Mar-
On Certain Forms of the English Verb. 97
lowe’s Tamburlaine the Great, Dido Queen of Carthage, Hero and Leander, and minor poems; five of Ben Jonson’s plays contained in the folio of 1616; forty-four of the fifty-two plays of Beaumont and Fletcher ; and the whole of the Shakespeare folio of 1623. For some of the works here cited I rely on the excellent reprints of Mr. Arber; but the works of Wyatt, Sur- rey, Sackville, Bacon, Marlowe, and Beaumont and Fletcher, have been consulted in modernized editions, and all statements in regard to their usage are necessarily lacking in complete- ness and accuracy, so far as original editions are concerned.
Obviously in a discussion of this kind, the argument can only be weakened, at least it will not be strengthened, by the citing of illustrations at all doubtful. It has accordingly been made a point to discard all forms in which there should fail to be a strong preponderant probability, if not an actual cer- tainty, that the author designed to make use of the number and person ascribed to him. This has necessarily led to the rejection of all such passages—and they are exceedingly numerous—as the following: .“ Whom God and man does hate.’’ (Spenser’s Fairy Queen, book i. canto i. 18): “‘ My mis- tress and her sister stays for you,” (Shakespeare, Comedy of Errors, act i. sc. i.): “The priest and his wife sitteth with the Siphogrant,’’ (More’s Utopia). For the rule so constantly and confidently set forth in English grammar that two or more subjects in the singular connected by and require a verb in the plural has never been observed by English writers, at least down to the close of the seventeenth century. On the other hand the verb in phrases like “There hath been men,” or “There has been men,” is treated as a plural, though the frequency in later writers, of the singular form after there with a plural subject renders this a matter of doubt.
With this restriction an examination of the works above- named show in the case of the third person plural the follow- ing results, though it has not been deemed important, except in the case of Shakespeare, to specify the passages. The poems of Wyatt show six plurals in -s and four in -th ; those of Surrey, four in -s and none in -th; Ascham’s Toxophilus, twelve in -s and thirty-six in -th; his Scholemaster, none in -s
18
98 T. R. Lounsbury,
and four in -th; Latimer’s Sermons show nine in -s and six- teen in -th; Bale’s Kynge Johan, four in -s and twenty-six in
-th; Udall’s Roister Doister, two of each; Sackville’s poems, five in -s and none in -th; Robinson’s translation of More’s Utopia, none in -s and five in -th; Whetstone’s Remembrance
of Gascoigne, shows four in -s and one in -t& ; while Gascoigne
himself has none of either ; Gosson, has four in -s and one in. -th ; Sidney’s Apologie for Poetrie, none in -s and one in -th ;
Marlowe, three of each; Spenser, two in -s and three in -th ;
the Leicester Correspondence shows none in -s and thirty-six
in -th; Bacon’s Advancement of Learning, one in -s and six
in -th; Ben Jonson, four in -s and one in -tk; Beaumont and Fletcher, seventeen in -s and eight in -th; and the Shakespeare
folio of 1623, one hundred and sixty-eight in -s and forty-
six in -th.
There are some peculiarities about this list. Thus the Tox- ophilus of Ascham, published during his life, abounds in both the Northern and Southern forms, while his Scholemaster, a posthumous publication, contains no plural in -s and four only in -th. Robinson in his translation of the Utopia never em- ploys the ending in -s, not even in the third person singular, though that was then in very general use. It is hardly neces- sary to say that in all these works the modern forms are much the most numerous. __ | :
The following are the instances in which these forms appear in the Shakespeare folio of 1628. On account of the great number of them, only the verb and subject will be given, be- ginning with the third person plural in -s :—
Tempest. Cares. . . roarers, a. i.sc.1; Bones aches, a. iii. sc. 3; Wraths ... which... falls, ibid.; Lies . ... ene- mies, a. iv. sc. 1; Tears runs, a. v. sc. 1.— Two Gentlemen of Verona. Comes. . . praises, a. li. sc. 4.— Merry Wives of Windsor. There has been knights, a. ii.sc. 2. There has been earls, ibid. Swine eats, a. iv. sc. 2.— Measure for Meas- ure. Doubts ... makes, a. i. sc. 5; Knows... laws, a. li. sc. 1.— Comedy of Errors. Ships puts, a. iv. sc. 3 ; Bones bears, a. iv. sc. 4; Clamors . . . poisons, a. v. sc. 1.— Much Ado about Nothing. Events stamps,a. i. sc. 2; Comes... reckonings, a. v. sc. 4.— Love’s Labor Lost. Appears...
On Certain Forms of the English Verb. 99
tears, a. v. sc. 2; Wits makes, ibid.; Contents dies, ibid.; Parts . . . forms, ibid.— Midsummer Night’s Dream. Sighs - .. that... costs, a. iii. sc. 2; Lions roars, a. v. sc. 2.— "Merchant of Venice. Dealings teaches, a. i. sc. 3; Times puts, a. iii. se. 2; Locks which makes, ibid. — As You Like It. Des- tinies decrees, a. i. sc. 2; Likelihoods depends, a. i. sc. 3; Fools
. that makes, a. iii. sc. 5.— Taming of the Shrew. Winds that shakes, a. ii. sc. 1. Things that belongs, ibid.; Wants
. Junkets, a. iii. sc. 2; Stars allots, a. iv. sc. 5.— All’s Well that Ends Well. Dispositions . . which . . makes, a. i. sc. 1; Bones looks, ibid; Bloods . . . stands, a. ii. sc. 3; Addi- tions swell’s, ibid.; Twigs that threatens, a. iii. sc. 5; Oaths that makes, a. iv. sc. 2; Desires who . . . recovers, ibid. ; Letters which makes, ibid.; Letters . . that sets, a. v. sc. 3.— Twelfth Night. Does . . ..lives, a. ii. sc. 3; Some achieves, a. ii. sc.5; Goes. . . griefs, a. iii. sc. 4; Feet hits,
a. iii. sc. 4.— Winter’s Tale. Needs .. . hands, a. ii. se. 3; Seas hides, a. iv. sc. 8; Lies... ‘secrets, ibid. — King John. Spirits . . . craves, a.ii.sc. 1; Wrongs . . . draws,
ibid.; Reasons makes, a. iii. sc. 4; Tidings comes, a. iv. sc. 2.— King Richard I. Hills and ... ways ... draws ... and makes, a. ii. sc. 8; Lies two kinsmen, a. iii. sc. 38; Comes . . . gardeners, a. iii. sc. 4. — King Henry IV. Part I. Lives... men, a.ii.sc. 4. Part ZZ. Comes... swaggerers, (occurs twice) a. ii. sc. 4; Comes none, ibid. — King Henry V. Wrongs ... gives, a. i. sc. 2; Qomes the English, a. ii. sc. 4; Lips blows, a. ii. sc. 6.— King Henry VI. Part TI. Stands lords, a. i. sc. 4; Commendations as becomes, a. v. sc. 8. PartII. Humors fits, a. i.sc.1; Them... that enjoys, a. ii. sc. 4; Curses . . . turns, a. iii. sc. 2; Enters . . Citizens, (stage direction) a. iv. sc.5; Them... that knows, &. V. SC. 1; Intends . . . forces, ibid. Part UL. Sands . . . that makes, a. i. sc. 4; Passions moves, ibid.; Strokes. . . hews down and fells, a. ii. sc. 1; Coals that fires . . . and burns, a. ii. sc. 5; Minutes ... makes, ibid.; Hours brings, ibid.; Wounds that kills, ibid.; Commixtures melts, a. ii. sc. 6; Who shines, ibid.; The scatter’d foe that hopes, ibid.; Means that keeps, a. iii. se. 2; Riddles sorts, a. v. sc. 5; Shows, such as
100 T. R. Lounsbury,
befits, a. v. sc. 7.— King Richard III. Deeds... provokes, a. i. sc. 2; Sorrows bids, a. iv.sc. 1; After-hours gives, a. iv. sc. 3; Children quits, a. v. sc. 3.— King Henry VIII. Women ... makes, a. i. sc. 4; Points . . . drives, a. ii. sc. 4.— Troilus and Cressida. Knots... infects... and diverts, a. i. sc. 3; They .. that speaks, a. iii. sc. 2; Rhymes ... wants, a. iii. sc. 2; Farewells goes, a. iii. sc. 3; Trumpets sounds, a. v. sc. 9; Armies separates, ibid. — Coriolanus. They... follows, a. 1. sc. 4; Blows ... craves, a. iv. sc. 1; Bosoms seems, a. iv. sc. 4; Places yields, a. iv. sc. 7; Conspirators . . . kills, a. v.sc. 0 (stage direction). — Titus Andronicus. Sons speaks, a.i. sc. 2 (stage direction); Years wants, a. ii. sc. 1; Horse
. runs, a. ii. sc. 2; Parcel ... which dreads, a. ii. sc. 3; Begins . . . sorrows, a. v. sc. 4; Letters . . . which signifies, a. v. sc. 1.— Romeo and Juliet. Masks . . . puts, a. i. sc: 1; Rem- edies ... lies, a. ii. se. 3; Wits faints, a. ii. sc. 4; Thoughts which... glides, a. ii. sc.5; Legs excels, ibid.; Gossamers that idles, a. ii.sc. 5; Comes the Capulets, a. iii. sc. 1; Eyes... that makes, a. iii. sc. 2; Howlings attends, a. iii. sc. 3; Fears comes, a. v. sc. 3. — Timon. Instruments... that keeps, a. i. sc. 2; Needs .. . feasts, ibid. ; These owes, a. iii. se. 3; Drops pays, &. iii. sc. 4; Laws ha’s, a. iv. sc. 8; Gods ha’s, ibid. — Julius Cesar. Affairs... rests, a. v. sc. 1.— Macbeth. Natures lies, a. i. sc. 7; Words . .. gives, a. ii. sc.1; Heav- ens... threatens, a. ii. sc. 4; The times has been, a. ili. sc. 4; Means that makes, a. iv. sc. 3.— Hamlet. Some says, a.i. sc. 1; All that lives, a.i. sc. 2; Seems . . . uses, ibid.; Thanks as fits, a. ii. sc. 2; Brains ... puts, a. ill. se. 1; Depends and rests ... lives, a. iii. sc. 3; Sorrows comes, a. iv. sc.5; Houses... lasts, a. v. se. 1; Griefs bears, ibid. — King Lear. Regards that stands, a.i. sc. 1; Faces . . . stands, a. li. sc. 2; Germens... that makes, a. iii. sc.2; Means... graces, a. ili. sc. 7; Manners urges, a. v. se. 3.— Othello. Wars which . . . stands, a. i. sc. 1; Minerals that wakens, a. i. sc. 2; The young affects, a.i.sc. 3; Do’s.. . acquaintance, a. ii. sc. 2; Wars that makes, a. iii. sc. 3; Words that shakes, a.iv.sc.1; They . . that wins, a.iv.sc.1; Messengers... stays, a. iv. sc. 2; Eyes ... drops, a. v. sc. 2.— Antony and
a e.2e7 9 ®
On Certain Forms of the English Verb. 101
Cleopatra. Knaves that smells, a. i. sc. 4; Menacrates and Menas, famous pirates, makes, a. i. sc. 4; Needs . . . words, a. ii. ec. 7; The people knows, a. ili. sc. 6; Friends that does, a. ili. sc. 6; Rhymers ballads, a. v. sc. 2. — Cymbeline. Graces ... that charms, a. i. sc. 7; Springs . . . that lies, a. ii. sc. 3; Testimonies ... lies, a. iii. sc. 4; Seas brecds, a. iv. sc. 2; Follows the two young Leonati, (stage direction) a. v. sc. 4; Comes. . . staggers, a. v. sc. d.
The following are the third person plurals in -th : —
Comedy of Errors. Fits hath, a. v. sc. 1.— Merry Wives of Windsor. Successors ... hath, a. i. sc. 1.— Much Ado about Nothing. Wrongs doth, a. v. sc. 1.—Love’s Labor Lost. Hairs... hath, a. iv. sc. 83; Inches doth, a. v. sc. 2.— Midsummer Night’s Dream. Fogs which... hath, a. ii. sc. 1; Kindred hath, a. iii. sc. 1; Eyes doth, a. iv. sc. 1; Minds ... Witnesseth (?), a. v. sc. 1.— Merchant af Venice. Men ... doth, a. ili. sc. 2.— All’s Well that Ends Well. Wars hath, a. i. sc. 1. — Winter’s Tale. Encounters hath, a. i. se. 1; Changes hath, a. i. sc. 2; One three . . . hath, a. iv. sc. 3. — King John. Stones that ... doth, a.ii.sc.1; Arms... hath, ibid. ; Affections ... doth, a. v. sc. 2.— King Richard IZ. Wars hath, a. ii.sc.1; Them . . that . . hath, a. iii. sc. 2; Hours hath, a. iii. sc. 4. — King Henry IV. Part I. Beds of sweat hath, a. ii. sc. 8. Part II. Examples .. . hath, a. iv. sc. 1.— King Henry V. Spirits that hath, Prologue. — King Henry VI. Part I. Hath ... Frenchmen, a. ii. se. 2; Rea- sons bindeth, a. v. sc. 5. Part I. Dreams ... doth, a. i. sc. 2; Traitors hateth, a.iv. sc. 4. Part IZ. Cheeks... presenteth, a. ii. sc. 5; Looks doth, a. iii. sc. 2; Words doth, ibid.; Dislikes . . . doth, a. iv. sc. 1.— King Richard II. These . . . hath moved, a. i. sc. 1.— Troilus and Cressida. Contents . . . doth, a. i. sc. 2; Wounds doth, a. v. sc. 8. — Coriolanus. There hath... men, a. ii. sc. 2; Ministers that doth, a. iii. sc. 3; There hath. . . insurrections, a. iv. sc. 3 ; Examples . . . hath, a. iv. sc. 6; Spoils .. . doth, a. v. sc. 5. — Titus Andronicus. Hands hath, a. ii. sc. 5.— Romeo and Ju- liet. Griefs ... doth, a. iv. sc. 5.— Macbeth. Blows and buffets .. . hath, a. iii. sc. 1; Evils ... hath, a. iv. sc. 3. —
102 T. R. Lounsbury,
King Lear. The best and soundest ... hath, a. i. sc. 1. — Othello. Such things ...as doth, a. i. sc. 3. — Antony and Cleopatra. Contempts doth, a. i. sc. 2.
It is not impossible that some examples existing in the folio and in the other works examined may have been overlooked ; that in some instances the examples here given may have been misprints; that a few may be susceptible of a different explan- ation from that which implies the continued use of a plural in -s or-th. But conceding this, the number of illustrations will be little affected by any one of these possibilities or by all of them combined. The examination of the literature of the time is of course only partial. But assuming that a fuller in- vestigation will show results not materially different, it seems fair to draw the following conclusions:
First. That while some writers doubtless never made use of either the third person plural in -s or in -th, both were nev- ertheless considered legitimate and proper, just as at the pres- ent time the third person singular in -th, though never under any circumstances used by many, is admitted to be correct by all. |
Secondly. That the plural in -A was much more common about the middle of the sixteenth ‘century than towards its close: that during the latter half of that century it was dying out, and at the beginning of the seventeenth was almost en- tirely confined to the present tense of the verbs do and have. In connection with this point it is important to observe that in Shakespeare, with the exception of a possible plural in the Midsummer Night’s Dream, which is probably a singular, there are no plurals in -th but Aath and doth, with the excep- tion of King Henry VI. where the plurals bindeth, hateth, and presenteth are found. There is only one case of this plural to be found in Ben Jonson, so far as his works have been exam- ined. This ts seen in the line, “ The choicest brain the times hath sent us forth,” taken from the first act of ‘ Every man in his Humor,’ as it appears in the folio of 1616. In Gifford’s edition the ath is changed to have. The most marked excep- tions to the general conclusion that has been drawn occur in ‘The Leicester Correspondence,’ as published by the Camden Society, in which letters appear from the Earl of Leicester, and
On Certain Forms of the English Verb. 103
Thomas Dudley, and from the Queen’s ministers, Walsing- ham and Burghley. In this, besides doth and hath which are very frequently met with, are also found as plurals, continu- eth (Let. xvi.), keepeth (Let. xxi.), commeth, hopeth, writeth, falleth (Let. xl.), pleaseth (Let. Ixv.), goeth (Let. xcix.), groweth (Let. cili.), keepeth (Let. cxi.), termeth (Let. cxvi.), and impeacheth (Let. cxxxiv).
Thirdly. That there is no proof of the decay of the plural form in -s during the period under examination. Indeed, if there were any change at all the evidence of the facts so far collected seems to show that the use of it rather increased than diminished. Yet as it is most prevalent in the drama, and as the drama represents more fully and accurately the speech of common life than any other kind of writing, there is little doubt that the plural in -s, while much employed at this time in conversation, was disappearing from use in the literary language. Its infrequency in a work so carefully edited as the Ben Jonson folio of 1616, with its constant re- currence in one so carelessly edited as the Shakespeare folio of 1623, seems to bear out this conclusion.
In the second person singular the northern form in -s, as thou lovés, maintained its ground during this period along with the regular form in -st. Though the examples of this usage are far from being as numerous as in the case of the third person plural, there are still enough of them to justify the assertion that this form was regarded as allowable at least, if not perfectly proper. Some of the writers above referred to furnish no instances of this usage. Of those that do, Wyatt has the form twice; Latimer, three times; Gosson, four times ; Spenser, six times ; Ben Jonson, once; Beaumont and Fletcher, five times; and the following twenty-eight examples occur in the Shakespeare folio: — Likes, Zwvo Gentlemen of Verona, a. ii. sc. 7; Splits, Measure for Measure, a. ii. sc.2; Exists, ibid., a. iii. sc. 1; Calls, Taming of the Shrew, a. iv. sc. 1; Affects, Winter’s Tale, a. iv. sc.3; Disputes, Love's Labor Lost, a. v. sc. 1; Requests, ibid., a. v. sc. 2; Torments, Aing Rich- ard IT., a. iv. sc. 2; Thinks, King Henry V., a. iv. sc. 1; Got’s, King Henry VI. Part I, a.i.sc.4; Lends, ibid., Part ZZ, a. i. sc. 1; Starts, ibid., a. iv. sc. 1; Mistakes, ibid., a. v. sc. 1.
104 T. R. Lounsbury,
Requests, King Richard IIT., a. ii. sc. 1; Loves, Romeo and Juliet, a. i. sc. 2; Woulds, ibid., a. iii. sc. 1; Expects, ibid., a. ili.sc. 5; Counterfeits, ibid.; Girdles, Timon, a. iv. sc. 1; Puts, a. iv. sc. 8; Comforts, a. v. sc. 2; Revisits, Hamlet, a. i. sc. 4; Ha’s practised, King Lear, a. iii. sc. 2; Lusts, ibid., a. iv. sc.5; Looks, Othello, a. ii. sc. 3; Dafts, ibid., a. iv. sc. 2; Makes, a. v. sc. 2; Refts, Cymbeline, a. iii. sc. 3; Makes, ibid., a. ili. sc. 4.
There are a number of instances in the Shakespeare folio, in which the first person singular ends in -s, as it does to this day in the language of low life. They may possibly be mis- prints, but this is hardly probable. Still there are not enough of them to enable us to draw from them alone any satisfactory conclusion. It has, however, been deemed best to give the passages in which they occur, which are the following :—Make him proud to make me proud that jests, Love’s Labor Lost, a. v.sc.2; Nor God nor I delights in perjured men, zbzd.; I be- seech you, hear me who professes Myself your loyal servant, ... yet that dares Less appear so, Winter’s Tale, a. ii. sc.38; So I does, Troilus and Cressida, a.i. sc. 2; My name is Thomas Mowbray, Duke of Norfolk, Who hither comes, engaged by my oath, King Richard IT.,a. i. sc. 8; On me that halts and am misshapen thus, King Richard III, a. i. sc. 2; I for a Clar- ence weeps, so doth not she, ztd., a. ii. sc. 2. Such forms are common enough in later dramatic writers, where the lan- guage of uneducated men is represented; but this is nota peculiarity belonging to any of the above examples. Myself and thyself are also found sometimes as the subjects of verbs ending ih -s: but in this case it may be that self is looked up- on as a substantive.
There is no more singular characteristic of the inflection of the verb during the sixteenth century than the occasional ex- tension of the ending -th to the first and second persons singu- lar of the present tense, as, for illustration, in the following lines from Wyatt and Shakespeare:
“ But I that truth Hath always meant Doth still proceed to serve in vain.” ‘Cassio hath beaten thee, And thou by that small hurt hath cashier’d Cassio.”
On Mistaken Notions of Algonkin Grammar. 105
There are several instances of this usage, and it may, per- haps, be regarded as an indication of a tendency then existing to discard all the distinctive endings of the verb. Still little stress can be laid upon these examples. They are at present too few in number, and many of them too doubtful in charac- ter, to be regarded as justifying the formation of any theory. Closer examination and wider investigation may enable us to state with certainty what is the actual fact in regard to these forms; but such a statement it is not safe to make with our present knowledge.
VIII.— On some Mistaken Notions of Algonkin Grammar, and on Mistranslations of Words from Eliot’s Bible, gc.
BY J. HAMMOND TRUMBULL.
John Eliot’s version of the Bible in the language of the Indians of Massachusetts has been characterized as ‘a rich mine of Indian philology,” from which “a complete grammar and valuable dictionary might, with labor and perseverance, be extracted.”* Scholars like Pickering and Gallatin have now and then really worked a vein or two of this mine, with — moderate success; but for every such one there have been fifty who were content to glean a few surface-specimens and spare themselves all trouble of assay or analysis. The rich- ness of the mine considered, it is surprising that so much worthless ore has been brought out of it and that so much whiclt was intrinsically good has been made worthless in the smelting process to which it was subjected to prepare it for filling the molds of comparative vocabularies, for bracing up an unsound hypothesis, or for pinning together some lin- guistic structure which was not quite strong enough to stand alone. Ifan Algonkin placo-name is to be mis-interpreted, the mis-interpretation is usually made on the supposed author- ity of Eliot. When his version is referred to for the purpose
* Duponceau’s Notes to Eliot’s “ Indian Grammar Begun,” in Massachusetts Hist. Collections, 2d Ser., vol. ix. p. ix.
14
106—Ctit J. H. Trumbull,
of finding an Algonkin word corresponding to one in the English text, the chances are that an affix or formative is mis- taken for the rvot.
There are few writers on American languages who have not somewhere been led into error by relying on statements made on the alleged authority of Eliot’s Bible or of Zeisberger’s Grammar of the Lenni Lenape (Delaware) language. It is not surprising that distinguished European philologists, who could consult these authorities only at second-hand, have been thus misled. They are excusable for adopting and giv- ing currency to the false notions of Indian syuthesis, the worthless etymologies, and the mis-translations, which had received the endorsement of American scholars of high repute and passed unquestioned from this side of the Atlantic.
I propose in the present paper to call attention to a few of these errors, and to show that some of the best accredited dicta concerning the Algonkin languages rest on very slight foun- dations —or have no foundation whatever. They may be divided in two classes,—as they belong to the grammar, or to the vocabulary. Of the former, I mention first, —
The alleged existence of a definite article, in certain Algon- kin languages, especially in the Massachusetts aud the Lenni - Lenape.
Mr. Duponceau was the first to announce the discovery, in the Natick (Massachusetts) dialect, of ‘“‘a part of speech which had not been noticed by grammarians in the Indian lan- guages”. In a note appended to Pickering’s edition of Eliot’s Indian Grammar Begun (1821), he wrote as follows: —
“It is remarkable, that this language appears to possess a definite article, although no mention is made of it in this Grammar. This article is mo, contracted from monko, and properly signifies it... .. This pronoun when used as an article is still further contracted into m, which, when followed by a consonant, Eliot connects with it by the English short w, according to his method, and sometimes by shorte. Thus he writes metah, ‘‘the heart,’ which should be pronounced m’tah. It is evident, that the m stands here for an article, because the personal affixes ‘my’, ‘thy’, ‘his’, are n, k, and w; nuttah or n’tah, ‘my heart’, kuttah or k’tah, ‘thy heart’, wuttah or w’tah,‘his or her heart’. . . . In the translation of the Bible,
On Mistaken Notions of Algonkin Grammar. 107
this article frequently appears: Kesteah pahke metaH “ Create in me a clean heart”. Ps. li. 10.— Pohqui kah tannogki METAH “A broken and contrite heart.” Ibid. 17. Several words are also found in his [Eliot’s] Grammar, in which this article is prefixed, though not noticed as such. .... This article exists in several of the Indian languages,” &c. (pp. xiv. Xv.)
To this note was appended the copy of a letter received from Mr. Heckewelder, assuring Mr. Duponceau that “the article ‘mo’ for ‘a’ or ‘the’, which he had discovered in the language of the Naticks is the same in the language of the Lenape.”
In the translation of Zeisberger’s Delaware Grammar, pub- lished in 1826, the statement that “there is an article in the Delaware language” is repeated; and reference is made (p. 36,) to the translator’s discovery of this article ‘in the Massa- chusetts language.”’
Again, in the well known Mémoire sur le Systeme Grammat- tcal des Langues de quelques Nations Indiennes (Paris, 1838), Mr. Duponceau asserts that “les langues Algonquines ont Particle. . . . Les grammairiens Eliot et Zeisberger ne |’ont pas méme apergu, c’est pourquoi ils n’en ont pas parlé”’; but, “des Indianologues plus récens ont enfin découvert son existence”, etc. (p. 148).
In Mr. Gallatin’s “ Synopsis of the Indian Tribes’’ (1836), Mr. Duponceau is credited with “the discovery of an article mo; as m’hittule ‘a tree’ or ‘ the tree’,” Cp. 220) and allusion is made (p. 163) to “ the initial m often prefixed to the noun in the Knisteneaux and the Chippeway” languages, as “‘ seem- ing to corroborate the existence of a definite article mo, discov- ered by Mr. Duponceau in Eliot’s translation of the Bible.’’
And so the definite article, — unknown to Eliot and Zeis- berger, disbelieved in by ““M. Heckewelder lui-méme ... . jusqu’a ce qu’il fat convaincu du contraire par les recherches des philologues” ,— took its established place among the parts of Algonkin speech. :
Yet it may easily be shown that the m’ prefixed to certain ‘ classes of Algonkin nounsis not a definite article, — that it does not stand for mo,— that mo is not a contraction of monko, — and that monko does nof signify ‘it’, in Eliot’s Bible or elsewhere.
108 J. HA. Trumbull.
M6 or mo is put by Eliot (Ind. Grammar, 21) among “ad- verbs of denying’’, “‘sometimes signifying not’. Thus he writes mo teag and mo"leag ‘nothing’ (Isaiah xl. 17; xli. 17): mo teag ohtédéu ‘he hath nothing’ (Prov. xili. 4), and mo tea- guas ohtééu (Prov. xx.4). But he more frequently uses this particle as the sign of the preterit, to denote completed and terminated action or being, — that which was and ts not, — or as a substitute for the past tense of the substantive verb. It has this meaning in the verses cited from Eliot’s version by Mr. Duponceau, and in many others. ’Nnih or wnnth means “it isso’, and mo nnih (Genesis i. 15) ‘it was so’; wunnegen ‘it is good’ (Ps. lii. 9), and mo ahche wunnegen ‘ it was very good’ (Gen. i. 31) ; na mm pharisaé wosketomp ‘ there wae a Pharisee man’ (Juli iii. 1), and matta mo wosketomp ‘ there was not a man’, literally, ‘ not was man’ (Gen. ii. 5); wequaz [there is] ‘light’, and mé wequat ‘ there was light’ (Gen. i. 3), ne mo wequat ‘ that was [the] light’ (John i. 9); ken mo wut- tinneumin ‘thou wast a servant’ (Deut. v.17); a mo kesukod ‘there was a day’ (Matt. viii. 26). In avery few instances— nearly all of which occur in the first chapter of Genesis, at the beginning of Eliot’s work of translation — he employed the questionable synthesis ménk6 nnih for ‘it was so’ (vv. T, 9, . 11, 24, 30): ménké having been formed, apparently, from mé and ké, to signify ‘ was and continues to be’.*
Mr. Duponceau having mistaken the sign of tle past tense for a pronoun transformed the supposed pronoun into a definite article. But the office of the prefixed m’ (as in Mass. m’tah ‘heart’) was just the reverse of that of a definite article. Primarily a negative or a privative —always undefinitive — it was used not with all nouns but with a few only,— with the names of the body and its members, of articles belonging to or generally associated with the person, of terms expressing rela-
* The particle ké6 or koh denotes continuance or progression. As an auxiliary, it refers to a past time action or being not yet completed or terminated, — when what now is ‘began to be’ or ‘once was’— or affirms present as related to prior ac- tion or being. Eliot occasionally employs it for the verb substantive, as in Job . xiv. 10, kah uttoh ké wutapin? ‘and where ts he’; noh koh mé, noh koh, kah noh paont ‘who was, and ds, and is to come’ (Rev. iv. 8); and ken nukoh [ = noh koh], kah ken nukoh mé, kah ken padan, ‘thou who wast, and art, and art to come’ (Rev. xi. 17).
On Mistaken Notions of Algonkin Grammar. 109
tionship, and some others: and it served to divest these of all personal and individual relation or appropriation. For cxam- ple, when an Indian spoke of ‘body’ or ‘person’ he usually employed a possessive pronominal prefix,—‘ my body’, ‘thy body’, ‘hts body’ (Mass. n’hog, k’hog, w’hog) : but if he found it necessary to speak of ‘ body’ or ‘heart’ in the abstract, or divested of its natural associations, he substituted for the pos- sessive and personal the negatire and impersonal prefix, m’. WM hog (mithhog, Eliot,) denotes ‘body not mine, yours or his’ —some body, regarded as without appropriation or personal relation: m’tay (métah, El., mtee, Zeisberger,) ‘heart’, not my heart (tay), nor yours (k’tay), &c.”
Another modern discovery in Algonkin grammar was that of a vocative case of nouns. Eliot had stated (in his Indian Grammar Begun, p. 8) that nouns in the Massachusetts lan- guage are ‘“‘not varied by cases, cadencies and endings,” — except that ‘‘there seemeth to be one cadency or case” of animale nouns, corresponding to the Latin accusative. But Zeisberger found terminations in the Delaware which “ex- press the vocative”. He gave several examples of these in his Grammar of that language (p. 37), and Mr. Duponceau, in his Notes to Eliot’s Grammar (p. xiv), pointed out ‘different terminations of the same word, in various parts of Eliot’s translation of the Bible’”’,—of which “the termination zn in the vocative singular and unk in the vocative plural’ could not, he thought, be accounted for consistently with Eliot’s _ “positive statement that substantives arc not distinguished by cases.” He cited: Zeisberger’s authority for the fact that * the Delaware has a vocative case, which generally ends in an.” Mr. Gallatin (Synopsis, p. 173) repeats: ‘ There is a vocative case in some at least of the Algonkin-Lenape languages, ter- minating, in the singular of the Delaware, in an, and of the
* Howse (Cree Grammar, p. 245) has pointed out the mistake of ‘some writers who have considered the clement of me- (and w- or we-) prefixed to certain nouns, as equivalent to the European Article.” This element, he says, is found in the Cree “‘only in the names of the body and its parts, . . . in those expressing rela- tionship, as ne-gduwee ‘my mother’, me-gduwee ‘a mother’ &c., —with a very few others.”
110 J. H. Trumbull,
Massachusctts in an; in the plural Delaware, in enk, ‘when coupled with the pronoun our.’ (Zeisberger, p. 99.) The same termination eunk is used generally for the second person plu- ral in the Massachusetts.” Dr. Pickering in his paper on ‘Indian Languages,” in the Encyclopedia Americana, adopted Zeisberger’s statement that “in the Delaware, in two cases, the vocative and ablative, there is an inflection,”’ — the former being “expressed by the termination an’, &c. On so excel- lent authority the Delaware vocative in an and the Massa- chusetts vocative in in and eun/ have been received, without question, into the Algonkin grammatical system.
Without affirming or denying the existence of a vocative form in some Algonkin languages, but considering only the evidence on which it has been engrafted on the dialects of Massachusetts and Delaware, — I assert that Eliot’s Bible will be searched in vain for a vocative singular in in or for a “ ter- mination ewnk used generally for the second plural plural’, and that among the examples given by Zeisberger there is not one of a noun in the vocative case ending in an or enk, but that all these examples are verbs or participles of the suffix- animate form or, as Heckewelder (in his Correspondence with Duponceau, p. 416) termed it, the “ participial-pronominal- vocative form.’”” The supposed Delaware vocative in an is a verb in the conditional (subjunctive) mood, 2d pers. singular of the subject with 1st pers. singular of the object, and the form is nearly the same in the Massachusetts language as in the Delaware. Zeisberger’s “ Nihillalian, O thou my Lord!” is, literally translated, ‘Thou who ownest (or, art master of) . me’, i. e. ‘Thou as owning me’; ‘‘ Pemauchsohalian, O my Saviour!’ is ‘Thou as giving life to me’, &c.* Eliot has nowaan ‘thou that sayest’ (thou as saying), and mdéskowdan ‘thou that makest thy boast of’, Rom. ii. 23; ken wadohkean ‘thou that dwellest’, Ps. xxx. 1, &c. The supposed vocative in -enk, in the Delaware, is the 2d person singular of the sub- ject with the 1st person plural of the object; “ Nihtllaliyenk, O thou our Lord!”’ (Zeisb. Gram. 116) is ‘Thou who ownest
* Howse, Cree Grammar, pp. 310, 311, has shown that Zeisberger’s vocatives “have verbal endings” and are all “in the Subjunctive or Subordinate mvuod.”
On Mistaken Notions of Algonkin Grammar. 111
(or, as owning) us.” When the subject is plural, and the ob- ject in the 3d person or the verb intransitive, Eliot uses a participle or verbal formed from the second person plural of the subjunctive by adding -ish: e. g. kenaau wonkandgish ahtomp ‘ye that bend the bow’, Jerem. |. 29; kenaau quoshé- gish ‘ye that fear’, Ps. cxv. 11; kenaau kdkobsudgish ‘ ye deaf’ (i. e. ye as not-hearing), kenaau pogkenumdgish ‘ ye blind’, Is. xlii. 18. But this form is not distinctively vocative, for it is found with the pronoun of the first person, as in I. Thess. iv. 15, 17, nénawun pamontamdgisy kah apedgisy ‘ we which are alive and remain’,and Hebr. iv. 3, nenawun wanamptamagisH ‘we who believe.’
In his search for vocatives in the Massachusetts language, Mr. Duponceau was “surprised to find different terminations of the same word, in various parts of Eliot’s translation of the Bible’, some of which he was at a loss how to explain, “ other- wise than by the conjecture that our author might have had recourse to different Indian dialects in translating.” (Notes on Eliot’s Grammar, xiv.) He gave the following examples :—
Wuttaunou Zion, *‘ Daughter of Zion’. Lament. ii. 8. Woi Jerusalemme wuttaunin, ‘O daughter of Jerusalem’, v. 18.
Woi kenaau Jerusalemme wuttaun EUNK,‘O ye daughters of Jerusalem’, Solom. Song, il. 7.
Kah ompetak wutténeu, ‘And she bare a daughter’,—as Mr. Duponceau translated it, but which in the verse cited (Gen. xxx. 21) stands for the words “and afterwards she bare a daughter”. He mistook the adverb ompetak ‘after- wards’ for a verb meaning ‘to bear’, and wutténeu (mis- ‘printed, wuttaneuw) —a verb in the 8d pers. sing. indicative present (aorist), meaning ‘she bare a daughter’, for a noun ; remarking that the termination “eu in the accusative governed
* When Duponceau wrote his Mémoire sur le Systeme Grammatical &c., pub- lished in 1838, he had learned that the terminations which Zeisberger regarded as belonging to the vocative were verbal forms; but he was still persuaded that the words receiving these forms were nouns not verbs. ‘‘ Au lieu du vocatif”’ — he says (AMfémoire, p. 159) —on emploie une forme verbale qu’on applique au nom substanti/; elle varie selon les nombres. Ces formes, qu’il est inutile de préciser davantage, tiennent la place du verbe étre: ainsi, lorsqu’on dit: O mon dieu! c’est comme si on disait: O toi gui es mon dieu!” &e.
112 | J. H. Trumbull,
by an active verb’? “cannot be accounted for’, —which is quite true.
Of the three forms Wuttaunoh, -Wuttaunin, and Wuttaun- eunk, he remarked that “ the first is correct.” So it is, —but not for the reason he assigns, that “‘it is a proper nominative of this word.” If it were a nominative, it would stand in apposition with Zon, and the translation must be ‘his (or her) daughter Zion.’ But the termination -oh, with the pro- nominal prefix wu’, marks the governing noun (as in the Hebrew construct form), —‘the daughter of.’
Wuttaun-in is a proper nominative, its termination marking it as 2 noun-animate indefinite, ‘a daughter’ or ‘any daughter.’ That this termination -in is not “in the vocative singular” may be shown by reference to other verses in which the same form of the word occurs,—as a nominative, in Micah vii. 6, wutlaunin ayeuhkonittué ohkasoh ‘the daughter, against her mother’, and in Numbers, xxxvi. 8, nishnoh wuttaunin noh ahtunk oht6onk ‘every daughter that possesseth an inherit- ance’, —and after a governing preposition, Levit. xii. 6, wutch wunnaumondin asuh wuttaunin ‘for a son or a daughter.’
The termination of Wuttauneunk,—“‘unk in the vocative plural”, as Mr. Duponceau regarded it,—is that of a collective noun, without reference to case or person. Wuttaun-eiinie, in the verse cited, means ‘the daughters’ collectively, the daugh- terhood; 80, in Judges xxi. 21, we find Shiloe wuttauneunk ‘the daughters of Shiloh’, the Shiloh daughterhood. Nouns of this form are of frequent occurrence in Eliot’s version. Thus we have wémonwk wmweemattinneunk ‘love ye the brother- hood’, I. Peter, ii. 17; wutwehinneunk ‘the fathers’ or the fatherhood, Numb. xxxi. 26; I. John, ii. 13; wunnaumonain- neunk ‘the children’ collectively, Luke, xvi. 8.*
We are now in @ position to sum up the evidence on which
* Molina (History of Chili, American translation, vol. ii. p. 303) mentions sim- ilar nouns collective in the Araucanian language, and classes them with abstract terms formed by adding gen (representing the verb ‘to be’) to adjectives or verbs. Thus, “instead of saying pu Huinca ‘the Spaniards’, they commonly say, Huin- cagen ‘the Spaniolity’,—‘amén cutagéen ‘your trio’, that is, you other three,” &c. See Pickering’s notes on Edwards’s Observations &c., in Mass. Hist. Coll., 2d S., x. 120.
On Mistaken Notions of Algonkin Grammar. 113
philologists have agreed to recognize a vocative case-ending of nouns in the Massachusetts language. We have only Mr. Duponceau’s misinterpretation of two words employed by Eliot. He mistook the termination of a noun indefinite for that of the vocative singular, and made a vocative plural out of a noun collective.
The fact that’ no Algonkin language has an independent verb-substantive —a fact denied by Cass and Schoolcraft, and which has been questioned by many writers on American lan- guages, — may now be regarded as established. Much of the discussion on this subject has turned on the precise mean- ing of the phrase by which Eliot translated ‘‘I am that I am”, in Exodus, iii. 14, — Nen nuttinnitn nen nuttinniin.
Heckewelder, in reply to a question from Duponceau, could only say that this ‘‘ could never be a literal translation of the text,” and that “if it means anything, it must be ether “I am aman, I am a man,” or “I do so, I do so.”” Duponceau, “after much consideration and study of the subject, inclined to the opinion that Mr. Heckewelder is right in his last con- jecture” (Notes on Eliot’s Grammar, xlii.); and in his Mé- moire (p. 195) he unhesitatingly accepts this translation, as deciding the question of the existence of the verb ‘to be’ in Algonkin languages. ‘On a trouvé’’— he writes, — “le moyen de la décider d’une maniére qui ne laisse plus de doute. On a cherché dans la Bible indienne d’Eliot, la traduction du célébre passage: ego sum qui sum (Exod. iii. 14), et on a trouvé nen nuttinniin nen nuttinntin ; on a cherché aussi dans le méme livre, la traduction du passage ego [sum] sicut vos, dans l’épitre de saint Paul aux Galates, ch. iv., v. 12, et on a trouvé nen neyane kenaau; on a envoyé ces deux passages ainsi traduits aux missionnaires les plus instruits dans les langues Algonquines, et ils ont trouvé que le premier signifiat: je fais, je fais; 6t le second: nous nous ressemblons ou je vous ressemble.”
Duponceau’s dictum — founded, as we have seen, on a guess of Heckewelder’s — was authoritative. Since the publication of the Mémoire, “I do, I do,” has been the accepted transla- tion of Eliot’s nen nultinniin nen nuttinniin,—and has been
15
114 J. H. Trumbull,
pointed to as a proof of the poverty of American languages.* No one apparently has taken the trouble to re-examine the text or to analyze the synthesis Eliot employed, — though this might easily have been done without other help than his version of the Bible itself affords. |
To supply the want of a verb-substantive every Algonkin dialect has several verbs to express the where and the how of being,— modal and conditioned existence. Those which most frequently occur in Eliot’s version are, —
1. Ohteau ‘it has itself’, the intransitive form of ohtau, ‘he has’, ‘owns’, ‘possesses’. Used only when the subject is inanimate: e. g., ayeuonk ohteau ‘the place is’, Judg. xviil. 12; pish ohteau ‘it will be’, Gen. xvii. 18; suppositive or con- junctive, ohtag,‘ if Cor, when) it is’, Matt. v.14. Chippeway, “‘ até, there is of it; it is’ (Baraga) ; ‘ atta, to be” (School- craft).
2. Appu (Chip. abi, Baraga; Cree, apt, abt, Howse ;) ‘he sits’, ‘is at rest’, — hence ‘ he remains’, ‘ abides’ ; and so, ‘he is’ or ‘ continues to be’ — in a state of rest or inactivity is im- plied. With an adverb of place, wutappin; as na wutappin ‘he sat down there’, Ruth, iv. 1, ‘he was there’, John, v. 5; yeu wutappin ‘he is here’, John, vi. 9; toh kutappin? ‘ where art thou ?’ Gen. iii. 9.
3. Ayeu (Chip. ahyah, Jones ; tau ‘he is’, Schoolcraft — who has given a paradigm of it, as the Chip. verb ‘ to be’,) ‘he is in some place’ designated ; ‘he is there’, John, xi. 30; hence, ‘he dwells’ or ‘inhabits’. Noh ayeu kah appu ‘he dwells and abides’, Job, xxxix. 28: imperfect, nut-ai-up ‘I was there’, Acts, xi. 5: conjunctive, dytit, atyit (Chip. akyod, Jones), noh dyit machemotagtt ‘ he that inhabiteth (i. e. as inhabiting) eternity’, Is. lvii. 15. The 2d person conjunctive (dycan, Eliot,) of this verb is found in various Algonkin versions of the Lord’s prayer; “ who art in Heaven”,.Moheg. ne spum- muck oieon (Edwards) ; Old Abnaki, spemkik aitan; Old Pas- samaquoddy, spemkik éhine (Vetromile, from Rasles?), Ma-
* Mr. Farrar intrgduces it (Chapters on Language, p. 54), to illustrate of the ‘primordial and unbroken barbarism of the North American Indians”’, etc., — and again, in his Lectures on Families of Speech, p. 183, to show the “almost imbecile deficiency of abstraction,” which characterizes Amcrican languages.
On Mistaken Notions of Algonkin Grammar. 115
reschit, — eyane (Ib.) ; Chip. tshpimingk eatiin (Testament), &c. Eliot’s version omits the verb; ‘“‘ Our Father in Heaven.”
4. ’Nnth, Unnth, ‘it is so’ or (aorist) ‘it was so’, Gen. i. 7, 9,15. Eliot uses this word for the phrase ‘it came to pass’ or ‘comes to pass’. Imperat. 3d pers. sing., ne naj, ne natch, ‘be it so.’
do. Neane, Neyane, ‘it is like’ or ‘the same as’; as in the passage cited by Mr. Duponceau, Galatians, iv. 12, nen neyane kenaau ‘I [am] as ye [are]’. The imperative 2d pers. plural (with Ist person sing. object) and the adverbial form are found in the same verse: unniyegk neyanté ‘be ye as I [am]’. The conjunctive participle nedunak (or -nag) used as a noun, ‘that which is like’ or ‘ being like’, stands for ‘likeness’, ‘ ap- pearance’, ‘ color’, ‘fashion’ of, &c.: nedunag yeu muttaok * the fashion of this world’, I. Cor. vii. 31.
6. Wuttinniin ‘he is of the kind of’ or ‘is such as’. This verb cannot be exactly translated in English. It expresses the relation of an individual to a species or a class, the appro- priation of its subject to an object expressed or understood, a belonging-to, —not merely external likeness or relation. It is conjugated in the present indicative as follows:
nuttinniin, 1 am of the kind of, I am such as,
kuttinniin, Thou art of the kind of,—such as,
wuttinniin, He is of the kind of, —such as.
It occurs not unfrequently in Eliot’s version; e. g., Prov. Xxiii. 7, nedne unnantog ut wuttahhut, ne wuttiniin ‘as he thinketh in his-heart so is he’, i. e., of that kind is he; I. Sam. xxvii. 11, ne pish wuttinniin ‘so will be his manner’, i. e., that will he-be-of-the-kind-of; and Is. xxiv. 2, neanwit wuttinneumin, ne wuttinniin wussontimomun ‘as with the servant, so [of that kind is] his master.’ In Exodus, iii. 14, nen nuttiniin nen nuttinniin means, literally, ‘I myself am of the kind of I my- self am of the kind of’ or ‘I am such as I am such as’ — Eyo sum talis qualis ego sum, for the “‘ Ego qui sum”’ of the Vulgate and the “‘I am that Iam” of the English text. Marked em- phasis is given to the pronoun of the first person by using both its forms (independent and prefixed) with each verb,—nen n’-, ego ipse’. |
116 J. H. Trumbull,
In the first edition of Eliot’s Bible (1663), ne ‘that’ stands in the place of the second nen. This was corrected on revis- ion, because ne, the inanimate demonstrative, cannot properly be employed to denote the subject or object of a verb animate.
The very general use of transitional forms of conjugation, in which the pronoun of the object as well as of the subject is combined with the verb, has led some distinguished writers on American languages to infer that the Indian verb cannot be divested of its pronominal suffix. Edwards (Observations on the Muhhekaneew Language, p. 13) states, that the Mohe- gans ‘‘never use a verb transitive without expressing both the agent and the object, correspondent to the nominative and accusative cases in Latin. Thus they cannot say, ‘I love’, ‘thou givest’, &c. But they can say, ‘I love thee’, ‘thou givest him’, &c. viz. Nduhwhunuw ‘1 love him or her’; nduhwhuntamin ‘T love it,’ &. Mr. Cass, in an article on the Indian Languages, in the North American Review (for Jan- uary, 1826; vol. xxii. p. 80) made a similar statement; ‘‘The pronouns, actor and subject, are associated with the verb. One is prefixed, and the other is suffixed; and the latter is generally inseparable in its form. The active verbs cannot be used without this personal association. An Indian cannot say J love, I hate, I fear, abstracted from the operation of the verb upon the object.” Mr. Bancroft repeats this, sub- stantially, in his observations on the synthetic character of the American languages (Hist. of the U. States, vol. iii., 12th ed., p. 261): “An Algonkin cannot say J love, I hate; he must also, and simultaneously, express the object of the love or hatred. . . . Each active verb includes in one and the same word one pronoun representing its subject, and another repre- senting its object also.”
Dr. Edwards was wrong —as the very examples he used for illustration show: but his error is less apparent because it is restricted to a denial of the use, by the Stockbridge Mohegans, of transitive verbs without a pronoun-objective. Mr. Cass’s denial extends to all active verbs and to all Algonkin Jan- guages. Nothing can be farther from the fact. There is no Algonkin dialect in which an Indian may not say ‘I love’ or
On Mistaken Notions of Algonkin Grammar. 117
‘TI hate’, without denoting by a pronominal suffix the objeet loved or hated. He has for this the choice of three or four verbs; (1) strictly intransitive, affirming the existence of affection, ‘I am in love’ or ‘I feel lovingly’; (2) animate- active intransitive (the adjective-verb form, as some gramma- rians term it) —affirming the exercise of affection, —‘I am loving’ or ‘I am a lover’; (8) active-transitive absolute, — the forms of which vary (but not by a pronominal suffix) as the implied object of affection belongs to one or the other of the two great classes of Indian nouns, animate and inanimate, the former class including not only allliving beings but many tn- animate objects held in special regard by the Indians. These forms serve, respectively, for the affirmations ‘I love some person, animal or object of the class animate’ (a bow, a kettle, or tobacco, it may be,) or ‘I love something’ not of that class. Either may receive in addition to the formative proper a pro- nominal suffix,— but each is complete without it.
It is true that a savage’s conception of ‘love’, subjective or objective, differs from that of a Christian, and missionaries by whom the Algonkin languages have one after another been reduced to writing have not all agreed in the selection of the word which comes nearest to the meaning of the English verb to love or the French azmer. Eliot in Massachusetts and Roger Williams in Narragansett’ employed a verb the precise meaning of whose root (wém, waum) is not ascertained. The Roman Catholic missionaries have generally adopted another, more common among the northern and western Algonkins, from the root sdg, saug, ‘to cling’ or ‘hold fast’. With this explanation, the following examples are enough to show how ‘I love’ may be expressed in the principal languages of this family:
Massachusetts: na@-womantam, v. i., ‘I love; am love- minded.’ To verbs of this form, “expressing a disposition, situation, or operation of the mind”, Zeisberger assigns a special conjugation (the third) in his Delaware Grammar (pp. 00, 89). In the Chippeway, they end in -endam (Baraga, p. 154). Examples may be found on almost every page of Eliot’s version; e. g. musquantam ‘he is angry’, literally ‘bloody-
118 J. H. Trumbull,
minded’ ; nut+ishantam ‘I hate’, ‘I feel hatred or abhorrence’ ; no-wabesuontam ‘I fear’; nut-chepshontam ‘I am frightened’, &c. All these verbs may be used, with the appropriate suffix, as transitive inanimate, ‘he loves it’, ‘he hates it,’ &c. Chippeway: nin sdgia (Baraga), ne saugeaw ‘I love a per- son’ (Schoolcraft),— but Baraga, more exactly, translates ‘I love him, her, or it’, remarking that, in this form, “the object upon which acts the subject of these verbs, is always contained in the verb itself.” (Otchipwe Grammar, 200.) With the pro- noun: o sdgian (Bar.), 00 zdhgeahn (Jones), ‘he loves him’. Cree: ne-sdkehewdn ‘I love some one’ (indeterminate) ; ne sdkechegan ‘I love something’ (indefinite) ; ne-sdkehewdywissin (adj.-verb, active-intransitive.) ‘I am loving’ or, as Howse analyzes it, “I am love-someone-ing’’. Cree Grammar, 105,114. Northern Algonkin of Canada: nz sakidjike‘I love’. This form is ‘“‘sans régime, exprimant un sentiment’; ne sakiton means ‘I love it’; nz saktha, ‘I love him’.* Micmac: “ kejalmei, j’aime,” is placed by Maillard (Gram. Mikmaque, p. 56) among verbs “qui ne regoivent aucun régime dans leur acception’’,— “ verbes sans régime’’.
Passing now to the consideration of another class of errors, — those which concern the vocabulary, including mistransla- tions, false analyses, and mistakes in the identification of words in Eliot’s version corresponding to those in the English text,— our first example shall be taken from that “immense monument of historical research,” the Mithridates of Adelung and Vater. In the third part of this work Professor Vater gave (8te Abth., p. 388) a list of words in the language of the “ Na- ticks, from Eliot”. One of these words is ‘‘ Chequikompuh’’, standing as the Natick name of the ‘Sun’. Balbi, borrowing these words from the Mithridates reproduced them in his Atlas Ethnographique (Tab. xli.), where Chequikompuh appears as ‘¢ Massachusetts or Natick” for ‘Sun’. Now the Massachusetts name of the Sun—wnepduz (Narr. nippdwus, R. Williams, ) occurs at least a hundred times in Eliot’s version. In Joshua,
* Etudes philologiques sur quelques Langues Sauvages de l’Amérique (Mon-
. treal, 1866), pp. 50, 55, 60.
On Mistranslations of Words from Eliot’s Bible, gc. 119
x. 13, for the words: “ the sun stood still’’, of the English text, we have “‘nepduz chequnikompau.” Mistaking the order of the words, Prof. Vater sets the (mutilated) verb instead of the noun against the word ‘ Sonne’ of his vocabulary.
In the same volume of Mithridates (2te Abth., p. 349), the learned author notes the resemblance of “‘ cone’’, as a New Eng- land word for ‘Sun’, to the Tatar kun. Unfortunately, cone (as Roger Williams wrote it ; kan of Eliot and Cotton) means ‘snow’, not‘sun’. The same error is found in an earlier work of Vater’s, (Untersuchungen iiber Amerika’s Bevdlkerung, Leipzig, 1810, p. 51), whence more than one comparative philologist has taken it as evidence of the relationship of American and Asiatic languages. :
A similar mistake was made by Mr. Duponceau, in a list of words “selected from Eliot’s translation of the Bible,” and incorporated by Dr. Pickering with the verbal index to his edition of Eliot’s Indian Grammar Begun.* Among these we find Sohstimdonk, as the Massachusetts word for ‘Forest.’ Eliot’s version has for ‘forest’, towohkomuk, (literally, ‘desert place’, ‘wilderness’,) from which was formed the adjective touohkomukque. Sohstimdonk, a verbal from sohsumm ‘it shines forth’, was employed for the translation of the word ‘glory’, —literally, ‘a forth-shining’. In Isaiah, x. 18, for ‘the glory of his forest’ we find wut-touohkomukque sohsiméonk ‘his forest glory’, the English order of words being inverted, in accordance with the laws of Algonkin synthesis. Hence, doubtless, Mr. Duponceau’s mistake.
Of all explorers of Eliot’s ‘rich mine’ Mr. Schoolcraft was perhaps least successful. In the first volume of his magnum opus, “Information respecting the History &c. of the Indian Tribes,” he gave (pp. 288-299) a vocabulary of nearly 300 words ‘extracted from Eliot’s translation.” How the ex- traction was effected, and what is the real value of the vocab- ulary as a contribution to comparative philology, a few speci- mens will show.
The first word is Manitoo, for ‘God’, with a reference to Gen. xxiv. 26 (by misprint probably, for 27). This should
* Massachusetts Historical Collections, 2d Scries, vol. ix. p. liii.
120 J. H. Trumbull,
be Manit, and should have been accompanied by the remark that it was not usually employed by Eliot as a name of the Supreme Being. Mr. Schoolcraft was wrong in saying (p. 287) that in Eliot’s version “the words God and Jehovah appear as synonymes of Manito” or Mantt. Those names were generally —‘ Jehovah’ was always transferred to the Indian text; not translated by Manz. The form Manito (or -w) combines with the noun the representative of the verb-sub- stantive, and means ‘ Manit is’. The plural, mantttwog (or -t60g), is used for ‘ gods’ of the English version; as in I. Cor. vili. 5, manitwog monaog ‘gods many.’
‘612. Husband, Muntmayenok”,— for which Gen. xxx. 15 is cited. In that verse, keneemuntmayeuonk nahsuk stands for “thou hast taken away my husband”. Mr. Schoolcraft mistook the verb for the noun; and rejecting the pronominal prefix —and something more, for nee belongs to the root, — he made, by help of a misprint, munémayenok !
“13. Nunawmonittumwos. Wife. Job, xxxi. 10.” For ‘ wife’ Eliot has mittamwussis or mittamwas. Nun-naumon is ‘my son’, which Mr. Schoolcraft somehow contrived to mix up with nummittamwos, ‘my wife’, in the verse cited.
‘“47. Kon, Bone.” The references are to Job, xxx. 30, xxxi. 22. In the former verse, nuskonash stands for ‘my bones’; in the latter, wutch wuskonit for ‘from its bone.’ The root wskon ‘bone’ cannot be used without a prefix ; nuskon ‘my bone’, wuskon ‘his bone’, or (indefinite) muskon ‘any bone’. There is no such word as Kon.
“77, Noonshoonum, Boat. Acts, xvii. 16,” — an error for Acts, xxvii. 16, where namshanun — a verb in the first person plural (with its prefix) — means, “‘ we came by boat”. The noun m’shon (mushwn, mishon) ‘a boat’ is used in John vi. 22, Acts, xxvii. 80, &.
“79. Omoquash, Sail. Acts, xvii. 17,” — another misprint, for Acts, xxvii. 17, where pungwémuhquash ‘ quicksands’ happens to stand next to nwkakinnumwog ‘they strake sail’ (lit. ‘they let it down’). The word for ‘sail’ is sepdghunk ‘that which is stretched out.’ |
“81. Hunkaueehtaeaug, Oar. Ezek. xxvii. 6.””> The man-
On Mistranslations of Words from Eliot’s Bible, gc. 121
gled remains of suttuhhunkanéhteaog, ‘they made thy oars’, — a causative verb formed from wuttuhhunk ‘oar’ or ‘ paddle’.
“172. Taskookau, Thistle.” No reference is given; but as taskuhkau is the 3d pers. sing. indic. present, of a verb mean- ing ‘to tread upon’, and as in 2 Chron. xxv. 18, taskuhkauau kigkéunogkohquohhouoh stands for “he trode down the thistle’’, we may infer that Mr. Schoolcraft again mistook verb for noun.
“225. Nunneem, Pigeon. Levit. xv. 6.” The word ‘ pigeon’ (Mass. eouskuhwhan) does not occur in the verse cited, but it may be found in vv. 14 and 29 of the same chapter, as the object of the trans. anim. verb neemunau ‘he takes’. This verb also occurs in v. 6 of ch. xiv. in the form wunnemunoh (‘he takes it’). ‘ Nunneem” is, I suspect, a misprint for Wunneem — the first two syllables of wunneemunoh.
And so on, — through the whole vocabulary. Prefixed to it are some observations on the ‘“‘ Massachusetts Indians’ and their language, in which we find a curious mistake, — unsur- passed by any in the vocabulary itself. The language of Eliot’s version is said (p. 287) to be ‘“‘a well characterized dialect of the Algonkin’’, but Eliot found in it, “it appears, no term for the verb to love, and introduced the word ‘womon’ as an equivalent, adding the Indian suffixes and inflexions, for person, number, and tense.”
Mr. Schoolcraft ought to have known that this word was not of Eliot’s invention or introduction. The intransitive, womantam ‘he loves’, the animate-active intrans. (or adjective, verb) womoausu ‘he is loving’ or ‘a lover’, and the trans. animate womonau ‘he loves (some one)’, with their derivatives, are much used in Eliot’s version; but forms from the same root may be found in Roger Williams’s Indian ‘ Key’, printed in 1643, twenty years earlier: e. g., waumatsu ‘loving’ (p. 140); cowdmmaunsh [in Eliot’s orthography, ka-womon-sh] ‘I love you’; cowammatinuck ‘he loves you’; cowdmmaus ‘ you are loving’ (p. 8), &c. Earlier yet, in Wood’s rude “ Nomen- clator” (appended to New England’s Prospect, 1684), we have “wawmauseu, an honest man” (for ‘a kindly disposed’ or a ‘loving’ man), and “ noewammawause, I love you.’
16
122 J. H. Trumbull,
This story of Eliot’s manufacture of an Indian verb ‘to love’ from the English word ‘woman’ will always find believers. It belongs to the same class with that of the mistake made in the translation of Judges, v. 28, “‘ The mother of Sisera looked out at a window and cried through the lattice’’, — where, it is said, for ‘lattice’ Eliot used an Indian word which really means ‘eel-pot’. This story has been printed scores of times, —and will continue to be printed, for it is ‘too good to be lost’. There are only two exceptions to be taken to it: (1) that the Indian eel-pot was of ‘ lattice work’ and that its name would not be a mistranslation of ‘lattice, though hardly a sufficient translation; and (2) that in the verse in question Eliot did not translate the word ‘ lattice’ at all, but transferred it from the English to the Indian text, adding only the locative suffix: “ papdshpe lattice-ut, through the lattice.”
Eliot’s work has not been appreciated, even ‘by scholars, as highly as it deserves to be. Mr. Howse—the author of a valuable ‘ Grammar of the Cree Language” (London, 1844,) —remarks in his Introduction, that “from the circumstance of Eliot’s having translated the Bible into the language of the Massachusetts Indians, or rather from his being the re- puted translator, (which 18 a very different thing,) it has been erroneously supposed that he was thoroughly versed in their language:’’ Mr. Howse was “ much inclined to think, however, that grammatically considered, it is an imperfect perform- ance,” and that, “if correct, it was formed only by the assist- ance of a half-breed interpreter.”’ A half-breed interpreter co-operating with the good Apostle to the Indians, in Bible- work, in puritan Massachusetts, and before 1660!
But “the most unkindest cut of all’? at the Wunneetupana- tamwee Up-Biblum was given by a chip thrown from Max Miiller’s German workshop. This eminent scholar, in a paper (first published in 1862) on the Abbé Brasseur de Bourbourg’s translation of the Quiché Popul Vuh,* mentions “ the transla- tion of the Bible in the Massachusetts language” as a specimen of picture-writing, and informs his readers that “ the verses from
* Chips from a German Workshop, vol. i. (1867), p. 320. The list of symbols stands between quotation marks, but Prof. Miiller does not give his authority for the statement.
Contributions to Creole Grammar. 123
25 to 32 in the thirtieth chapter of Proverbs are expressed by ‘an ant, a coney, a locust, a spider, a river (symbol of motion), a lion, a greyhound, a he-goat and king, a man foolishly lift- ing himself to take hold of the heavens’. No doubt these symbols would help the reader to remember the proper order of the verses, but’? — observes Prof. Miller, and I shall not venture to differ with him on this point, —* they would be perfectly useless without a commentary or without a previous knowledge of the text.”
IX.— Contributions to Creole Grammar. BY ADDISON VAN NAME,
LIBRARIAN OF YALE COLLEGE.
The Creole dialects which have grown out of different European languages grafted on African stock, though inferior in general interest to even the rudest languages of native growth, are in some respects well worth attention. The changes which they have passed through are not essentially different in kind, and hardly greater in extent than those, for instance, which separate the French from the Latin, but from the greater violence of the forces at work they have been far more rapid, and, what constitutes the peculiar interest of the case, the languages from which they have sprung are still living and are spoken side by side with them. Under ordinary conditions these changes proceed at so slow a pace as to become appreciable only at considerable intervals of time, but here two or three generations have sufficed for a complete trans- formation. The process has to be sure been mainly, but not altogether, one of decay; the extent of the loss has made some compensation necessary, and we find, if not many new forma- tions, numerous instances of old material put to new uses. It is scarcely necessary to remark that these languages are not _ of mixed blood, half African and half European, for languages
124 A. Van Name,
do not mingle so readily as races. Even in the Creole vocab- ularies, the proportion of African words is very small. They have borrowed in general less from this source than from other foreign sources; less, for example, than the Creole French from the Spanish, a fact sufficiently accounted for by the great confusion of dialects among the imported slaves. Redpath enumerates in Hayti no less than thirty different tribes, and Oldendorp found in the Danish islands representatives of an almost equal number. Still more remote must be the influ- ence of African on Creole grammar. It is rather in the pho- netic structure of the Creole,.in the dislike of an accumula- tion of consonants, the preference, especially marked in the Negro English of Surinam, for a final vowel, that such influ- ence may with more likelihood be traced. The simplification - of grammatical structure, on the other hand, which is charac- teristic of the Creole, arises rather from the general conditions of its growth than from the nationality of those who speak it. Of the causes which have contributed to the formation of these dialects the chief are: first, the mature age of the slaves, who were brought from Africa at a time of life when the vocal organs are no longer flexible, and when the intellectual effort necessary for the mastery of a new language is even under the most favorable circumstances very considerable, and here quite out of the question; secondly, the fact that they constituted the great body of the population, the whites being in a minor- ity seldom as large as one-fourth. The language spoken by the first generation of tlacks was a broken French or Spanish, as the case might be, which, in the course of time, developed into a well defined Creole. It is a matter of surprise at first view that while the French Creole is so widely spread, in the Spanish islands, Cuba, San Domingo, Porto Rico, and Mar- garita, we find no Creole; but the difference in the relative numbers of the two races, the African and the European, affords a ready explanation; the blacks are here outnumbered by the whites. Pichardo, in the preface of his Diccionario provincial de voces Cubanas, 2 ed., Habana, 1849, makes the following statement: “A corrupt dialect is spoken in all parts of the island (Cuba) by the Negroes born in Africa,
Contributions to Creole Grammar. 125
which is uniform, no matter what the difference of race, and is retained through life, unless they came very young.” He describes it as ‘‘a, distorted and mutilated Spanish, without concord, number, declension, or conjugation, without strong r, final s or @, frequently exchanging ii for fi, e for t, g for v, &c.; in fine,a jargon the more confused the more recent the arrival. It can nevertheless be understood by any Spaniard, apart from a few words, common to all, which require trans- lation.” This description accords nearly enough with the Creole Spanish of Curacgoa to show that we have here the be- ginning of proper Creole, but, for the reason given above, it has failed of development. Pichardo adds that the Creole Negroes, i. e., those born on the island, all speak the Spanish.
While these are the general conditions under which the Creole dialects have been formed, there seems to have been a difference in tlhe readiness with which the several languages have taken on the Creole character. The French has taken root most easily, and is found not only in all the West India islands either now or formerly in the possession of France, but in some, ec. g., Nevis and Montserrat, where its presence cannot be thus accounted for. The Spanish Creole is found only in the island of Curacoa and its dependencies, Aruba and Bonaire, which were colonized by Spain, but for more than two centuries have been in the possession of the Dutch. This isolation from the influence of the mother tongue has not un- likely had a favorable influence on the growth of the Creole, since we find in Surinam, under circumstances somewhat similar, the only English Creole which deserves the name. The greater number and fullness of the vowels in Spanish, as compared with the French, which give the syllables a structure more nearly resembling that of the African languages, by making the Spanish easier to acquire, may have been less favorable to the Creole tendency, just as the fact that the English is already so thoroughly creolized in its grammar has undoubtedly been an obstacle to further progress in that direction.
The better to exhibit the nature of the Creole tendencies, and to show the working of them under a change of circum-
126 A. Van Name,
stances and upon different material, I have attempted to sketch very briefly the principal features of Creole grammar. The survey includes five dialects of the French Creole and the Spanish, to which I have added for the sake of comparison brief notices of the Dutch Creole, and the Negro English of Surinam. For the study of the Portuguese Creole, spoken by one tribe of Bush Negroes on the Upper Surinam, no printed materials exist. At this distance from the field in which the languages are spoken, very small addition to the stock of posi- tive knowledge concerning them could be expected, and I have not been successful in all cases in obtaining the material which has been already collected. The more important part of the printed works I have perhaps had, and also in the case of three of the French dialects, the Spanish, and the Dutch Creole, the opportunity of gathering, from one or more indi- viduals who spoke them, new material, so that though neces- sarily very incomplete the results reached may not be altogether without value.
FRENCH CREOLE.
This is not only the most widely spoken, but also in a phi- lological point of view the most interesting of the dialects. “It is spoken not only by the Negroes and the colored people, but also by most of the white natives of Hayti, St. Bartholo- mew, Guadaloupe, Deseada, Dominica, Martinique, St. Lucia, St. Vincent, the Grenadines, Grenada, Tobago (in one dis- trict only), and Trinidad.” (Mr. J. J. Thomas, in Tribner’s Lit. Record, Dec. 31, 1870.) To this list should be added Montserrat, Nevis, part of St. Martin’s, and a considerable por- tion of the city population in St. Thomas. Besides this cen- tral territory it occupies in the east Mauritius, and in the west a part of Louisiana. The wider geographical limits of this dialect have given room for the growth of some varieties which are not without interest. The following is a list of the dialects included in the present sketch, with the authorities for each, printed and oral.
1. Hayti. Manuel des habitans de Saint-Domingue, par S. J. Duceur-Joly, Paris, 1802. Tome il. (pp. 282-355) contains a Vocabulaire Frangais et Créole, according to the author the
Contributions to Creole Grammar. 127
first ever published, followed (pp. 357-391) by Conversations Créoles. This work is of especial interest as it furnishes in some sort a measure of the more recent progress of the Creole.
Guide to Hayti, edited by James Redpath, Boston, 1861, contains (pp. 181-185) a scheme of the Creole conjugation, and some general statements.
I have had also the opportunity of repeated conversations with a young man, who between the age of seven and four- teen (he is now eighteen) lived in Hayti, and has a good command of the Creole.
2. Trinidad. Theory and Practice of Creole Grammar, by J. J. Thomas, Port of Spain (Trinidad), 1869, 8vo, pp. 184; the most extended and valuable work on the grammar of the language.
3. Martinique. Les Bambous: Fables de La Fontaine, tra- vesties en patois Créole (by M. Marbot). Nouv. éd. Fort-de- France (Martinique), 1869. 8vo, pp. 144.
Victor Schelcher, in Des Colonies Frangaises, Paris, 1842, gives (pp. 417-434) a collection of proverbes et locutions négres, which represent the dialect either of Martinique or Gauda- loupe, or perhaps both, not taking account of minor differen- ces. Mr. Thomas also, in his grammar, makes an occasional observation on these two dialects.
4. St. Thomas. For this dialect my authority has been a young man, Frederico Antonio Camps, recently from St. Thomas. He was born in Havana, but since his sixth year has lived in St. Thomas, and speaks, besides the French, also the Dutch and Spanish Creole, all three with great readiness. He has also visited nearly all the West India islands, and can speak from personal observation of the limits of the several dialects.
5. Louisiana. Here my materials have been gathered mainly from three individuals. Two of them until the recent war were slaves, one in the parish of St. Charles, about thirty miles west of New Orleans, the other about the same distance east, while the third had lived mostly in New Orleans. Of the first two the Creole was the native language, and one of them, up to the time when he left the South, spoke only that.
— 128 A. Van Name,
For a few facts respecting the Mauritius Creole I am in- debted to a notice of that dialect by Mr. Walter Besant, in the Atheneum of Dec. 31, 1870.
While the Creole of each dialect approximates to a uniform standard, there is, as we should expect, far more of variation than in languages which have had a more natural growth. Especially the contact with the French, which has never en- tirely ceased, has been a disturbing influence and a source of corruption, for the standard of Creole purity is to be sought at the farthest remove from the French, where the Creole ten- dencies are most fully developed. From the presence of these intermediate forms, representing different stages in the Creole progress, a brief outline like the present must necessarily be a very incomplete statement of the forms actually in use.
In the matter of orthography, there being no settled usage among the few writers of Creole, I have tried to keep in view the fact that the Creole is purely a spoken language, and final consonants which have become silent I have usually not writ- ten, especially where they convey to the eye (like the plural s) distinctions which no longer exist. I have also taken the liberty of modifying the orthography of the authorities used, in order not to separate by a difference of spelling, forms which are identical in sound; but in doing this, there has been, I hope, no loss of accuracy. Apart from the use of ch and j to represent the English sounds of ch and 7 in choose, join, the occasional use of &, and of 6 for the open sound of 0 (like ow in ought, but somewhat shorter), only the ordinary resources of French orthography have been employed.
PHONETIC CHANGES.
1. Vowel Changes, of which the following are the more im- portant. The Creole has a special dislike of the indistinct e muet, and where the syllable is not dropped altogether, sub- stitutes for it usually 7; thus dimain (demain). Not unfre- quently it suffers attraction from the vowel of the following syllable; thus /évé (lever), soucouré (secourer) ; sometimes the dialects follow a different course; thus Mart., Trin., dow- boute (debout), elsewhere diboute; Mart. bouwsowin, elsewhere
Contributions to Creole Grammar. 129
bison. Before v it may pass into ow; thus douvant (devant), chouval (cheval) Louis. goil. |
UY and wi usually become i; thus pli (plus), ri (rue), la (luc), dibri (du bruit) ; but touyé (tuer), Trin. choulé (reculer), elsewhere Kilé ; Trin. jiré (jurer), Mart. jouré, Louis. zouré ; Hayti couzine (cuisine), elsewhere kizine; lowile (huile).
Eu and its orthographic variation ew at the end of a sylla- ble become é; thus difé (du feu), pé (peut) ; but where a fol- lowing r has become silent, é; thus pé (peur). Before a con- sonant it is €é ore; as nef (neuf), bef (baeuf).
Oi sometimes retains its French sound; as in dibois (du bors), but more commonly becomes ove or oué; thus douetie (doigt), chouési (choisir) ; sometimes e, as drette and drouette (droit).
A final vowel preceded by a nasal consonant, n or m, fre- quently becomes itself nasal. This tendency is strongest in Trinidad ; thus aimén, menén Caimer, mener) ; it appears also in Mart., as mennein (mener), nein (nez), and in Louis. after — nm mouillé (which is not heard as a consonant, but only makes the preceding vowel nasal) ; thus ganien (gagner), conien (cog- ner), peinien (peigner), panien (panier), and a faint nasaliza- tion in some other cases, as limein and laimé (aimer). Even in Hayti, it is found, as in most of the dialects, in the pers. pron. moin (moi). In this tendency to nasalization we may perhaps recognize an African trait, since it has found its way into the Spanish, English and Dutch Creole, where a final or m frequently has lost its consonant power, making the pre- ceding vowel nasal. In one class of words, however, the vowel has lost the nasal character which it possessed in French, namely, in most French verbs in -ndre, thus vanne (vendre).
Consonant Changes. French ch and 7, and g before e and -i, in Louisiana, especially in the parish of St. Charles, and in Mauritius, frequently exchange the sounds sh and zh for s and z; thus, St. Charles, gappé (échapper), cose (chose), cumin (che- min); Mauritius, civé (cheveur), gaque (chaque). Before an a or o vowel the change docs not always take place; thus chaud, chien, charbon ; some words, as ¢apeau (also pronounced cha- peau), appear to be in a state of transition. The change of7
180 A. Van Name,
and g (before e and 7) to z is more uniform; thus, St. Charles z6rdi (aujourd’ hui), fromaze (fromage) ; Maur. zardin (jar- din). In Hayti and Trinidad, g of the termination age passes into the vowel 7; thus fromai, baggai Cfromage, baggage). Only one of my Louisiana authorities made the above changes at all regularly, the others only in individual words.
S before an 7 vowel (1, u, &c.) passes occasionally, in Louis. jana into sh (Fr. ch); thus dichic (sucre) ; in chongé (songer) the change is produced by assimilation to g (zh) in the follow- ing syllable. English ch (tsh) and 7 (dzh) appear in the Creole, developed out of a palatal (and in some cases a dental) mute followed by an i vowel. This change is most frequent in Trinidad, where itis found in chuite (cuite), chinze (quinze), chéchin (quelqu’un), and other words which elsewhere are free from it. In Louisiana but a few words are thus affected ; ‘thus chiré (tirer), chué (tuer), chilotte (culotte). It is found in all the dialects in one word chumbo or chémbé (from tiens bon or tiens bien), and in St. Thomas this is the only instance of its occurrence that I have detected. In the remaining dialects also the examples are few. The corresponding sonant j ap- pears in all the dialects in jéle (gueule), and in some other ‘words less uniformly.
These sounds, ch and j, we shall distinguish in the present paper from the French ch and 7 by the use of a heavy-faced type. In the Spanish Creole also the same sound of 7 will be similarly represented.
# is a weak letter which suffers more or less change and loss in all the dialects, but on the whole shows most weakness in Trinidad, and least in Louisiana and St. Thomas. Asa final it is-heard faintly in some words in Louisiana, but appar- ently in none of the other dialects. In Trinidad when it fol- lows a labial and is itself followed by any other vowel than o it passes into ow; if followed by o it may be lost altogether ; thus pout (prix), foté (frotter). In Hayti and Martinique it shows very nearly the same weakness. 2 sometimes suffers’ transposition ; thus dromi (dormir), crobeau (corbeau), Louis. ; rade (hardes), Hayti and Mart. ; radi (hardi) ; deré (dehors), Mart.
Contributions to Creole Grammar. 131
Contrary to the general tendency of the Creole as. well as of the French to cast off final consonants, the Creole has in some cases recovered a final consonant, especially ¢, which in French is heard only when the following word begins with a vowel; thus, ratte (rat, the ¢ of which does not regularly pass to the following vowel), Hayti, Mart., St. Th.; Jette (lait), Louis., Hayti; douwette (doigt), Trin., Hayti, Louis.; diboute, Louis., | douboute, Mart. (debout); canote, Hayti; valette, Mart. In Mart. we find also trope (trop). To this same preference for a stronger ending is due the choice of the feminine instead of the masculine form of several adjectives; thus coute (court), drette or douette (droit).
The Creole has an apparent dislike of an initial vowel, and in nouns almost always avoids it by prefixing some form of the article. In verbs the same tendency sometimes shows itself; thus, aimer is in Louisiana laimé or limein, in Hayti (DuceeurJoly) haimé; ouvrir in Hayti is louvout.
In compound words, especially in verbs of three or more syllables, a prefix preposition frequently falls away ; thus rété (arreter), touné (retourner), mandé (demander).
GRAMMATICAL FORMS.
ARTICLES. The French definite article has entirely lost its force in the Creole, though it remains attached to many nouns as an inseparable prefix. In its place the Creole has made of the demonstrative adverb /@ (from Lat. ilac, as Je from ille,) a postpositive article, which however retains somewhat more of the original demonstrative force than Je, and is therefore not used in all cases where the French employs the definite article. To this prevailing usage the dialect of Mauritius forms an ex- ception. Here the French definite article, though an insepa- rable prefix of many nouns as in the other dialects, has also retained its position and force, but has only a single form li (le) for both genders and numbers.
The indefinite article, also less used than in French, is yon in Trin., Mart. and St. Th.; ein (with the sound of ezn in hein or of 272 in vin) in Louis. and Mauritius; youne (Duceeur-Joly, nion) in Hayti. The numeral ‘one’ isin Trin., Mart., St. Th.,
132 A. Van Name,
yone; in Louis. and Hayti it is not distinguished from the in- definite article.
Both the indefinite and definite articles are invariable in respect to gender, but the latter, in some of the dialects, by composition with the personal pronoun third plural forms a plural which will be considered hereafter.
Nouns. With the loss of gender in the article, grammatical gender has disappeared also in the noun. Difference of sex is expressed in some cases by independent words, though the number of such words employed is much smaller than in French, and less in some dialects than in others; thus nomme (homme), femme, a few words denoting human relationships, as fré, sé (frére, seur),and afew names of animals. In the case of most animals a word denoting sex is prefixed, usually male, fimelle, but sometimes a more specific word; thus bouc- cabouite ‘he-goat’ (Trin.), macou-chatte ‘tom-cat’ (Trin., St. Thos.). In Trinidad, where the female has borne young, maman is used in place of fimelle; as maman-chatte ‘she-cat’. In Hayti, besides vache, we have bef-lette (beuf @ lait) ‘cow’. Feminine terminations have preserved their vitality only in a few nouns; thus négue, negresse; milate, milatresse; cousin, cousine, in all the dialects; dansé (danseur), fem. dansése in Trinidad ; but in Hayti and St. Thomas, dansése, and in Louis- lana, danse, are each of common gender, and we must distin- guish by prefix, nomme dansése, femme dansése. This ending ése has in Trinidad supplanted in two or three cases the end- ing were; thus lavandése (lavandiére), which is found also in St. Thomas. Somewhat similar is Trin. amise (amie). In Hayti from codenne ‘turkey-cock’ (cog-d’inde) we have a feminine codine ‘ turkey-hen.’
Plural inflections have disappeared from the Creole as they have in great measure from the French, where the ordinary plural ending s is heard only when the following word begins with a vowel. In the Spanish Creole also, and in the Negro English of Surinam, the plural has been lost, although neither the Spanish nor the English ending s shares the weakness of the French. Practically the spoken French, which alone has influenced the Creole, distinguishes the plural by the article
Contributions to Creole Grammar. 133
les when definite, des when indefinite. The Creole gives up the distinction in the case of indefinite nouns, and for definite nouns employs in Trinidad and Martinique a construction which is purely French; it places, namely, the weak demon- strative ces before and Ja after the noun; thus, nomme la ‘the man’, ces nomme la ‘the men.’ But in Hayti, St. Thomas, and Louisiana, we have a more original method; the pers. pron. 3. plur. (eux) is added to la, giving in St. Thomas yo la, in Hayti layo, and in Louisiana layé or lai; thus nomme layé ‘the men’. It is a very interesting fact that in the Spanish and English Creole, and for the most part in the Dutch Creole also, we have the definite plural formed in a similar way, by means _ of the personal pronoun.
Among the most curious features of the Creole are the numerous cases of agglutination of the French article with a following noun, the same which we have in Fr. lierre from Lat. hedera. Nouns having an initial vowel offer the readiest attachment to such a prefix, and very few such nouns are without it. The desire to escape the initial vowel-sound seems in fact to have been one of the chief motives of it. We find thus employed both the indefinite and definite article, and the latter both in the masculine and feminine, singular and plural. The choice is of course mainly determined by the frequency with which one or other form has occurred in con- nection with the noun, but partly also by phonetic considera- tions. The different dialects also have individual peculiarities. Instances of such a union of the indefinite article with the noun are not common; nomme (homme) is found in all the dialects ; nanné (année) in Trinidad and Louisiana (but Hayti, Mart., lanné); name (dme) Trin., Mart., St. Th.; ninime (énigme) Trin.; namé (armée) Mart. ; nouvrier (ouvrier), nab (arbre), népingue (épingle), nejui Caiguille) Louis. (else- where usually zépingue, zégui or zéjui). Ducceur-Joly gives in several cases two forms, as nomme, zomme; népron, zépron ( éperon); and in one case three, lain, nain, zain (ain). In none (oncle), for which in Trin. we have mounonc, as in Eng. nunele, n comes from the poss. pronom. adjective, mon oncele.
In the case of the singular definite article the distinction
134 A. Van Name,
- between the masculine and the feminine vanishes before an initial vowel, and the Creole in these composite forms shows no preference for one gender above the other; thus largent (argent), louse Cours), lidé (1dée), lowile (huile). Before an initial consonant the case is different, and here, while the feminine da is very frequent and found in all the dialects, the masculine li (le) appears only in Mauritius; thus lapl: ( pluie), laline (lune), laté (terre), and (Maur.) lichien (chien). The masculine article is avoided here doubtless because it would coincide in form with the pers. pron. 3. sing. Ji (lut). Mr. Thomas gives a list of seven masculine nouns which in Trini- dad have the fem. article prefixed; as, /a restant ‘ remainder’. In the partitive form du, the masc. article appears in the fol- lowing words in most if not all the dialects; divin (vin), dithé (thé), difé (feu), dirt, dourt (riz) ; and in Louisiana in quite a number of others; as, distc or dichic (sucre), dichou (chou), dibri (bruit), &ce. Of the fem. sing. partitive the only in- stances observed are dlo or dolo (eau), and (in Louis.) dilouile (huile); before a consonant the additional syllable required, dila (de la), would be too burdensome.
The plural article is of frequent occurrence before an initial vowel, to which the final consonant only of the article attaches itself; thus zami (amis); zombi (ombres) ‘ghost’; Louis., Maur., zozo, Hayti, St. Th., Mart., zowézo (orseaua) ; zhé (yeur). + So far as the Creole form is concerned the z might be referred equally well to les or to des; in most cases we may admit a joint influence, with some words les, with others des being oftener used. We must also add doubtless in some cases the pronominal adjectives mes, ses, so that zamz, for example, rep- resents les amis, des amis, mes amis, &c. In Louisiana a few nouns take the full prefix des; thus dézef (auf); dézo (08) ; dézoi (ove) ; déra (rat). Two or three examples will illustrate the difference of form which in the case of some words pre- vails in the different dialects and sometimes within the same dialect. Thus we have for Fr. étoile, in Louis. latuelle and létoile, in Hayti and St. Th. zétuelle; for Fr. euf, in Hayti zef, zé,in Mart. zé, (Scheelcher, zef), in Louis. dézef, zé; for Fr. rat, in Louis. déra, in Mart., St. Th., Hayti, ratte; for Fr. 08,
Contributions to Creole Grammar. 135
in: Louis. dezo and zos, elsewhere zo. Nouns having an initial vowel show in some dialects a preference for the plural, in others for the singular article. The following enumeration is only approximatively correct, and an examination of the entire vocabularies of the dialects might give a quite differ- ent result. The preference for the plural is most strongly marked in Hayti. In Ducceur—Joly they outnumber the singu- lars in the ratio of 10 to 1 (40 and 4), but additions obtained orally reduce the ratio to that of 4 to 1; in Thomas’s Gram- mar (Trinidad) the ratio is 2 to 1 (30 and 15), and about. the same in St. Thomas; in Louisiana I have observed about an equal number of singulars and plurals, and in “Les Bam- bous” (Martinique) the singulars are in excess, 35 to 20. A comparison however of a definite list of words found in all the dialects would show less divergence. The instances in which the fem. sing. Ja is attached to an initial consonant are most numerous in Martinique (40 in “Les Bambous”) and Louis- _iana, fewest in Hayti and St. Thomas.
The French article when it has thus become a constant element of the Creole noun loses, of course, not simply its definite force but also its power of marking number and gen- der; thus, in Louis. déra, when indefinite, is either ‘rat’ or ‘rats’; ein déra, dé déra, ‘one rat’, ‘two rats’; but déra la, ‘the rat’, déra layé, ‘the rats.’
ApJecTivEs. ‘There is,” says Mr. Thomas, “a distinct though ill sustained attempt at gender inflection, especially in the case of adjectives describing the qualities of human beings.” The number of such inflected adjectives is however exceedingly small and confined apparently to certain deriva- tives in which the force of the termination is still felt. Mr. Thomas instances: (a) French gentilics in -ais, -aise, Creole é, ése; thus anglé, anglése, fouancé, fouancése; but potijé écossé, tlandé, are not usually inflected in Trinidad, evidently because they are less used and the distinction is less impor- tant. In St. Thomas these are all inflected, but not the less strongly marked feminines of americain, italien, &c.: (b) Ad- jectives in -eur, -euse, Creole é, ése; thus, flatté, flattése: (c) Adjectives in -2, -ine; thus cochin, cochine, (coquin, coquine): (d) Adjectives in -z, -se; thus -malhéré, malhérese.
186 A, Van Name,
In one or two cases I have found parallel forms elsewhere, e. g. coguin, coquanne, in St. Thomas, but in general the dis- tinction is lost, and the adjective practically without inflec- tion. )
Where the French has distinct forms for both the masculine and feminine adjective the Creole has chosen sometimes the one, sometimes the other as the common representative. The following appear in all the dialects in the masculine form; bon, f6 (fort), grand, gros, Trin. goud, gras, gris, lou (lourd), nef (neuf). The following are uniformly in the feminine: belle, coute (courte), drette, douette (droite), frette, fouette (froide), latde, nette, soude (sourde). Others are divided in form; thus, blanc in all the dialects except that of Hayti, where blanc is ‘white man’ and the adj. ‘white’ is blanche; pouése (Epais) only in Hayti, elsewhere épé; dow and douce in Martinique, elsewhere only dou; molle, Trin. and St. Th., mou, Hayti and Louis.; frat, Trin. and St. Th., fraiche, Hayti and Louis. ; toute, Trin., Hayti, tow, Louis., in Mart. and St. Th., both forms; longue, Trin. and St. Th., long, Hayti and Louis. In the following cases a different meaning is assigned to each of the two forms occurring in the same dialect; Trin. fin ‘ fine’, fine ‘slender’, in Hayti and Louis. both forms are used indifferently, in St. Th. and Mart. fin; Trin. sec ‘curt, crisp’, chéce ‘dry’ (not wet), St. Th. see ‘dry’, chéce ‘dry chips’, Hayti chéche ‘thirsty’, Mart. chesse ‘ thirsty’, Louis. sec ‘dry’; pitt, frequently shortened to ¢z, is ‘small’, but in St. Thomas we have also pitze ‘child’.
Mr. Thomas observes that where a noun and adjective have been taken from the French not separately but in combination, the adjective may retain its French form, and instances tabe oudnde ‘round table’, chandelle oudmaine ‘Roman candle’, the adjectives having elsewhere the form oudnd, oudmain.
CoMPARISON OF ADJECTIVES. The comparative is formed as in French by ple ( plus) and moénce (moins), the latter not often used. The objects compared are connected generally by passé, sometimes by qui or qué (que). Where passé is employed the positive alone is sufficient; thus, pli grand passé ga or grand passé ga, in all the dialects. In place of passé
Contributions to Creole Grammar. 137
we may have in Hayti and Louis. gui, in St. Th. qué (que). The comparative of bon is in Trin. méié or pli bon; in Mart. we have more commonly the double comparative pli mzé, in St. Th. pli méié, in Louis. pli bon, pli méié, in Hayti miyér, pl miyér, pli bon. The superlative is cither not distinguished from the comparative or is expressed by some such method as the following: (St. Th.) ga c’est pli méié passé toute, or qué toute.
PersonaL Pronouns. 1. sing. Trin., Mart. moin ; St. Th., Maurit., mo; elsewhere two forms are employed, one for the subject when not emphatic, the other for the emphatic subject and the object; in Hayti (Ducceur-Joly) mo, emphat. moué, (Redpath) mon, m’, emphat. moué, (oral) moin, m’ ; in Louis- iana mo,emphat. moin. The final vowel is made nasal through the influence of the preceding m.
2. sing. Trin. ow; Mart. to or ou; St. Th. ow; Hayti (Duc.- Joly) to, emphat. towé, present usage ow; Louis. to, emphat. tot, or ou ; Maurit. to, vou.
8. sing. (both masc. and fem.) i, Trin. also ’é.
1. plur. now. 3
2. plur. Trin., Mart. 26¢; Hayti ou; St. Th. ow or 2dt; Louis. vou, vouzét. |
8. plur. Louis. yé, elsewhere yo.
These are derived from Fr. moi, toi, lui, nous, vous, vous autres, cur.
The choice of the accusative instead of the nominative forms of the personal pronoun is characteristic also to a great ex- tent of the other Creole dialects. In the French itself we have the beginning of this usage in the employment of mot, tot, luz, eux, in the emphatic repetition of the subject. The motive for this change lies in the fact that the subject pronoun came gradually to be regarded as a necessary adjunct of the verb, and with the loss of independence was also weakened in form. The process has recommenced in the Creole, where moi, tot, are, When they stand before the verb, shortened to mo, to, and before a vowel to m’, ?’.
Instead of the possessive pronominal adjectives, the Creole, except in Louisiana, and perhaps Mauritius, employs the per-
18
138 A. Van Name,
sonal pronouns. When used attributively, they are in Trin., Mart., St. Thomas, simply placed, like any other attributive case, after the noun; thus, liv li ‘his book’. This is also the present usage in Hayti, but Ducceur-Joly inserts the preposi- tion @ ; thus zamid@ li yo ‘his friends’. This latter method is employed also, according to Mr. Thomas, in Guadaloupe; in Scheelcher we find both forms; thus, bouche d ow ‘your mouth’, daite ou ‘your finger’. In Louisiana, however, the French pronom. adjectives have kept their place, though hardly dis- tinguished in form, except in the 3. sing., from the pronouns. They are mo, to, son or 80, no, vo, ye; thus, mo liv, ‘my book’.
The predicate form is in Trin. cela-moin, in Mart. Guadal. ta-moin (abbreviated from c’est @ mot) ; in Hayti, pa moin and pou moin; in St. Thomas, pa mo and cela mo; in Louis. pou moin. Powis the preposition pour ; pa is perhaps part, since we find in Ducceur-Joly avila part ayo ‘theirs’. He uses also another form of the predicate, which he writes both tien @ toué and guien @ toué, probably to represent the sound chien, evi- dently the last syllable of appartient, as pa used above may possibly be the second; thus, mal @ vou li arten coté tien ad li ‘your suffering is nothing in comparison with his’.
Demonstrative pronouns and adjectives. The distinction be- tween the nearer and more remote demonstrative, between ‘this’ and ‘ that’, is not in general clearly marked in Creole. In Trin. we have for the first, cela-la or gala following the noun, and for the plural the same preceded by ces; thus, zombi cela-la ‘this ghost’, ces zombi cela-la ‘these ghosts’; for the second, only the weak demonstrative which serves also as the definite article; thus, nomme la ‘that man’, ces jou la ‘those days’. In Hayti (Ducceur-Joly) we have for the adjec- tive ‘this’ and ‘that’ /a, plur. layo, for the pronoun ga or cila- la, plur. cila-yo; I have found also ¢a or cela, plur. cela-yo, used for the adjective. In St. Th. we have cé liv ‘this book’, cé liv yo ‘these books’, liv la ‘that book’, liv yo la ‘those books’; in Louis. Z’v ga and liv cila-la ‘this’ or ‘that book’, liv layé ‘these’ or ‘ those books’. In Martinique we find tala ‘that’ (abbreviated from aut la?) and the plural as in Trin- idad ; thus, cee gene In ‘ those persons.’
Contributions to Creole Grammar. 139
Relative and Interrogative Pronouns. The relative ‘who’ is commouly gui; ‘he who’, ¢a qui; ‘which’, ga or gui; ‘ that which’, ‘ what’, ga, ga gut, and in St. Th. also gui ga. The passing of a demonstrative into a relative is of common occur- rence in other languages, and is seen, for example, in Germ. der, Dutch die, Eng. that. In the case of ga it arose probably from the use of the compound form qui ga, the gui being after- wards dropped.
The forms of the interrogatives are not usually distinct from the relatives, but gui moune (monde) is very generally used for ‘who’ ?, quilece (quel est-ce) Trin., Mart., St. Th., for ‘which’? (a is also interrogative ‘what!’; thus, ga ga yé ‘what is that?’ (Mart.).
For the reflexive pronoun corps is frequently employed in place of méme ; thus, jidé (aider) corps-ou Epis léte’a jidé ou ‘help yourself and others will help you’, (Hayti). Méme is still employed to add emphasis to the subject; thus, i méme fai li ‘he did it himself’, and in Louis. and St. Th. is preferred also for the reflexive.
Verss. The verb offers the best field for the development of the Creole tendencies because the French verb as com- pared with the other parts of speech has preserved a large number of inflections. The destructive, and by way of com- pensation, the constructive power of the Creole is therefore here best seen. In general it may be said that the verb has lost all inflections and is reduced to a single form. The distinction of person and number is left to the subject, that of tense and mode is expressed solely by means of auxiliaries. The choice of this form has been determined of course not by any abstract consideration of the fitness of one form above another, e. g. the infinitive, to represent the idea of the verb without limitation, but simply by the frequency of its occurrence and the strength of the impression it has made on the ear. The usual statement, which Mr. Thomas also adopts, is that the Creole verb is the French infinitive, but though more weight is doubtless to be ascribed to this than to any other one form of the verb it is hardly correct to claim for it the whole influence. In the most numerous class
140 A. Van Name,
of French verbs, those in er, the past participle, one of the forms of most frequent occurrence from its use in the com- pound tenses, coincides with the infinitive, as would also in the Creole pronunciation certain parts of the imperfect and past definite. In the regular verbs in zr the case is very much the same. Final r has become silent generally in Creole (just as in French infinitives in e7), and here also the iufini- tive, participle, and past definite would be identical in form. In verbs in re, whether regular or irregular, the infinitive and past participle are somewhat. widely separated, and here the Creole has chosen sometimes the latter, but commonly the former, a choice determined in part perhaps by the general agreement between the present indicative and the infinitive in these verbs. To these considerations must be added another of more positive character, the fact that the simple tense of the Creole verb is, with a few exceptions, not a present but a past ; mown mangé is not ‘I am eating’ but ‘Tate.’ These ap- pear to be sufficient reasons for assigning to the participle a considerable influence upon the Creole verb, and apart from phonetic considerations it is better to write mangé rather than manger because the former may more fitly represent both in- finitive and participle. Verbs of the first French conjugation, in er, present scarcely any irregularities. Mr. Thomas gives two instances of what he calls conversions of an indicative into an infinitive, in verbs of this class, viz. in bailler ‘give’, which except in Louisiana has taken the place of donner, and travailler ‘ work’, the Creole forms of which are ba or bai and travai. But the shortening is here evidently caused by the / mouillé, just as in the Spanish Creole cavallv becomes cabai. From aller, besides the regular form allé we have va and ’a as auxiliaries of the future, and the imperative annon, anno (allons), shortened sometimes to an; thus, an nou fai ga ‘let us do this.’ Expressions like tempoui (je Ven prie), which the speakers no longer analyze, are hardly to be considered exceptions. In the specimens of Mauritius Creole given by Mr. Besant, however, such forms as pouss’, guett’, occur along with guetté, but these specimens are
Contributions to Creole Grammar. 141
hardly to be received as pure Creole according to the defini- tion of the term before given. |
Verbs of the remaining French conjugations in re, ovr, and tr, the majority of which are strong, or irregular, especially in the participle, receive a less uniform treatment. We have sometimes two forms in use in the same dialect, and between the different dialects there is frequent want of agreement. The following are uniform in all the dialects, i. e. they are everywhere derived from the same part of the French verb, and differ, if at all, only in the minor phonetic peculiarities which belong to one or the other dialect.
a. Forms which combine the infinitive and participle, and, in some cases, still other parts of the French verk: di (dire), Geri (Ecrire), li (lire), ri (rire), bouilli (bouillir), court, cououi (courir), ddmti, dromi (dormir), menti (mentir), parts (partir), senti (sentir), servi (servir), sortt (sortir), vint (venir).
b. Forms in which the French infinitive and certain parts of the indicative present are united: attanne (attendre), craine (craindre), fanne (fendre), fone Cfondre), joéene (joindre), plaine (plaindre), péne (pondre), ranne (rendre), repéne (repondre), batte (battre), coude (coudre), boi, boue (boire), cot, coué (croire) (Ducceur Joly gives the form erére), oué, voue, oua (voir). The above verbs in ndre lose both d andr; the x however recovers its consonant power. But in pren (prendre) the vowel remains nasal, unless perhaps in Trinidad. (Thomas writes both pouende and pouend.)
c. Indicative forms: faw (faut), vau (vaut), vlé, Louis. woulé (voulez), pé (peut), doé (dort).
The following table exhibits most of the important varia- tions found in the form of individual verbs. In the dialects of Martinique and Hayti, where Schcelcher and Ducceur-Joly use a second form along with that given by the other authori- ties, it is marked by sor D; where they use only a different form it is enclosed in a parenthesis.
TRIN. Marr. Harti. St. Tsos. Louis. Mauvrit. connaitre connaite connaite coné connaite connaite conné (S. connait) cuire ehuite cuite couite couite choui
D. coui chouite
142 A. Van Name, faire fd, faite fd, faite fe sz) fe fd défaire défaite difaite difaite difaite entendre tanne tanne tendé tanne tendé tendé tendé frire foui fri foui friyé friyé fri mettre métté metté mété mette mété . mette metté promettre promette promette promi premette promi D. promette promé prometté) commettre commi commette démettre démi démette démetté démi mordre modé médé modé mdédé médé6 mode S. méde moudre moulé moulé peindre peinti peintiré _peintiré perdre pédi addi édi pdde perdi E péde tp. perde) suivre souive sivé sivé souive s’asseoir assige assise chita assise assise Bizé (S. chita) (D. sitta, sizé assite synta) savoir sa save savé save save 8a savé, sai couvrir couvé couvé couvoui couvé couvri (couvout) : ouvrir ouvé ouvd louvoui ouvé ouvri (ouvoui) souffrir soufie souffri souffoui souffri souffri souffoui motrir m6 mouri moori mouri mouri md md md tenir tini, tni, _—tini, ni, tini tini tienne tienne tien
Where two forms are found in the same dialect there is sometimes, though not usually, a distinction in the use; thus in Trin., Mart., St. Th., ¢inz is ‘to have’ while tienne preserves the proper signification of tenir ‘to hold or tokeep’. In Hayti and St. Th. mé has hardly more than an adjective force, ‘dead’. In St. Th. tendé is present (probably from the fre- quent use of the phrase entendez-vous?), while tanne is past.
Such forms as médé, metté, soucouré, which are found in the dialect of Trinidad, and which correspond neither to infin- itives or to participles, Mr. Thomas regards as imperatives (mordez, &c.), but when we add other similar forms from the preceding table, us friyé, moulé, promé, peintiré, sivé, sizé,
Contributions to Creole Grammar. °* 148
and still others, as choisé, conténé (St. Th.), rompé (Louis.), couyé (cuerllir) (Hayti), cheré (traire) which has in Louis. become identical in form with chiré (tirer), the explanation, though admissible perhaps in some cases, does not meet all. These forms are for the most part due rather to attraction on the part of the more numerous class of verbs in é (first Fr. conjugation, in ev), a change similar to that of the strong to the weak conjugation, of Lat. cedere and corrigere to Fr. céder and corriger.
In the foregoing table we find regular participial forms; as commit, démt, couvé (couvert), ouve, souffe, pédt (perdu), mb (mort); a few feminine participles, as assize, cute, faite, difaite (the two last are possibly imperatives).
Peimtt and peintiré are perhaps to be referred to peinture rather than peindre. A similar form confitc ‘to preserve’ (confiture), is used in St. Thomas. Chita and assite are ap- parently asszeds-toi fused into one word. Another curious verb, similarly formed, runs through all the dialects. Chwmbo in Louis. and Maurit., elsewhere chembé, ‘ to catch hold, hold fast’, is the imperative tiens bon.
The simple form of the Creole verb performs the duty of a preterit or perfect and an imperative. The other tenses are formed by auxiliaries. Apé, the auxiliary of the present in Hayti and Louisiana, is the preposition aprés, not in its ordin- ary signification ‘after’, but in the more etymological sense ‘at’ (Lat. ad pressum, Ital. appresso), as in the phrase elle est aprés sa toilette, and the vulgarism 7 est aprés lire. This tense corresponds closely with our English progressive present ; thus, Papé li ‘he is reading’. Prefixing té (été, était), we have an imperfect; thus, li ?apé li ‘he was reading’. In Mauritius, according to the statement of Mr. Besant, apres forms only the imperfect ; thus, guand mo apres faire ga ‘ while I was doing that’.
In Trin., Mart., St. Th. we have, instead of apé, the auxiliary ca, which has also a somewhat wider use, expressing repeated and habitual as well as progressive action. It is used not only where a present or imperfect is required but in the state- ment of general truths, proverbs and the like; thus, malhé pa
144 ° A. Van Name,
ca chagé con laplt ‘accidents do not threaten like rain’, (Trin.). So far as I have observed apé has a more limited, a strictly progressive or continuative force. In his Grammar Mr. Thomas did not venture an explanation of ca, but more recently, in Triibner’s Record for Dec. 31, 1870, he has ex- pressed the opinion that ca and da, which performs a similar office in the Negro English of the West Indies, are of African origin, and from the same root. It is however extremely im- probable that while the African element, even in the Creole vocabulary, is so small, a word having such an important grammatical use should have been borrowed from this source. The Creole auxiliaries in general, Spanish and Dutch as well ’ as French, from their less independent character, have suffered more than usual change, and the original form is not always easily recognizable. Thus in Hayti and Louisiana the con- nection of apé with apres is no longer felt, and other etymolo- gies have been assigned to it. -Aprés in the sense of ‘after’ is pronounced in Louisiana as in French, in Hayti apoué, though Duceeur-Joly writes the auxiliary aprés, from which we may perhaps conclude that in his day the two forms had not yet separated. Sooner than abandon the attempt to explain ca from the French, we should be disposed to accept one of the following etymologies, no one of which we freely admit is very obvious, and which are offered simply as conjectures. They are, quand, or comme (Creole con), the use of which in clauses expressing contemporaneous action, as comme wl farsatt ga ‘just as he was doing this’, may possibly have furnished a starting point for the auxiliary. The change required will perhaps seem less improbable when we consider that the Creole has converted the conjunction et puis ‘and then’ into an in- strumental preposition ‘ with’; thus, /¢ talé (étalé) léte lad té Epis you coude bourque ‘he felled the other to the ground by a blow with a brick’ (Trin.). Courir is another possible ety- mology, supported by the analogy of the Dutch Creole, where lo, from loopen ‘to run or go’, is the auxiliary of the present. The simple form of the Creole verb serves as a preterit and perfect, and an imperative. The pluperfect is formed, just as the imperfect from the present, by prefixing té to the preterit.
Contributions to Creole Grammar. 145
Mr. Thomas makes, quite unnecessarily it seems to us, two Creole conjugations, one with, the other without ca. The latter, a small class, comprises such verbs as aimein (aimer), connaite, tini (tenir) ‘to have,’ vlé ‘to wish,’ and the like, all of them naturally continuative in meaning, and therefore not needing ca. The corresponding words in English do not take the progressive form, ‘I am loving,’ &c. In such cases atmem is the present, ¢¢é azmein the preterit.
The auxiliary of the future is va, frequently shortened to 7a. In place of this, callé (ca allé) is frequently used in Trin., Mart., St. Th.; apé allé in Hayti; allé, cour: in Louis.
The auxiliaries of the conditional present and past are the following: Trin., sé, past sé va and ¢té va; Mart., sré, past aré té; (Scheelcher, té, past ta); St. Th., sé, past té va and sé va; Hayti, ta, past té va and ta va (DuceurJoly, seré, past seré te and ¢é seré) ; Louis., té and sé, past ta. Sé is contracted from serazt, ta from té va, so that in ta va (Hayti) va is repeated.
‘Can’ is expressed in Trin. by sa (savoir) and pé (peut), in Mart. usually by pé (Scheelcher also uses sa), in Hayti by eapabe (capable) (Ducceur-Joly also savé), in St. Th. by pé and capabe; in Louis. by connai and capave. The past, — ‘ could,’ is formed in each case by prefixing té. In Trinidad capabe in an affirmative sentence has a bad sense, like the Eng. ‘ capable of’, e. g. capabe mentz ‘ capable of lying,’ but in the negative sentence this shade of meaning disappears. ‘Must’, ‘ought’, are variously expressed by faut, doué (doit), pou (pour).etc.; thus, pou li gagné sentrment ‘he ought to be ashamed,’ (St. Th.). |
Of étre the parts in use (not including the auxiliaries ¢é and sé,) are, pres. cé or yé, pret. cété, té and occasionally the infinitive éte. The copula is however sparingly used and especially when the predicate is an adjective, the present cé is regularly omitted. In Ducceur-Joly we find instead of cété, the compound form ¢é yé.
Avot not only as an auxiliary, but also in the sense ‘ to have’ has disappeared from the Creole. It is found however in davoé (d@ avoir), used in Trin. as a conjunction, ‘ because,’
19 :
146 A. Van Name,
after blamé and words of similar meaning ; also in napoént (il n’y a point). Its place has been supplied in Trin. and Mart. by tinz (tentr), in Hayti and Louis. by gagné ; in St. Th. both gagné and fini are in use. These verbs are also used for the French idiom y avoir; thus, pa gagné moune la ‘there was nobody there’ (Hayti). .
By the loss of avoir and étre, French nouns and adjectives are converted sometimes into Creole verbs; thus, pé ‘to fear’ (avoir peur de), btsoin ‘to need’ (avoir besoin de), content ‘to like’ (étre content de). Others pass directly into verbs; thus, marron ‘ to run away,’ plein ‘ to fill.’ Still other anom- alous verbs are vaumié, simté ‘ to prefer’ (vaut mieuz, serait mieux), found in Trinidad, to which may be added from Hayti pito (plutét), with the same signification.
The Creole relative and interrogative adverbs of time and place are especially interesting, the latter being in some cases distinguished by an interrogative prefix.
The relative ‘ when’ is in Trin. temps, lé (Vheure), inter- rogative not given; Mart. rel. quand, interrog. qui temps ; St. Th. rel. temps, quand, interrog. qui temps ; Hayti rel. 16 (Vheure), interrog. guilé ; Louis. rel. quand, interrog. équand (est-ce quand). .
‘Where’ rel. is in Trin. 6t: (on est-il), interrog. 6ti, cété, qui cété; Mart. rel. 6ti, qui cété, interrog. not given; St. Th. rel. 6ti, cété, interrog. 6ti ; Hayti, rel. cété, interrog. cété ; Louis. rel. o& interrog. é-ou, cbté, ou-ce-qué.
Mr. Thomas explains 6¢ as a contraction of on es-tu ; so péti, used in questions with the force ‘ can possibly,’ for peux- tu; vienti-vati ‘a gadder about,’ for viens-tu, vas-tu ; but in all these cases the pronoun 1s doubtless 7, not tw.
The Creole is simplified in syntax quite as much as in grammatical forms. The loss of inflections has narrowed the freedom of position, just as for a similar reason the French order is more rigid than the Latin. Little is in fact left be- sides position, to determine the relation of words in Creole. Thus the subject must always precede the verb, and the sen- tence cannot be made interrogative by inversion. Where there is no interrogative pronoun or adverb, the question is in-
Contributions to Creole Grammar. 147
dicated by the tone of voice, or introduced by éce (Trin.), &cequé (Louis.), éceqgui (St. Th.), Fr. est-ce gue. The prepo- sitions @ and de, which play so important a part in French syntax, are very sparingly used in Creole and mostly in cer- tain phrases having the fixed character of compounds. In most cases they have simply dropped away and the gap has. been left unfilled. Thus the attributive relation is expressed simply by placing the dependent after the governing noun ; e. g. chapeau papa tt fi Id ‘ the hat of the father of the girl’ (Trin.) Verbs which in French take in addition to a di- rect object an indirect object with d, in Creole generally omit the preposition, the indirect object preceding the direct ; thus, mown di zami moin ga, ‘I told my friend so ;’ but in Mart. we have also di pou ‘ say to’ (pour).
Verbs which in French have only the indirect object: either dispense with the preposition or exchange the weak @ or de for a stronger; thus, lz ri moin, or li ri aprés moin, ‘he. laughed at me’ (Louis.), Fr. de mot. So also where the com- plement is an infinitive ; thus, yo fouémt pou oué, ‘they shud- dered to see’ (Trin.) ; mot connai comment pou fai ga, ‘1 can do it’ (Louis.) Reflexive verbs sometimes give upthe pronoun without change of meaning; thus, iz sauvé, ‘he ran away’ (Hayti) ; mo voulé servi li, ‘I wish to use it’ (Louis.)
In Trinidad the verb ba or bai (bailler) ‘ to give,’ may be used to mark the dative; thus, i: pdté touwment bai famie ’s ‘he brought trouble to his family.’ In the Negro English of Surinam gi ‘ give’ is used in a similar way.
The Creole not unfrequently combines prepositions for greater strength, or forms new ones; thus Mart. quant @ pou moin ‘as for me,’ Louis. en haut laterre ‘ on the ground,’ for which we have in Trinidad the equivalents en-lair and la sou (dessus) ; but these combinations are simple in comparison with Fr. au dessus de.
For the conjunction e¢ when used to connect individual words we have in Louis. avé (avec), in Hayti acque (avec), in Mauritius semd (ensemble). In Trinidad épzs (et puis) has passed through the opposite process. Employed first to connect clauses, then individual words, it passed into a pre-
148 A. Van Name,
position denoting accompaniment, and finally the instrument, and is used interchangeably with évec (avec).
After sét (sortir) ‘to go out’ and pren (prendre) ‘ to take away’ we have the preposition nans (dans) employed; thus, lt s6tt nans cabane it ‘ he has just left his bed;’ & pren li nans main moin ‘he took it away from me.’ A similar idiom is found in the Negro English ; thus, kKomoto na ini hoso, literally ‘come out'to-in (within) house,’ i. e., to come out of the house. Sé6ti answers to Fr. venir de; thus, moin sdts contré ‘T have just met’ (Zrin).
The adverbs até ‘at once’ (Mart. ) and éni ‘ only’ (Trin.) furnish a good illustration of the amount of change which it is possible for a Creole word to undergo. Atd, as the inter- mediate forms ator, astor (Ducceur-Joly) and astére (Louis. ) show, is for @ l’enstant. For éni we have in Mart. ant, anique, in Hayti (Duc.-Joly) necque, in St. Thomas ennique, from which to unique is not a difficult step.
A few examples, in conclusion, of the power of the Creole to modify the meaning as well as the form of words:
(a.) Creole of Hayti. Bougé ‘to dwell.’ This change to & meaning opposite to that of Fr. bouger is explained perhaps by the circumstance that in such phrases as i ne bouge pas des spectacles, pas or point is not usually expressed (Dict. de l’Acad.), and the Creole negative being pas, not ne .... pas as in French, the weak ne goes for nothing.
Délagué, ‘to untie, loosen ;’ the prefix de repeats the nega- tive idea which is already in the simple verb darguer.
Chonger (songer) ‘to remember’ (also in Louisiana).
Campé ‘ to stand’ (also in Louisiana).
Miel ‘ honey-hee,’ ‘ honey’ being siwo miel (sirop).
Tount (tout nu) ‘ naked ; thus, pié touni ‘ barefoot.’
Grand gout ‘hungry.’
Hin pé (un peu) ‘a little ;’ grand pé ‘a great deal.’
(b.) Creole of Louisiana.
Senti ‘ to smell’ (also Mart.) ; compare Eng. scent.
Court ‘to go’ and galopé ‘to run’ show the influence of slavery in quickening the natural pace.
Tatllé ‘ to whip.’
Contributions to Creole Grammar. 149.
Rompé (rompre) ‘to beat soundly.’
Désert ‘ field.’
Bitin ‘ thing’ Fr. butin ‘ plunder,’ used also in St. Thom- as and Guadaloupe ; in Hayti and Trinidad, baggai (baggage).
Cabane, which in the West Indies has the significant ‘ bed’, in Louisiana has the usual French meaning ‘ hut.’
SPANISH CREOLE.
Of this dialect there exists so far as I have been able to. learn no published grammar or dictionary. The materials for the present sketch have been drawn from two sources: . 1, a brief Dutch grammar in the Creole dialect, published in 1849, by J. J. Putman, a Roman Catholic clergyman in Curagoa, followed in 1853 by a second part containing Creole and Dutch conversations ;* 2, the oral testimony of Camps, who had excellent opportunities for acquiring the language and speaks it with great readiness. The master-workman with whom he served a six years’ apprenticeship in St. Thomas and his fellow workmen were natives of Curagoa and made constant use of the language, and Camps has himself also spent a little time in Curagoa. Between the language as written by Putman and as spoken by Camps there is a sub- stantial agreement, not only in general structure but also in- the form of individual words. There are minor points of difference in which uniformly the Creole of Camps shows a progress in the natural direction of the language, towards pho- netic decay, and since the variation is not towards but away from the Spanish, his native language, I have not hesitated to accept it as a genuine variety.
Putman employs in writing the Creole the Dutch orthogra- phy; for this the Spanish has here been substituted with a few modifications mentioned below. 7
* Proeve eener Hollandsche spraakkunst ten gebruike der algemeene armenschool in de gemeente van de h. Rosa op Curacao. Eerste stukje. Santa Rosa, 1849, 16mo, pp.. 48. Tweede stukje, gemeensame zamenspraken. 1853, pp. 64. Putman has published at least one other little work in the Creole: Bida dt Hesce Kriestoe. (Life of Jesus Christ.) Santa Rosa, 1852.
160 A. Van Name,
. PHONETIC CHANGES.
The principal phonetic changes from Spanish to Creole are the following:
1. Vowels. Accented vowels are subject to few changes. The dipthongs ze and we usually return to e and o their Latin originals ; thus, seru (sierra), bon (bueno), sofio (suefio); but bientu (viento).
Unaccented vowels are more variable but the changes can- not be reduced to rule. Very frequent is the change of final o to u; thus, malu (malo), pretu (prieto), the derivative suffix mentu (mento). Some changes are apparently due to attraction by an accented vowel in a preceding or following syllable; thus, afta (afio), caya (calle), bicht (bicho), bint (venir).
2. Consonants. The peculiarities which distinguish the. - Cuban from the Castilian pronunciation are shared by the Creole, i. e. U1 has the sound of y, v the sound of }, and z the sound of s, as has also ¢ before e, but before 7 like s in the same position it passes into the English sh; thus, yama (llamar), bifia (vino), serebes (cerveza), shelu (cielo), dushi (dulce), shete (siete). |
Other changes are peculiar to the Creole. A final n fre- quently loses its consonant power and only makes the prece- ding vowel nasal, a characteristic feature of all the Creole dialects; thus, nan, tin, shon. So also the Creole fi is not resolvable like the Spanish #% into n followed by y consonant, but with the preceding vowel (like gn in the French Creole, ) it makes a nasal vowel followed by y consonant. WV between two vowels sometimes becomes fi; thus, bisifia (vecino), gafia (gana). It is sometimes inserted before a dental; thus, co- minda (comida), landa (nadar). In hunga (jugar), nenga (negar) the palatal mute is changed to the corresponding nasal. ’ |
A mute followed by a liquid is a combination which in the middle of a word after an accented vowel the Creole avoids ; / and r in this situation are transposed with the following vowel; thus, homber (hombre), cuater (cuatro), milager (mila- gro), culpabel (culpable). Sometimes, but not usually, an
Contributions to Creole Grammar. 151
initial mute and liquid are thus separated; thus, purba (prar bar). In drwmi (dormir) the change is in the opposite direc- tion. Instead of transposition we have sometimes an inserted vowel; thus, tabula (tabla), temperan (temprano), cologa (col- gar). #& is transposed to avoid hiatus in hari (reir), kere (creer). £ andr are occasionally interchanged; thus, carson (calzon), lastra (arrastrar). |
Tin the middle of a word passes sometimes to d, the regu- lar change from Latin to Spanish ; thus, hende (gente), mondi (monte) ; but in putri (pudrir) we have the opposite change.
Di before e becomes j, (dzh), as in jes (dies), jente (diente).
Final ad, with d which has become final by the loss of a fol- lowing vowel, either passes into r or is dropped. Thus, tur (todo), mitar (mitad) are found both in Putman and Camps; salur (salud), criar (criado), and many participles in ar, er, . tr (Sp. ado, ido), e. g. habrir (abrido), pisar ( pesado), are given by Putman, but Camps drops the final syllable altogether and pronounces salé, crid, habrt, pisd. Merdia (medio dia) has the sanction of both.
Gu before a and o loses the g; e. g. awa (agua), warda (guardar).
J initial and in the middle of a word before an accented vowel retains its aspirate sound ; after an accented vowel it is lost or passes into y; thus, abdu (abajo), cangréo (cangreo), conséyo (consejo), but fica (fyar).
Initial 2 is dealt with quite after the cockney fashion. Be- fore the dipthong we, where in Spanish it 1s strongly aspirated, in Creole, as also in the Cuban, it is silent; thus webu (huevo), wesu (hueso), werfanu (huerfano). Before other vowels it is silent in Spanish but generally aspirated in the Creole. In the following words (an incomplete list), it is aspirate; heru (hierro), huma (humo), hostia (hostia), higra (higado), hom- ber (hombre), hilu (hilo), humilda (humilde), horta (hurtar), hast (hacer), herebe (hervir), haya (hallar), hala Chelar), hereda (heredar). In ora (hora), onor (honor), yu (hyo), awe (hoy) and some others it is silent.
152 A. Van Name,
. His also prefixed to some words beginning with a vowel ; thus, haltu (alto), hanchu (ancho), harafia (arafia), hufia (ufia), habri (abrir), hisa (izar), hole (ole), hunta (untar).
More striking are the changes occasioned by the loss of entire syllables, initial, medial, or final, not protected by the accent, a loss caused in part no doubt by the increased strength of the accent in Creole. As in the French Creole, prefix prepositions fall away very readily; thus, bisd (avisar) sifid, (ensefiar), derd (enterrar), splicd (explicar). The e which - the Spanish prefixes to s followed by another consonant, is also dropped; thus, spiritu (espirito.)
The loss of a medial syllable is less common, e. g. punéird (preguntar), dirtt (derretir), drecht (derecho), hermentu (herra- mienta), trahd ‘to make’ (trabajar), but trabdu ‘ work’ (traba-
40); 80 camnd ‘to walk’ (caminar), but camina ‘ road’ (cam- _ 3no).
' ‘The decay is naturally greatest at the end of the word, and the extent of this in the Creole is shown by the entire loss of verbal inflections. Instances in other parts of speech are, sombre (sombrero), pida (pedazo), cabai (caballo), caché, Put- man cachor, ‘dog’ (cachorro), limpt (iimpto).
An initial and a final, or a medial and a final syllable may be lost in the same word ; thus, costuma (acostumbrado), cord, Putman coral (colorado).
The only final consonants in use in the Creole are /, m,n, r, and s. The Spanish finals d, z, 2, are not found; m, which does not appear in Spanish, becomes final by the mutilation of a word; thus, bam ‘come!’ (vamos), com ‘how’ (como).
A few words ending in a consonant add in the Creole a vowel; thus, clina (clin or crin), reda (red), bigesa (vejes), solo (sol), ayera (ayer).
_ Dutch words, which constitute a not very large, but doubt- less increasing portion of the vocabulary, are subjected neither to the Creole, nor to any uniform laws of change. Verbs are sometimes shortened as in the Dutch Creole; thus, stop (Dutch stoppen), fluit (Dutch fluiten) ; sometimes they are made uni- form with the Spanish; thus, lezé ‘to read’ (Dutch lezen), konopd ‘to tie’ (Dutch knoopen.) Dutch nouns are frequently
Contributions to Creole Grammar. 158
adopted in the diminutive form, in some cases apparently for the sake of the softer vowel-ending; thus, buki ‘book’, brefi ‘letter’, stulcht ‘chair’; or a vowel may be assumed; thus santu, (Dutch zand.)
InrLtections. These have gone the way of all Creole in- fiections. The article, adjective, and noun are invariable in _Tespect to both gender and number.
Gender, which can no longer be expressed by termination, is marked, wherever necessary, by adding homber, ‘man’ and muhé (mujer), ‘woman’; thus, ‘ruman homber’ ‘brother,’ ruman muhé ‘sister’ (hermano,na); mucha homber ‘boy’, mucha muhé ‘girl’ (muchaco,-cha); patu homber ‘drake’ (pato). Homber and muhé may, however, be omitted when it is already sufficiently clear which of the sexes is referred to, and ruman, mucha are then used indifferently for both.. Yu ‘child’ is the representative of both -Azjo and hija. In the place of padrino,na ‘godfather’, ‘godmother’, we have only padrina ‘godfather’; ‘godmother’ is madrina, as in Spanish. For abuelo,-la ‘grandfather’, ‘grandmother’, the Creole has ewela ‘grandfather’ and wela ‘grandmother’.
The idea of plurality is expressed only when the noun is definite, and then, as in the other Creole dialects, by adding the personal pronoun 3. plural nan; thus, cas (casa) ‘house’ or ‘ houses’, cuater cas ‘four houses’ ; but e cas nan ‘ the houses’, ¢ cas grandi nan ‘the large houses.” The only instance which I have discovered of the Spanish plural ending ¢ is in the adjective barios ‘ various’, (vario), where the prevailing use of the plural has left its impress on the form. No corres- ponding singular dario is in use. |
Adjectives, as has been already remarked, are indeclinable. The comparative is formed as in Spanish by mas, and accord- ing to Camps the superlative by e mas, but Putman gives di meas; thus, grandi, comp. mas grandi, superl. e mas grandi or di mas grandi. Bon (bueno) and malu (malo) are thus compared ; bon, miyon, mas miyon; malu, piyo, mas piyo. For maiyon (mejor) Putman writes mehor; the final vowel is made nasal by the initial m. Mens (menos) ‘less’ forms a superla-
20
154 A. Van Name,
tive mas mena ‘least’. The termination isimo, used as in Spanish only in an absolute sense, occurs in muchisimu.
The numerals from eleven to fifteen differ from the Spanish in employing the separate forms jes un, jes dos, &c. instead of once, doce, &c. Dos and tres, according to Camps, retain the final s only when they stand alone; before a noun they become do, tre. Only the first four of the ordinals are in use; the cardinals take the place of the others.
Pronouns. The following table exhibits the personal pro- nouns.
Singular, Plural,
1. mi 1. nos
2. bo 2. boso, bosonan 8. el, e. 3. nan.
The objective forms are less numerous here than in the other. dialects; the 1. sing. m is the only unmistakable one. The 2. sing. 60 is for vos ; boso for vosotros, vosotras. Putman employs only 4080, Camps only bosonan, a plural similar to that of nouns, literally ‘you-others-they’. In the 8. sing. Putman uniformly gives el, while with Camps the regular form is e, el appearing only before the auxiliary of the preterit a; thus, e bisa ‘he says’, but el a bisa he ‘said.’ The follow- ing vowel accounts here for the preservation of the final Z, which has elsewhere disappeared. Of the origin of the 3. plur. nan I can give no satisfactory explanation.
The possessive pronominal adjectives are in the 1. 2. sing., 1. 8. plural the same as the personal pronouns; the 3. sing. is su (suyo); the 2. plural is like the 2. singular bo. When used attributively they precede the substantive, and they are made predicative by the preposition dz, as dz mt‘ mine’, &c. After ta ‘is’, the d is elided and we have tat mi for ta di mi, &c., but in the 3. sing. di before e passes as usual into J, in spite of which we have not ta je but tat je, i. e. ta di die. The possessive becomes substantive when the definite article, or demonstrative e (Putman es), is prefixed to the predicate form; thus, e di mi, to which corresponds the Spanish el de él ‘his’. Both the definite and indefinite articles are invaria- ble. Camps employs for the definite article e, which is prob-
Contributions to Creole Grammar. 165
ably to be referred to the Span. demonstrative ese rather than to the article el. Putman uses no form distinct from the weak demonstrative es, which is however stronger than the e of Camps. Zs is both the nearer and the more remote demonstrative, both ‘this’ and ‘that’, but it may be made more definite, as in Spanish, by aguf ‘here’, and ai or ayd (ahi or alld) ‘there’; thus, es liber aqui ‘this book,’ es liber nan ayd ‘those books’.
The indefinite article is wn.
The Span. article e/, Ja, appears in two or three cases insep- arably joined to the noun; thus, laire (Putman laria) (el aire); lama (Putman lamar) ‘sea’, e lama ‘the sea’; alafin ‘finally’.
The reflexive is formed by means ef curpa (cuerpo); thus, su curpa ‘himself’; while mes (mismo) is employed to give emphasis to the subject; e. g. el a hast e mes, ‘he did it him- self’. This corresponds with the prevailing usage in the French Creole.
The relatives and interrogatives, as in Spanish, have the same form;
Who, gueng (quien), quende (que gente).
What, qui (que), quico (que cosa).
Which? quz, cual.
THe Vers. Putman distinguishes two forms of the verb, one corresponding to the Span. infinitive, but without the final r ; the other with 7, which has come however from d of . the participial ending do; thus pasd, pasdr; conosé, conostr (pasar, pasado; conocer, conocido). Skirbi and pone, however, which in Spanish have strong participles, escrito and puesto» in the Creole are weak, skirbir and poner, — the latter also retaining the vowel of the infinitive, e, instead of the 1, which we should expect (ponido). This form in r is not used, like the Spanish participle, in the compound tenses of the active voice, but has only a passive sense; thus, nos mester pone ‘we must put’, but es mester ta poner ‘it must be put’. I have found in Camps no trace of the above distinction; the verb is reduced as in the other Creole dialects to a single form,— the first of those given. The final r, always a weak letter
156 A. Van Name,
except where it becomes final by transposition as in homber _ (hombre), if once sounded has now disappeared.
The verb estar which in Spanish has largely taken the place of ser, in the Creole has quite supplanted it. The only trace of ser which I have observed is in podese ‘perhaps’ (puede ser). Ta ‘to be’, the Creole form of estar, is conjugated as follows :
Pres. sing. 1. mi ta I am, Plur. 1. nos ta we are, &e. 2. bo ta thou art, 2. boso ta 3. eta he, she, it is. (osonan ta, Camps.) (el ta, Putnam.) 3. nan ta.
Past (imperf., perf., plupf.), mt tabata (Camps, often, m¢ taba).
Fut. (and pres. condftional), lo mz ta.
Past conditional, lo mi tabata or lo mi taba.
Ta is, of course, Sp. esta; taba, Sp. estaba ; so that in tabata we have the verb doubled. The position of lo before the sub- ject would seem to indicate that it is not of verbal origin ; it is probably the Span. /wego ‘ presently’, the final syllable being lost and the first suffering the regular change of vowel, we to o. In the past tense, only ¢abata is found in Putman, but the examples are too few to establish a rule; Camps’ usage is not uniform, but apparently while tabata is both pluperfect and imperfect, taba is only imperfect.
Tin ‘to have’ (and also, like tint in the French Creole, ‘there is’) (tentr), which has supplanted haber, is similarly conjugated; pres. tin, pret. tabatin, fut. lo (mi) tin. In the signification ‘to hold’, the fuller form tinz is used.
The conjugation of the regular verb is illustrated in the fol- lowing paradigm of duna (Sp. donar) ‘ give’: —
Pres. mt ta duna I give, or am giving, Impf. mi tabata duna I was giving,
Pret. mia duna I gave,
Plupf. mt a caba duna I had given,
Fut. lo mi duna I shall give, Conditional (past) lo mia duna _I should have given, Immed. Fut. mi ta baiduna Iam going to give,
Imper. duna Give.
Contributions. to Creole Grammar. 187
A few verbs, as in the French Creole, form the present without an auxiliary, e. g. mi sabi ‘1 know’; the imperfect is in this case naturally wanting. The auxiliary of the preterit, a,is the only relic of haber which the Creole retains. Caba isthe Sp. acabar, and is in use both as a verb, ‘to finish’, and an adverb, ‘already’; thus, mz a caba hasi e caba, ‘I had done it already’. The following sentence will illustrate the use of the conditional ; s7 50 bis’ esai’ atrobe (otra vez) mi ta dal bo, or lo mi dal bo ‘if you say that again [ shall strike you’; si ela bis’ esat’ atrobe, lo mi a dal e ‘if he had said that again I should have struck him’.
‘Can’ is expressed by por (Camps, except before a vowel, po) Sp. poder; ‘could’ by por a; thus, mi por hasi e ‘I can do it’, mi por a hast e ‘I could have done it’.
‘Must’ is variously expressed; thus, mi tin pa (para) bai ‘I must go’; pa mi hasi e or mi debi hasi e‘I must do it’.
Laga (Putman larga) Sp. largar is the causative, as well asthe hortatory ‘let’; thus, laga hasi ¢e ‘let it be done’, laga nos hast e ‘let us do it’.
The three Spanish conjugations are generally traceable in the Creole verbs; the final vowel being protected by the accent has only in a few cases suffered change. Most common is the change of e to 7, of which the following are all the cases ob- served: hast (hacer), sabi (saber), bari (barrer), debi (deber), cai (caer), priminti (prometer), gradisi (agradecer). So, also, when Latin verbs pass in Spanish to a different conjugation, the change is usually from er to ir; e. g. battuere (batir), fer- vere (hervir), which in the Creole herede ‘boil’ returns to the Latin vowel e. In biba (vivir), prefera (preferir), sambuya (zabullir’)) we have the change from 2 to a; in mula (moler), from e to a.
The Spanish verbs have nearly all become weak, and as a natural consequence we find among the Creole verbs but few irregular forms. Besides those among the auxiliaries, already considered, there are the following: —
Dal ‘to strike’ (dar, which in the sense ‘to give’ has been replaced by donar). The / which here appears belongs possi- bly to the suffix pronoun, Sp. darle.
158 A. Van Name,
Que, before a vowel quer, (Putman quier), Sp. querer, ‘wish’, ‘ will’: e. g. e quer a bai na cas ‘he wished to go home’; bo quier ta asina bon di bisa mi ‘ will you be so good as to tell me’.
Bai ‘to go’ and bam, in exhortations, ‘come!’ from vaya and vamos, parts of the defective ir.
Duel, in the phrase ta duel mi ‘I am sorry’, may perhaps be the pres. indic. 3. sing. of doler, but more probably the noun duelo used as a verb.
When immediately followed by the pronoun or article e, the final a and e of verbs of the first and second conjugation are lost; thus, el a 673’ e ‘he told him’ ; el a hal’ e cabuya, (this last ‘is a word of Indian origin; v. Pichardo s. v.).‘he pulled the rope.’
The negative no is generally so closely attached to the pre- ceding word as to lose its vowel; thus, en tin e ‘he has’nt it.’
Reflexive verbs give up the pronoun without change of meaning: thus, parse (parecerse) ‘to resemble’. As has been already observed, in Putman the past participle has preserved a distinct form, and the verb can be conjugated passively, but Camps makes no such distinction. The usual resource is to turn the construction into an active form, but in some cases the verb is used intransitively in the preterit in place of a pres- ent passive ; thus, e cas a caba ‘ the house 7s finished; e palo ta kibra ‘the tree is breaking’, but e palo a kibra ‘the tree is broken’. | '
The only derivative suffixes which retain vitality enough for new formations are the diminutive endings -tcu, -2tu (-tc0, -ito) and -mentu (-mento) ; thus from the irregular verb dal ‘to strike’ we have dalmentu ‘a blow.’
The prepositions in most common use are the following: dé (de) marks, as in Spanish, the genitive relation, na (Dutch na, naar) which has taken the place of a, the dative; the stronger den (dentro) has supplanted en ‘in’; con retains its full form before e; thus cone ‘with him’,—-but is elsewhere eu; ‘under’ is bau dz (debajo de), ‘above’, ariba di.
‘When’ interrogative is guitem (d que tiempo), sometimes cuando; relative, ora.
Contributions to Creole Grammar. 159
‘Where’, both interrogative and relative is unda (en donde) ; thus, unda.bo a sali ‘where have you been ?
A few of the more peculiar, either in form or meaning, of the Creole words are the following: —
(a.) Peculiar in form. Cuminda ‘to greet’ (cumplimen- tar); drinta‘to enter’ (entrar modified by the preposition den Sp. dentro); grabata ‘to scratch’ and also ‘to itch’, (garra- patear ‘to scribble’); hiba ‘to carry’ (Camps yeba, the regular form) Sp. Uevar; lubida ‘to forget’ (olvidar) ; lamantar ‘to rise’ (levantarse) ; trese ‘to bring’, used along with the regular form trae (traer); papia ‘to talk’, whence the name papia- mento given by the Spaniards to the Creole dialect, possibly from hablar; até, atd ‘there it is!’ for alld ta e(?); feneta ‘pin’ (alfiler); shon Sp. seftor, sefiora, (shon mosa (moza) is used for sefiorita); numa ‘only’ (no mas); anto ‘then’ (entonces) ; pasoba ‘ because’.
(b.) Peculiar in meaning. Farifia ‘corn meal’ and hartfia ‘flour’, both from Sp. harina ; pretu ‘black’, Sp. prieto ‘of a dark color’ (negro being the distinctive name of the black race another word for ‘ black’ was found necessary) ; mzsa ‘church’, properly ‘mass’; dia domingo ‘Sunday’, Sp. domingo; buta ‘to put’ (botar ‘ to throw’) ; bringa ‘to fight’ (brincar ‘to jump’ ‘fret’); masha ‘very’, ‘greatly’ Sp. demasiado ‘too much’ (in the Negro English of Surinam, we have too mooshi and in the negro dialect of our Southern States too much, used in @ similar way to express only a high degree of a quality) ; tata ‘father’ (used also in the Dutch Creole, vader being in ordin- ary use applied only to God,) probably of African origin. Jata in some of the Italian dialects is used, especially by chil- dren, for ‘father’, but in Spanish we have tato, tata ‘younger brother’, ‘younger sister’. In several of the Western African dialects tata is ‘ father.’ :
DUTCH CREOLE.
This dialect, spoken in the Danish West Indies, St. Thomas, St. Croix, and St. John, is confined mostly to the plantations. Until within a few years the Moravian missionaries have
160 A. Van Name,
preached in this language to the blacks, but they have now abandoned it for a broken English. English is in St. Them- as the language of the public schools.
My authorities have been the Creole New Testament pub- lished at Barby, in 1802; the Psalm-Boek, Barby, 1784; a grammatical sketch in Oldendorp’s Geschichte der Mission der evangelischen Briider auf den Inseln S. Thomas, S. Croix and S. Jan, Barby, 1777 (pp. 424-436), but more especially the oral testimony of Camps. That the language of the New Testament and Psalm-book is in great measure artificial, and no fair representation of the spoken language, one who has studied the other Creole dialects would easily discover, even without Oldendorp’s statement that the missionaries were compelled to borrow from the Dutch and German, not only many words for the expression of ideas wanting in the Ore- ole, but also a passive conjugation. The hope which he -en- tertained that the spoken language would gradually be eleva- ted to this level, has not been realized. Oldendorp approaches much nearer the spoken idiom of the present day. The Dutch Creole is more fortunate than its sisters, in preserving in nouns a plural inflection. Both the Dutch plural endings, en and s are found, the former quite frequently, the latter but seldom ; e.g. berg, pl. bergen; broeder, pl. broederen; neger, pl. negers. But in a majority of cases these endings have been replaced by the pers. pron. 3. plur., sende, which, how- ever, like the corresponding plurals in the other Creole dia- lects, can only be employed when the noun is definite ; thus, boom, definite plur. die boom sende. Certain words which are apparently in process of transition add the pronoun to the plural] termination ; thus, vroww, pl. vrowwen and vrouwen sende ; schip, pl. skepen and schepen sende.
The noun has no case endings, and the article and adjective (except when used substantively, e. g. die dooden) are en- tirely without inflections either of number, gender, or case. The genitive when it follows the governing noun is connec- ted with it by the prep. va (van); when it precedes, the pos- sessive pronoun shi is inserted; thus, de boek va Jan, or Jan shi boek ‘ John’s book.’
Contributions to Creole Grammar. 161
The personal pronouns are :
Singular. Plural. 1. mi 1. ons 2. joe 2. jende, jen (Oldp. jender, jen)
3. am (Oldp. em). 8. sende, sen (Oldp. sender, sen).
All these except the 38. plur. are objective forms, corres- ponding to the Dutch mz, u, hem, ons, u andere, zij andere. The 2. and 8. plur. correspond precisely with the Creole Fr. z6t, (vous autres), whereas the Dutch adds lieden to the singu- lar pronoun where it is necessary to distinguish the plural, though more often the distinction is not marked; thus 2. sing. gi7, 2. pl. giz, or gij eden; 3. sing. fem. 277, 3. pl. fem. zu or zt) lieden. The forms of the personal pronouns also serve for the possessive pronom. adjectives, except in the 3. sing. where sht (zijn) is employed. When attributive they elther precede the noun, or follow it with the preposition va (van) ; thus mz boek or die boek va mt; with sende the latter form is preferred. Va mi, ya joe, etc., are the predicate forms.
In Oldendorp’s time, both the comparative and superlative endings er, st were retained in a few adjectives, though most were compared by means of meer, meest ; thus, groot, grooter, grootste. But now only the superlative ending is preserved, the weaker comparative having disappeared ; thus, groot, mee groot, grootste. The only instance observed of the compara- tive ending er is in beter, of which the positive in use is fraat and the superlative beste or mee beter. On the other hand the superlative can no longer be formed by meest, and such adjectives as have not the ending ste do not distinguish the superlative from the comparative ; thus, lee (/aag), comp. and super. mee lee. Mee kleentje, the comparative of kleen, shows the diminutive ending we curiously attached to an adjective. The substantive beetje ‘ a little,’ forms a compara- tive mee beetye, ‘less,’ ‘a smaller quantity.’
The Creole definite article die is to be referred to the Dutch demonstrative die rather than to the article de. Die retains also in the Creole its demonstrative use and is, moreover, as in Dutch, frequently relative.
_ The Creole verb is, according to the statement of Olden- 21
162 _ A. Van Name,
dorp, the Dutch infinitive deprived of the termination en; a better statement is that it is the verb stem, stripped of all terminations and frequently of a portion of the root in addi- tion. Thus from staan, where an belongs to the root of the verb, we have the Creole stan, but from saan, root slag, Creole sla. The final consonants of the root are, in almost all cases, written by Oldendorp and found in the Creole New Testa- ment, but are, now at least, frequently not heard; thus, lo (loopen) ‘to go,’ ki (kijken) ‘to see,’ are written by Oldendorp loop, kik.
In the case of the irregular or strong verbs, where the vowel of the preterit or participle differs from that which appears in the present, the infinitive, and the imperative, the Creole choses the latter. The only exception noticed is vlo ‘ to lose’ (verliezen, verloor, verloren).
As an illustration of the Creole conjugation, we take val (vallen) ‘ to fall.’
Pres. mi lo val. I am falling.
Impf. mi a lo val. I was falling.
Pres. mia val. I fell.
Perf. mz ka val. I have fallen.
Plupf. mi aka val. Thad fallen.
Fut. mi sa val. I shall fall.
Past Condit. mi a sa val. I should have fallen.
The auxiliary of the present Jo (Oldp. fe) is apparently the same with the verb lo (loopen), ‘to go’, to which it is pre- fixed to form the immediate future tense ; thus, mi lo lo val ‘I am going to fall.” The preterit a is a weakened form of hebben ‘have’ (Creole ha), and sa future, is from zal. Ka is not so readily explained. It is perhaps for gehad, g being made surd by the following 4; possibly from kom (kommen), the Creole representative of worden, or from kaba (Spanish acabar) which, both as a verb ‘to finish’ and an adverb ‘ al- ready,’ has found its way into all the Creole dialects.
The verb ‘ to be’ (Dutch zijn, wezen) is conjugated as fol- lows: Pres. bin, mi, na; Pret. a wees; Fut. sa wees. Mi and na are apparently corruptions of bin, arising from rapid and indistinct pronunciation, but the three forms are not used in-
Contributions to Creole Grammar. 163
differently ; thus, to the question, Wie dte bin da? ‘who is there ?’ the answer is, Na mi ‘it is 1;’ mi mi fraat ‘I am well,’ &c. Oldendorp remarks that ka, the auxil. of the per- fect, is often used in place of ben; thus, mi ka moe ‘I am tired,’ but it appears in such cases to denote always a result- ant state, ‘1 have become tired.’
The passive conjugation found in the Creole New Testa- ment as has been already remarked, is purely artificial; such forms as sal word genammt ‘shall be called,’ have no trace of Creole character. The Creole, however, makes limited use of a passive formed by means of kom or ka; thus, die hoes ka bouw ‘ the house is built.’
Curious changes in the form or meaning of words are less common in the Dutch than in the other Creole dialects. The following are examples: estan (laat staan), Creole ‘ leave,’ is also uaed in exhortations ; thus, lestan ons lo ‘let us go.’ The separable prepositions op and toe are used as verbs, the former for opsiaan, the latter for toemaken; thus, wat tit you hopo voevroe? ‘at what time do you get up in the morning ?’ toe joe mon ‘shut your mouth.’ Voevroe is a case of redupli- cation, so common in the Negro English of Surinam; an- other is pat-pat (Spanish pato) ‘ duck.’
THE NEGRO ENGLISH OF SURINAM.
To complete the survey we add the more important charac- teristics of this dialect as it appears in Wullschlagel’s Neger- englisches Wérterbuch, Lobau, 1850, and in the Negro Eng- lish New Testament of the British and Foreign Bible Society, Lond., 1829.
The grammar and the chief. part of the vocabulary are of English origin, but the long Dutch possession of the country has introduced, as in the case of the Spanish Creole, a strong Dutch element, and the additions now made to the vocabulary are from this source, while many English words are falling into disuse. From former contact with the Portuguese Creole it has taken up also many words from this source,and we not unfrequently find synonyms from two or even from the three languages ; thus for ‘ nothing’ we have Eng. nutz, Dutch niksi,
164 A. Van Name,
Port. nada. In outward appearance the language has under. gone more change, become more Africanized in the structure of its syllables, than either of the other Creole dialects, but the grammatical changes are less important. Very seldom are any finals except a vowel or a nasal, m or n, tolerated, and as @ consequence a vowel is very frequently assumed. The fol- lowing proverb from Wullschlagel will serve as an illustration : Te Gado wani pai Dagoe, a gi hem wan soro na mindri hede; fa a sa doe kisi hem vo lekki? literally ‘Time God want pay dog he give him one sore in middle head; fashion he shall do catch him for lick ?’ i. e. ‘When God wants to punish a dog he gives him a sore on the back of his head; how will he con- trive to lick it?’ (We retain the established Dutch orthogra- phy, but to bring it into correspondence with the English it will be generally sufficient to bear in mind that the vowels have the Italian sound, except that ve is used for u, ze for long i, ee for long e, and that 7, #7, dj and v are used for Eng. y consonant, ch, 7 and f respectively). Combinations of conso- nants are frequently reduced, by the dropping of one or more, toa simpler form; thus, tan (stand), port (gpoil). Rand l are often interchanged ; thus, srefi (self), ripi and lepi (ripe). Final m and n after any vowel but 2 frequently make the vowel nasal. JD is the usual Creole, as it is the Dutch equivalent of th, both surd and sonant, but the change is not without ex- ceptions ; thus trowe: (throw away), tifi (teeth). The definite article is da (that), the indefinite wan. The personal pronouns are,
Singular, 1. mi. Plural, 1. wi. 2. joe. 2. oenoe, oen, 3. a, hem. 3. dem.
In the 3. sing. a is the unemphatic subject, hem both subject and object. The 2. plur. oenoe, which I am not able to ex- plain, appears in the dialect of the Port Royal islands (Prof. W. F. Allen, in the introduetion to Slave Songs of the United States), in the form oona or ona, used for both singular and plural, but only between friends. Mr. Thomas gives for the Negro English of the West Indies the following pronouns ; me, you, ’e, awe, ayou (all you ?), dem. The personal pronouns
Contributions to Creole Grammar. 165
serve also for possessives, when used attributively either pze- ceding the noun or following it with the preposition va or vo (Dutch van); thus, mi hoso or da hoso vo mi‘ my house.’ The predicate form is, as in the Dutch Creole, di vo mit, ‘ mine.’ 7
Nouns have lost the plural ending and are made plural when definite, by prefixing the 3. plur. pronoun dem; thus, dem hoso, ‘the houses,’ Only the connection determines whether dem has a plural or a possessive force, whether dem hoso is ‘ the houses’ or ‘ their houses.’ In the Port Royal dia- lect dem is similarly employed, though possibly with more of a demonstrative force; thus, sing. that cow, plur. dem cow. In a few cases the Creole has adopted a noun in the plural form, using it however equally for the singular ; thus, jest (ears), ekst (eggs), tifi (teeth), leden ‘limb’ (Dutch ld, pl. leden). In didiebri ‘ the devil,’ the article has become a per- manent prefix ; pl. dem didiebri ‘ the devils.’
Gender is expressed usually by the prefixes man, oeman ; thus, man-hasz ‘ horse,’ oeman-hast ‘mare.’ As a suffix man is sometimes of ccmmon gender ; thus, dansi-man, ‘ dancer’ whether male or female.
The comparative and superlative are formed by moro, thus oeree ‘old’, comp. moro oeroe, super. da moro coerce. Boen ‘ good’ forms the comp. betre, moro betre, or moro boen ; super. beste, moro beste, or moro boen.
The numerals from eleven to one hundred (except twints, 20), differ from the English form; thus, tin-na-wan, ‘ eleven’ (ten to one); dritemtin, (three-times-ten) ‘thirty.’ The iteratives are formed by the suffix tron (turn); as toetron ‘twice ;’ the reduplicatives by doblo (double) ; as dri doblo, ‘ three-fold.’ |
Of the verb ‘to be’ the present is de, the past den be, th fut. sa de. De is apparently the adverb there, the use of which in the combinations ‘ there is,’ ‘there are,’ etc., fur- nishes possibly a link of connection between the adverb and the verb. The usual Creole form for ‘there’ is dapee, (that place) or dédapee (there that place), but de is also used ; thus, a de ‘he is there,’ the copula being omitted, or included in the predicate.
166 A. Van Name,
The verb is reduced to one invariable form, which in the case of the irregular verbs is that of the present infinitive and imperative, seldom that of the preterit or participle, as in last ‘ to lose’ (dost). In the Port Royal dialect, where there is the breaking up of English forms and constructions without the settling down into a uniform usage, which is the second stage in Creole development, the different parts of the verb are employed in a confused way.
The auxiliary of the progressive present is de, which is per- haps the same with de, ‘ to be,’ The analogy of the English ‘he w reading,’ and of the Spanish Creole where the progres- sive tense is formed by the aid of ta, ‘is’ favor this view. De may however be a corruption of do, the use of which in the emphatic and interrogative forms of the English verb would furnish a point of departure for the Creole progressive. In the Negro English of the West Indies the auxiliary is da, in the Port Royal dialect, da or do. Ben is the auxiliary of the preterit, sa of the future ; de go forms an immediate fu- ture; thus, areen de go fadém, ‘ it is going to rain.’
Among the most interesting of the new forms of the Creole are the reduplications. These abound, and are either formed for the occasion, when emphasis is required, as nyoe-njoe ‘brand new,’ or are standing forms, as goesoe ‘ shoe,’ poest- poest ‘ cat,’ wawan ‘alone’ (wan ‘ one’). 3
There are some curious cases of composition. Dasnoti (that’s nothing) ‘pardon’ is a good Creole noun, and may stand as the object of a verb; thus, gi hem dasnoti, ‘ grant him pardon.’ So also od: (how d’ye?) and in the Port Royal dialect huddy ; thus, mi takki odi gi Urbanus, ‘I salute Urbanus,’ literally ‘talk how d’ye to Urbanus.’ Gi (give) is here, like 6a in the French Creole the sign of the dative. Fadém, sidém, lidém are the Creole representatives of the simple verbs, fall, sit, lie. Tra ‘other’ prefixed to tamara. ‘to-morrow, gives irdtamara ‘ day after to-morrow.’ |
Njam ‘ to eat,’ njam-njam ‘food,’ are perhaps from yam, Sp. igname, the tuber which forms a chief article of food. Njam zondei is ‘to keep Sunday.’ rin (clean) is ‘ light’ used both as a noun and adjective ; thus, krin moe kom ! ‘ let
Contributions to Creole Grammar. 167
there be light!’ Cliterally ‘light must come’); poeloe dati moro krin gi wi ‘explain (pull more clean) that to us.’ So in Port - Royal day-clean is used for ‘ day-break.’ For ‘heaven’ the usual Creole word is tapo (top), for ‘ earth’ gron-tapo (ground top).
As in the other Creole dialects, with verbs of motion the direction of the motion is indicated rather by the verb than by the preposition ; thus, komoto na hemel fadém ‘to fall from the sky ;’ hopo na gron ‘to rise up from the ground,’ (hopo from up, or Dutch op ; it is also the Creole for ‘ open,’ and in the Dutch Creole it unites these two meanings).
The relative and interrogative pronouns and adverbs ex- hibit some of the same peculiarities noticed in the French Creole. The demonstratives disi (this) and di (Dutch die demonst. and rel.) are also the relative pronouns, but are frequently omitted; thus, da soema disit,‘he who’, sani dist ‘that which’ are frequently shortened to disoema, sant; soema (some one?) being the Creole for ‘ person’ and sant, for ‘ thing.’ The interrogative ‘who?’ is hoesoma, ’osoema or simply soema ; ‘what’? hoesant or sani; ‘where?’ hoepee or pee (place) ; ‘when ?’ hoetem or tem.
PROCHEDINGS.
AMERICAN PHILOLOGICAL ASSOCIATION.
PRELIMINARY MEETING.
As the need has been felt for some years of frequent meetings ameng the professors, friends, and patrons of linguistic science in America, to give opportunity for mutual exchange of views, for forming more intimate personal acquuintance, and for the general promotion of. philological studies, a correspondence was opened during the summer of 1868, with reference to the organization of a permanent national philological association.
This correspondence resulted in the issuing of an informal call for a preliminary meeting of the friends of philological studies who reside in New-York City and vicinity, for the purpose of discussing the desirableness and feasibility of forming such an association. This preliminary meeting was held on November 13th, 1868, in the chapel of the University of New-York. Over fifty of the prominent linguists, educators, and literary gentle- men of New-York and the adjacent.cities were present. Rev. Dr. Ferris, President of the New-York University, was elected - President, and Professor Harkness, of Brown University, was elected Secretary. Dr. Ferris, upon taking the chair, welcomed the members of the meeting to the university, and expressed his strongest sympathy with the purpose for which the meeting was convened. He then called upon Professor Comfort to state more at length the objects of the meeting. |
“ Professor Comfort commenced his remarks by alluding to the important resulta which are accomplished in Europe, in every branch of learning, by the many societies and associations of both local and national character, which have now been formed in almost every European nation. He spoke especially of the many societies for the promotion of linguistic studies, speci- fic and general, local and national, which exist in the different cities and states of Germany, and of the important influence these societies have exercised in making Germany the home and centre of modern philology.
“From the uniform tenor of the correspondence which had been conducted, he believed there exists a very general desire to have an association formed in America, which shall give opportunity and occasion for those interested in
6 American Philological Association.
philological studies and investigations to meet together at stated periods. Of all the European societies, the ‘Sammlung der deutschen Philologen und Schulminner’ seems, more nearly than any other, to offer a model for us to follow. This association was organized in the year 1837, upon the occasion of the celebration of the hundredth anniversary of the University of Gottingen. It holds yearly sessions during the long summer vacation of the educational year, and is attended by several hundred of the professors of language and philology in the univergities and gymnasiums of Germany and Switzerland. The papers that have been read at the sessions, and published in the journals of this association have contributed greatly to the promotion of philological Science, and the discussions upon linguistic pedagogy have exercised a great influence in elevating in Germany gymnasial and university instruction in the ancient and modern languages and literatures to their present high stan- dard. In France, a somewhat similar society, but confined to one single de- partment of philology, the ‘ Association pour l1'Encouragement des ftudes Grecques,’ though founded but two years ago, gives good promise, both from - the large number of distinguished classical scholars who form its member- ship, and from the high order of its publications, of exercising a powerful influence upon classical education in that country.
“In our own country, a similar work has been accomplished in some other branches of learning by the American Association for the Advancement of Science, which has devoted its labors to the promotion of the natural sciences and mathematics.
“What seems to be adapted to the present wants of America in regard to philological science, is an association which shall be open for membership to all professors of language of respectable standing in our colleges, universi- ties, theological seminaries, and other schools of high education, and to others interested in the promotion of philological studies ; which shall hold its ses- sions annually during the long summer vacation, so that all may be in attendance without being forced to leave the duties of their chairs; and which shall hold its sessions in such different cities in different parts of the country as may, from time to time, be found convenient.
“ The association should embrace in its scope the whole field of philological investigation and instruction. This would necessitate eventually the separa. tion of the work into a number of sections, of which the following might serve as a schedule:
“1. The science of language, and history of philology.
“2. Oriental languages and literatures.
“3. Classical (Latin and Greek) languages and literatures.
“4, Modern European languages and literatures.
Ҥ. English language and literature.
“6, American aboriginal languages.
“7. Linguistic pedagogy.
“ At first, in all probability, one general session would suffice for all the work that would come before the association. The division into sections would take place as the extension of the work of the association may from time to time demand. So broad is the domain of philology that in America, as in Europe, the general association would never remove the necessity, nor could it accomplish the work, of local and specific societies. Indeed, the
American Philological Association. {
American Oriental Society has already by its proceedings and publications gained an honorable reputation in Europe as well asin America. It would be desirable, if possible, to have the first regular session of the association during the coming summer.
“ Professor Comfort closed by reading some of the letters which had been sent by persons who could not be present at the meeting, as Dr. Barnard, Professor of Columbia College ; Dr. Cattell, President of Lafayette College ; Professor De Vere, of the University of Virginia ; Professor Evans, of the University of Michi- gan ; Prof. Tyler, of Amherst College ; Rev. Dr. Peabody, of Harvard College ; Prof. Thatcher, of Yale College, and others, in all of which letters was expressed a strong approval of the project to found a national philological association, and a confidence in its success if conducted with wisdom and energy.”
A general discussion followed, in which Rev. Dr. Crosby, of New-York; Professor Whitney, of Yale College; Hon. J. H. ‘Trumbull, of Hartford; Dr. J. Hart, of Trenton; Rev. Dr. Prime, of New-York; Rev. Dr. Brown, of New-York, and others took part.
Upon the motion of Rev. Dr. Crosby, it was resolved that Professor Comfort be appointed chairman of a committee of arrangements, with authority to add such persons to the com- mittee as he might think advisable, for the purpose of making the necessary preparations and arrangements for holding the first annual meeting of a philological association during the summer of 1869.
The meeting then adjourned.
CONVENTION OF AMERICAN PHILOLOGISTS.
PouGHKEEPSIE, N. Y., July 27, 1869.
THE convention met in the Mill street Congregational church, at three p.m., and was called to order by Professor Comfort. Pro- fessor J. R. Boise, of Chicago University, was elected Temporary Chairman, and Prof. Silber, of the College of New-York, was elected Temporary Secretary.
Hon. J. H. Trumbull, of Hartford, Ct.; Professor L. Kistler, of the North-Western University, Evanston, Il; and Pro- fessor A. H. Mixer, of Rochester University, Rochester, N. Y., were appointed a committee to nominate the permanent officers of the convention, and also to nominate a business committee.
When the committee had retired for deliberation, Professor Comfort was called upon to make some remarks concerning the organization and workings of philological societies in Europe, and especially in Germany.
Rev. Dr. Crosby was requested to read a paper which was sent by Professor Joseph Henry, Secretary of the Smithsonian Institu- tion, upon “The Operations of the Smithsonian Institution in regard to Philology,” of which the following is an abstract :
“ About one fourth of the publications of the Institution have been with reference to ethnology. In the department of ethnology which relates to language, the Institution has published a dictionary and grammar of the Dakota or Sioux language, and a dictionary and grammar of the Yoruba, one of the languages of Western Africa. .
“The Institution has also published: (1) A History of the Archeology of the United States, containing a digest of what had been done previous to 1854, relative to the Philology of North-America ; (2) Instructions relative to Ethnology and Philology of America ; (3) Vocabularies and dialects of different Indian languages and dialects ; (4) Library of American Linguistics ; (5) Gram- mars, vocabularies, and phrase-books of languages in New-Mexico; (6) the Institution is proposing to publish a dictionary of the Nisqually language of Washington Territory.”
President Samson, of Columbian College, Washington, followed in some happy remarks, in which he characterized the paper as an olive branch held out by Professor Henry from natural science to language. |
American Philological Association. 9
The committee upon nominations reported for the business com- mittee Professor George F. Comfort, of Alleghany College, Pa., ° Rey. Howard Crosby, D.D., of New-York, and Professor Oscar Howes, of Shurtleff College, Illinois.
During the retirement of the business committee, an interesting but informal discussion was carried on with reference to many philological questions, especially with reference to the pronuncia- tion of certain common and proper names in the English, French, and German languages.
Upon the recommendation of the business committee, the con- vention adjourned, to meet at eight o’clock in the evening.
EVENING SESSION, JULY 27.
The committee upon business reported, as the order of sessions of the convention, that there will be three sessions daily ; from 9a.M. to 12.30 p.m.; from 3 to 6 P.m.; and from 8 to 9.30 P.M.
The report of the committee upon permanent officers for the convention, which was read, accepted, and adopted, presented the following officers :
President, Professor William D. Whitney, of Yale College.
Vice-Presidents, Professor A. Harkness, of Brown University, and Dr. B. W. Dwight, of Clinton, N. Y.
Secretary, Professor George F. Comfort, of Alleghany College.
Prof. W. B. Silber, of New-York College, was appointed Assis- tant Secretary. |
As the chairman of the local committee, C. J. Buckingham, Esq., was unable through temporary illness to be present, his address, welcoming the convention to the hospitalities of the city of Poughkeepsie, was read by Benjamin J. Lossing, Esq., the historian. Mr. Buckingham spoke of Poughkeepsie, with its more than two thousand students from abroad in its various institutions of learning, as the Athens of the Hudson. He assured the mem- bers that the citizens of Poughkeepsie would spare no pains to make the session of the convention in the city agreeable and pleasant. Poughkeepsie will feel itself honored in receiving as guests so many of the distinguished philologists of America. We are glad also to see in your number representatives from other countries, as Germany, Russia, France, and also Americans, whose labors in investigating the languages of the Indians have been an honor to our country. Mr. Lossing added some interesting re-
10 American Philological Association.
, marks with reference to the historical associations connected with
Poughkeepsie and vicinity.
Professor Whitney, the President of the convention, replied to the address of Mr. Buckingham, acknowledging the kindness with which the convention had been welcomed, congratulating it upon the happy circumstances amid which it had assembled, and ex- pressing the hope that its sessions and transactions would result in advantage to the interests of sound learning and education.
Short speeches were then made by a number of the members of the convention.
Dr. 8. H. Taylor, of Andover, Mass., spoke of the desire he had felt for many years of conventions and associations among the teachers of linguistic science. He gave an account of some meet-
ings which have been held by the professors of language in Massa- chusetts, and closed by expressing his great gratification at now seeing face to face so many professors of national reputation whom he had never before had an opportunity of meeting, adding the hope that this would be the first of a regular series of annual meetings, and that the influence of these meetings might extend to all the schools in which language is taught,
Honorable H. G. Squier, of New-York, gave an interesting sketch of the character, grammatical structure, and richness in vocabulary of the language of the ancient Incas in Peru.
: Mr. P. C. Bliss, late of the American Embassy to Paraguay, gave an account of the Indian languages in Paraguay, and the northern part of the Argentine Confederation.
Mr. B. J. Lossing spoke upon the variety in the ways of spell- ing and pronouncing the name of Poughkeepsie, and upon which of these is correct.
, Speeches were also made by Rev. C. H. Brigham, of Ann Arbor, Mich., Professor J. R. Boise, of Chicago University, Pro- fessor L. Kistler, of the North-Western University, Dr. H. von Holst, of Heidelberg, Germany, and Dr. Crosby, of New-York.
Letters were read from Dr. Tayler Lewis, of Union College; Dr. W. S. Tyler, of Amherst College; Hon. Charles Sumner, of U. 8. Senate; Gen. Garfield, M.C.; Col. W. T. Higginson, of Newport, R. I.; Rev. Dr. E. O. Haven, of Michigan University, and others, expressing regret at not being able to be present, and a desire to codperate in the promotion of the objects of a phi- lological association, should one be formed.
Upon motion of Rev. Mr. Wheeler, of Poughkeepsie, resolutions
American Philological Association. 11
were passed of condolence with Professor S. F. B. Morse, of this city, who was prevented by a severe and serious accident from attending the sessions of the convention, as had been his inten- tion and desire.
Mornine Session, WEDNESDAY, JULY 28.
Professor C. M. Mead, of Andover Theological Seminary, was appointed additional secretary.
Upon the motion of Professor Comfort, a committee, consisting of Hon. J. H. Trumbull, Rev. Dr. Crosby, and Professor Had- ley, was appointed to prepare a plan for organization of a perma- nent society.
The business committee announced that the morning session of the day would be devoted to subjects connected with instruc- tion in the ancient classical languages, and the afternoon session to the modern languagés.
The first paper of the morning was read by Professor J. B. Feuling, of the University of Wisconsin. It treated of “The Best Method of Pronouncing the Latin and Greek Languages.”
“Professor Feuling considered the main point in pronouncing Latin and Greek to be the quality of the vowels, whether they are long or short. The methods in use in different countries differ ; and especially, s0 various are the methods followed in America, that it will be difficult to establish any arbitrary rules. In addition to the so-called English and continental methods, he spoke of another which might be termed the American method, being a mixture of elements taken from both of the others. He urged the continental method as the least open to objection. He advocated also very strongly the obser- vance of the written accent in pronouncing Greek, and that the quantity of the vowel should be observed in both Latin and Greek.”
Professor 8. S. Haldemann, of St. John’s College, Maryland, ’ followed with an essay upon the same subject.
“He agreed with many of the views presented by Professor Feuling. Under the influence of the increased facilities for travel, modern opinion favors the pronouncing of proper names of persons and places in accordance with the rules of the language to which the names belong to an extent which would have been considered some years ago to smack of affectation. We are thus becoming acquainted with various ways of pronouncing the same letter, and it will not seem so strange and unnatural as formerly to adopt another than English mode of pronouncing the vowels in Latin and Greek. He criti- cised at great length the ordinary method of explaining long and short vowels, as the terms are used by English grammarians, showing that the terms open
12 “American Philological Association.
and closed should frequently be used where now long and short are employed: and that the length of a vowel should refer only to the duration of the sound. He illustrated the distinction between these terms by many examples drawn from the English, French, German, Latin, and Greek languages. He insisted upon giving distinct vowel sounds to both vowels in diphthongs, in the Latin and Greek languages. He objected to the claim which is made by many of the advocates of the continental system of pronunciation, that this system contains the original vowel sounds of those languages. The most important advantage which would come from adopting that system would be that it would secure uniformity in the different schools of the country.”
Professor Whitney then read a paper which had been prepared by Charles Astor Bristed, upon the same subject.
“Mr. Bristed treated at length of the pronunciation of certain vowels. He was persuaded that, as it is now impossible to ascertain the original pro- nunciation of Latin and Greek vowels, it will be useless and unimportant to attempt to establish any uniform method. He criticised the neglect of the study of prosody, which is so general in American schools and colleges. The written accent he regarded as having no binding force upon pronunciation, and as never having had any such significance.”
After reading the paper, Professor Whitney remarked that it would without doubt be much more imperative to adopt a uni- form system, if we could be sure that we understood the sounds of the Greek and Latin languages, so that a Greek or Roman could understand us when reading or speaking either of those languages. But, upon the other hand, students and professors will the more willingly yield any system which they may have been following, in order to secure uniformity, from the fact that all the systems are in reality arbitrary. He objected to Mr. Bris- ted’s views about Greek accent, believing that it should be ob- served in pronunciation. He held it also to be very important to give the proper quantity to the vowels,
Dr. A. N. Arnold, of Madison University, Hamilton, N. Y., read a translation of a treatise by Mr. Rangabd, the late represen- tative of the Greek government to the United States, upon the pronunciation of classical Greek, and treating especially upon the conflict of prosody and accent.
“Dr. Arnold then made an earnest argument in favor of pronouncing classical Greek in the same way as modern Greek is pronounced. He argued that, as the Greek has never ceased to be a living language, there is every probability that the pronunciation of the vowels in modern Greek is much nearer that in classical Greek than any other system we can adopt. Again, many educated Americans are constantly visiting Greece, and many Greek
American Philological Association. 13
merchants are establishing themselves in our chief commercial cities. Thus it is becoming more and more easy for the modern Greek pronunciation to become diffused through America. He alluded to the fact that Mr. Rangabé attended the commencement of an American college, where speeches were made in English, French, German, Latin, and Greek, all of which speeches Mr. Rangabé could understand, except the one in Greek.”
Dr. Crosby opposed the views advanced by Dr. Arnold, on the ground that it is impossible to read ancient Greek rhyth mically by the system of modern Greek pronunciation.
Professor Feuling remarked that the Greek professors at Athens were anxious to introduce the system which is followed in the German universities, and which is essentially the same as that which is termed in America the continental system.
Rev. H. M. Colton, of New-York, expressed the hope that some uniform system would be agreed upon, as the teachers in acade- mies and preparatory schools felt the embarrassments extremely while preparing students for different colleges, in which different systems of pronunciation are followed.
Dr. Dwight alluded to the fact that in every nation Latin is pronounced according to the system of pronunciation of its own language. He sympathized strongly, however, with the desire to have some one uniform system adopted in all the colleges and classical schools of America. He moved, therefore, that a com- mittee of five be appointed to take the subject into consideration, and to present a report expressive of the sense of the convention.
This motion was discussed at length by. Rev. Mr. Anderson, of Danbury, Ct.; Professor Mixer, of Rochester University ; Messrs. Bisbee, Weston, Gallup, and Raymond, of Poughkeepsie ; Profes- sor Boise, of Chicago; Dr. Taylor, of Andover; Professor Magill, of Swathmore College. Before a vote was taken, the convention adjourned.
» AFTERNOON Session, JULY 28.
The discussion was resumed by Rey. Dr. Brown, of New- York, President Samson, of Washington, and others. An amend- ment offered by Professor Harkness was adopted, authorizing the committee to defer making their report till next year. The mo- tion to appoint the committee was then carried. The president then appointed the following persons upon the committee: Dr. Dwight, of Clinton, N. Y.; Dr. Taylor, of Andover, Mass., Pro-
14 . American Philological Association.
fessor Feuling, of the University of Wisconsin; Dr. Arnold, of Madison University, and Professor Hadley, of Yale College.
Professor Harkness, of Brown University, then read a paper upon “The Best Method of Instruction in the Classical Lan- guages.”
“ After tracing the characteristic features of the various methods in vogue of teaching the classical languages, Professor Harkness pointed out the excellen. ces and deficiencies of many of them, and then spoke of an ideal system of instruction, of which the following are the salient points :
“The study of the classical languages should, like all other studies, be con- ducted in such a way as to develop the intellect of the student in the most profitable way. In America, regard must be had to the peculiar bent and habits of the American mind ; hence the folly of introducing the German, or any other foreign system of studying the classics. It is impossible for a stu- dent, at the age when scholars begin the study of the classics, to learn Latin or Greek in the same way in which they, learn their mother tongue. Upon the other hand, it is not wise for the student to begin by crowding the mind with all the details of grammatical rules. The knowledge of grammar and of words must progress in equal proportion. Instead of training the student to make translations, and make references mechanically to the grammatical rules, the teacher should initiate the student into the spirit and general structure of the languages, and into the principles upon which the gramma- tical rules are based. As to words, the student should not only learn their meanings, but also the modification of signification which the word receives in the particular passage. The former is inherent in the word ; the latter de- pends upon the context. It isan invaluable exercise to read Latin or Greek passages through understandingly, without going through the formality of translating them. _ It is also useful to commit to memory choice passages from the classical authors. - Composing in Latin and Greek may be useful to a cer- tain extent; but this exercise should be regarde] as a means and not as an end. In the study of the classic authors, it is doubtful whether any impor- tant change can be made to advantage in the selection of works to be read, as they now stand in the college curriculum. To study the work of an au- thor successfully, the student must acquire a thorough knowledge of the lan- guage, geography, history, and mythology of the country. He must also have a delicate appreciation of beauty of form and expression of language. In taking up a new author, the student should acquaint himself with the chief facts connected with the life of the author, with the contemporary history of the country, and the relation of the author to this history. Much of this pre- paratory knowledge will necessarily have to be given by the professor in lec- tures. The professor can also suggest side courses of reading, as may be ap- propriate to the particular subject, and the attainments of the particular class of students. Attention should be specially paid to grammatical analysis, to the study of words, and to the logical scope of thought in the mind of the author. Much insight can be given into the nature of the classical languages by ox- plaining the meaning and use of such words as have no synonyms in the Eng- lish language, and by drawing the attention of the student to points of simi-
American Philological Association. 15
larity and difference in the grammatical structure of the ancient and the English languages. One of the most beneficial exercises in the study of the classical languages is to make elegant and precise translations into English.”
Rev. H. M. Colton, of New-York, then read a paper upon the same subject.
“He spoke of the importance of clear views as to method of instruction in language in general, as to the method peculiarly applicable to the study of the classical languages, and, finally, of the discretion which every teacher must exercise with reference to the modification of his method with particular classes.”
Professor E. H. Magill, of Swarthmore College, Pa., read a paper upon “The True Foundation for a Course of Linguistic Studies.”
“ Professor Magill argued that the first study of language should be of one’s own native tongue ; that the child should receive practical—not theoretical —instruction in some one modern language at: as early an age as when five years old, beginning a second modern language two years later; that the formal study of language should begin at ten years of age, when the child should begin the study of Latin, commencing Greek about two years later ; that the formal study of the modern languages should be commenced after the child has made some progress and proficiency in the study of the ancient lan- guages. This he considered the most natural, practical, philosophical, and effective method of studying language.”’
The subject was then thrown open for general discussion.
Professor Haldeman illustrated by many striking and entertain- ing examples his theory of the usefulness of learning the meaning and derivation of words in the classical languages, by associating them etymologically with words in the English language.
Professor Feuling described at length the method of instruc- tion in the classical and modern languages which is followed in the German gymnasia, and stongly recommended its adoption in the colleges and academies of America.
Professor Boise urged the importance of written exercises in the study of Latin and Greek.
After further remarks by several other members, Dr. Crosby, chairman of the committee which was appointed to draught a plan of organization for a permanent national philological society, an- nounced that the committee was prepared to report a constitution for the society. The following is the text of the constitution, which, after a short discussion, was adopted unanimously :
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 know- ledge. 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 first 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 de- termined 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.
ARTICLE IV.—MEMBERS.
1. Any lover of philological studies may become a member of the associa- tion bv 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. Thero shall be an annual fee of five dollars from each member, failure in payment of which for two years shall ipso facto cause the membership to cease. .
8. Any person may become a lifemember 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.
American Philological’ Association. 17
2. Publications of the association, of whatever kind, shall be made only under the authorization of the executive committee.
ARTICLE V1.—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.
Upon motion, it was resolved that at twelve o’clock on Thurs- day, July 29th, the convention should adjourn sine die, and that the members should meet immediately thereafter, and form the American Philological Association.
It was voted that all members of this convention who shall sub- scribe their names to the above constitution, and who shall pay five dollars before twelve o’clock on Thursday, shall be ipso facto the founders and first members of the new association.
Professor Kistler was appointed to receive the fees of member- ship until a treasurer shall be elected.
EVENING SESSION, JULY 28,
Dr. Dwight, of Clinton, N. Y., read a paper upon “ The Desira- bleness of thorough Classical Study to the Attainment of the Ends of the Higher Education.”
“He stated the claims of ‘The New Education,’ and characterized them as assumptive. The people at large should be furnished with the best educa- tion for their circumstances. The great question is, What kind of education shall be given to the favored few, whose advantages, time, and circumstances give them the position of leaders in the intellectual movements of the world ? He held that there are three kinds of training which demand our attention— the practical, scientific, and classical. The American college system is the outgrowth of our own social state, and is better adapted to our wants than any European system can be. There is a great popular demand for a modifi- cation of the college course. If the course of college study could include an additional preparatory year, and also be extended another year in the higher studies, many modifications for the better might be made. Our democratic institutions require especially a high standard of national education, or we shall degenerate as a people ; and the colleges must continue in the future, as they have done in the past, to furnish the leaders in society and in the state. Two questions force themselves upon us: the true end to be sought for in the higher education, and in what way our classical institutions shall secute these ends. These ends are found in the development of intellectual power, securing the best using of this power, and the capacity of patient, per- severing labor, directed to worthy objects. The culture which will produce .
18 American Philological Association.
this discipline requires time and a varied course of training. The scientific course, Dr. Dwight held, can not produce this. On the other hand, the clas- sical course has done so in the past, and can do so in the future. He consi- dered scientific and elective studies useful adjuncts, but poor substitutes for classical studies.”
The reading of the paper was followed by an animated discus- sion, in which Rev. Mr. Maury, Professor Feuling, Doctor Von Holst, Rev. Mr. Brigham, Professor Kistler, and others took part.
Doctor Crosby offered the following resolutions, which were passed :
“ Resolved, That this convention tender its hearty thanks to the trustees of the Mill street Congregational church, of Poughkeepsie, for their generous offer of their beautiful and commodious edifice for the use of the convention ; to the people of Poughkeepsie for their large-hearted hospitality ; to the sexton of the church for his faithful attention and codperation ; and to the local com- mittee, (and especially to its chairman, Mr. C. J. Buckingham,) for their elabo- rate and munificent preparations, which have contributed so largely to the comfort and success of the convention.
“ Resolved, That we gratefully acknowledge the value of the labors of Pro- fessor Comfort, in initiating and maturing the design for the establishment of the Philological Association, fully aware of the perplexity and manifold details of such a work so thoroughly performed.”
Professor Harkness offered a resolution, which was passed, re- turning the thanks of the convention to the publisher, Mr. Draper, of Andover, Mass., for his gift to each member of the convention of a copy of the Poema Admonitorium of Phocylides, edited by Professor Feuling ; also that the thanks of the convention be tendered to Professer Feuling for dedicating the book to this con- vention.
MorninG Session, THurspay, Juty 29.
[As nearly all the members of the convention accepted the invi- tation of the trustees of Vassar College to visit that institution, the session did not commence till ten o’clock.]
Upon a motion being passed to that effect, the president ap- pointed Doctor Raymond, Professor Harkness, and Mr. P. Bliss a committee to report upon a project with reference to phi- lological investigation, which Mr, Leurio, the originator, had brought before the notice of the convention.
Doctor Crosby read a paper by Doctor Schliemann, of Paris, in which it was urged that at least one half of the time of the colle-
American Philological Association. 19
giate course should be devctel to the study of languages, and vuai but a single Janguage, in addition to the student’s mother tongue, should be studied at 1 time. Dr. Schliemann also recom- mended that a chair of the American Indian languages be added to the faculties of our universities. |
Doctor Crosby then gave the substance of a paper sent by Rev. Mr. Burnham, of Minnesota, in which many of the common errors ot grammarians and teachers of langnage were pointed out.
Professor Schele de Vere, of the University of Virginia, read a paper upon “ The Critical Study of the English Language :”
“ Professor De Vere began by speaking of classical education as the founda- tion of all true culture, even though it is in so many cases out of sight, like the foundations of buildings, although supporting the superstructure of later and practical culture and training. It is an oft-repeated error that we are to learn our mother tongue without effort from our parents and from those with whom we come in contact. There is then the utmost importance in educat- ing the rising generation, who are to be the parents of the coming gene ration, so that they will give good models of language to their children. And in addition, we are receiving yearly half a million of people from Europe, who are but grown-up children as to their knowledge of the English language. And besides teaching our own young children, and our grown-up adopted children, we are called upon now and then to give a lesson to our ancestors across the water. It is an acknowledged fact that the English is spoken in greater purity in America than in England. What is there the privilege of the educated classes, is here the common heir-loom of all. A certain spirit of fashion in the use of words also prevails in England, by which the language is much more often perverted than in America. Hence the strange result that the new country becomes the guardian of pure old English. Our authors, as Hawthorne, make good old English words once more familiar to English ears. We aro then in a sense the conservators of the English language, We are to kcep it pure and in a healthy growth by having intelligent ideas as to its nature, mechanism, structure, and growth, and also of its beauty, grace, and highly spiritual character. Professor De Vere then alluded to the circumstances which have Jed to the prominence of the Latin element in the English language. He drew the line between the Latin and the Ger- manic elements in the English language, and he urged that the student and child be taught to look with reverence upon his mother tongue, and that in- structors give the language the prominence it deserves in our educational system.” |
- Mr. J. Pierson, Principal of Belvidere Academy, N. J., read a paper upon “The Critical Study of the English Language in our
Academies and High-Schools.” ‘Mr. Pierson alluded first to the increased attention that is now paid to the
study of the English language and literature in ourcolleges. He then spoke at length concerning the necessity of a better method in those institutions
20 American Philological Association.
which are immediately below the college—the academies and high-schoola. For it is in schools of this grade that the great mass of the population of the country get whatever they do acquire of a critical knowledge of their native language. Most especially is this true of the female part of the population, as they are not admitted to colleges. Mr. Pierson advocated the plan of tak- ing up short extracts from authors, and examining them with the greatest minuteness, and elucidating all points connected with the history, geography, the customs, laws, and political condition of the country, and of the times in which the authors lived and wrote. There should also be a critical analysis of all the philological relations which are involved in the few passages that are read. The derivation and history of all the words should be carefully traced. This method of critically examining a few passages from an author, it was urged, would be found far more advantageous than to run rapidly through many pages, and simply parse the words. It was also urged that more should be made of the lecture system in academies and high-schools. The students should have opportunities to hear three or four lectures a week, upon branches which they are studying. This would best increase the value of their recita- tions.”
Doctor H. N. Day, of New-Haven, indorsed most fully the views of the two previous speakers. The English language is in a “formative period ” to-day, as are all living languages. For it is the nature of language to be ever changing. Changes are taking place in thé meaning given to words in daily use. New words are constantly coming into use, and old words insensibly pass out of use. Our orthography is changing, in most cases for the bet- ter. The laws of syntax, though more permanent, are still also undergoing modification. It is a great mistake to have the Eng- lish language taught, even to the young, as though it were per- manent in form and unchangeable in laws, like mathematics or other abstract sciences.
Professor Haldemann made some very entertaining remarks, illustrative of the changes in pronunciation which have taken place within the last thirty years.
President Samson urged the importance of studying the English language, even in the academies and high-schools, in a strictly philological method, showing clearly what are the Saxon and what the Latin elements, both in the grammatical forms, the syn- tax, and the vocabulary. )
Upon motion of Professor Harkness, resolutions of thanks were passed to the trustees, the treasurer, and the president of Vassar College, for the opportunity given the members of the convention to visit Vassar College. |
The committce to which was referred the consideration of the
American Philological Association. 31
best method of pronouncing Latin and Greek announced that, after several sessions, they found themselves able to make a par- tial report at this meeting of the convention, and asked permis- sion to present the report which they had prepared. Upon mo- tion, such permission was voted unanimously.
The report was as follows :
“While the committee recognize the fact of wide diversities of opinion among American scholars concerning the mode of pronouncing the classical languages, and while among the members of the committee themselves there are considerable differences of judgment, they agree in stating that they deem some uniform system of pronouncing the classical languages to be greatly to be desired, if possible to be obtained.
“They would favor, as at least one feasible step toward such a result, the adoption of the continental system of pronouncing the vowels in both Latin and Greek. |
“It is also their opinion that tt ts desirable to observe the written accent tn reading Greek, and also to mark the quantity of vowels in reading both Greek and Latin.”
The report of the committee was greeted with loud and conti- nued applause. Upon motion, the report of the committee was adopted as an official expression of the convention upon the question of how the classical languages should be pronounced.
The hour of twelve (noon) having arrived, at which time it was voted yesterday that the convention should be dissolved, upon the motion of Dr. Crosby, the convention adjourned sine die.
AMERICAN PHILOLOGICAL ASSOCIATION.
THURSDAY, JULY 29.
Tae Association was called to order at twelve, (noon.) Dr. S. H. Taylor was elected Temporary Chairman, and Prof. William B. Silber was eiected Temporary Secretary. Upon the motion of Professor Magill, it was resolved that the officers of the conven- tion which had just adjourned be elected as the permanent officers of the association.
The officers were, therefore, as follows :
President, Professor W. D. Whitney.
Vice-Presidents, Dr. B. W. Dwight, Professor A. Harkness.
Secretary and Curator, Professor G. F. Comfort.
Honorable J. H. Trumbull was elected Treasurer.
Professor W. B. Silber and Professor C. M. Mead were elected additional secretaries for the present session.
A resolution of thanks was tendered te Hon. E. G. Squier, for the presentation to the Association of a printed bibliographical list of books upon South-American literature. |
A resolution of thanks was also voted to Dr. Schliemann, of Paris, for a volume of the transactions of the “ Association pour Encouragement des Etudes Grecques.”
A committee of three, consisting of Doctor Samson, Professor Mead, and Professor Magill, was appointed to nominate the re- mainder of the executive committee.
Mr. J. E. Munson, of New-York, then read a paper upon “ The Relation of Phonetics to Philology.”
“ He introduced the subject by tracing the progress of opinion in England and America in favor of phonetic spelling. He quoted especially from the writings of Mr. Max Miller, Dr. Latham, Mr. Pitman, Professor Whitney, and Professor Haldemann. Probably the study of other languages, especially of the Sanscrit language, has contributed largely to the change of sentiment among scholars upon this subject. Mr. Munson then compared the different phonetic alphabets. He gave the preference to the one employed by Mr. Alexander Melville Bell, in his system of ‘visible speech.’ By adopting this system, learning to read would be so simple that it would be acquired of
American Philological Association. 23
an ordinary child in a few days. Learning to spell would be robbed of its terrors, and the time that is now worse than wasted in learning absurd sys- tems of spelling would be applied to gaining useful knowledge. Some of the greatest and earliest obstacles to acquiring an education would be removed. That this is no fancy picture is shown by the testimony of many travelers who have been in India, and have seen there ths similar method of writing which is adopted by the Hindoos. The plan of teaching children to read first phonetically, and then to read by the ordinary method, has been frequently tried in England, and found to consume less time than to begin by the usual method. A system has also been invented whereby the phonetic form of the word is stamped upon the word as spelled by the ordinary method. Another of the most obvious results of learning first by the phonetic method is the check that this method would have upon the tendency to change the sound of words and letters. As it is now, changes are introduced by speakers ; not being recorded, they go on increasing until the spelling of a word often gives no idea of its pronunciation. By the phonetic method, the first variation of a public speaker from the standard pronunciation would be recorded, and thus the history of changes could be traced.”
The reading of the paper was followed by a discussion upon the possibility and desirability of adopting phonetic spelling in the English language.
The nominating committee reported the names of the following - persons for filling the remaining number in the executive com- mittee, all of whom were unanimously elected: Dr. 8. H. Taylor, Dr. H. Crosby, Professor Schele De Vere, Professor J. R. Boise, C. J. Buckingham, Esq.
AFTERNOON SEssion, JULY 29.
Professor A. H. Mixer, of Rochester University, read a paper upon “The True Position of the Modern Languages in our College Curriculum.” :
q
“The chief struggle in systems of education for the last three centuries has been with reference to the position which should be accorded to the study of the classics. In England, the classics have so far retained their ground. In France, they have yielded to the demands of natural science and modern Jan- guage to such an extent as to endanger the effectiveness of the national edu- cational system. Germany alone has been little moved by the cry for reform. There the classical languages hold essentially the same prominence which they have held for the last two centuries. In America, the proportion of young men in our colleges to the whole population of the country is less than formerly. Some attribute this to the conservative spirit in the colleges. It is rather owing to the avenues of wealth that are opening to young men, calling them off from any system of study whatever. Still, some concession must be made to the sciences and the modern languages. But this must not
24 American Philological Association.
be done at the expense of the study of the classic languages. For the study of these languages is the real basis of all true education. After the mind has become trained to right habits of thought, it is prepared to receive the facts of | natural science and training in the modern languages. Thereshould be a unity of plan in all the instruction in language. In the study of Latin or Greek, words should be traced from their earliest appearance, down through the modern languages, to their entrance into the English language of to-day. In the study of the modern languages, words should be traced back to their inost primitive radical forms, and forward to their latest modification of form and meaning. The modern languages, French and German especially, may be employed as the channel of communicating a knowledge of many branches of learning. French or German text-books may be employed in the higher mathematics, in science or history, especially in the history of literature. The student will thus be familiarized with these languages in connection with his other collegiate studies, and the modern languages wi-] have a prac. © tical value which can not be realized in any other way. So grent, indeed, is the quantity of literature upon every branch of knowledge in the French and German languages, that no man can call himself liberally educated to-day who has not a knowledge of these languages. As to the proportion of time, in a liberal system of education, which should be given to the study of lan- guage, the speaker thought that one half of all the work in academic and college curriculums should be thus employed. Of this half of the entire time, two thirds should be given to the ancient and one third to the modern Jan- guages. But an equally rigid philological method should be followed in the study of both these classes of languages.”
Rev. Mr. Maury, of Cold Spring, N. Y., read a paper, in which he defended the classical languages, and the prominence which they now occupy in our educational system.
Professor Comfort spoke upon “The Importance of Post-Gra- duate Instruction in Language.”’
“We have four grades of schools—the primary, academic, collegiate, and post-graduate, (or the university.) In the post-graduate schools, only medicine, law, theology, and some branches of natural science are taught at present. In the universities of continenta? Europe, all branches of human learning are carried equally far. In the University of Berlin, there are fifteen professors of philology and language, who give over seventy courses of lectures each year upon language and literature. In America, we have no means of giving any higher instruction in language than that which is to be had in the colleges. The discussion with reference to the position of language in education can never be settled in our country until we provide means for post-graduate instruc- tion in language and philology. This can be accomplished by establishing schools of philology in connection with existing colleges, like the School of Mines of Columbia College ; by having separate institutions, like our medical and theological schools ; or by having philology form one department of a great post-graduate university. Teachers for the common schools are usually educated in our academies ; teachers for the academies are educated in the colleges ; 80 we must have universities where professors for colleges can get
American Philological Association, 25
the education and training which are necessary for their position and duties. It would be very easy to select from our own number ten or fifteen profes- sors who would man the philological faculty of a post-graduate university quite as successfully as do the professors in the faculties of our scientific schools. The cost of such a university in America would be very great, but it would not be incommensurate with the great present and prospective wealth of our country. The Duchy of Baden, in Germany, with one eighth the territory, one third the population, and probably less than one fourth the material wealth of the State of New-York, has two (post-graduate) universi- ties, with a hundred and fifty professors and twelve hundred students. So great is the solidarity among nations to-day, that it is as impossible to keep the institutions of high culture which abound in every state of Europe from being soon established in America, as it is to keep telegraphs and railroads from being established in Japan. It may bea generation before such univer- sities are founded in America, but they are needed to-day.”
In the midst of the discussion which followed, and which was participated in by Dr. Crosby, Dr. Dwight, Dr. Samson, Profes- sor Feuling, Dr. Raymond, and others, Dr. Dwight made a mo- tion that a committee of five be appointed to consider the prac- ticability and desirableness of securing somewhere in the land, at an early date, the establishment of a school of instruction of the highest and best kind in linguistic science, corresponding in grade and in the style of its advantages to the schools of science lately established in connection with several of our colleges. [In the evening session, the discussion was continued, and, upon mo- tion of Dr. Raymond, the motion was laid upon the table till next year.] .
The committee appointed to consider the proposition of Mr. Leurio made the following report, which was adopted : “That the raising of funds for such purposes does not fall within the pro- vince of this association.”
Evenine Sxsston.
Hon. J. H. Trumbull, of Hartford, Ct., read a paper on “The True Method of Studying the North American Languages.”
“ As preliminary to the consideration of the means and agencies to be em ployed ‘to secure from destruction the languages of the Indians of America,’ Mr. Trumbull offered some suggestions as to the method which should be ob- served in the collection and arrangement of matorials for the study of these languages. He pointed out the disadvantages of exclusive reliance on the use of standard vocabularies of English words, for the most part concreto names, as the frame-work for collections and for the exhibition of results.
26 American Philological Association.
Comparative vocabularies are valuable for the classification of provisional languages by families or groups. The work of the linguistic scholar begins when this provisional and tentative process of the word-collector ends. It is desirable to know not merely what a language ts like, but something of what tt 18.
“From fundamental differences in the grammatical structure and in the plan of thought of the American and the Indo-European languages, Mr. Trumbull argued that it is nearly if not absolutely impossible to find any Indian name or verb which can be exactly translated by an English name or verb. The standard vocabularies are framed on the mistaken idea that such translation ts possible. They assume that English analysts may be adequate- ly represented, word by word, by Indian syntheats.
“The aim of the collector and student shouid be to resolve synthesis by analysis. Every Indian bunch-word is a sentence, of which the translation should be sought, and such a translation can not often be cast in the mould of an English vocabulary. The Indian languages must be studied in their primary words and roots, not in their polysyntheses.
“To what extent suc: analysis can be prosecuted successfully, is not the question. Every step toward it is something gained, and without it no real advance can be made. In collecting materials, special attention should be given to the most simple forms, that is, generally, to the shortest words. Every concrete name should be analyzed and translated, not merely set ina prepared vocabulary against the English name of the same object, animate or inanimate.
“In closing, Mr. Trumbull protested against the generally received notion of the Indian process of word-making. Duponceau adopted the statement of Egede that words are formed ‘ by taking and joining together a part of the radical words which are to be combined.’ Heckewelder and Schoolcraft help- ed to give this statement general currency, and the latter believed that the arrangement of the syllables depends largely ‘on the will or the skill’ of the word-maker. This is not only without foundation in fact, but it tends to es- tablish a doctrine of Indian synthesis which is directly opposed to the true one.”
The Rev. Thomas Hurlburt, of Caistorville, Canada, read a paper “On the Structure of the Indian Languages,” which he introduced by an interesting account of his life and labors among the Cree and Ojibwa Indians.
“ He had been for nearly forty years a missionary to these tribes, in British North-America ; had preached in the Cree and Ojibwa languages for thirty years, and for some time had published an Ojibwa newspaper, setting the types himself, and doing the work of printer as well as editor. He had be- come so familiar with the language that he was accustomed to dream in it, ‘and the Indians insisted that he was not really a white man, but ‘an Indian in a white man’s skin.’
“ He spoke of the Indian languages as the most exquisite pieces of mechan. ism ever presented for human study. In their grammar we have a transparent structure, through which we can study the process of word-making, as we may
American Philological Association. 27
watch the building of cells within a glass bee-hive. The number of roots or primaries is small, but the number of words which may be formed from these by prefixes and suffixes, by combination, etc., is almost infinite. In tlic Ojibwa language, a single verbal root may receive 200,000 modifications, and each of these may again be conjugated as a verb, through mode and tense forms. He estimated the number of words possible to be formed from a sin- gie Ojibwa root as not less than 17,000,000. Some of the peculiarities of the grammar were illustrated by examples.
“Mr. Hurlburt presented to the association copies of the New Testament in the Cree and Ojibwa languages, and a manuscript Ojibwa grammar, on which he had been more than fifteen years at work. He proposed to revise and perfect this grammar for publication, and hoped to accomplish the task in about two years.”
Mornine Session, Fripay, Juty 30.
Upon the recommendation of the executive committee, it was voted that the next meeting of the association be held at Roches- ter, New-York, on the 26th of July, 1870. |
Upon the motion of Dr. Crosby, it was resolved that the asso- ciation adjourn to-day at half-past eleven a.m.
Rev. H. M. Colton, of New-York, remarked that the text-books which are prepared for use in academies and preparatory schools are too difficult for the students to use profitably. He urged the general adoption of easier text-books in preparatory schools,
Professor Comfort gave an abstract of a paper.which was sent by Dr. Roehrig, of New-York, upon the languages of the Dakota Indians.
Hon. Mr. Trumbull, of Hartford, read a paper by Mr. George Gibbs, of New-York, with reference to the question, “ What more efficient measures can be taken to preserve from destruc- tion the aboriginal languages ?”
“The object of the paper was more particularly to point out the bearing of philology upon the classification of the Indian tribes, and the importance in an ethnological view in preserving their languages. After stating briefly what had already been done in this respect by different nations, and that with few ex ceptions the vocabularies of the languages north of Mexico were sufficient only to ascertain the most palpable relations among them, it suggested an_ enlarged standard of comparative philology, to comprise, first, a vocabulary of not less than fifteen hundred selected words, as nearly as possible radicals, arranged according to subjects, and having reference to geographical condi- tions ; second, a large number of well-digested phrases based upon these words, calculated to draw out the different forms of speech, from which their gram- matical structure could be deduced ; and finally a succinct and popular state-
28 American Philological Association.
ment of the most striking grammatical peculiarities of languages already known, as a guide to the study of others. This work to be distributed widely as a hand-book. As a general thing, the collection of material would be the task of a different class of men from those who are ultimately to dissect and arrange it, and it is only by popularizing the subject that we can hope to accumulate that material. In the mean time, however, it is desirable that the general principles of these languages should become a part of our college studies as a branch of universal grammar. The paper closed by urging the adoption of the Roman alphabet, as recommended by Professor Whitney, and adopted by the Smithsonian Institution as the most suitable and intelligible form for writing.”
Professor Haldemann read a second paper upon the same sub- ject, by Hon. E. G. Squier, but referring more especially to the languages of Central and South-America:
“ Under the Spanish conquest and dominion, in spite of the beneficent and humane efforts of the ‘Council of the Indies,’ many Indian tribes, and, indeed, entire Indian nations, were exterminated entirely, or were incorporated into other nationalities. The Roman Catholic missionaries gathered and sent to Spain and Rome lengthy and complete accounts of the customs, habits, and religions of the Indians. They also acquired their languages, and translated into them the Ave Maria, Pater Noster, and other prayers. The missionaries also gathered vocabularies, sometimes forming quite complete dictionaries, and also compiled quite complete grammars of the languages. They also — developed a literature, almost entirely religious in its character. Several of the Indian languages were also taught in the colleges and universities in Central and South-America, a chair having existed in the University of Lima till within a few years. We have thus abundant material recorded for the ‘study of many of the languages of Central and South-America. But there are many subordinate dialects which are either lost entirely, or which, being imperfectly recorded, have often been taken for parent languages. In the municipal libraries and archives, there are many documents in diaJects which are now nearly or entirely extinct.
“Of the ‘measures. which should be taken to preserve from destruction the languages of the aborigines,’ one of the first is to gather in some central and convenient place originals or copies of all existing documents, before indiffe- rence, neglect, or time shall have destroyed the records forever, so that they may be preserved, and also that they may be accessible to the American stu- dent of philology. Our ministers, consuls, merchants, travelers, and corre- spondents in Central and South-America can be enlisted to codperate with the association in both gathering documents that are already printed or written, and in securing additional and valuable records with reference to languages which are not yet recorded. To illustrate what may be done, I will not affect a modesty in saying that in Central America alone I have collected, without assistance, more than four thousand pages of original vocabularies and gram- mars. One of these is a dictionary of twenty-seven thousand words, another is trilingual. Some of the languages, I have reason to believe, would have
American Philological Association, 29
-been utterly lost, had I not secured these vocabularies and grammars. The following measures I would suggest :
“1, The designation of some safe and accessible depository of material re- lating to American aboriginal languages.
“9, The preparation and wide diffusion of circulars to travelers and others, indicating to them what is wanted and where it may probably be obtained, with instructions for transmitting all material that is gathered.
“3, To obtain copies of all unprinted original vocabularies, dictionaries, grammars, otc., of American languages, in cases where the original can not be secured.
“4, To obtain as rapidly as possible whatever is already printed that has bearing upon these languages.
“The conservation of the material thus collected would, doubtless, be will- ingly undertaken by any one of several institutions which have library facili- ties ; and patrons will doubtless bé found to defray whatever expense is in- curred in carrying out this project, so important in its bearings upon the philological study of the languages of America.”
Mr. Porter ©. Bliss, late U.S. Consul in Paraguay, offered some remarks on the languages of South-America,
“From a careful examination of eleven languages, he had become satisfied of the inaccuracy of much that has been written respecting them. He gave the outlines of what he regarded as a true classification of the Indian tribes of this region, and promised to preparo for a future meeting of the association a more elaborate paper on the subject. He stated that four fifths of the inha- ditants of Pera and Bolivia are of unmixed Indian blood, who still speak the languages of the Incas of Peru. Mr. Bliss confirmed the positions presented in Mr. Trumbull’s paper, and gave numerous illustrations in support of them.”
The Rev. Joseph Anderson, of Waterbury, Ct., spoke of vari- ous localities in which the Indian languages are still spoken by remnants of tribes. He read a letter from Rev. Asher Wright, of the Seneca Mission, describing the present condition of the Troquois nations, with remarks upon their languages. Mr. An- derson thought the best way to preserve from destruction the Indian languages was to do what is possible to preserve the In-
_dians themselves.
Rev. Dr. McCosh, President of Princeton College, addressed
the Association upon the “ Relation of Language to Education.”
“Language he regarded as an original endowment of the human mind from the Creator, quite as much as memory or imagination. Thought pre- codes language, but language is its most important instrament. Facility in learning language is much greater in childhood than in later years, A child should begin the study of some language other than its vernacular at the age of nine or ten years. Our college courses should be so arranged that stu-
30 American Philological Association.
dents can enter at fifteen and graduate at nineteen or twenty years of age. The proposition to found a post-graduate institution was well deserving of consideration. But till that was accomplished much could be done to promote high scholarship in language by establishing fellowships in connection with our colleges, whereby students could be supported while continuing their stu- dies for one or two years after finishing their present college course. It is gratifying to see that the 1ight spirit prevails in this body with reference to the study of the ancient and modern languages. They should go hand in hand, and thus mutually aid and supplement each other.”
At half-past eleven a.m., the Association adjourned, to meet for the second annual session in Rochester, N. Y., on July 26th, 1870.
SECOND ANNUAL SESSION, HELD AT ROCHESTER, N. Y.,
JULY, 1870.
The Association assembled agreeably to notification in the Brick Presbyterian church, at three p.m., with the President, Professor Whitney, in the chair. The Secretary having been de- tamed by an accident on the Erie Railroad, Rev. Dr. Howard Crosby was appointed Secretary pro tem. The roll was then called and the following new members were received, according to the provisions of the constitution: Professor W. W. Goodwin, of Cambridge, Mass.; Dr. M. B. Anderson, Dr. R. J. Buckland, ' Dr. A. C. Kendrick, Professor J. H. Gilmore, Professor E. H. Wilson, and Professor N. W. Benedict, of Rochester, N. Y.; Mr. T. R. Lounsbury, of New-Haven, Ct.; Mr. T. H. Norton, of St. Catharine’s, Canada; Mr. A. B. Evans, of Lockport, N. Y.; Mr. H. L. G. Brandt, of Clinton, N. ¥Y.; Professor L. D. Hill- mann, of Carlisle, Pa.; Professor F. A. March, of Easton, Pa. ; Professor G. R. Bliss, of Lewisbury, Pa.; Miss M. B. Flint, of Monticello, N. Y.; Dr. E. G. Robinson, of Rochester, N. Y.; Dr. A. B. Hyde, of Meadville, Pa.; Prof. C. G. Hudson, of Lima, N. Y.; Professor J. C. Overheiser, of New-York; Professor B. P. Mackoom, of Cedarville, Ky.; Professor N. White, of Canton, ' N. Y.; Professor A. Winchell, of Ann Arbor, Mich.; Professor M. E. Gates, of Albany, N. Y.; Dr. H. G. Warner, of Roches- ter, N. Y.
The treasurer’s report was read and was referred to an auditing committee consisting of Professor S. S. Haldemann and Pres. G. W. Samson.
Professors J. Hadley, A. H. Mixer, and A. B. Hyde were ap- pointed a committee to nominate officers for the ensuing year.
- After a short recess, the Association reassembled, and the Nominating Committee presented the following nominations: Dr. Howard Crosby, New-York, President.
4 PROCEEDINGS OF THE
Professors J. R. Boise, Chicago, Ill., and W. W. Goodwin, Cambridge, Mass., Vice-Presidents.
Professor G. F. Comfort, New-York, Secretary and Curator.
Hon. J. Hammond Trumbull, Hartford, Ct., Treasurer.
Additional members of the Executive Committee: Dr. A. C. Kendrick, Rochester, N. Y.; Professor Louis Kistler, Evanston, Ill.; Professor A. C. March, Easton, Pa.; Professor C, D. Morris, Peekskill, N. Y.; Professor W. D. Whitney, New-Haven, Ct.
Upon motion, the nominees were unanimously elected by the Association.
Upon the motion of Professor Whitney, two additional secre- taries were appointed by the chair, as follows: Professor A. Winchell and Professor J. C. Overheiser.
The Auditing Committee reported that they had examined the accounts of the treasurer and examined the vouchers, when the whole was found to be correct.
Upon motion, the members of the Executive Committee ez- officio, were constituted a business committee for the present session.
Professor Haldemann moved an amendment to Article II, . Section 3, of the Constitution, as follows:
Instead of “ All the above officers shall be elected at the first session of each annual meeting,” substitute, ‘ All the above officers shall be elected at the last session of each annual meeting.”
The amendment was laid upon the table for action at the next meeting, as required by the constitution.
EVENING SESSION.
President Anderson, speaking in behalf of the local committee and of the citizens of Rochester, welcomed the members of the Association to the hospitalities of the city.
Dr. Crosby, the President-elect, returned the thanks of the members of the Association for the cordial welcome expressed by Pres. Anderson. :
Professor Whitney, the retiring President of the Association, then gave the annual address:
Professor Whitney stated that he addressed the society by the request of the executive committee. He could not but begin by expressing the gratifi- cation felt by all at the successful organization of a national philological association—a thing which, a year or two ago, had seemed well-nigh imprac-
; AMERICAN PHILOLOGICAL ASSOCIATION. 5.
ticable to many, himself among the number. The real success of the enter- prise, however, was yet to be assured by devoted and persistent labor. The scientists have long had a pleasant and useful organization of the same kind. The advantage popularly ascribed to them in the range of their subjects and the rapidly progressive character of their methods and results, is wont to be greatly overrated. Philological studies are bearing their full share in the progress of the age. The scientific study of language has opened innumera- ble new points of view, and multiplied the value of all linguistic material. The circle of classical languages, and of antiquities to be studied, has been indefinitely extended. Egyptian, ancient Persian, Mesopotamian, Chinese, Sanskrit, are branches newly opened or immeasurably developed. Even old- established studies, like Greek and Latin, have their methods revolutionized, in every part, till they seem almost a creation of the most recent time. The modern languages, the English itself, have for the first time taken their proper places in the philological field. These are the conditions which fill the philologist’s mind with enthusiasm, and prompt him to more engaged effort, and which make the formation of an association like ours a necessity of the time.
Of the general objects which we seek to attain by association we are more fully conscious than of the means by which they are to be gained. Nor are too definite plans of action possible or desirable. The association is to be just what its members shall make it, and will not bear much managing or mastering. It-must discuss the subjects that are interesting American philolo- gists, and with such wisdom and knowledge as these have at command. The repressive powers invested in the executive committee must be sparingly wielded, and only as sustained by an overwhelming public opinion among the members. In every such free and democratic body, things are brought for- ward into public which might better have been kept back. We desire to discuss the living questions of the day, in a way to help their settlement— but our own living questions, and to their settlement among us, which may involve the spreading of light eleewhere won, as well as the bringing out of new light. Our best welcome, at any rate, will be reserved for actual addi- tions to general knowledge, and such will receive first admission into our published transactions, while more popular and less original papers may be not less acceptable at our meetings. A working society, which we aim to be, is worth in the last analysis what it brings forth for universal use. The classics, of course, will occupy the leading place ; that department will be most strongly represented, and will least need fostering, while it will call for most careful criticism. The philology of the American aboriginal languages: on the other hand, demands, as it has already begun to receive, the most hearty encouragement. Circumstances, and our duty toward the races whom we are disposeessing and destroying, make American philology and archseo- logy our especial responsibility, and it is our disgrace asa nation that we have been unfaithful to it. Educational subjects are also closely bound up with philology, and will necessarily receive great attention ; yet there should be a limit here ; our special task is to advance the interests of philology only, confident that education will reap its share of the benefit. We shall need to consult brevity and point in papers and discussions, repressing the national disposition to too much talk, (sometimes wrongly attributed to the over-pur-
6 PROCEEDINGS OF THE
suit, instead of the under-pursuit of philology,) and frowning particularly on papers which undertake to grapple with subjects for which a volume would be insufficient, and which involve a host of debatable points. The character of the audience we address must be borne in mind, and popular and elemen- tary explanation cut short. General exposition and defense of the merits of philology is also out of place before philologists. Not less offensive is the de- preciation of any other department of study. The especial duty of philology is appreciation, full and generous, of every part and parcel of human knowledge.
Education is an exclusively human process. Its basis is the sum of human knowledge, accumulated and accumulating. This sum is much too vast for any individual to possess ; acquisition of its most valuable part, and the being placed in apprehensive sympathy with the rest, is culture, the aim of general education ; but not the sole aim of education, which has equally in view the advancement of knowledge, and the equipment of the individual for his special work in life. The process of education, which should last as long as life, is also divided in respect to time, the first part being chiefly preparatory, or “disciplinary.” For discipline is, in its essence, preparation; that is a disciplinary study which prepares the way for something to follow. There is no discordance, but the closest connection, between discipline and the gain- ing of valuable knowledge ; but the value of knowledge is relative; and the disciplinary method implies that the instructor, overlooking the whole body of knowledge, brings before the pupil’s mind the right kind, at the right time, to secure the best final result. The process of education is a tentative one, necessarily involving much waste and failure, from the deficiency of human wisdom and foresight. No one system is to be rigidly held and im- posed upon all. Not all minds will reap the essentials of culture off the same part of the great field. ‘‘ What study is disciplinary?” is a false ques- tion ; we should ask what kind of discipline each study affords ; what prepa- ration it requires, what it yields.
The acquisition of our mother-tongue is the first step in education. Anda mother-tongue like English is itselfa door to the chief treasures of human knowledge, a possible means of the highest culture. We have no right to look down upon the man who knows English only as necessarily half- educated ; for he may have gained in and through it more than an equivalent for the more varied linguistic acquirements of others.
Sooner or later, the taste and choice of the pupil has to be consulted by his educators. Downright constraint answers only with children ; a training which is felt mainly as drudgery throughout is a failure, as leading to the cessation of study when the constraint is removed. The pupil must leave the hands of his disciplinary instructors with a generous capital of valuable knowledge, of which he feels the value; realizing something of what there is in the world worth knowing, craving to know it, and trained in the ways in which it is to be learned. Mere intellectual gymnastics is to be ruled out entirely. The judgment is competent to deal only with matters in which it is actually versed. Information, positive knowledge, fact, is the sound basis of all fruitful intellectual activity.
These are the principles by which the study of philology, in general and in particular, is to be judged and its value determined. And it willstand every
AMERICAN PHILOLOGICAL ASSOCIATION. q
test of usefulness, lower or higher. There was never a time when studies in language had such absolute claim upon the age as now. The same is true of the classical division of those studies ; although its position is, of course, relatively other than at the revival of letters, when the classics were almost the sole sources of knowledge and means of discipline. We fejoice in the wonderful growth of other departments of Knowledge, and acknowledge that, as we have more and more of human history behind us, the comparative importance of any one part of it is diminished. What we most need is the wisdom that consists in knowing how little we know, and, as its result, that humility and charity which shall lead us to estimate at its full value what is known by our fellow.
Mornina Sression, WEDNESDAY, JULY 27.
After the transaction of some miscellaneous business, a paper by Professor James Hadley, of New-Haven, on “The Theory of the Greek Accent,” was read :
Professor Hadley’s paper, as he stated at the outset, had already been pre- sented as a communication tothe American Oriental Society. The following is a summary of its contents:
The Greeks distinguished one syllable in each word by sounding its vowel on a higher key ; this higher key was represented by the acute accent. The ordinary lower key was not represented in writing. But when it followed the higher key on the same long vowel, it was represented by the grave ac- cent, which then united with the acute to form the circumflex. And when a high-tone ultima, followed by other words in close connection, dropped down to a lower Key, it was written with a grave accent instead of the acute. The melodic character of the Greek accent Professor Hadley illustrated from Dionysius Halic., (De Comp. Verb., 12,) who calls the interval between the higher and lower keys a fifth, (three tones and a semi-tone.) That there was any difference in stress (or force of utterance) between accented and unac- cented syllables, is not intimated by the ancient writers ; that such difference, if it existed, can not have been great, is made probable by the total disregard of accent in ancient verse. The question has been raised whether any dis- tinction was made among the lower tones; whether there was any middle tone intermediate between the highest and the lowest. Some ancient writers speak of a middle tone; but the statements are not so definite as could be wished. G. Hermann (De Emend. Rat. Gramm. Grec.) recognized a middle tone in the grave accent where it takes the place of an acute on the ultima. G. Curtius (Jahn’s Jahrb., vol. 72) recognized it also in the grave accent where it forms part of the circumflex. “Recently, F. Misteli, (Kuhn’s Zeitsch. vol. 17,) founding on the analogies of the Sanskrit accent, holds that the high tone, (acute accent,) where it was not final, was always followed by a middle tone. Professor Hadley set forth a theory based on that of Misteli, but with additions and modifications of hisown. In the undivided Indo-European, as in Sanskrit, there was no restriction on the place of the accent ; it might fall — on any syllable of the longest word. Hence, the high tone with the follow-
8 PROCEEDINGS OF THE
ing middle tone might be separated from the end of the word by a succession of low-tone syllables. If, now, there came to be a prevailing dislike for such & succession, an unwillingness to hear more than one low-tone syllable at the end of a word, the result would be to confine the accent to the lust three syl- lables. This result, as it is found both in Greek and in Latin, may be referred to the time of Greco-Italican unity. But for the Greek we have to assume also a subsequent restriction ; the final low tone must not occupy the whole of a long syllable ; if it came upon a long vowel, the first half of that vowel ‘must be sounded with middle tone. Thus “high tone, middle tone, short low tone,” became a prevailing cadence for Greek words, and was brought in wherever it could be attained without throwing back the accent. The lead- ing rules of Greek accentuation—no accent allowed before the antepenult ; only the acute used on that syllable, and not even this if the ultima is long; an accented penult must ‘take the circumflex if it has a long vowel and the ultima a short one; an accented penult must take the acute in any other case ;—all these are explained by this cadence, being all necessary to secure it. As for throwing back the accent to obtain this cadence, (or as much of it as possible,) one branch of the Greeks, the Molians of Asia Minor, did so ; whence Aolic forms like ydAeroc, yaléruc, AeAvnorec, for which the common Greek has yaderic, yaderde, AcAvxorec, with the primitive accent.
The Latin took a different though analogous course. It allowed the final low tone to have either quantity, but would not allow the middle tone before it to occupy the whole of a long syllable, whether long by nature or by posi- tion. Hence, the cadence, “ high tone, short middle tone, low tone,” which the Latin procured, or as much of it as possible, in all words, even by throw- ing back the accent like the Molic Greek. In this way all the varieties of Latin accent—légeres, légeret, monéres, monéret, legéndua, vizit; réa—may be easily accounted for.
In conclusion, Professor Hadley referred to the hypothetical character of this theory, pointing out the unproved assumptions contained in it; but re- marked that these assumptions are so natural in themselves and furnish s> simple an explanation for 80 many seemingly unconnected facts, that it is difficult to believe them wholly unfounded. |
The next paper was by Professor Whitney upon “The Sanscrit Accent.”
Professor Whitney stated briefly the main features of the system of Sans- krit accentuation, by way of analogy and support to the principles of Greek accent, a8 laid down in the preceding paper. Our sources of knowledge on the subject are the writings of the ancient Hindu grammarians themselves, They teach three accents, corresponding to the Greek—acute, circumflex, and grave ; and give definitions of them which prove their identity in character with the Greek. The acute (udétta, “ elevated”) is described as uttered ina high tone ; the grave, (anuddtta, “ not elevated,”) as uttered in a low tone; the circumflex, (ssarita, precise meaning doubtful,) as combining the two tones of acute and grave. And all the phenomena of accentual change are such as to prove these definitions accurate. The range of use of the Sanskrit (independent) circumflex is very different from that of the Greek, and much
AMERICAN PHILOLOGICAL ASSOCIATION. 9
more restricted ; the accent is mostly found on syllables whose vowel is pre- ceded by a y or zw that represents an original acute ¢ or 1—thus, kwd, nadyds, for ki-a, nadi'-as ; apart from these, only on long vowels, (by nature,) in certain special cases, where an acute and a grave vowel have been blended into one syl- lable. As to the place of the accented syllable in the word, the Sanskrit knows no restriction whatever ; no tendency. to a particular cadence, or other general tendency, has hampered the freedom of position which we must suppose to have prevailed in the original Indo-European period.
The Hindu grammarians recognize a second, an enclitic, circumflex, as re- gularly following an acute syllable—implying that the voice, instead of pass- ing directly from the raised tone of acute to the ordinary level of pitch, comes down in the course of the succeeding syllable. It is an open question, perhaps, whether they might not more correctly have apprehended this enclitic tone as a middle tone than asa slide; but their authority, at any rate, is entirely in favor of the latter value: they completely identify, as regards essential character, the two kinds of circumflex.
After the discussion which followed these papers, Professor Comfort called attention to the fact that the Committee upon the Pronunciation of Latin and Greek had made but a partial report last year, and that two members of the committee were not present at this meeting. Professor Morris and Dr. Kendrick were appointed to fill the vacancies in the committee.
AFTERNOON SESSION.
Professor A. Ten Broeck, of Ann Arbor, Mich., Professor H. L. Baugher, of Gettysburg, Pa., Mrs. E. A. Weston, of Antioch, Ohio, were announced as new members by the Executive Com- mittee.
Professor W. W. Goodwin read a paper upon “The Construc- tions allowed in Greek, after "Owes, “Owes 7, and the double ‘ne-
o99
gative Ov p7.
The so-called Canon Davesianus, in its later form, declares the first aorist subjunctive active or middle a solecism after all these words. What is true in this rule arises from the simple fact that, wherever the two constructions of the future indicative and subjunctive were allowed, an author naturally avoided those forms of the subjunctive which differed from the future indica- tive only by the quantity of asingle vowel, (that is, the first aorist active and middle,) and when he wished to use the subjunctive at all, he would prefer the more decided form of the second aorist. But this can not be applied to con- structions like pure final clauses, in which only the subjunctive was allowed in good Attic prose. Here, therefore, we can not expect the rule to hold, and examples like Thuc. II. 60 and VII. 39 ought to have the aorist subjunctive, as the Mss. require. In the construction in which é7w¢ or étw¢ py is used with the future (as it is always printed) by an ellipsis of a verb meaning see
10 PROCEEDINGS OF THE
to it, or the like, there is a strong reason why the subjunctive (even the first aorist) should have been tolerated, even although it is rare (or even impos
sible, as some consider it) in the full form, when thé leading verb is used. There the more familiar form of prohibition, 4) worgonc, made éxw¢ pe) momnoge quite as natural as the future to Greeks who never thought of the ellipsis. So in the case of of 27, the same familiar form of prohibition, 4) motjogc, made ob} p? mowjaoys perfectly natural; and where there is sufficient Ms. authority for this, it should not be emended to suit a mere theory. The doctrine of Elmsley, that this construction of od «7 is always interrogative, (so that ob pe? Tovro mrojoecc ; Will mean, Will you not not do this? that is, Do not do this,) is one cause of the supposed necessity of insisting on distinct forms of expres- sion in this and in the other more common construction. In the latter, there. can be no doubt that od 4) yevyrac (or yevgcerat) is a strong negative, meaning simply, i¢ will not happen ; and here all scholars allow both subjunctive and future indicative. The strongest reasons can be urged for explaining both on the same principle, and if it appears that both subjunctive and future indicative are used in the prohibitive construction, the strongest reason for Elmsley’s distinction disappears.
A collation which the speaker had recently made of the passages containing ov py and duc 27, in the Clouds and Frogs of Aristophanes, in the two Vene- tian and in ten Paris Mss., shows that there is generally better authority for the first aorist subjunctive here than for the future, while the future stands in all modern editions. It seems, therefore, that the subjunctive ought to be restored in all cases in which it has the superior Ms. authority; certainly where it has a2 the authority, as in vs. 293 of the Clouds. To meet the usual objection that copyists constantly confused ec and y, as well as e and 9, in the classic Mss., he inspected in all these Mss. passages of a different character, in which these letters could not have been interchanged without an obvious blunder, and in all these the copyists used consistently the correct form, with no more exceptional mistakes than twelve modern copyists would make in copying as many passages of English.
Mr. Porter C. Bliss, Secretary of the United States Legation in Mexico, gave an account of an Inscription in one of the native Indian languages, which has been discovered in Central America.
Mr. T. R. Lounsbury read a paper upon “ Certain Forms of the English Verbs which were used in the Sixteenth and Seventeenth Centuries.”
The object of this paper was to show that the forms of the present tense ending in 8, belonging strictly to the Northern dialect, and the forms in th, belonging to the Southern, were in much more extensive use in the literary language during the sixteenth and seventeenth centuries than is generally supposed, at least in the second person singular and the third person plural. Proof of this was furnished by more than five hundred examples drawn from writers who flourished in the middle or latter half of the sixteenth and beginning of the seventeenth century. The quotations made seem to show clearly, first, that during that period, the forms both in s and th
AMERICAN PHILOLOGICAL ASSOCIATION. 11
for the third person plural were not simply met with occasionally, but were in constant and common use, and by authors in every style ; secondly, that the same is true of the second person singular in ¢, thirdly, that the third person plural in ¢h disappeared rapidly after the middle of the sixteenth century, and toward the end of it was confined almost entirely to the verbs to do and to have ; fourthly, that of the forms in « of the same person and number, there is no evidence of decay within this period, and if there were any change whatever, the use of them seemed rather to increase than diminish ; and finally, that forms in th for the first and second person plural, forms in 8, for the first and second persons, both singular and plural, were occasionally to be met with, and that even forms in ¢h for the first and second person singular are to be found, though under too doubtful circumstances to au- thorize the formation of any theory in regard to them.
Professor S. S. Haldemann read a paper upon what is termed * Pennsylvania Dutch,” in which he traced the grammatical basis of this language back to native German dialects, especially those of Suabia and Switzerland.
The Association having accepted an invitation to attend a soirée given by the Hon. Freeman Clarke, no session was held in the evening. ‘
Morntne Session, THurspay, JuLy 28.
Professor Comfort, the Secretary, announced the donation to the Association of several books, as follows: From Dr. Heinrich Schliemann, of Paris, Ithaka, der Peleponnes und Troja ; from Professor S. S. Haldemann, (1) Jnvestigation of the Greek ¢, by means of Phonetic Laws ; (2) Trevelyan Prize Essay, Analytic Orthography—an Investigation of the Sounds of the Voice and their Alphabetic Notation, including the Mechanism of Speech and its Bearing upon Etymology ; (3) Etymology as a Means of Education ; also the Annuaire de [ Association pour 0 Encou- ragement des Etudes Grecques en France, 4 Année, from the secre- tary of that Association.
Dr. Dwight moved that a committee of five be appointed to consider the subject of placing the Association upon a more per- manent and effective basis, with a view to having corporate pri- vileges. By motion, the subject was referred to the Executive Committee. |
Upon the motion of Protessor Whitney, the proposed amend: ment to Article IL, Section 3, of the Constitution was taken from the table by-unanimous consent and adopted by an unanimous vote.
12 PROCEEDINGS OF THE
Professor E. R. Ruggles, of Hanover, N. H.; Professor G. McMillan, of Hillsdale, Mich. ; and Hiram W. Sibley, Esq., of Rochester, N. Y., were announced as elected members of the Association.
Professor Hyde, presented some points to illustrate “‘ The Disuse of Passive Forms in Languages of the Aryan Family, and their Replacement by Reflexive Forms.”
The first appearance of a passive in the Sanecrit is very noticeable, from its employment of the root ¢, or ya, for its expression. This means “ to go,” and the simple verb of existence, “to be,” does not appear in the passive forma- tion. This use of the verb, “to go,” (so singularly repeated in the like usage of the negro dialect,) is very natural and reasonable. Amatus sum, ‘I was loved,” express a condition very fully, an action very meagrely. The verb go conveys the sense of action. We have reason to think that in the mother Aryan, the genius of Schleicher would have decided for a similar form.
«Passing to the Greek, we recognize the Sanskrit root formative of the passive in the aorist ending 7-v, and the future 7-0 oua:. These are thus the only true passives of the Greek. In the Latin verb, there are no true passive forms. The formative root occurs in few verbs, as veneo, compared with vendo, pereo with perdo.
All these deficiencies of the true passive are supplied by the increased use of the reflexive forms. Thus, “Aeirera:,” amatur, are shown by the root-analysis to be reflexive. The Greek gives the reflexives a somewhat regular classifi- cation, as middle; the Latin declares them passive, excepting the deponents. The modern languages of our family, with scarce an exception, have no semblance of a passive form. The passive idea is expressed adjectively by the verb fo be, and the participle denoting finished action, as, “he is loved.” Only the German expresses distinctly the idea of a state as separate from an action. These modern tongues make avery copious use of the reflexive form, which, however, has come to be periphrastic. Even the English, whose idiom inclines least to reflexives, tends to follow, as far as possible, the usage of its neighbors. In general, it may be said that proper passives have ceased to exist in our family of languages, and their place has been assumed by the reflexives.
One might reasonably inquire what feeling or fashion of the mind has caused this phenomenon. Language, in all ita developments, is but the utterance of the inner conception, and takes its shape therefrom. If, now, we examine the use of the reflexive in the French, why may we not have a clue to its historic tendency from the beginning? The chief quality of the French reflexive is not 80 much its reflexiveness as its indirectness, its non- committalism. This may even be called its charm. This quality commends it to an expression aiming at cautiousness or politeness. In the French phrases, very beautifully does their reflexive guard or soften our blunter English talk. If this is the clear effect of the use of the reflexive to-day, it is not unreasonable to affirm that such it may ever have been, and in this the notable phenomenon of the loss of passive forms and the prevalence of reflexives may find its true explanation. Some features of the Semitic verbal forms indicate a similar tendency.
AMERICAN PHILOLOGICAL ASSOCIATION. 13
A paper upon “ Contributions toward a Grammar of the Creole Dialects of Hayti and Louisiana,” by Addison Van Name, was read, in the absence of the author, by Professor Whitney.
In this paper, suggested by Mr. Thomas’s recent grammar of the Creole of Trinidad, a sketch was given of the more prominent characteristics of two dialects, the one of which is in common use in Hayti, among the uneducated classes, and the other spoken by a considerable portion of the former slave population of Louisiana. These dialects, like that of Trinidad, to which they are closely related, are descended from the French, with but a slight admix- ture of foreign elements, whether African or European. The process by which they have come to their present form is not unlike that by which the French itself grew out the Latin; the changes are not only of the same general nature, but often in the same direction, a tendency which appears in the French being carried still further in the Creole. There are not wanting also new formations which establish the claim of these dialects to be regarded as something more than mutilated French.
After noticing the phonetic differences between Creole and French, both’ the uniform changes which certain sounds undergo, and others of a more vio- lent nature, the grammatical forms were considered. The French definite article has entirely lost its grammatical force, but remains attached, as an in- separable prefix, to many substantives. In its place, the Creole has converted the demonstrative adverb /d into a post-positive article, and by composition with the personal pronoun evz has even formed a plural, layo or layé. In- stead of the weaker forms of the personal pronouns, je, tu, iJ, ils, the stronger moi, toi, lui, eux, (Creole moin, toi, li, yo or yé,) alone are used, and serve, at the same time, for the pogsessives. Nouns and adjectives have lost their few remaining inflections, and are invariable both for number and gender. The verb has suffered still more. The twenty or more inflections of aimer which are distinguishable by the ear gre in Creole reduced to one, aimé, which may be considered the joint representative of the infinitive and participle, since in all regular and most irregular verbs these would, in the Creole pronuncia- tion, be identical ; where they differ, the infinitive is generally chosen. The different tenses are formed solely by means of auxiliaries, among which, however, avoir does not appear. The tendency of the strong or irregular verbs to become weak or regular, which has spread so widely in the Indo- European family, is noticeable here also. The syntax of the Creole is simpli- fied in a corresponding degree. The genitive relation is indicated by plac- ing the governed immediately after the governing noun. Many verbs which in French are reflexive, or which have the régime tndirect, take, in Creole, the régime dire.
Hon. J. H. Trumbull read a paper upon “‘Some Mistakes con- cerning the Grammar, and in Vocabularies of the Algonkin Lan- guage.”
This paper pointed out some of the errors that have resulted from cursory readings, or mis-readings, of Eliot’s version of the Bible in the language of the Indians of Massachusetts. Some of these concern the grammatical
structure of the Algonkin languages; others are found in the vocabularies. Of the former class, Mr. Trumbull! mentioned :
14 PROCEEDINGS OF THE
1. The alleged discovery of a definite article in the languages of Masaa- chusetts and Delaware. This discovery was announced by Mr. Duponcean, and on his authority has been affirmed by distinguished American and Euro- pean philologists. It rests only on two or three mistranslations of verses from Eliot’s Bible, and a false inference. Mistaking (mo) the sign of the past tense for a pronoun, Mr. Duponceau derived from this pronoun the prefix (m’) which he supposed to be the definite article, but which is in fact a privative and tndefinitive, employed only before a few inanimate nouns.
2. The supposed vocative case of Indian nouns. In verses cited from Eliot, the terminations which Mr. Duponceau, Mr. Gallatin, and others regarded as belonging to the vocative singularand plural (in and ewnk) are, respectively, those of indefinite nouns-animate and collective nouns, irrespective of case; as is shown by reference to other verses in which nouns of these forms were used by Eliot, as nominatives and accusatives.
8. Various false analyses of Eliot’s translation of the name in Exodus 3 : 14, “7 am that I am,” as bearing on the question, whether a verb of simple existence can be found in any Algonkin language. Without entering into the discussion of that question, Mr. Trumbull showed that the Indian verb, used in the verse cited, affirms the relation of an individual to a species, or of like to like; ‘“‘to be of the kind of,” or “to be such as ;” and that nan nutten niin nén nuttinniin signifies, literally, “I myself am such as I myself am such as;” go talis sum qualis ego sum.
Errors of the second class—those that are found in the vocabularies—are more numerous. Many of these come from mistaking the order of words in the Indian text. Such was Professor Vater’s, who put into his list of words in the language of the “ Naticks, nach Eliot,” (Mitiridates, vol. iii., pt. 2, p. 338,) for the name of the sun, a verb meaning, “ he stood still ;” taking it, probably, instead of the noun which precedes it, from Joshua 10: 18, “The sun stood still.”
Twice, at least, the same distinguished scholar pointed out the resemblance between cone, another “ New-England” name of the sun, and the Tatar kun, as an indication of the relationship of American and Tataric languages. Unfortunately, cone (koon, Eliot) means “ snow,” not “ sun.”
Dr. Pickering incorporated with his verbal index to Eliot’s Grammar a few words selected by Mr. Duponceau from the translation of the Bible. ‘Among these, sohsumo'onk stands for ‘‘ forest.” It means “ forth-shining,” and, as used by Eliot, “glory.” Mr. Duponceau found it in Isaiah 10: 18, and was misled by the position of words in the Indian text, where the order of the English— “ glory of his forest ’—is inverted.
Mistakes of this kind, and of all kinds, abound in a list of nearly three hun- dred words “extracted from Eliot,” printed in the first volume of Dr. School- craft’s “ Information, etc., respecting the Indian Tribes.” For “husband,” he gives, from Genesis 30 : 15, a mutilated fragment of a verb signifying, “thou hast taken away ;” for “ boat,” stands the verb “to come by boat,” from Acts 27:16; and so on through the whole vocabulary, which, to the comparative philologist, is worse than worthless.
AMERICAN PHILOLOGICAL ASSOCIATION. 15
AFTERNOON SESSION.
Dr. H. McCartee, of China, gave an account of the Languages and Dialects of China.
Dr. McCartee had observed, even in the writings of Max Miller, state- ments which allude to the Chinese as if it were one language, instead of being a collection of dialects. The written Chinese language differs also from the spoken language; it being uniform throughout the empire. Each word is represented by a character, which is capable of inflection or infliction. In reading these characters, a different pronunciation is employed in different parts of the country ; just as the date of the present year, 1870, though represented by the same figures in English, German, and French, is pro- nounced differently in each of those languages. There are, indeed, as great differences between the dialects of China as between the languages of Europe. Many words, as some of the nouns and verbs, are common to the written and spoken languages. Other words, as the adjective “good,” are nearly the same in all the dialects. But the word for “ bad,” for example, varies in every dialect and province ; though a word which is an equivalent to “not good” would be everywhere understood. The pronouns differ in the various dialects, with the single exception of the first person singular of the personal pronoun.
There are two court or Mandarin dialects; one being spoken to the north, the other to the south of the Yangtse River.
What was the original dialect or language, it is impossible to say. The Chinese are not the aborigines of the country. Of these only a wild, savage remnant still exists) But since the emigrations of the Chinese into the country, (which took place in two directions ; one by the way of Hoang-Ho, the other by the Yangtse-Kiang,) both the customs and the languages of the people have gone on diverging, until the empire is fitly represented to-day in its languages and customs by the continent of Europe. When summoned, a short time since, by the coroner in New-York, to act as interpreter between tivo Chinamen, he found that they could not talk together in their own language, though one was from Canton and the other from a place not a hundred miles distant from that city ; but that they were holding a frag- mentary conversation in broken Spanish.
In regard to the “pigeon-English.” which is regularly taught in the schools of Canton, and which is often supposed by the English to be pretty good Chinese, and by the Chinese to be good English, this mongrel language is composed mostly of English words with Chinese endings, and which are used according to the grammatical and idiomatic construction of the Chinese language.
The written language is a sort of phonetic representation of the spoken. They have no alphabet, but use a character for each word. There are con- sequently many ten-thousands of these characters ; but the acquisition of six thousand will give a person a good facility in using the language. One of the greatest difficulties in learning Chinese arises from their use of inflections in two words which are pronounced alike ; a different inflection upon each giving to them altogether different meanings. The number of tones or in-
16 PROCEEDINGS OF THE
flections employed can be reduced to five. The spoken language is more syllabic than the written, the words of which are nearly all monosyllables, In the dialect of Ningpo are many words of three or more syllables, some of which are compound words, while others appear to be simple and incapable of analysis. +
It has been said that it is impossible to print the spoken dialect in Roman characters. But this is really easier than to use the Chinese characters. The whole of the New Testament, and many other books, have been printed with Roman characters, and they are readily intelligible to the Chinese.
Hon. J. H. Trumbull read a paper “On Algonkin Names of the Dog and the Horse.”
In every Algonkin language, we meet with two names for the dog: one of which belongs to that animal exclusively ; the other is given him in com- mon with other domestic animals.
The former, which may be called his proper name, seems to be derived from a primary verb signifying “to command,” and in the passive, “to be commanded,” hence, “to obey,” “to serve,” “to be a helper.” Among the northern tribes, the dog was “the helper,” not only in the chase, but as a draught-animal.
The other name describes him as “the belonging ”’ or “ the live property ” of his owner—his chattel ; suum peculium. The Indian had neither flocks nor herds. The dog was the only animal with which he associated the ideas of exclusive possession and of personal property ; just as in the Aryan languages these ideas were associated with the cow and the ox.
When other species of domesticated animals were introduced, the Indian gave them also this general name, of “ live property,” or the like, in addition to the specific and descriptive name he invented for each. Sometimes this specific name was formed by prefixing a qualifying word to the general. The Chip- pewa spoke of his horse and his dog alike, as o’dt-un, “his domestic animal,” . “his live property ;” but he had for the latter a specific name, annimoosh, and he described the horse by another, as paipaizhikégazhi, “‘ the one-nailed,” or “the beast having undivided hoofs.”
In the Delaware, the specific name of the European dog marks him as “wolflike.” The specific name of the horse is “the beast who is accustomed to carry upon his back a living burden.” There was a general name for both—employed also as a prefix to the name of every other kind of domestic animal—meaning, as in other Algonkin languages, “live property,” or “servant.” The Wyandot name of the horse, though of different composition, has nearly the same meaning as the Delaware ; “the slave-animal that carries upon his back.”
The paper gave other examples of the names of domestic animals in several American languages, with an analysis of each. In conclusion, it suggested to the compilers of vocabularies the importance of ascertaining not merely what each animal ¢ called, but the meaning of the name, and whether it is specific or (more largely) appellative. In spoken languages, the constancy of Indian roots, and the inviolability of the laws of verbal growth, make analy- sis comparatively easy. It is possible that a thorough investigation of the
AMERICAN PHILOLOGICAL ASSOCIATION. 17
composition of animal names used by any American tribe would give results not without value to philologists who are prosecuting similar investigations in other families of language, or are working their way toward the very origin of speech. ,
Dr. A. C. Kendrick spoké upon the “Translation into English of the Greek Aorist Participle:”
The Greek future and perfect participles correspond to the English, leav- ing only the English present participle to answer to the Greek present and aorist. The Greek present denoted continued action; the aorist, a simple act. The aorist naturally throws itself into the past, because a temporary act is ordinarily expected in the past rather than the present. Hence the statement that the aorist participle is to be translated by a form denoting a time anterior to that of the verb with which it is associated. In point of fact, however, the aorist participle in this connection has a logical rather than a temporal connection with the verb. The aorist does not necessarily express an act anterior in time to that of the finite verb with which it is associated ; and when it does express a previous act, the fact of priority is never empha-. sized. Even though we may, in certain connections, find it impossible to translate the aorist participle by any thing else than the English perfect, we thereby inject into the Greek an idea which it does not contain. The English present participle, according to the speaker, had not only a present force, but an aoristic force in which it corresponds strictly to the Greek aorist. Which force the English present participle may have in any given case, must be de. termined by the context. One mode, then, of translating the Greek sorist participle is by the English present participle—not as a matter of accommo- dation, and because we can do nothing better, but because the English has the precise force of the Greek. Other modes of rendering are by a finite verb. Sti other modes are by the English present participle—(Query? verbal noun)—in connection with an appropriate preposition. As a last resort, the English present participle may be used. As we have no aorist participle, eo nomine, in English, we have come to use our perfect participle in somewhat closer approximation to the Greek aoristic sense. As the Greeks seem to have had an especial fondness for the aoristic form, they may have used it where the perfect would have been more natural—especially since the perfect participle was longer and more awkward in form than the aorist. Of course in such cases the rendition of the aorist by the English perfect is desirable. The speaker wished to strike a blow at the current teaching that the Greek aorist participle should be uniformly rendered by the English perfect. Crosby’s Grammar took this ground. Taylor’s translation of Kithner took this ground. Hadley’s Grammar—the author of which sat just before the speaker—was the only one which he knew that made a correct statement on this point.
EVENING SESssIOoN.
Dr. Dwight, Chairman of the Committee upon the Pronuncia- tion of Latin and Greek, presented the following report :
The committee see no middle ground between the general principles laid down in the report of last year and the taking up of the whole matter in
18 PROCEEDINGS OF THE
detail. They are not prepared, therefore, to recommend, in a formal way, any thing additional to the former report.
The members of the committee are unanimous, however, in stating that, individually, they prefer the pronunciation of the diphthongs (1), au (av) like ow in now ; (2), ou (ov), like 00 in moon; (3), eu (ev), like win duty. They also prefer, in reading Greek verse as such, to regard the rhythmic instead of - the written accent.
By a motion at a subsequent session, it was ordered that the report of the Committee on Pronunciation, which was adopted last year, be reprinted in connection with the above supplemental ‘report. The report adopted last year was as follows:
While the committee recognize the fact of wide diversities of opinion among American scholars concerning the mode of pronouncing the classic languages, and while among the members of the committee themselves there are considerable differences of judgment, they agree in stating that they deem some uniform system of pronouncing the classic languages to be greatly to be desired, if possible to be obtained.
They would favor, as at least one feasible step toward such a result, the adoption of the continental system of pronouncing the vowels in both Latin and Greek.
It is also their opinion that t is desurable to observe the written accent in reading Greek, and also to mark the quantity of the vowels in reading both Latin and Greek.
Professor Whitney read a paper upon “ The present State of the Discnssion of the Origin of Language.”
Professor Whitney said that he had no intention of discussing the vast and difficult question of the origin of language, but wished only to straiten the field of discussion a little, and to point out in what direction further la- bor would be likely to produce the most valuable result.
In the first place, he claimed that the question was a legitimate scientific one, and to be treated by purely scientific methods. This, on the one hand, excludes all traditional or historical evidence, since the period of origin in- contestably lies far beyond the reach of tradition or history. It forbids the mingling together of scientific and scriptural argument; when scientific re- search has reached a result, then is the time to compare it with scriptural statements, and see how they agree or are to be reconciled. On the other hand, it excludes the assumption (such as has been made especially by one popular authority, M. Miller) of a different human nature, a special faculty or instinct. To make such an assumption is to quit entirely the scientific basis.
In the second place, we are to distinguish clearly between what is already done, and what remains to do. To a certain extent, the question is historical ; we arrive by actual historical inquiry at roots, usually or always monosylla- bic, not parts of speech, grammatically unformed, as the concrete beginnings of speech. All authorities, worthy of attention, agree in this; no further work is progressive that does not build on this as a foundation.
AMERICAN PHILOLOGICAL ASSOCIATION. 19
In the third place, the point of widest bearing and highest consequence next pressing for settlement is, whether tho first inducement to specch was from within, an impulse toward expression, for the relief and benefit of the speaker, or whether it was from without, a desire to communicate with another—whether speech was pushed out or drawn out. On this, authorities are greatly at variance, and the one view or the other is commonly assumed without argument. If it were well settled, opinions would be vastly nearer accordance on the whole question.
One more point of prime consequence was noted: Does the conception pre- cede the word, or the word the conception, or are the two inseparable in ori- gin, so that the conception can not exist without the word? This, too, is one upon which opposite and apparently irreconcilable opinions are held by highly- esteemed authorities.
These are the four tests by which every investigator in this department may be tried. If he is not sound upon the first two, his work is out of har- mony with the present condition and spirit of linguistic science, and will not be heeded ; if he leaves out of sight the other two, or is unclear respecting them, his work will add nothing essential to what has been already done— and, it may be claimed in conclusion, if he takes the wrong side of them, he will never reach a valuable result.
President Samson read a paper upon «“ Embryological Method in the Study of Language.”
The method called embryological, because it proposed the traces of organic development from the germ, has been employed only in modern times in phy- sical science, since an instrument like the microscope is essential to minute observation. In metaphysical science, the tracing of spiritual development from the first manifestation of sensibility in the yet unconscious infant was pursued by the Brahmins of India as truly as by Leibnitz and Hamilton in re- cent days. A
Human speech belongs to physical science, as Miller and the German writers argue, inasmuch as the law of its accumulated growth in languages long ela- borated is beyond the control of the individual will. At the same time, on the lip of every master speaker, language is an instrument shaped as well as wielded ; in its origin, it is an invention of human ingenuity : and hence its investigation belongs to metaphysical science. : |
For a double reason, then, the method of embryological study may bo ap- plied to the analysis of language. The first utterances of childhood, the in- fant-like effort to gain a foreign tongue, the fixed dialect of rude tribes, the vocabulary of technical artisans, the interjectional utterances of persons under strong excitement, and the elaborate periods of finished speakers as compared with the language of ordinary conversation, furnish so many distinct fields in which the germs of development in the linguistic communication of ideas may be traced.
Language is addressed to the eve in both sign and written symbols, and to the ear in vocal utterances. Sign-language, prominent in children and un- cultured nations, an accompaniment of all speech, and in the master orators more expressive then vocal utterance, is now specially studied and elaborated
20 PROCEEDINGS OF THE
for the purpose of higher instruction given to deaf-mutes ; and hence its ger- minal development as a human invention can be readily traced.
In sign-language, there are three elements of address—mimic, tropic, and phonetic illustrations of thought. To represent plain or rolling Jand, smooth or agitated water, the deaf-mute moves his hands and arms in horizontal or undulating, smooth or notched lines ; thus employing mere imitation in ex- pression. To picture the abstract ideas of fear or jealousy, of faith or charity, he gives the concrete expression always manifested in the features and move- ment of the person possessed by these emotions ; and thus, by tropes or sym- bols, communicates intellectual conceptions. Yet again, to indicate reference to a particular place or person, only to be recognized by his name uttered to the ear, or spelled in letters for the eye, the deaf-mute uses phonetic signs, re- presentative of the letters of the :.]phabet.
This latter presupposes the existence in the mind of the deaf-mute of the forms of written language. In this element, now—the phonetic—a striking analogy is remarked between the principles of sign and written symbolism. Champollion’s first clue to the system of Egyptian hieroglyphics was the ob- servation that where, on the Greek inscription of the Rosetta stone, a proper name occurred, an oval inclosed certain figures ; indicating that a phonetic element entered into this mode of conveying thought. Future study, both of Clement, who, in the first century of our era, described the hieroglyphic sys- tem, and examination of the monuments of Egypt themselves, has develop- ed classes of symbols, mimic and tropic, as well as phonetic. The _ hiero- glyphic representations of water rippled, and of land undulating, is just the notched and waving line of deaf-mutes; and as the great study of the deaf- mute teacher is now the invention of new tropical representations of concep- tions, so the great effort of the decipherer of hieroglyphics is to divine the ancient inventor's art in tropical symbols,
The Chinese written characters are now classified into the same three divi- sions; the mimic being gradually systematized into simpler tropic signs; and the phonetic signs, quite distinct in both character and history, being used with a uniting bar to link them in forming proper names. The newly- discovered Moabite stone gives historic confirmation to the previously-existing inductive suggestion, that the old Phenician alphabet, on which the Hebrew and Semitic and also the Greek and Indo-Huropean alphabets were formed, was originally composed of the same three elements. This old and somewhat primitive specimen, as Colonel Rawlinson has shown, specially illustrates the mimic element.
As now sign-language accompanies and aids to illustrate all vocal utterance, the eye supplementing the ear, and as written language is an invention trans- lating auricular into ocular signs, the suggestion is a natural one, that an analogy exists between the elements of vocal and written signs. Both seem alike to have been the invention of human genius; and it is legitimate, in watching the first utterances of childhood and the exclamations of men ‘essaying a new language or speaking under excitement, to seek in these ger- minal types of human speech the same analysis found to enter into written language.
A careful examination reveals the fact that the vocal elements are few in
AMERICAN PHILOLOGICAL ASSOCIATION. 21
number ; the effort of written language being to find a symbol for each of them. The vowels, or vocales proper, are but about fifteen in number, being formed by fixed positions of the lips, teeth, palate, and larynx ; while the consonants, which are, as their name indicates, but transitions between the sounds proper, do not exceed twenty-five in all the known tongues. A com- prehensive study indicates that the lip-vowels and consonants, in which the movement of the organs is conspicuous, are found in all the rude and simple dialects ; that the harsher sounds are heard among energetic and especially in warlike tribes ; and that softened utterance grows up with culture. From the age of Grecian philosophy, an analogy between sounds addressing the ear and lines and angles agreeable in their proportions has been traced in varied arts. Plato, in his Cratylus, makes Socrates trace virtually the three classes of elements observed in sign and written symbols. Vocal signs, or ut- tered words, he argues, stcceeded to sign-language ; some of the earlier words were imitations of sound given forth by natural objects either animate or inani- mate, but could not imitate forms or other conceptions. Other words, virtually tropes, suggested indirectly by their sound objects addressing the other senses, and also shadowed mental conceptions, outside of all representations, ap- pealing to the mind through the senses. Yet, thirdly, in proper names, while some represented certain attributes or characteristics of the person bearing the designation, many also must have arisen in mere arbitrary phonetic signs invented to designate different persons in whom no special characteristic could be noted.
The paper presented numerous facts in the structure of abariginal as con- trasted with cultured tongues, illustrating the analogy hinted ; and also traced the history of investigations made since the days of the Greeks, in the early ages of Christianity, in Charlemagne’s era, and in yet later times ; all of which indicate that the origin and dev@opment of vocal utterance can be phi- losophically studied, as truly as that of a plant, or as the growth of the useful and the fine arts developed by human genius.
MorninG Session, Fripay, Jury 29.
President Anderson invited the members of the Association to visit the University in the afternoon.
A motion was made and carried unanimously, that the next (third) annual meeting of the Association be held in New-Haven, Ct., to commence July 25th, 1871. |
The following report from the Executive Committee was
adopted:
With reference to the proposition for this Association to establish a Philo- logical Institute, which was referred for consideration to the Executive Com- mittee, this committee, after due discussion, passed the following resolution :
Resolved, That in the judgment of the Executive Committee, it is nos expedient for the Association to commit itself at present to the project to establish a Philological Institute.
22 PROCEEDINGS OF THE
Messrs. Willis C. Gaylord and William F. Lush, of Rochester, were announced as members of the Association.
Prof. 8. 8. Haldemann, Mr. S. J. Buckingham, and Prof. Schele de Vere were appointed a committee to nominate officers for the ensuing year.
The following resolutions were offered by Prof. Schele de Vere, and were unanimously adopted :
Resolved, That the cordial thanks of the Association be tendered to Mrs. Freeman Clarke for her courteous attentions to the members, who have highly enjoyed and fully appreciate her hospitality.
Resolved, That the thanks of the Association be tendered to Mesars. Ell- wanger & Barry, and Mr. James Vick, for the beautiful flowers with which they decorated the place of meeting of the Association.
Resolved, That the thanks of the Philological Association be tendered to the trustees of the Brick Presbyterian Church, of Rochester, for the use of their building and its numerous conveniences.
Resolved, That the thanks of the Association be tendered to the local com- mittee for its very thorough preparation and liberal outlay ; also, °
Resolved, That the local committee be respectfully requested to convey to the citizens of Rochester the high appreciation and cordial thanks of the Association for their generous hospitality extended to its members.
The report of the Committee upon Renominations, recommend- ing the reélection, for the year 1871, of the officers of the present meeting, wus unanimously adopted, and the officers were declared duly elected.
B. W. Dwight read a paper on “Considerations in favor of the thorough Reconstruction of Latin Grammar on Philological and Analytical Principles.”
He expressed his surprise that Latin grammar had never yet been placed in England or America on its true historical and philosophical basis. Of the three forms of lingual analysis—morphological, verbal, and sentential—he stated that only the latter had received any adequate scientific treatment in Latin. Any analytic comprebension of the real genetic structure of Greek forms, nominal and verbal, was impossible to the grammarians at Alexandria in their day of rude beginnings in linguistic study. And the Latin scholars who copied their ideas and analyses had only such conceptions as they fur- nished them of the relational elements of human speech, and of the special forms of their own language. The idga on which they acted in arranging declension-forms and conjugation-forms was as much the military one as any other, of making a bold front, behind which all exceptional and doubtful matters could be easily stowed away from view, except as they were searched for.
It is as necessary to the higher ends and uses of thorough lingual study to comprehend the anatomical structure of verbal and grammatical forms, as it is, for accuracy and effect in drawing and painting, to understand well the in-
AMERICAN PHILOLOGICAL ASSOCIATION. 28
ward conformation, part by part, of the things represented. Not until with- in a very recent date have the materials existed for the proper and final read- justment of Latin grammatical forms on a completely scientific basis; and what has been done so well in Germany and America in respect to Greek grammar, ought now to be done in respect to the Latin.
The structural elements of the Latin verb are the verb-stem and tense- characteristic, (which, when associated, make a special tense-stem in each case ;) with the personal or pronominal endings used for flexion ; and the union-vowels, (¢, $, and u,) employed to connect verb-stems and tense-stems with their personal terminations ; and, in the forms of the passive, the passive voice-sign. In the structure of the Latin noun, there is the same arrange- ment of word-stem and flexion-ending as in the verb; and, in the case-form, the idea of its number is carried plainly in each case as a general fact. Both nouns and verbs should be classified according to their elemental sub- stance, or their actual stems ; and in separate divisions according to the origi- nal or changed forms of those stems. Contraction isa mach more leading force in determining grammatical and verbal forms in Latin than in Greek ; although this fact is so little hinted at in our best grammars and dictionaries.
Latin nouns are divisible into two great clasees.
I. The consonantal declension, (the present third,) or those having a con- sonantal stem.
II. The vowel or contract declensions—all having vowel-stems, and all con- tract—five in number, A, E, I, O, U.
Exceptions, so named in our school grammars now, will disappear in large numbers under a right classification of Latin nouns, and a true presentation of their structural and pathological history. The reasons also for quite a number of prosodial rules will become manifest.
I. The consonantal typal or normal declension-form. It comprises all nouns that have a stem ending in a consonant.
Here the author gave, with references case by case to the corresponding Sanskrit-forms, the proper fiexion-endings of this declension and as the ort- ginal norm of all Latin declenston-forms ; arranging the cases according to their analytic relationship to each other as, nominative, vocative, accusative, genitive, dative, and ablative.
Under this declension he classified nouns as
1. Those having the pure unaltered stem in the nominative ; (1.) With the gender-sign, 8, affixed ; (2) Without it.
2. Those having an altered stem with or without the gender-sign.
1. Those of pure unaltered stems he subdivided into: (1) Labial stems ; (2) Dental stems ; (3) Liquid stems; with examples and remarks under each head. | 2. Those of altered stems in the nominative comprise the great mass of nouns in this declension of labial, dental, liquid, and guttural stems.
(1) Nouns having labial stems altered in the nominative, (done always by affixing the gender sign, s, to them,) are few in number: (3) As to those of dental stems, the four following rules prevail : 1§. They always drop a final dental before the gender-sign, as lapie for lapids, stem lapid: 28. They al- ways (with the exception of caput and its compounds) drop a dental when
94 PROCEEDINGS OF THE
final without the gender-sign, as poema, stem poemat, and lac, stem lact : 8§. They always drop one of two letters of the same kind, although radical, which would be final in the nominative, as in os, stem os¢: 48. They some- times lengthen the final vowel of the stem, or make a stronger vowel of it than before, as in miles, stem miltt.
(83.) As to liquid stems,
1. They have a manifest tendency to assume a strong form in the nomina- tive, as appears, |
§ 1. In strengthening the short vowel ¢ of the stem into e before n final, as in fumen, stem flumin.
§ 2. In taking the still stronger vowel o as a final vowel, when rejecting a radical » from the end of the nominative form, as in tmago, stem tmagin.
§ 8. In changing short o before r final in the nominative into w, asin robur, stem robér. In corpus, stem corpor, we have, besides this change, the change also of final ¢ into the stronger letter s. In such forms as fedus, stem fader, we have similar compound changes to those in corpus—only here it ise that is deepened into 2.
2. Liquid stems reject, like dental, the final letter of the stem before the gender-sign, as in pollis, stem pollin ; glis, stem glir - mas, stem mar.
8. In fel, mel, and fur—stems fell, mell, and farr—one of two repeated letters is dropped in the nominative, (as in the dentals as and bes.)
4, S final in the nominative of liquid stems is sometimes the gender-sign with the final letter of the stem absent, as in cints and pulois, stems ciner and — puloer ; and is always to be so analyzed when the given word is of the mascu- line or feminine gender. But it is sometimes substituted, for greater ease of utterance for an original 7 radical, as in aes and crus, stem aer and crur. It is always to be so analyzed when the stem is neuter.
(4.) As to guttural stems: they are never found unattended by the gender- sign: they are never of the neuter gender.
- Mr. Dwight then traced at length an analysis of the vowel or contract de- clensions. |
1. The vowel A declension. It includes all nouns whose stems end in a. Here belong also a few Greek nouns ending in a, as, and es—the final s in each of these forms being the masculine gender-sign. The Greek ¢ nouns, always placed here, should be contrarily placed with the ¢ declension.
2. The vowel E declension, (the present fifth.) In Greek grammar, nouns having ¢ stems are placed with those of @ stems, because so much alike in both their full and contracted forms. They should be in Latin in juxtaposition. Here the present and earlier forms of this declension were presented together, and their contractions were explained.
8. The vowel I declension. This is a new declension-form and not arbitra- rily chosen, but necessitated by the fact that nouns of the ¢ stems can be properly placed nowhere but in an ¢ declension. They are such as amnis, avis, civis, finis, ignis, ovis, unguis, etc., all whose stems end in ¢. Some | nouns belonging to this declension have the stronger vowel e in the nomina- tive, as in ades, nubes, rupes, etc., stem edi, nuln, etc.
4. The vowel O declension, (the present second.) It includes all nouns whose stems ending. Either the final o of the, stem is changed into wv in the nominative before the gerider-sign 4, and before m, the sign of no-gender ; or
AMERICAN PHILOLOGICAL ASSOCIATION. 25
the o syllable is dropped from the word, which is thereby so much shortened in form, asin gener, (for generus, for genero—s,) stem genero, and oir for virus, stem viro. In the Latin-Greek noun, heros, stem hero, (which should be placed under this declension,) the original mode of declining o nouns in Latin, appears but little changed from its typal form.
5. The vowel U declension. This includes all nouns whose stems end in 4. The changes in this declension are less positive from the normal form than those of the other declension-forms.
How absurd appears the statement, in the light of these facta and princi- ples, that nouns are to be divided into five declensions in Latin according to their various genitive endings ; and that the stem itself is to be found by cut- ting off, in the first and second declensions, the last letter of the genitive, and in the fourth declension the syllable @s, leaving only the fifth declension to be a vowel-declension.
The verb-forms of the language were analyzed by the author in the same way as those of the noun. The Latin verb was treated as if a wholly organic structure; and it was dissected and reconstructed according to the principles of its own organic constitution. Verbs were divided, like the nouns, into,
I, The consonantal, typal or normal conjugation-form.
II. The vowel conjugations.
1. The A conjugation. 2. The E conjugation. 3. The I conjugation, 4. The U conjugation, (a new conjugation, made necessary to include words of u stems, as acuo, loguor, pluo, ruo, spud, etc.)
Our Latin grammars should be fashioned so as to present to the student all that is now known of the structure of the language, according to the most thorough morphological analysis. Latin grammar is, however, after almost two thousand years of the study of its elements in all civilized nations, ina crude, unscientific, and unartistic shape, like the agriculture of the eastern world, in respect to both its principles and its instruments ; like the natural sciences in all countries at the beginning of this century; like all our dictionaries, Latin and Greek, and, within less than six years, English also.
Mr. H.'M. Colton, of New-York, read a paper, in which he pointed out some of the difficulties which are practically met with in securing uniformity of pronunciation of the Latin and Greek languages. This paper was followed by a long discussion, at the close of which, upon the motion of Prof. Buttz, the mem- bers of the former Committee upon the Pronunciation of the Classic Languages were appointed a committee “to devise further means to secure uniformity in the pronunciation of the classic languages.”
Prof. Comfort exhibited a folio volume, which was presented to the Association by Dr. McCartee, entitled, Zhe Monument of Yu, and containing a reduced copy of an inscription upon a tablet at Si-ngarfu, in the Province of Shensi, in China. An antiquity of over three thousand years is claimed for the inscription, and is
26 AMERICAN PHILOLOGICAL ASSOCIATION,
accepted as correct by Bunsen and some others, The monument is interesting as showing the tadpole character or form in which the ancient Chinese classics were written, before the present characters were adopted.
The minutes of the Association were read, and the Association adjourned at twelve m.
MEMBERS OF THE AMERICAN PHILOLOGICAL ASSOCIATION.
Rev. William Arthur, Albany, N. Y:
Prof. W. F. Alien, University of Wisconsin, Madison, Wis.
Pres. M. B. Anderson, University of Rochester, Rochester, N. Y.
Rev. J. Anderson, Waterbury, Ct.
Prof. N. L. Andrews, Madison University, Hamilton, N, Y.
Rey. Dr. A. N. Arnold, Chicago, Hl.
Prof. J. G. Barton, College of City of New-York,
Prof. W. F. Bascom, Washington, D. C.
Prof. H. L. Baugher, Pennsylvania College, Gettysburg, Pa.
Dr. E. L. Beadle, Poughkeepsie, N. Y.
Prof. N. W. Benedict, Rochester, N. Y.
Otis Bisbee, Poughkeepsie, N. Y.
Rev. Dr. Nathan Brown, 37 Park Row, New-York.
Prof. G. R. Bliss, Lewisburg University, Lewisburg, Pa.
Prof. J. R. Boise, Chicago University, Chicago, Ill.
Miss Alice R. Boise, Chicago, Il.
Dr. E. C. Bolton, Poughkeepsie, N. Y.
Miss Mary L. Booth, 89 Madison avenue, New-York.
Rev. J. C. Boyce, South Baptist Theological Seminary, Greenville, 8. C.
H. C. G. Brandt, Clinton, N. Y.
Prof. F. P. Brewer, University of North-Carolina, Chapel Hill, N. C.
Rev. C. H. Brigham, Ann Arbor, Mich.
Rev. Dr. R. J. W. Buckland, Theological Seminary, Rochester, N. Y.
B. Brannan, New-York.
Prof. H. A. Buttz, Drew Theological Seminary, Madison, N. J.
Porter C. Bliss, Washington, D. C.
Charles A. Bristed, Westminster Hotel, New-York.
C. J. Buckingham, Poughkeepsie, N. ¥.
8. M. Capron, Principal of High-School, Hartford, Ct.
Pres. Alexig Caswell, Brown University, Providence, 8. I.
A. B. Chailly, Poughkeepeie, N. Y.
Prof. H. F. Clark, Berea College, Berea, Ky.
Prof. C. J. Chandler, Columbia College, New-York.
Prof. Elisée Charlier, Principal of Charlier Institute for Young Ladies, New- York.
Pres. N. E. Cobleigh, East-Tennessee Wesleyan University, Athens, Tenn.
Prof. W. P. Coddington, Genesee College, Lima, N. Y.
Rev. H. M. Colton, Yale School, 1300 Broadway, New-York.
28 MEMBERS.
Prof. G. F. Comfort, (care of Harper Brothers, Franklin Square, New-York.) Rev. G. W. Cook, Poughkeepsie, N. Y. Prof. William B. Corbyn, St. Paul’s College, Palmyra, Mo. Prof. A. Crittenden, Packer Collegiate Institute, Brooklyn, N-Y. Pres. Howard Crosby, University of New-York, 302 Second ave., New-York. Prof. E. P. Crowell, Amherst College, Amherst, Mass. Prof. H. N. Day, New-Haven, Ct. Prof. H. A. Dearborn, Tuft’s College, Malden, Mass. Prof. Schele De Vere, University of Virginia, Charlotteville, Va. Dr. B. W. Dwight, Clinton, N. Y. Miss E. H. Denio, Albion, N. Y. Asher B. Evans, Lockport, N. Y. Prof. J. B. Feuling, University of Wisconsin, Madison, Wis. Prof. W. M. Fisher, Independence, Mo. Miss M. B. Flint, Monticello, N. Y. Prof. Norman Fox, William Jewell College, Liberty, Mo. Miss Helen M. French, Mount Holyoke Seminary, South-Hadley, Mass. 8. T. Frost, Amenia, N. Y. Prof. N. E. Gates, Albany, N. Y. W. C. Gaylord, Rochester, N. Y. Miss E. L. Geiger, Vassar College, Poughkeepsie, N. Y. George Gibbs, 260 Greene street, New-York. Prof. @. H. Gilmore, University of Rochester, Rochester, N. Y. W. Goodrich, Poughkeepsie, N. Y. Prof. W. W. Goodwin, Harvard College, Cambridge, Mass. Prof J. B. Greenough, Harvard College, Cambridge, Mass. Prof. J. H. Hadley, Yale College, New-Haven, Ct. Prof. 8. 8S. Haldemann, (St. John’s College,) Columbia, Pa. Prof. L. H. Hammond, Lebanon Valley College, Armville, Pa, Prof. A. Harkness, Brown University, Providence, R. I. Prof. C. S. Harrington, Wesleyan University, Middletown, Ct. William Haskell, Bucksport, Me. Prof. B. J. Hawthorne, West-Tennessee College, Jackson, Tenn. Prof. H. W. Haynes, University of Vermont, Burlington, Vt. Col. T. W. Higginson, Newport, R. I. Prof. 8. D. Hilman, Dickinson College, Carlisle, Pa. Prof. Oscar Howes, Shurtleff College, Alton, II1. Prof. C. G. Hudson, Genesee College, Lima, N. Y. Prof. M. W. Humphreys, Washington College, Lexington, Va. Rev. Dr. A. B. Hyde, Allegheny College, Meadville, Pa. Rev. Dr. A. C. Kendrick, University of Rochester, Rochester, N. Y. Prof. A. C. Kimber, Jr., Annandale College, Annandale, N.Y. _ Prof. Louis Kistler, North-Western University, Evanstown, II. J. R. Leslie, Poughkeepsie, N. Y. Rev. W. T. W. Lewis, Poughkeepsie, N. Y. Pres. A. A. Livermore, Unitarian Theological Seminary, Meadville, Pa. Aaron Lloyd, 117 Nassau street, New-York. Prof. Ch. L. Loos, Bethany College, Bethany, W. Va. Benson J. Lossing, Poughkeepsie, N. Y.
MEMBERS. 29
Prof. T. R. Lounsbury, Yale College, New-Haven, Ct.
Prof. O. F. Lumry, Wheaton College, Wheaton, II.
William F. Lush, Rochester, N. Y.
Prof. E. H. Magill, Swathmore College, Philadelphia, Pa. Prof. F. A. March, La Fayette College, Eaton, Pa.
Prof. D. Marvin, Racine, Wis.
Rev. M. Maury, Cold Spring, N. Y.
Prof. J. H. McDaniels, Geneva, N. Y.
Rev. C. M. Mead, Andover Theological Seminary, Andover, Mass. Prof. G. M. Millan, Hillsdale College, Hillsdale, Mich.
W. J. Milne, Lockport, N. Y.
Prof. A. H. Mixer, University of Rochester, Rochester, N. Y. Bela P. Mockoon, Cedarville, N. Y.
Prof. W. L. Montague, Amherst College, Amherst, Mass,
Prof. C. D. Morris, Mohican Lake, Peekskill, N. Y.
Pres. F. A. Muhlenburg, Muhlenburg College, Allentown, Pa. James E. Munson, 117 Nassau street, New-York.
Rev. Dr. J. H. Myers, Milton, N. Y.
Bennett H. Nash, Boston, Mass.
C. W. Nassau, Lawrenceville, N. J.
Prof. E. North, Hamilton College, Clinton, N. Y..
T. H. Norton, St. Catharine’s, Canada.
Prof. F. W. A. Nots, Muhlenburg College, Allentown, Pa. Pres. Cyrus Nutt, Indiana University, Bloomington, Ind.
Prof. J. C. Overhiser, New-York.
J. Person, Belvidere, N. J.
Prof. Oval Pirkey, Christian University, Canton, Mo.
Prof, Samuel Porter, College of Deaf-Mutes, Washington, D. C. Rey. E, L. Prentice, Poughkeepsie, N. Y.
Prof. A. J. Quinche, University of Mississippi, Oxford, Miss. Pres. J. H. Raymond, Vassar College, Poughkeepsie, N. Y. Prof. J. A. Roberts, Vassar College, Poughkeepsie, N. Y.
Prof. J. A. R. Rogers, Indiana Asbury University Greencastle, Ind. L. L. Rogers, New-York.
Prof. E. B. Ruggles, Dartmouth College, Hanover, N. H.
Pres. G. W. Samson, Columbian College, Washington, D. C. Prof. A. J. Schem, 114 Nassau street, New-York.
Dr. Henry Schliemann, Paris, France.
Prof. J. B. Sewall, Bowdoin College, Brunswick, Maine.
Pres. Joseph Shea, St. John’s College, Fordham, N. Y.
H. W. Sibley, Rochester, N. Y.
Prof. W. B. Silber, Albion College, Albion, Mich.
Rev. W. C. Smith, Poughkeepsie, N. Y.
Gen. A. B. Smith, Poughkeepsie, N. Y.
Prof. J. W. Stearns, University of Chicago, Chicago, Il.
Prof. William Steffins, St. John’s College, Annapolis, Md. Prof. W. A. Stevens, Dennison, Granville, Ohio.
Prof. W. F. Swahlen, McKendree College, Lebanon, I11.
Prof. D. 8. Talcott, Bangor Theological Seminary, Bangor, Me.
30 MEMBERS.
Rev. Dr. 8. BR. Taylor,* Theological Seminary, Andover, Maga.
Prof. E. A. Tanner, Dlinois College, Jacksonville, 1.
Hon. J. Hammond Trumbull, Hartford, Ct.
Prof. J. C. Van Benschoten, Wesleyan University, Middletown, Ct. Dr. H. Von Holst, 206 West Thirty-sixth street, New-York,
Addison Van Name, Librarian of Yale College, New-Haven, Ct.
Miss Julia E. Ward, Mount Holyoke Seminary, South-Hadley, Mass. . Dr. H. G. Warner, Rochester, N.Y.
Prof. J. B. Weston, Antioch College, Yellow Springs, Ohio,
Mrs. A. E. Weston, Antioch College, Yellow Springs, Ohio.
Prof. Nehemiah White, St. Lawrence University, Canton, N. Y.
Dr. William D. Whitney, Yale College, New-Haven, Ct.
Prof.E. H. Wilson, University of Rochester, Rochester, N. Y.
Prof. Alexander Winchell, University of Michigan, Aan Arbor, Mich. G. W. Winslow, Evanston, [11.
ADDITIONAL, TO JUKE 1, 1871.
F. L. Balchelder, Stafford, Conn.
H. R. Blaisdell, Maysville, Ky.
Rev. J. H. Blakely, Bordentown, N. J.
Prof. W. C. Crane, Baylor University, Independence, Texas. Pres. E. 8. Dulin, St. Stephen’s College, Columbia, Mo.
Prof. 8. A. Theo. Jobe, St. John’s College, Little Rock, Ark.
E. C. Parsons, Derry, N. H.
Prof. E. Snyder, Illinois Industrial University, Champaign, UL Hon. E. G. Squier, 189 West Thirty-fifth Street, New-York. Prof. F. Stengel, Columbia College School of Mines, N. Y. Prof. E. H. Twining, University Minnesota, St. Anthony, Minn. Prof. Lyman Coleman, Lafayette College, Easton, Pa.
Prof. Wm. Carey Crane, Baylor University, Independence, Texas. Prof. 8. J. Young, Brunswick, Me.
TRANSACTIONS
AMERICAN
PHILOLOGICAL ASSOCIATION.
1871.
PUBLISHED BY THE ASSOCIATION. PRINTED BY CASE, LOCKWOOD & BRAINARD, . HARTFORD,
1872.
CONTENTS.
I. Studies in Cymric Philology. By Professor E.W. Evans. - . . . 1 II. On the so-called Attic Second Declension. By Professor Freperic D. ALLEN. . 18 IIT. Strictures on the Views of August Schleicher respecting the Nature of Language and kindred subjects. By Professor Wittiam D. WHITNEY. . 85 IV. On English Vowel Quantity in the Thirteenth Century and in the Nineteenth.
By Professor James HaDLey. . . . 65 V. Anglo-Saxon and Early English Pronunciation. . By Professor Francis A. March... . . 108 VI. Some Notes on Ellis’s Early English Pronunciation. By Cuartes Astor BrRIsTeD. . . «. 114 VII. On Algonkin Names for Man. By J. Hammonp TRUMBULL. . 1388 VIIL On some forms of Conditional Sentences in Latin, Greek, and Sanskrit. By Jamres B. GREENOUGH. - 159
PROCEEDINGS :—Third Annual Session, New Haven, 1871.
TRANSACTIONS
OF THE
Awertcan PurtotogicaL ASSOCIATION. 1871.
I. — Studies in Cymric Philology.
By E. W. EVANS, PROFESSOR IN CORNELL UNIVERSITY.
My object in this paper is to commence a series of notes on questions of Cymric philology, some of which are discussed or suggested and uthers left untouched in the great text-book on this subject, the Grammatica Celtica of Zeuss. My referen- ces will be to the second edition, in which some errors of the original work have been corrected, and some important addi- tions made, by the learned Ebel. I shall also refer frequently. (by the abbreviation Myv.) to a class of documents not rauch used by Zeuss or his editor, the old and early-middle Welsh poems, as they appear in the Myvyrian Archaeology, edition of 1801. Among other documents referred to will be the Beitrige zur Sprachforschung (Beitr.), Berlin, 1858~65,— the oldest copy of the Welsh Laws, known as the Venedotian (Leg. Ven.), referred to the twelfth century, — and the oldest copy of the first part of the Annales Cambriz, known as the Chron- icum Cambriew (Chron. Camb.), and referred to the latter part of the tenth century.
I.
A few preliminary statements in regard to the history of consonant changes in Welsh may conduce to the better under- standing of some things that follow.
2 KE. W. Evans.
In comparing old Welsh, as seen in the ancient glosses and fragments published by Zeuss and Stokes, with modern Welsh as seen in all compositions dating from the Reformation down, we perceive that there has been a general infection of conso- nants not initial, as follows: Old p, ¢, and ¢ have become, re- spectively, 5, d,and g; old b and m have become f (pronounced as English v): old d has become dd (pronounced as English th in the); while old g-has in some cases passed into ¢, y, or close e, and in other cases disappeared. Exceptions regularly appear, however, in certain combinations, e. g. in st, rt, and né.
Extant manuscripts of the twelfth century show that these changes in consonant sounds had already taken place, for the most part, in the transition from old to middie Welsh. The most prominent exception is, that in middle Welsh there was © more or less fluctuation between final p, ¢, c, and 8, d, g.
The changes above described I shall designate as the de- pression of consonants, in order to distinguish them from other kinds of infection, known as the aspirate and the nasal.
While initial consonants have, in passing from old to mid- dle and modern Welsh, been persistent in the radical forms of words, the complex Welsh system of initial inficctions Cif we may so designate a system by which words undergo initial changes when placed in certain syntactical relations) received considcrable increments during the middle period.
Il.
There has been some room left for doubt as to when the change from old d to modern dd took place in pronunciation. Until about the year 1400 there was no distinctive notation for the latter sound; see Stephens’ Literature of the Kymry, p. 453. I find the clue to it, however, in earlier documents, by comparing two modes of spelling. In some of them d is used to represent this, besides its more usual sound, while in others, strangely enough, the same secondary office is assigned tot. By observing what places occupied by d in the one class are regularly assigned to ¢ in the other, it may be seen that even in early-middle Welsh (aside from such cases of initial
" Studies in Cymric Philology. 3
inflection as were not yet common) the subvocal dd sound generally obtained where it now does.
The nearest approach that I have seen to a recognition of this test is in the second edition of Zeuss, where it is stated that ¢ when final sometimes represents the infected d (dd), but hardly when internal. Examples are given from the old- est copy of the Laws— where a mixed orthography prevails in this particular as in others. I therefore deem it important to call attention to the fact that in the majority of the poems of the twelfth century and the first half of the thirteenth, printed in middle Welsh orthography in the Myvyrian — in- cluding those taken from the Black Book of Caermarthen — the use of ¢ to represent the dd sound is quite regular, not only when final but in other positions as well. As test exam- ples I give the following words in which we cannot, consist- ently with what is known of their history or etymology, suppose a mute ?¢: Old Welsh Griphiud (Chron. Camb.), modern Gruffudd, Griffith, in one class of middle Welsh documents is Gruffud (Myv. I. 365), in the other, Gruffut (ib. 290); old Welsh dodin, gl. ‘turma,’’ modern byddin, is in one class bydin (ib. 202), in the other bytin (ib. 387) ; modern Dafydd, David, is in one class Dafyd (ib. 198), in the other Dawit (ib. 836); modern bardd, bard, (compare the Bapdoc Of Strabo,) is in one class bard (ib. 337), in the other bart (ib. 280); modern Gwyddel, Irishman, (compare old Irish Gaedal,) is in one class Gwydel (ib. 174), in the other Gwitel (ib. 80); modern heddiw, to-day, (from diw or div, day,) is in one class hediw (ib. 415), in the other hetiw (ib. 165) ; modern ymddiddan, conversation, (from diddan,) is in one class ymdidan (ib. 173), in the other ymtitan Cab. 265) ; so also ymtial, revenge. (ib. 79), modern ymddial, from old Welsh digal (Chron. Camb.). The list might be extended in- definitely.
It detracts nothing from the force of the argument to say that there are exceptional instances of variable spelling in the same document. The evidence, then, goes to show, what we should expect from analogy, that as a general fact the infec- tion of old d@ took place in the transition to middle Welsh.
4 E. W. Evans.
But whether the dd sound was altogether unknown in old Welsh is a question which I do not at present discuss.
Hil.
Zeuss observes that since the quantity of vowels is not marked in British MSS. it must be “ determined by compari- son,” that is, by comparison with Irish and Latin words, lat- inized British and Gallic names, etc. Doubtless the conclu- sions to which he is thus led are generally correct so far as old Welsh is concerned; but he often falls into error in assuming the persistence, in later Welsh, of original short vowels.
In considering the quantity of Welsh vowels I leave unac- cented syllables out of the account, because the tendency of the modern language is to make them all short without regard to their origin — diphthongs, of course, excepted. Thie accent, it should be observed, is almost always on the penult. In regard to the quantity in accented syllables and monosylla- bles I have two general facts to state which do not seem to have been observed.
1. In monosyllables and accented syllables the vowel is regularly made short when followed by two (or more) counso- nants. This statement must be understood as referring to the inherent quantity of the vowel itself; for on account of the time required for the distinct utterance of two consonants the syllable may still be called long. When a long vowel is thrown into such a position, by composition, derivation or other grammatical process, it is shortened: thus cryfder, strength, from ergf, strong ; undeb, union, from in, one ; porfa, pasture, from pdri, to graze; etc. The rule holds good even when the second of the two consonants isz or w: thus moliant, praise, from molt, to praise ; gweddwon, widows, from gwéddw ; etc. But we must avoid the error of treating w as a conso- nant in the diphthong wy (pronounced very much as French out in bouillon) ; thus gwé slwyd, was seen, from gwé led, to see. The exceptions to the rule are very few, and arise from syn- wresis; thus gwndnt, they do, older gwnadnt. We must ex-
Studies in Cymric Philology. D
clude from this rule, as properly belonging to the following, such words as ofn, cwbl, gwobr, and others ending in two con- sonants the last of which is J, n, or r; for they are really dis- syllables (formerly sometimes written ofyn, cwbyl, gwobyr, etc.), although on account of the very short quantity of the last syllable they are treated in verse as monosyllables.
2. Vowels followed by only one consonant (in monosylla- bles and accented syllables) are, as a very general rule, long, when the consonant is b, d, g, f, or dd,— that is to say, when it belongs to the class of depressed consonants, or those which - have undergone the change before mentioned as marking the transition from old to middle Welsh. A very few words are excepted,— ag and nag when not emphasized, rhag, ab, and possibly one or two others, that have not occurred to me. All the examples given in Zeuss under the head of “ Vocales Britannicae Breves’’ conform, in their modern forms, to the rule; that is, the original short quantity has been lengthened: thus, mdb, son, old Welsh map; cdd, battle, old Welsh cat; llaéfar, speech, old Irish /abar; méfl, disgrace, Irish mebul ; gof, smith, Gallic gob; ébol, colt, from primitive ep, horse ; cog, cook, old Welsh coc, Lat. coqguus ; rhyd, ford, old Welsh rit; llydan, broad, old Welsh litan ; byd, world, old Irish bith ; Difed, Demetia; etc.
That the depression of consonants and the lengthening of preceding short vowels were chronologically connected will appear evident when we consider, further, that before single consonants not depressed the vowel (in monosyllables and ac- cented syllables) is very frequently short — always so in the cases where p, ¢, c, and m remain: thus llac, loose; llyffant, toad ; calon, heart; cyllell, knife; crwm, bent; gwan, weak ; llong, ship; copa, top, summit; gyru, to drive; cusan, a kiss ; cetyn, a piece, a bit; chwythu, to blow; etc. But before / cases of short quantity are rare; ch regularly lengthens the preceding vowel; so also in monosyllables does 8s. The con- dition is here implied, of course, that the consonant is not fol- lowed by another. Formerly consonants not depressed, except / and those represented by two characters (ll, th, ng, etc.) were often doubled, to indicate the short quantity of the
6 KE. W. Evans.
preceding vowel; thus, gwann, gyrru, cettyn, etc. It should be added that ng — which generally represents an original ne — always shortens the preceding vowel, as if it were still two consonants.
Long words have a secondary accent, preceding the primary. It is subject to the two foregoing rules of quantity, except where it falls on the prefixes cyd and di; in that case the vowel.remains long even before two consonants ; thus, eydsyl- weddoldeb, didreftadu.
IV.
In treating of the derivation of substantives and adjectives in Welsh, Zeuss makes no mention of the termination -ing, which in the early poets, (old and early-middle), occurs not in- frequently. In the oldest Welsh MSS. g is used to represent (besides its more usual sound) the sound now represented by ng: thus in the Black Book of Caermarthen, Freige, modern Ffraine, the French, or France; Tegigil, modern Tegeingl,.a local name (Myv. I. 578). Hence -ing is usually disguised as -ig; thus Hrgig, modern Erging, a local name (ib.). In late-middle copies of the earlier writings the spelling -tng prevails.
This termination is often used with a patronymic force. It appears also in the names of certain districts, most of which are known to be derived from personal names.
Thus in Gwalchmai, a poet of the twelfth century (Myv. I. 194), Cynan Coeling, Cynan, of the race of Coel : in Cynddelw, also a poet of the twelfth century (ib. 232), Kynverching werin, the people of the race of Cynvarch ; 0 vonet Coelig, of Coelian stock; 0 Vaelgynig (rhyming with 2g, that is ing, dis- tress), of the race, or country, of Maelgwn; roted ardunyant ar Dinodig, honor has been bestowed on Dinoding, that is, on the country of Dunawd (the Dinoot of Beda). Of the old poets, Taliesin (ib. 71) has Coeling; Meigant (ib. 159) has Cadelling, of the race of Cadell, and Dogfeiling, of the race of Dogmael ; Golyddan (ib. 157) has Glywysyg, which in a copy of Nennius referred to the tenth century is spelt @levis- ing. This is the ancient name of some district in South Wales,
Studies in Cymrie Philology. 7
derived from the personal name Glewys (see Stevenson’s Nennius). Price, in his Hanes Cymru, erroneously retains the early-middle spelling, Glewysig.
Does -ing represent the -inews, -inca, -incum (Z. 807), of Gallic personal and local names ?
V.
The Juvencus gloss, “ istlinnit, profatur,” (Beitr. iv. 392) is mentioned in the second edition of Zeuss as if it were the only example of the preservation in Welsh of the third singu- lar present indicative active in -it; compare Irish -id and Latin -é.
In the‘pld Welsh poems, which although they come to us in a corrupt form — that of imperfect translations into middle Welsh — yet often preserve archaic features, I find frequent examples of the use of this termination, generally depressed, however, in the later copies, to -id. Thus in the Elegy of Cynddylan, by Llywarch Hén (Myv. I. 109):
Eryr Pengwern pell gelwid heno;
Ar waed pwyr gwelid. The eagle of Pengwern calls afar to-night ; Over the blood of men it watches.
In Llevoed Wynebglawr: golut byt eyt dydau (Myv. I. 154), worldly wealth goes, comes; guae drut ny chretit (ib. 155), woe to the presumptuous one that does not believe ; difrys guanec, diffustit traeth (ib.), the wave hastens, it beats the shore.
In the “ Englynion Cain Cynnwyre,” of unknown author- ship but undoubtedly old : gorchwythid gwynt uwch aber, strong blows the wind over the estuary ; cyrchid carw dan vrig derwen, the stag seeks the shelter of the oak; anrheithid rhywynt anial, the tempest ravages the forest. (Myv. III. 142.)
Among the proverbs, which, it should be observed, often bear internal evidence of having been drawn from old Welsh sources, many examples in point occur, including the follow- ing: anghwanecid mevl mawrair, boasting adds to the disgrace (Myv. III. 147); eltd bryd yn ol breuddwyd, the fancy goes
be
8 E. W. Evans.
according to the dream (ib. 155); gwnelid anghelfydd an nerth, weakness makes unskillful (ib. 161) ; Uyvid y cc y gwaew y brather ag ev, tho dog licks the spear wherewith he is wound- ed (ib. 163) ; rhetid maen oni gafo wastad, the stone rolls till it finds the plain (ib. 176).
Aside from the proverbs I have found no examples of the use of this termination in prose. It occurs two or three times (doubtless as an archaism) in the poetry of the twelfth century, and then disappears. None of the Welsh gramma- rians, so far as I know, have recognized it at all. Translators have gencrally confounded it with the imperative active -it or -id, which sometimes occurs, later, for the usual -et or -ed.
Vi.
In treating of the Welsh passive conjugation, Zeuss gives the present (and future) indicative ending -27; to which, in the second edition, the less frequent -awr is added. No men- tion is made in either edition of the very important forms -ator, -etor, -itor (sometimes, -otor, -cator, -etawr, -ttior, -itiawr) ; compare Irish -ithir, -ither, in passives, and -adar, -edar, -idir, in deponents ; also, Latin -atur, -etur, -itur.
These passive endings (occasionally depressed in our copies to -ador, -idiawr, etc.) occur frequently in the old Welsh poets, and sometimes in the earlier poets of the middle-Welsh period : e. g. in Llywarch Hén (Myv. I. 107), cenau Cyndrwyn cwyn itor, the offspring of Cyndrwyn is bewailed ; in the Gododin, gweinydiawr ysguydawr yngweithen (ib. 7), shields are pierced in the combat ; in various old poems which have been attrilb- uted to Taliesin, gwelattor arwyddion (ib. 83), signs are seen ; golchettawr et lestrt, bid gloew et vrecct (ib. 39), his vessels are washed, his wort is clear; hyd tra fwy fyw-. crybwylletor (ib. 70), a8 long as I live he shall be commemorated ; cath! gwae canhator cylch Prydain amgor (ib. 75), the song of woe is sung - round Britain’s borders; in Llevoed Wynebglawr (ib. 154), pob llyvur llemityor arnau, every coward will be trampled upon ; in Gwalchmai (ib. 197), ef gwr gwelitor, le is seen (appears as) a man; in Cynddelw (ib. 205), arwyrain Owain cain
— Studies in Cymric Philology. 9
cenitor, the praise of Owain is (or, will be) fitly sung. Again in the proverbs, clywitor corn can ni weler (ib. III. 151), a horn will be heard though it be not seen ; telstor gwedi hal- awglw (ib. 177), there is paying (lit. it is paid) after false swearing. Besides the two last I have found no examples in
prose. |
Like -ir, these endings are present or future, singular or plural, according to the connection. But unlike -ir and the other passive endings, they are used only in the third person. At least I have failed to find a single example of their use in the first or second person, in the whole mass of documents published in the Myvyrian. Yet the pronouns of the first and second persons occur so frequently in the early poetry that we have a right to expect such examples, if they were not precluded by usage. I must therefore dispute the correct- ness of the statement made by Zeuss and others, that the Welsh language preserves no remnant of the personal conju- . gation in the passive voice.
Dr. Owen Pughe, who is a very unsafe guide in early Welsh, calls verbs in -ator or -iator gerunds, and verbs in -ttor or -etor supines, translating thus: ‘‘ adedator, in building,” “ adeil- ttor, to be building.” It is to be regretted that these fictions are reproduced in the Welsh introduction to the second edi- tion of the Myvyrian, lately issued. Iam not aware, indeed, that the real character of these verb-endings has ever been
pointed out. VIL.
Zeuss derives Cymro, Cambrian, from cyn-, synonymous with Latin con-, and bro, region, Gallic brog. The name would thus mean compatriot. The plural, Cymry, might come by umlaut from Cymre, after the analogy of ffyn, staves, from Sfon.
But the feminine of Cymro in middle and modern Welsh is Cymraé’s. This points to Cymra as the earlier masculine form, which, again, might give Cymry as the plural by umlaut after the analogy of bustych, steers, from bustach. °
An earlier Cymra is also indicated by the name of the lan- guage, Cymraég (middle Cymraéc, Myv. I. 272); thus
2
10 E. W. Evans.
Gwyddeleg, the Irish language, from Gwyddel, Irishman ; Gwenhwyseg, the Gwentian dialect, from Gwentwys, etc.
It is Cymra (as opposed to Cymro) that is indicated, again, by the adjective Cymréig, Cambricus; compare Ffrengig, Gal- licus, from Ffranc, Gallus ; gwyrenig, from gwyran; gwledig, from gwlad, etc.
There are no analogies whatever for deriving any one of these words from the form Cymro; we should have, instead, Cymroé's, Cymroé'g and Cymroig, which forms never occur.
In view of these facts I cannot but regard the etymology of the name Cymry as still unexplained. Ido not discuss the theory of its identity with the Cimbri of the Romans, except so far as to say that any argument against that theory based on Zeuss’s account of the origin of the word would be worth- less.
We have an analogous case in middle Welsh Cornaw, Corn- wall (Myv. II. 267). The more frequent form Cernyw (as a derivative whose ending begins with a slender vowel) indi- cates the root Carn (as opposed to Corn). This we accord- ingly find in the Latin Carnabii.
The orthographical distinction between Cymry, as the name of the people, and Cymru, as the name of the country, (pro- nounced alike,) is a late one. In early-middle writings both are spelt with a final y, usually Cymry; in the oldest copy of the Laws, referred, as already stated, to the twelfth century, the spelling is Kemry (Leg. Ven. 2); in the Black Book of Caermarthen, also referred to the twelfth century, Kimry (Myv. I. 578).
The following early readings are also worthy of notice: Camaraé's, a Welsh woman (Leg. Ven. 96); Aymeraé‘c, the Welsh language (Brut Gruffudd ab Arthur, Myv. ITI. 155). But however the early-middle scribes varied the spelling in other respects, they never doubled the m in Cymro or any of | its derivatives. This shows that they had no idca of its being compounded of cyn- and bro. Dr. Owen Pughe adopted the spelling Cymmro in order to make it agree, as he supposed, with his theory of the etymology (cyn, first, and 6ro, which should really give us cynfro, however): and the remarkable
Studies in Cymric Philology. 1]
Cynmraeg of Zeuss must be another accommodation of the same sort, taken from some erratic modern writer ; it is judi- ciously left out of the second edition.
Meilyr, a poet of the close of the eleventh century and the first half of the twelfth (Myv. I. 191), has clas Cymreyt (Cymreijdd, as the rbyme shows), which Dr. Owen Pughe translates “the region of connected mountains,” intending Wales, but assuming the name Cymreydd to be compounded of cyn-, and bre, height. I cannot doubt, however, that it is another plural of Cymro — indicating Cymra again as the ear- lier form; compare glenydd, banks, from glan; gwledydd, countries, from gwlad; trigfeijdd, abodes, from trigfa, etc. I give the passage with a translation.
Edewis eurwas clas Cymreyt, Canawon Mordai, mynogi ryt, Dytwyreo Owain Eingl didudyt.
The illustrious one of the land of the Cymry (the race of Mordai, of lavish generosity) did promise that an Owain should arise, the expeller of the English.
Mordai was a country of the northern Cymry, celebrated inthe Gododin. To the northern Cymric heroes, much lauded in the old poets for their generosity, the Welsh of the middle ages were fond of tracing their pedigrees. The reference in the passage is evidently to a reputed prophecy of Myrddin, still extant (Myv. I. 144), in which it was promised that an Owain should reconquer England as faras London. Meilyr would fain see this prophecy fulfilled in the fortunes of his young contemporary, Owain Gwynedd.
That Pughe has entirely misapprehended the meaning of the above lines appears, further, in his making the verb edewts govern an indirect object, thus: “promised the sons of Mor- dai that,” etc. This, though good English, would be a gross solecism in Welsh; the preposition t, to, being required after the verb in such a case.
VG.
Zeuss mentions the Armoric adverb quet (pronounced ket), used in negative sentences like the French pas or point, as of uncertain origin ; and when it is used in affirmative sentences he seems to miss its precise force. Examples: ne tardomp
12 Ei. W. Evans.
quet ‘* ne tardemus ;” me carhe gouzout quet goude “ scire certe opto postea’’ (better, scire paululum opto postea) ; heb quet anam “sine ulla macula.” The word should be explained by the Welsh cat (diminutive cetyn), a piece, a bit. The above examples would be literally translated thus: Let us not delay a bit,— I wish to know a bit, after this,— without a bit of stain. .
The reader should understand that qu for & or ¢ is to be accounted for by the use, in Armoric, of French modes of of spelling.
IX.
In the second edition of Zeuss the following words are men- tioned as exhibiting in old Welsh the Celtic infinitive in -m (compare Irish -am, -em), to wit: dierchim, ad poscendum (Cod. Lichf.), modern ¢ erchi ; diprim, gl. “ essum”’, food, eating, Cornish dibbry, to eat ; molim, laudare (Cb.), modern molt. It would thus appear that in infinitives middle and modern -¢ represents old -im. .
Among the Luxemburg glosses is “‘ dowohinnom, austum,” that is, haustus, draught, drawing. In the second edition of Zeuss, modern “ gwehyn, exhaustio,’”’ is suggested in explana- tion of this gloss — with an unnecessary query, I think : com- pare gwehynydd dwfr, drawer of water (Bible) ; and as to the vowel-change compare mnouitiou, gl. ‘‘nundinae,” modern newidiau; bodin, gi. “ turma,” early-middle bedin (Myv. I. 85); kegin, “coquina” (Mab.); etc. Again, as to the change of do-guohinn to do-uohinn, here postulated, there are several other old Welsh glosses which show that initial de- pression, in composition, was already incipient (Z. xxvii). But what is-om? As do-, later dy- (ad) is usually a prefix to verbs, and the use of the infinitive as a verbal noun (like Eng- lish verbal nouns in -ing) is very common in Welsh, I am led to consider do-uohinn-om (q. d. dywehinno), as exhibiting an- other old Welsh form of the infinitive in -m. Through -auwm and -om, from an older -am, we may suppose the interchange- able infinitive endings -aw and -o of middle and modern Welsh to have come.
That -aw in infinitives represents an original -am, I pro-
Studies in Cymrie Philology. 18
pose to show by the rhymes in the Gododin. In our me- diaeval copies of that poem the infinitive -aw rhymes with llaw, taw, anaw, ffaw, gognaw and arnaw; see the text of Williams ab Ithel.
Llaw, hand, was originally lém; compare old Irish lém, band.
Taw, silent, quiet, was tam ; compare Ir. éamA, still, quiet.
Anaw, spirit, inspiration (not “harmony,” as Pughe has it), was andm; compare Irish anam, life, soul, in which, however, the quantity differs. To justify my definition of anaw I could cite many early examples; let the following from Gwalchmai (Myv. I. 198) here suffice: Owain anger- dawl, anaw anfeidrawl, aer-wrawl wrhydri,— the ardent Owain, of unbounded spirit, of battle-braving heroism.
Ffaw, glory, is the Lat. fama.
Gognaw, which seems to be used in the Gododin as a proper name, I pass over, because I am ignorant of its origin ; from analogy I would infer guocndm as the original form.
The compound arnaw or arno, on him, must be resolved into ar-n-dm; compare em (modern ef, he or him) in the Juvencus glosses; also compare the similarly compounded words ynddaw or ynddo (yn-d-dm), in him; rhagddaw or rhagddo (rac-d-dm), before or againsthim ; arnynt (ar-n-hwynt), on them; trwyddoch (trwy-d-awch), through you; ete. That the -aw or -o of this class of compounds was -am, in some dia- lect, at a time at least two hundred years later than that as- signed by critics to the composition of the Gododin, is shown by the example racdam, that is, rhagddaw, in the Juvencus glosses (Beitr. iv. 407), which are referred to the ninth century.
As au (aw) in those cases where it is interchangeable with o, regularly represents a primitive @ (Z. 94), we may infer that the quantity of the infinitive -am was long in Welsh, al- though it does not appear to have been long in Irish. The first change was doubtless to -aum or -om; in the Luxemburg glosses, o for the more usual az prevails; thus -ol for -aul, -oc for -auc, and -om for -aum. The next step was to -auv or -ov; thus dauu, that is dauv, son-in-law, in the Oxford glosses, for primitive ddm (Z. 1055), middle and modern daw ; so also
14 E. W. Evans.
lof, still preserved in the compound Jlofrudd, murderer, liter- ally red-handed, for primitive lém, middle and modern llaw. The infection of the m in -d@m took place, exceptionally, before the transition to middle Welsh; otherwise we should regularly find, in middle Welsh, -awf or -of instead of -aw or-o. As another indication that -am in infinitives was long (as well as in the other cases where it passed into middle Welsh -aw), it may be observed in the Gododin and the other old Welsh poems that it was never made to rhyme with the superlative ending -am, or with the verb-ending -am of the first person singular, both of which were short, and passed into middle Welsh -av (the modern -af). In Armoric the infinitive end- ings -im and -am passed, respectively, into -if and -af: thus dibrif, to eat, for diprim, and guisquaf, Welsh guisgaw, to clothe.
The dm postulated above, in arnaw and other compounds of that class, as another form for ém, he or him, (au or o for ef,) is preserved, regularly, in middle and modern 0, he or him ; thus gwelats 0, I saw him, gwelwyd 0, he was seen. In efo, he or him, we are to recognize ef-o (€m-dm) ; so also efe is ef-ef (€m-ém); compare hwynthwy (hwynt-hwy), they or them; Ayhi (hi-Ai), she or her; tydi (¢z-tz), thou or thee; etc. These doubled pronouns (analogous to Latin sese) are somewhat more emphatic than the simple forms, and are ac- cented on the last syllable.
X.
One of the most important of the Ogmian inscriptions is that found at St. Dogmael’s in Wales; see Stokes’ Three Irish Glossaries. It is bilingual; the Ogmian being Sagram- ni magi Cunatami, and the Latin, Sagrani fili Cunotami. The interpretation is (Zhe stone) of Sagranos the son of Cu- natamos ; the old Celtic masculine declension, -os, -t, being well established by Gallic inscriptions. Of the variations here seen in the forms of the two proper names, I take those ef which I have found the exact phonetic equivalents in mid-
dle Welsh. Sagranos in middle Welsh would regularly be Saeran ; com-
Studies in Cymric Philology. 15
pare Maelgwn for Maglocunus. I find the name Saeran in the Genealogy of Welsh saints (Myv. II. 51).
In magi we are to recognize Irish mac, Welsh map, son. Stokes infers makvos as the primitive form. Notwithstanding the usual correspondence of British p to Irish c, the form mac- ewy, youth, is found in old and middle Welsh writings ; e. g. in Llywarch Hén (Myv. I. 128) and in Cynddelw Cib. 252). There are several other words in which the Welsh has both the ¢c form and the p form; e.g. in talcen, that is, tal-pen, front of the head, forehead. This is probably an admixture arising from some ancient contact of British with Irish tribes.
Cunatamos would regularly be Cunadaf in middle Welsh ; and the name is found in precisely this form in the Triads of the War-horses (Myv. II. 21). The same name occurs also as Cunedaf and Cyndaf; compare the Cuneglasus of Gildas, which in middle Welsh is Aynlas (Myv. I. 85). In Liber Landavensis, which mixes old and middle forms, we find the name as Conatam (228) and as Condaf (132). In early Ar- moric it is Conatam (Z. 111).
As to the cun (cuna-, cune-, cuno-,) of this and other British names, e. g. Maglocunus, Cunobelinus, Zeuss compares Welsh cwn, summit; but in the early poets I often find the identical form cun in the sense of chief or captain: co. g. in Cynddelw (Myv. I. 210), rybydwn bencerd ben cun, I was the chief minstrel of the chief captain; also (ib. 233), un katkun val katki Aeron, one war-chief like the war-dog of Aeron. In a late-middle version of the ‘‘ Officium B. Mariae’”’ (ib. 559), occurs Duw ben cun, God the Supreme King.
If the tam in Cunatam were long, the name would mean, the silent chief; but in that case the middle Welsh form should be Cunadaw. Other evidence that it was short I find in the fact that in a poem attributed to Taliesin (certainly of old Welsh origin), the name rhymes with -af (old Welsh -am) of the first person singular of the verb; also with haf, summer (old Welsh ham, Cod. Lichf.), which is now long in conse- quence of the depression of the m, but was originally short (compare Irish sam). As to the meaning of tam, I have not yet satisfied myself; is it the Irish team, able ?
16 E. W. Hvans.
The name Cunadaf or Cuneduf, as the equivalent of Cuna- tam should, for historical reasons, be carefully distinguished from Cunedda, which represents the Cunedag of Nennius. The latter means the good chief, from cun, and dag, good, modern da; compare Irish deagh. The person designated by the name in Nennius is he that is celebrated, later, as Cun- edda Wledig. In the Triads of the Isle of Britain his name is writte Cunedda Wledig (Myv. Il. 10, 68). So also in later copies of the genealogies of Welsh Saints (ib. 34, 41); but in an early-middle copy, where the dd sound is regularly represented by d, it is Kuneda Wledic (ib. 23). In the tran- sition from old to middle Welsh a final g, following a vowel, is dropped. Thus the descent of Cunedda from Cunedag is perfectly regular, and a final f is entirely foreign to it.
Owing to the failure to distinguish between Cunedda and Cunedaf, there has been a troublesome dispute, in which Mr. | | Stephens, author of the Literature of the Kymry, has joined, as to the time in which Cunedda Wledig lived. Tho legend is that he came, with his sons, from a district of the northern Cymry called Manau Guotodin (the Gododin, better Gododd- in, of Aneurin), to North Wales, and expelled the Irish from some of his ancestral possessions in that region. Some writers, accepting the account in Nennius, in the genealogies of Welsh saints and other repositories of Welsh history and tradition, that he was the great-grandfather of Maelgwn Gwyn- edd (the Maglocunus of Gildas) who is known to have lived in the sixth century, naturally assign Cuncdda to the fifth. Others, quoting the testimony of Taliesin in the poem above mentioned, that a chieftain named Cunedaf was his contem- porary and patron, conclude, either that Cunedda should be assigned to the sixth century, or that the poem is spurious. It does not appear to have occurred to the disputants that Cunedda and Cunedaf might be very different names.
The confusion seems to have begun with the late-middle or early-modern scribe whose copy of the poem 1s printed in the Myvyrian (I. 71). For Cunedaf he erroneously writes Cun- eddaf, and in one instance Cunedda, which destroys the
rhyme, thus: Cyn cymun Cunedda Rym a fai biw blith yr haf.
Studies in Cymric Philology. 17
Restore the rhyme by substituting Cunedaf, and for cymun read cymyn, then translate: —
Before the slaying of Cunedaf I had milch cows in the summer.
Another fact that has contributed to the confusion is, that in some translations of an obscure passage of the poem, Cu- nedaf is called the son of Edern, while in the genealogies Cunedda is also called the son of Edern. Such a coincidence could have no great significance in its bearing on the ques- tion, in any case ; it can have none whatever after it has been shown that Cunedda and Cunedaf (elsewhere Cunadaf, as we have seen, ) represent two originals so different as Cunedag and Cunatam. There may have been many Ederns in the period of which we speak ; indeed we read of two, namely, the father of Cunedda Wledig, and that one of the sons or grandsons of the latter from whom the Welsh district of Edeyrniawn is said to have been named ; see Price, in Hanes Cymru. We might, without chronological difficulty, suppose the latter Edern to be the one mentioned in the poem of Taliesin. ~ I would not be understood to suppose that the Cunatam of Taliesin was the same person with the Cunatam of the inscrip- tion.- Indeed it must be conceded that the preservation of the old Celtic genitive in the inscription indicates an antiq- uity far higher than the sixth century; unless we suppose, what is not very probable, that there was a learned class in the sixth century who understood, and still used for special purposes, a language much older than that which was spoken
nd sung in their day. )
II.—- On the so-called Attic Second Declension. By FREDERIC D. ALLEN,
PROFESSOR IN EAST TENNESSEE UNIVERSITY.
THaT those nouns in we which form their genitive in w are but a subdivision of those in oc, genitive ov, could not escape the notice even of the ancient grammarians, who in their classi- fications always had an eye to outward multiplicity more than to inner unity. The Alexandrine Theodosius assumed six and fifty declensions of nouns, of which the words in question form the seventeenth. But while he brings them under a ' separate «army, he says in the main quite correctly of them : “ra eic o¢ Ovdpara peraroovrrec Arrexol cic we kai Ta Tapadfyorra Tay gwynivtwy cic € peraGBaddAovoy, kdivover b€ TavTa KaTa axofsoARy row o° G Xewe, rov Aew,”’ and so on (Bekker’s Anec. III. 984). Hence in our modern system of declensions this class of | nouns has properly been assigned to the second, where in distinction from the common type it is called Attic. How far this designation is in place we shall hereafter inquire.
I have found neither in special Greek grammars nor in the manuals of comparative grammar any satisfactory exposition of the origin of this form of declension. Even the best of the Greek grammars content themselves with little more than setting forth the forms in actual use, without any attempt at explaining their genesis. The most definite statement is that of Kiihner in the new edition of his Ausfihrliche Grammatik, 1869. The substance of his remark is, that these words are formed by contraction, but he adds very little by way of proof or illustration. We shall find that Kiihner’s notion is in the main well-grounded, though to one feature of his theory we shall be forced to take exception.
The first thing which strikes us in attempting the analysis of these words, is that in place of the o of the second declen-. sion stems, we have anw. We have therefore to deal with w-stems. But wstems of the second declension are at first view an anomaly. For the essential difference between the 1st and 2d declensions is one of quantity, uot of quality,
On the so-called Attic Second Declension. 19
of stem-vowel. If the w is original, that is, descended from a pre-Grecian ¢@, then these stems belong properly to the first declension, and the words are but a phonetic variation of the masculines, in ac and nc. Such a variation would not be in- conceivable, and would mainly coincide with the existing form of the words; yet this view cannot for a moment be enter- tained. For, to say nothing of the feminines of the Attic 2d decl., which do not vary from the masculines, the close rela- tionship of the whole class to the 2d decl. is conclusively shown by such duplicate forms as vadc, vedce. Accordingly we shall be prepared to find that these w-stems, as such, are not of high antiquity, but are secondary formations in some way derived from o-stems.
The explanation of this w lies close at hand in a large class of these words, those namely in ewe. Ade, vede are evidently, as Theodosius recognized, secondary forms of dace, vadc. The short o has been lengthened, and the long a has been shortened and attenuated: in other words there has been an interchange of quantity between the two juxtaposed vowels. Such a shifting of quantity is a well-known phenomenon in Greek. It takes place only in hiatus, and always so as to turn a trochee into aniambus. Examples are the Attic genitive vewc for vade, onic ’Arpeisew for Arpeidao, éwe while for Doric doc, ac, Epic joc (written eloc), méXewo for méAnoc, and many others. The hiatus is uniformly the effect of a consonant dropped out; this consonant is either F or7. In compensation for its loss either the foregoing or the following vowel is lengthened ; or what is more probable, the preceding vowel is in every case first lengthened, and this length may then, to facilitate pronunci- ation, be transferred to the following vowel.* The change of a into e which accompanies its shortening before the o-sound is a dissimilation for partial relief of hiatus. Precisely so are we to understand the change of rade into Awe. Kiihner’s explanation of these forms (Ausf. Gram., vol. i., p. 316) is
* Delbriick (Ueber Zws and réws Curtius’ Gram. Studien, vol. ii.) denies this transfer, and thinks that th. ejected semi-vowel operates directly on the subse- quent vuwel. But he fails to explain satisfactorily cases like ixérdo, Ixérew, and vdés (ndvis), ves, where the antecedent vowel is long by nature.
20 F. D. Allen,
singularly perverse. His idea is that the vowels were first — contracted into », and that this » afterwards took, for some reason, an « before itself; thus, rade, Adc, Aedc. What could have led Kiithner to this conclusion I am at a loss to conceive. Surely he could not have thought to explain the accent of MevéAewe etc., on such a supposition. MevédAewe simply follows the accent of the primitive Mevédaoc, just as wéAews that of xédnoc. As for the genitive Mevédew, its accent is in any case irregular (for Meveddov would give us, treated according to K.’s process, MeveAew)) and must be explained by the influence of the nominative npon the other cases. But it is not simply in contract forms (epirdovc, repixdov) that this influence is exerted ; it may undeniably be assumed in forms with shifted quantity as well; in proof of which we need only call to mind the gen. plur. réAewv. The reason of the accentuation MevéAew Theodosius comprehended perfectly : “‘ ca» ¢Sdon xporapotySjvac i) ebScia, waoa nréotc xporapotivera. This very unlucky thought of Kiihner scarcely needs refutation, and I should not have mentioned it, were it not that it emanates from so authoritative a source. The blunder is all the more surprising, since the author has elsewhere in the same volume (pp. 189, 180, 248) spoken with entire correctness of the shifting of quantity (which he calls Vokalverschiebung) in these and other words, of the synizesis frequent in them, and of their accent.
This premised, it will be well to pass in review all the words of the Attic second declension, and bring to light whatever we can respecting their formation. And to obtain as secure a foothold as possible, we shall begin with those which are easiest of analysis, that we may have the benefit of their anal- ogy in treating of the more obscure.
We saw in the examples of shifted quantity cited above that the hiatus was due to an ejected semi-vowel. We might nat- urally infer that the same would prove true of nouns in dog, For that in vad¢ the a and o were together from the begin- ning no one will suppose. In this case we have not to look far for traces of the consonant. The Aeolic form vaioc, found not only in Grammarians but in a Lesbian inscription, shows that it was F. That vaféc furthermore stands for vac-Foc 18
On the so-called Attic Second Declension. 21
rendered very probable by Curtius (Etymol., p: 294) who com- pares the root vec or voc, found in vé-opar and vde-roc. Thus the temple or dwelling-place is thought of as a place where. we may go home, which is not surprising when we reflect that olxoc is derived from a root wk ‘enter.’ If this supposition of Curtius be not true, then the root is af, and yvalw stands for vaFjw just as xaiw for caF-jw. Aadc people, with its com- pounds Mevédaoc, etc., has also dropped a F, as is plain from the form given to the latter in certain Latin inscriptions, where we read Menelavus, Archelavus, and others. Moreover if Savelsberg is right in considering the proper name ‘IdAaog, "IdAewc, 2S @ compound of Aadc, and in translating it “ qui populum coercet,”’ then we have the F attested in Greek as well, for on an Aeginetan vase the name occurs in full, Fcd- \aFoce: Priscian too testifies to a form AaFfoxédfwy. The root is Aaf as the uw in German lew-te indicates. Some find the same root in Bacwrstc, which they explain as “ people-leader,” like herzog.
A third well known word of the same form is the substan- tive dvwyewy hall, really only the neuter of an adjective avuyewc, to which eivyewe, icéyewe and other compounds are an- alogous. Here we have the collateral form dvéyanv. Even yi is but a shortened form of yaia. But yata itself, as the comparison of other languages shows, has lost a F and was once yafja. It is remarkable that the same original stem gav runs through several languages with the double meaning of earth and oz. The Sanscrit stem gavja, identical with ours in form, is an adjective meaning bovine, but the primitive from which it is derived, gé (i. e. gau, nom. gdus), means earth as wellas oz. That the Gothic gavi, stem gauja (Germ. gau), corresponding exactly in form and meaning with the Greek yafja, is a different word, I cannot bring myself to be- lieve, despite the unaltered g.* The history of our word in- side the Greek is at any rate dywyafjov, avwyaoy, dvwyaoy, avayewy.
But the language has more than one way of disposing of
*So Bopp, Vergl. Gram. I., p. 258. In the other meaning the initial consonant has undergone the regular change, hence our cow.
22 F. D. Allen,
the objectionable ending aoc. Take for instance the adjectives deuvaioc, rpipvatoc, etc., derived from pra, that is, most probably, pvaia (Ionic pvéa). The word is of Phoenician origin, but that makes no difference to us. Besides dipvewe, rpiuvewc, we have also, without the «, such forms as éipvwe, rpipvwe, well ac- credited.* Here instead of a transfer of quantity, contraction has taken place. Another example ‘is dzivwc ever-flowing, found in Attic poets instead the Herodotean aeivaoc. The stem of this adjective is de-raf-o, as the verb vaw with its ZHolic form yvaiw testifies. Ndéfw stands for ovaf-w: the real root is snu. ‘Ayfpwe free from old age is contracted from ayhpaoc. The consonant here ejected seems to have been oe, since the word is most naturally referred to the neuter sub- stantive stem ynoac. At any rate the a was short, and so an ayhpews Was impossible.
In these two classes we have traced the Attic forms to stems in ao, and this for the most part to afo. There are however others which point toa stem-form ofo. Such a one is the name of the island Koc with its uncontracted Homeric form Kowe-. It is extremely probable that this nameis identical with the Latin cavus. The root is ku, which means both to be hollow and to swell, significations by no means contradictory, since what is concave on the inside is convex on the outside. The Greek form of the adjective is cofoc, preserved in the forms «doc and koe, cave, hollow; this last by transfer of quantity gives us Kéwe and by contraction Kec. It is an instructive fact that the short form xéc was applied not only to the island, but also to a cave used as a prison in Corinth.f Kédwe is related to Kéoc as Epic opdwyrec to yedworrec. From Kaoc is derived the adjective Kéioc, Koc, Coan ; compare the Homeric gnpmiv dpea-Kyutot.
Similar to this is the word ydaAwe sister-in-law, which Homer has uncontracted, yadéswe. The suspicion which naturally arises, that a F has been lost here as in Kowe, and that yaAdwe stands for yadofoc, is confirmed by the Slavonic zelva; see Curtius, Etym. p. 164. The Greek word contains the suffix -
* Also dipvovs, t. €. dinveos. For all these forms sec Lobeck’s Phrynichus, p. 554 t On these words se: Steph. Byz., 3. v. «3. Compare also Hesychius,.s. vv cos and «60, and Strabo, VIII. 367.
On the so-called Attic Second Declension. 23
Fo; whether the preceding o is, as Curtius thinks, a mere help-vowel or not, is indifferent to us. The Homeric adjec- tive Zéc alive, contracted from the vulgar Zwdc, belongs here. The Hesychian Zovior i) Zovor shows that the earlier form was loFéc. Zor, that is 66Fwor, animal, is derived from this just as ixxcoc from ixwoc. The F in this word belongs, I am inclined to think, to the root, which in Sanscrit has the form g‘tv. The - accounts which I have seen of the origin of Zaw and Zwéc do not however satisfy me, and I am inclined to assume a fuller form of the original root gjav, Greek jaf, ZaF; and from it to explain faFu, 2éFw (fiw) and ZoFdc.*
Another word, which does not appear in the uncontracted form, but of the origin of which there can be no doubt, is the substantive rarpwc uncle. This word wavers between the 2d and 8d declensions. Which of the two forms was the ori- ginal is seen from the Latin patruws, which stands for patrovos just as tuus, swus for tovos, sovos (Curt. Etym. 532, Schleicher Comp. 219). Patravas is therefore the Indogermanic form, derived from patar by the suffix va. The Sanscrit petrivas ‘uncle’ is a derivative of this, and corresponds in form precisely to Greek xarpéioc, that is rarpoF-oc. There is therefore no question that zdrpwe stands for warpofoc, and we may safely assign to pirpwco mother’s brother a similar archetype parpoFfoc, and assume that it too once followed the second declension, although it has almost entirely passed over into the third.
The same change we see beginning in the feminine &dwe threshing-floor, though here the best usage is still overwhelm- ingly in favor of the second declension. A corresponding first declension form of this word is a\w4. In this, Curtius (Etym., p. 524) recognizes, guided by the word dev-pov, (yet more decisive would be the Hesychian gloss ddoud * hry, Kurpwr,) a stem Fadof-a. About drwc he does not speak so definitely, but the analogy of the above mentioned forms leaves no doubt that this stands for FadoF-oc. The root FadoF is that of the Latin volv-o : the o with which the ¢ in GAevpor is | identical, is an inserted help-vowel.
* Since the above was publicly read, I have received the last number of Car-
tius’ Studien (vol. iv., part 1), and am pleased to see that Brugman in his essay “de Gr. linguae productione suppletoria”” has hit upon precisely the same idea.
24 F. D. Alien,
So far our way has been tolerably clear. We have been able to trace a part of the w-stems to primitives in ofo and the rest to primitives in afo, or at least to ajo or aco. Let us now see how far we may be justified in similar conclusions respecting the remaining words of this declension.
Aayws hare has the collateral form daydc, used by Herodotus and even by Sophocles, and generally in the cov." Aayéde is the simplest possible formation. Whether the root ray be identical with the Sanscrit la"gh (Gothic latkan would seem to point to a primitive lag) and the hare be thus termed the “leaper,”’ or with Aay in Aayapds, and the word thus signify the “lank one”’ does not concern us. Here indeed it might seem as if the w in Aaydc were a mere lengthening of the o. But the Homeric daywdée disproves this supposition. This form, which was not confined to the Epic language, but was a favorite one with prose writers from Aristotle on, cannot possibly be the same formation with Aayéc- It contains cer- tainly an amplified stem. That this is Aayo-fo is suggested by the derived adjective Aaygoc, which looks as if it were Aayo-Fjoc. Aaywde 18 tO Aaygoc AS Lwde to yor. Of course we shall derive our Attic A\ayéc from Aaywee rather than from Aaydc.
"ASwe, the name of the well known mountain, is in Homer uncontracted 'A9dwe. Here Etymology leaves us altogether in the lurch ; we can only conjecture a primitive ’ASofoc, of which an additional hint is perhaps furnished by the adjective ’AS¢o¢, Aesch. Agam., 276, the orthography according to Eustathius. Hardly more perspicuous is the proper name Mtywe, which like xarpwe and Gdwe follows the third as well as the second declen- sion. Here we have not even an uncontracted form to guide us; but the adjective Miwioc would favor the conjecturejof a form Muvofoc. The name of the small island Miwa might well be a form of the same.
Returning to our words in ewe; the substantive repivewe stt- pernumerary, ship-passenger, leads us to a stem rep-vaf-o, in which we see the stem of vai-¢ augmented by a suffixo. A more serious variation of stem in the latter part of a com- “* Athenwus,ix.40. = = = |
On the so-called Attic Second Declension. 25
pound word meets us in tho adj. jdexpews well-flavored, having agreealle flesh. The substantive xpéac (Homeric gen. xpedv) is reduced by the aid of kindred tongues to the stem xpeFjac. Our jdixpewe shows however a shorter stem xpcefjo Colder xpaFjo), and that this stem is no imaginary one, but repre- sents in truth the older form of the word, is made evident by the Sanscrit krayja-m, ‘raw flesh,’ Lithuanian krauja-s, ‘blood,’ and Gothic hraiv, ‘dead body,’ the stem of which is hratva, : which stands for hrayja.* Perhaps indeed xpéa neut. plur. © regular in Homer is the plural of a xogov. With this jcvxpewe may be compared some compounds of xpéoc, as adisypewcs, which follow the Att. 2d decl. and show @ stem ypew instead of xpecc. Kiihner assumes that xpéoc stands for xpéFoc, on what evidence Iknow not. _
With regard to the Attic fw, Epic jac, I will here only re- mark that the primitive &foc is amply attested by the Aeolic and Doric forms. We shall speak more particularly of this word below. The Aeschylean adjective xparaiewe stuny is in- teresting in its relation to the well-known anomalous substan- tive Adac stone. Kparairewe leads us, according to the estab- lished analogies, to a nominative Ades, of which the genitive Adov is found in Soph. Oed. Col. 195. Comparing Actc, Acbw etc., we see the stem of the word to be Aafo. The vulgar forms Ada¢ (Adc), acc. Aday (Adv), gen. Adoc, nom. pl. edde ete. are difficult of analysis, but most of them may perhaps best be referred with Kiithner to a stem AaFac.
Of the proper names in -apewc some admit a conjecture as to their formation. The clearest is Turddapewc with its many collateral forms.’ The root is tud (Lat. tund-o) whence first Tuvdapeéc, formed like or:Bapdc, BrAaBepoc. This means of course striker, thruster, and would seem to have been one of the forms of the proper name ; its counterpart Zyndaruws occurs in Latin writers, and the patronymics Tu»dapic, Tuvdapidnc, are to be referred to it. Mostly however longer forms are in use ; Tuvdapnc, Tuvdapews, Tuvddpeoc, perhaps even Turdapeic. Turvddpewc and Tvvdapeoc are evidently identical, and point to a TuvddapeFo¢
* Just as stem haila for halja = Sanscr. kalja.
4
26 F. D. Allen,
or something similar. As Tuvddpewe from Tuvdapédc, so is Bprapewe formed from the adj. Bprapéc. "Aupiapewc, also ’Augedpaoc, Seems to be derived from apa curse, prayer ; that the suffix is Fo can hardly be proved.
The adjective wAéwe full finds its primary form in the Ho- meric wdeioc. Here there is no evidence that f ever was present. The « in its consonantal form 7 has, as in dipvewc, produced the effect of F in lengthening the foregoing vowel. Another adjective is ‘tAewe, poetic iAaog (also sometimes tAdoc). The Lesbians said tAAaog, the double \ having doubtless arisen from assimilation. But one etymology of this word has, so far as I know, been proposed ; ‘that of Gerth in Curtius’ Studien, I. 2.217, who assumes a form is-AaFfo-¢ and refers it to the root ts, ‘wish.’ The word would then signify well-wishing. To this may be objected, not so much the rough breathing in ‘ews, for that is often of secondary origin, but the « in the Aeolic é\Aad = thnS. Then too idapdc cheerful, gay, Gerth himself confesses to be nearly related, and its meaning does not accord well with the notion of wishing. Both adjectives would, I think, explain themselves better from the root ofeA glow, be bright (Curt. Etym., p. 503), and for tAewe we might suppose a formation oFe\ja-Fo-¢ or something similar. But this con- jecture rests on too little evidence to be worth much.° Little plainer is the proper name Iere#e in Homer, doubtless signifying rusher, hastener, from root wer; the form MHeraFo¢ is however presumable. The island-name Kéwe (Ionic Keo¢ ?) may be reasonably referred to a Kafoc; compare the adjective Kijioc, just like Kwioc. That Kéwe is merely a dialectic variation of Kéwe, Kec, I suggest, but do not assert.
The compounds of «épac, excluding those which take the r- form in the genitive, as dicepwe, dxépwroc, follow the Attic 2d decl.; tixepwe, inpixeow ; dxepwe, axeow; these may be plausibly explained from the adjective xepade, which is doubtless xepaFde
* Brugman in the dissertation above cited, p. 119, essays, following a hint of Ahrens (de Dial. 1., p. 284), to bring all these words, with Idcexopai, etc., into connection with salvus, salvere. He will have t\n$c = salve, and then, I suppose, DNdexcoSar rdv Ye6v will be to wish him good health! In this he will scarcely find
many followers. Ahrens was more cautious: it was only by separating fAAa% from the other forms that he proposed to connect it with otAs and salve. ©
On the so-called Attic Second Declension. 27
‘= cervus. The relation of this to xépac with its double stem xepac and xepar, I cannot with eertainty make out. Moreover the accent of these proparoxytone compounds is not quite clear. In this connection may be mentioned yéAus laughter and gpws love, which occasionally show forms of this declen- sion; as however some critics question their authenticity, or at least their antiquity (Bekker rejects them from Homer), they stand under some suspicion.
The proper name Tadws may be considered as a contraction (with change of accent) of Tadads, which itself appears as a proper name. TadaFfes is thought to be the old form. Tvowe whirlwind is also the name of a giant, who has besides the appellations Tugs and uncontracted Tugawy, whence we may fairly suspect that Tugwe is in like manner contracted from Tugadg = Tupafoc. The proper name ‘Idpwc* may be com- pared with the late Attic pews = iepeve;+ both have the stem ispefo = iapafo, namely the stem of iepeds augmented by ano. AS ispdc, iepevc, ‘Idpwe are to each other, so the three names of the same fish, oppoc, cppevc, cppac. dA€we, the name of a certain water-plant, is Attic for Ionic gAovc: one probably stands for ¢AaFoc and the other for gAofoc, and both are doubt- less to be referred to root gdv, in regard to which see Curt. Etym., p. 282. “Aypwe and a few similar compounds of xpéa stand alongside of axpovc etc. The adjective ode safe is plainly contracted from edoc, odoc, cdoc. There are no further indications, but we are naturally led to suspect safoc; cplw would then explain itself as coF-fw. Kadwe cable is Attic for vulgar «doc, and we may suspect that they sustain the same relation to one another as Aaydc, Aaywe; Turdapdc, Tuvddpews, ete. béwe a plant, Téwe the city-name, rad the name, probably of foreign origin, of the peacock, and a few obscure proper names complete the list so far as my knowledge goes.
The result of our inquiries is that the words of the Attic 2d decl. are to be traced, so far as their origin is traceable at all, to forms of the o- declension ending in ao¢ $ and woc; that
#*lapés is Doric for lepés.
+t Chocroboscus ad Theodosium, Bekker, Anec. HI 1197. |
tIn one or two cases nos (from esos or sjos) seems probable as intermediate step between aros or ajos and ews: thus dws from *\ijos, wAcjos, and only re- motely from xdajos.
28 «FD. Allen,
the forms in ewe and ows are derived from these by a shifting of quantity from penult to ultima, and those in we impure by contraction. We saw moreover that wos was uniformly trace- able to an older ofoc, and in like manner aoc, in most instances, to an older aFfoc, in a few to ajoc or aFjoc, in one to asoc. These words fall therefore to all intents and purposes into the same category as the contracts of the second declension. They are only the same fundamental forms differently treated. As from root zAv we get wAdFoc, rAdo¢, zAovc, 80 from root «v we have «éFo¢, koe, Kéc or Kéwe. "Ayfpwe has as good a right to a place among the paradigms of contract forms as apyupoig.
A word or two is all that is necessary on the inflection. The oldest authenticated form of the gen. sing. is seen in the Homeric Weredo for Meracio. Thee drops out after the long just as after the long a of the 1st decl., ’Arpeidao. In Il. 14, 489, the genitive of [nréAewe is according to the vulgar read- ing IInveAgoo; for this Bentley and Bekker read Hnvedéwo, which is probably right. Others of these genitives are cited by Choeroboscus.* The accusative plural, as ves, we shall of course not suspect of having lost a v, but derive directly from vew-ve. The neuter plural avwyew is for avwyea through the in- fluence of the other cases. The acc. sing. should of course properly end in wy; of the form without » we shall speak anon.
The accent of these words is in many points unsettled, the statements of the ancient grammarians being contradictory and the usage of the manuscripts inconsistent. All agrce that the ending ewe allows the tone to recede to the antepe- nult, and that the accentuation of the nominative of bary- tones remains in all the cases: MevéAcy, ay¥jpy, but Mevedrdy, aynpay. Notso however in the Epic genitive; Mnredéwo, ’Av- dpoyéwo. As to nouns in owe the general usage regards them as paroxytoues, though some write “ASowe, yadowc, at least in nom. and acc. Of the oxytones those in ews, it is certain, re- tained the acute even in gen. and dat.; why it is hard to say. Oxytones in we impure, though often accented like those just
* TIept révwv, Bekker, Anec. NII, p. 1223,
On the so-called Attic Second Declension. 29
mentioned ought probably to have the circumflex in the gen- ‘itive and dative; Aaywec, Aays, etc.: some ancient authorities would have us even write Aayéc, copsc in the nominative, in- correctly, however. Tvgéc may perhaps be right. The peris- pomena should retain their circumflex throughout. Here belong the monosyllables except é%¢ (from Zwdc), also rade. The accent of the numerous foreign proper naines in os, as Tapwe, Ivapwe, Nexws, is extremely uncertain.
But our treatise would by no means be complete without some notice of the many points of contact between the Attic second declension and other forms. As to the duplicates Aaywe Aaydc, KaAwe KaAoc, etc. we have thought ourselves justi- fied in assuming the Attic forms to have amplified stems, formed from the common ones by the addition of some suffix as Fo. Dut there are, beside these, interchanges of form be- tween the Attic second and the third declensions. It is now a well established principle that cases of heteroclisis and metaplasm have in general their origin in diversity of stem- formation ; that they are traces of a period in the history of the language when there was great luxuriance in the use of formative elements; a period of extended polymorphism of derivatives, such as every one must acknowledge to have been a leading characteristic of the pre-Hellenic tongue. The plainest proofs of this are to be found in the Homeric poems ; the multiformity of the Epic dialect finds its explanation only in the supposition of a still greater multiformity in earlier times. Thus, for example, the Fpic forms of vice point un- mistakably to a time when three nominatives, vidc, fic, visic, from three different stems, existed, each with full flexion, side by side. The difference of stem may consist in the inter- change of one suffix with another (decpo, deopar), or the addi- tion of suffix to suffix (you, yorFar). Thus, among other cases already considered, we saw the stem «peac alternating with a shorter stem «peo or cpew. Collision of vowel-stems with y-stems is less frequent ; yet a caso is ddwe, which has rarely a gen. Gdwvoc. But by far the most important of these substi- tutes for the Attic second decl. is the form of declension of the word jjpwe. In no less than three of the words which we
80 | F. D. Allen,
classed under the Attic 2d decl. we found the flexion in -woe, -wi,-wa alternating with the other; these were adwc, Mivwe, warpwc. We saw that in rarpwe the 2d decl. was certainly the original form, and it is presumable that the same is true of Gwe and Mivwe. Now the accusative. sing. of fpwe and the words declined like it often suffers contraction : jipw is quite as common in Attic as fowa. Just so the accusative adw, which occurs as well as dAwy, is a contraction of ddwa and is to be referred to the 3d-declension form of the word. Here evidently is the explanation of the numerous accusatives of the Attic 2d decl. without », as Aaya,* Kd, Kéw, “ASw, etc. These accusatives are to be regarded as having passed over into the third declension. Of course it is not necessary to presume in all these cases that a complete inflection like fpwe existed at any time: the analogy of adwe and words really admitting such inflection would favor the extension of these accusative forms to'‘other words, and so Aayd, Ka, etc., might be compared to Attic AnpoodSévny and Herodotean deordrea. |
But not only do words of the Attic 2d decl. thus show third-declension forms, but conversely, some words which from preponderance of usage are referred to the 3d declension still show in good writers forms of the Attic 2d decl. Mérpwe maternal uncle cannot possibly be separated from xrarpws, and though it very seldom exhibits forms of the 2d decl., yet they occur; the acc. pirpwy, and the dative (in Pindar) parpy. Of course we shall class the two words together; it is mere acci- dent that one has retained more fully than the other the earlicr inflection. Even jows, the representative word of its class, shows second decl. forms well authenticated ; the gen. fipw, dat. fipy, acc. fipwy. It is plain that the line is nowhere to be strictly drawn between the Attic second declension and the words like #pws of the 8d decl. Now what is the forma- tion of these third-declension words? The oblique cases fowo: etc. show that a consonant has been dropped from the end of the stem. I see no reason to doubt the conclusion already arrived at by Curtius,t with the word warpwe as a clue;
——.
* Athenaeus (9,400) says that this acc. was written “ ywpis rod v nai xepromwpé- yos”’. This confirms our theory, since the other acc. Aaya» was doubtless oxytone. t Erléuterungen zu meiner Schulgrammatik, p. 50.
¢
On the so-called Attic Second Declenston. 31
namely, that these words have f-stems. On what ground Kiihner rejects this theory I cannot imagine; he does not substitute any other. We see therefore that rarpws of the 2d decl. has a stem zarpofo; of the third, a stem rarpof ; that is, the stem of the second-declension forms differs from that of the third-declension forms by the plus of an o. Further- more it is evident that in the cases xarpws and pfrows the stems marpoFo and pnrpoFo are the original ones, of which zarpof and pnrpof are shortened forms. For that the stems in F are a for- mation existing independently of the longer forms in Fo, we are disinclined to believe, since this would necessitate the as- sumption of a secondary suffix F or v, a thing unheard of. Moreover if there were a. real original stem zarpof it is not easy to see why the nominative should not be rarpous like Bois. What is established for razpws and pirpws may be fairly as- sumed for Mivws, Gws and pws. It is therefore extremely probable that the stems Mivof, ddof, fpoF are in some way shortened from Mivofo, ddo0Fo, hpoFo; that is, that the second. declension forms are here too the original ones, the others being of secondary origin. This result immediately excites a suspicion that the few remaining words declined like jpus may have a like origin. Nor does this suspicion appear to lack confirmation. The remaining words are Tpws, duds, Sus. Of dues slave the sccond-declension form actually exists in Hesiod (Op. et Di. 480, £70) in the uncontracted state dpéos. Choeroboscus cites this, and also Tpéos. Apos is etymologi- cally clear: it is from root dap with fo; dpo-Fo-s. Ows jackal viewed in the light of these other words appears to be identi- cal with the adjective Sods swift, that is 9oFf-os, from root $v. The apparently irregular accent of the genitives Tpwy, duduwy, Swwy explains itself if we view them as from the nomina- tives Tpios, éudos, Sas.
‘We are thus led to the conclusion that jows and all the words declined like it are but mutilated forms of words of the second declension ending in ofo-s. They are as it were estrays from the Attic second declension. Such an abridgoment of a vowel stem is not unheard of in Greek: it has been shown to have taken place in other words, and must hereafter, I think,
32 F. D. Allen,
be assumed to a still greater extent when we come to study carefully the genesis of the numerous classes of consonant- stems in Greek. It occurs also in Latin. I refer to Bopp’s Vergl. Grammattk, III., p. 421, Schleicher’s Compendium, pp. 438 and 453 ; and in particular, Angermaun in Curtius’ Stw- dien, II]. 1, 117. Most frequently. final disappears: stem vucr, Indogermanic nakét: so the suffix ryr = Lat. tdt has arisen from téti. So too stem ypav, originally, according to Curtius, yepaFe, Still others in Latin, ment(¢), sacerdot(¢), etc. More difficulty is at first sight presented by the rejection of an open vowel, because the change in stem involves a material change in inflection. Yet examples even of this are not wanting. With regard to a number of «-stems, it can be shown by the evidence of the cognate tongues that the «has arisen from the suffix xo, originally ka. Examples are dAdrni, dprul, — pipunt 3 also pudad, the more primitive pvdAaxoc being here ac- tually preserved. iAaé and ¢iAacos are as duds and dueos. The feminine form corresponding to this class of nouns is that of the well-known words #,¢, re9d, etc. Curtius,* .tak- ing simply the dialectic forms of these feminincs as data, has assigned to them stems ending in of:, the feminine of ofo. Although other grammarians have dissented from this view, yet the grounds on which Curtius bases his conclusion seem to me jn themselves sufficient, and when we consider, what Curtius himself points out, that to rarpwe, pirpwe, fowe corres- ponding feminines really exist in the proper names Marpd, Mnrpo, Hod, the results separately arrived at in regard to the two classes of words are mutually confirmatory. ‘Hps, there- fore sustains (so far as formation of stem is concerned) the same relation to fpwe which Sovpec does to Sovpos, xopwri¢ to xopwyéc. Of course others of these masculines may form a feminine by changing fo into Fja: so Suph, that is duo-F7a. One more word claims our attention ; we have reserved it for the last, on account of its peculiar difficulty. It is the femi- nine 46; dawn, Attic fwc. In the Epic language as in nearly
Sha
all the dialects the word follws aicé; in declension, and so
* Erlduterungen, loc. cit.
On the so-called Attic Second Declension. 33
belongs unquestionably to the o-stems{ moreover the com- pound éwe-gdpoc leaves no doubt that the « belongs to the stem. The etymology of s#c is as clear as that of any Greek word. The root is us, from which, by suffix as, is derived in Sanscrit the feminine stem wshas “dawn.” This with the Guna amplification gives us ausas, Graeco-italic ausos.* The first s drops out in Greek, leaving stem atog (whence Aeolic abwe) or afoc, from which Doric afc, awe, Ionic wc. Now the Attic éwe distinguishes itself in three particulars from joc: it has a different accent, it has irregularly admitted the rough breathing, and it follows the Attic second declension except in the acc. sing., which is always é. It is I think commonly held by grammarians of the historical school, that the Attic forms of the word are due toa mere blunder, and that the word originally belonged only to the o-declension. But evidence is not wanting that other forms of the stem were known to the Greeks. There are pretty plain indica- tions that an afa, afac, etc., of the first declension was in use among the Boeotians, and a corresponding aia, atac among the Lesbians. On these forms see Ahrens, de Dialectis, I. pp. 121 and 206. It is not impossible that there should have been beside the a stem also an o-stem, afo, from which éwe would come as Aew from Aafo. The prototype of this stem aFo would be ausa or usa, and we find in reality a Sanscrit stem usha beside the ushas mentioned above. I am there- fore inclined to recognize in Greek three stems ; afa, afo, afoc.
Finally, we may remark that this form of declension is not peculiar ‘o the Attic dialect, but has received its name from its being most extensively employed by Attic speakers and writers. To the Aeolic dialects it is, so far as I know, quite foreign. The Dorians employed only a few of the contracted forms, and these chiefly in proper- names; Kae, Tddue, etc. The Epic language adinits a few contracts; ayipwc, ipixepwe, Gw:, oo: of the forms with shifted quantity it is very chary, admitting them only in proper names ; ’AyéAewc (also ’AyéAaoc), "ASdwe, Kowc, etc., and the single appellative yadcws. The
=— a
# Lat. aurora, i. e. ausdsa—Sanser. ushdsd.
5
84 - ‘FD. Allen,
usage of the Elegiac'and Iambic poets does not differ in this respect from the Epic. Herodotus has in proper names both classes of forms; Mivwe, “Awe, Urncikewc ; very rarely in ap- pellatives; xarpwy, X\edy. The Athenians exhibit on the con- trary a predilection for these forms. With reference to the words in ewc = aoc it has been shown that the form in ewe be- longed to the conversational every-day language, that that in aoc on the other hand bore with it a certain dignity and hence was affected by the tragic poets. Aristophanes uses the for- mer invariably in common dialogue, the latter in serious choral strains or in passages in which he travesties the stilted diction of the tragedy.
ADDENDUM.
In stating that xdzpwe is unknown in the uncontracted form, I overlooked the form warpwéc cited along with Aaywer in Arca- dius wepi révwy, lib. V., p. 42 of Barker’s edition. |
I take this opportunity of adding another passing remark to what I have said above (p. 25) ahout Adac stone. The gen. sing. Adoc, dat. Aai, a8 well as all the dual and plural cases, are never resolved, but the accent of the gen. and dat. shows them to have once been trisyllabic. The ancients so understood them (Bekker Anec. III. p. 1219; Arcadius, p. 128) and agree in representing Adaog etc. as the original forms. Ac- cordingly AaFac is probably the stem of these cases. But what of Aaac, Aday? These can come regularly from no consonant, stem. Curtius (Erlaut. p. 66.) derives them from stem AaF and thinks the short a to be merely ‘‘ ezngedrungen zur leich- teren Bildung des Nom. und Acc. Sing.” Is it not more rea- sonable, following the hint afforded by the form Adov, to consider them real 2d decl. forms, phonetic variations of Adoc, Adov? The Dorians made MevéAdc out of MevéAaoc, prob- ably through an intermediate stage MevéAaa¢ ; just so the suc- cessive forms Adoc, Ada, Adc, are possible.
IIT.—Strictures on the Views of August Schleicher respecting the Nature of Language and kindred subjects.
Br WILLIAM D. WHITNEY, PROFESSOR OF SANSKRIT AND COMPARATIVE PHILOLOGY IN YALE COLLEGE.
THE name of Schleicher cannot be uttered by any student of comparative philology of the present generation without respect and admiration. Especially now, when the memory of his early and lamented death is so recent, no one can desire to remember aught of him save his: immense industry and erudition, his ardor in the pursuit of the science to which his life was devoted, his critical acuteness, his liberal and in- dependent spirit, his love of freedom, and the many other excellencies of his character as man and as scholar. His part in the development of the historical study of language was no unimportant one. His manual of Indo-European comparative grammar" has been the convenient and instruct- ive text-book out of which many, in various lands, have drawn a knowledge and love of the subject ; and, being now in process of translation into English, its usefulness among English speakers will soon be largely increased. If 1, then, take the liberty to criticise and combat in this paper some of his fundamental views of language, I do it with no abatement of due respect to him, but because he stands forth as a very conspicuous representative of what I cannot but think a false and hurtful tendency in a part of modern linguistic science; and because his great and deserved reputation as a philologist, @ comparative student of the facts of language and their con- crete relations, gives a dangerous importance to his opinions as a glossologist, or student of the theory and philosophy of language. There is, unfortunately, no necessary connection between eminence in one of these characters and in the other ; many a great comparative philologist has either left untouched the principles and laws underlying the phenomena with which
* Compendium. der vergleichenden Grammatik der Indogermanischen Sprachen. Third edition, Weimar, 1870.
36 W. D. Whitney,
he deals, or has held respecting them views wholly superficial, or even preposterous and absurd. This state of things is one which marks the formative period of a science; there is every reason why it should now come to an end, and why certain fundamental truths, at least, should be accepted as so thor- oughly established that he who denies them shall have no right to be seriously reasoned with, and may be simply passed by as a humorist.
The views which 1 shall here criticise are put forth in two brief pamphlets, both published toward the end of their au- thor’s life. The first appeared in 1863, and is entitled ‘‘ The Darwinian Theory and the Science of Language.’” It is in the form of an “ open letter” to Prof. Hackel, the well-known zoologist, who, by dint of much urging, had persuaded its author to read Darwin on the Origin of Species. The work, once read, Had won Schleicher’s hearty and unqualified ap- proval; it seemed to him to be simply the natural and in- evitable next step forward in zodlogical science—in fact, the analogue of what had heen already done in linguistic science ; he had himself happened to state, at just about the same time, and in nearly equivalent terms, in his book on the German language,t the same conclusions respecting language which Darwin had put forth in attractive form respecting the history of animal life. And he goes on to draw out more fully the parallel between the two sciences, and to make the facts and principles of language demonstrate the truth of Darwinism. Now this parallelism has impressed many minds, and been used once and again, in the way of illustration or of analogical ar- gument, on the one side or the other; but no one, so far as I know, has hitherto attempted to make so much out of it as Professor Schleicher here does—to prove that one species of animals must have descended from another very unlike it, because a modern dialect comes from an exceedingly dissimi- lar ancient one; and that animals of higher structure must be
® Die Darwinsche Theorie und die Sprachwissenschaft. Offenes Sendschrci- ben an Herrn Dr. Ernst Hickel, a. o. Professor der Zoologie und Director des Zoologischen Museums an der Universitit Jena, von Aug. Schleicher. Weimar, 1863. 8vo. pp. 2%.
t Die Deutsche Sprache. Von August Schleicher. Stuttgart, 1860. 8vo. Second improved and augmented edition, 1869.)
On Schleicher’s Views of Language. « 87
developed from those of lower, because complicated tongues are derived from monosyllabic roots; and soon. Such rea- soning, of course, implics something like a real and substan- tial identity between an organized being, an animal or plant, on the one hand, and a language on the other. And this identity Schleicher is logical enough, and bold enough, to as- sume. His fundamental view of language he lays down In these terms (pp. 6, 7): ‘** Languages are natural organisms, which, without being determinable by the will of man, arose, grew, and developed themselves, in accordance with fixed laws, and then again grow old and die out; to them, too, be- longs that succcssion of phenomena which is wont to be termed ‘life.’ (Glottik, the science of language, is accord- ingly a natural science ; its method is on the whole and in general the same with that of the other natural sciences.”
Here, again, we have statements akin with those which are not seldom made by writers on language, only usually in less definite and categorical shape. Schleicher has put forth the theory of the independent and organic life of language in an extreme form, and has drawn from it extreme consequences, as if in order that we may be provoked to give it a thorough examination, and see whether it is a valuable guiding truth, or only a delusive figure of speech. |
Our author does not attempt any proof of his dogma, or even let us see clearly the grounds on which it rests in his own mind. For aught that appears, he regards it as self-evi- dent, or as sufficiently supported by the further expositions which he makes, and which involve it as an clement. This is to be regretted, as imposing additional trouble and per- plexity upon one who would fain test and, if possible, refute the doctrine ; since it may remain to a certain extent doubt- ful whether the considerations which were held to be of the most importance have been after all touched. But Schleicher gives us in his statement two hints which we are justified in taking up and dwelling on, as very probably indicating the grounds of his faith: languages are “not determinable by the will of man,” and their growth and change is “ according to fixed laws.”
38 . W. D. Whitney,
Of these two, the former is evidently the more important. If the voluntary action of men has anything to do with making and changing language, then language is so far not a natural organism, but a human product. And if that action is the only force that makes and changes language, then language is not a natural organism at all, nor its study a natural science. Let us, then, look first and especially into this.
If wé desire to understand the forces which are at work in language, we must be willing to examine their operations in petty and prosaic detail, not content with standing in admir- ing awe before their collective result. That language is a glorious thing, a divine gift, a characteristic of human nature, the sign and instrument of our superiority to the brute, and all that, is unquestionably true, and might be indefinitely en- larged upon, if pertinent to the present inquiry. Of some- what the same character is a Beethoven symphony, a Grecian temple, an Egyptian pyramid. But if I wish to ascertain whether a certain pyramid is a work of human art, or, rather, a stupendous natural crystal, indeterminable by the will of man, aud developed under government of the eternal laws of regular solids, I look to see how it is made up in its parts, and whether it is composed of independent stones, bearing the marks of human tools, and apparently fitted together by human hands; I do not stand at a distance and wonder at its regularity and immensity, contrasting these with the feeble powers of the men whom even a climb to its summit now ex- hausts. That no man can make a language, any more than he can make a pyramid ; that no man, unaided, can make any item of language, any more than he can move or set in place one of the stones of the pyramid; that no man, nor any number or generation of men, can affect the present of a language except as they have its past behind them, any more than they can lay the top-stone of a pyramid without having its lower courses beneath them, is all obvious enough; only, so far as I can judge, these and others like them have been the con- siderations that have led some people to deny human agency in language ;—for the equally reasonable purpose of disprov- ing it in the pyramids, I do not remember to have seen them adduced.
On Schleicher’s Views of Language. 39
Every one acknowledges that languages at the present time, not less than in the earlier stages of linguistic history, are in a state of constant change, or “ growth,” as it is often and properly enough called ; and it ouglit not to be impossible, nor very difficult, to recognize the forces which are effective in producing this growth, and then, by comparing the modes and results of earlier growth, to satisfy one’s self whether any other force or set of forces may or must be assumed as caus- ing the latter. Now the difference which separates any given language, modern or ancient, from its predecessor at any dis- tance in the past, is not a single integral thing, but rather the sum of a great number of particular items; and these items admit of being classified, in order to the better determination of the causes producing them. Let us briefly examine the classes, and see what kind of action they imply.
In the first place, the words of a language come to have a different meaning from that which they had formerly. Of all the modes of change, this is the most insidious and unavoida- ble in its action, and, in languages circumstanced like our own, the most deep-reaching and important in its effects. Every part and particle of every vocabulary is liable to it. And does it come about by an interior force, working in the substance of the spoken word ? Not the least in the world; it is simply a consequence and accompaniment of the growth of men’s knowledge, the change of men’s conceptions and beliefs and institutions. It isas purely extraneous to language as the fact that the name John Smith given to the puling infant 1s borne also by the tottering old man into whom that infant grows. The world-wide change in the value of priest, from the sim- ple ‘older person (elder)’ that it originally designated to its present sense of ‘consecrated (and, in some religions, half- divine) minister of God,’ is wholly subordinate to the change of men’s ideas as to the character of the official to whom it is applied. The words faith and love, and God itself, are, in the meaning we give them, indexes of the education in point of religion and refinement which our part of tle human race has enjoyed. The peculiar American sense of college, quite different from tie English, is due to the peculiar circumstances
40 W. D. Whitney,
which have governed the development of our educational sys- tem ; just as the names robin and blackbird have been applied by us, for the sake of convenience and under the government of old associations, to birds essentially unlike, and only superfi- cially like, those to which they belong in the mother country. That the name of arace, Slave, las become in Germanic speech the name of a bondman, has no other foundation than the historical circumstances which made so many Slaves bond- men of the Germans. The peculiar sacredness of association of home, the pregnant sense conveyed by comfort, have noth- ing to do with the phonetic texture of those vocables them- selves, but are what the habits and feelings of English speakers have endowed them with. Talent is a term borrowed from a parable by men who had read and studied the Bible, and is applied, in accordance with the significance of the parable, to designate the treasure of ability which one possesses, as it were by gift of the Creator. And there are hosts of words like light, and heat, and earth, and sun, which have been, not indeed changed in outward application, but indefinitely wi- dened and deepened in inner and apprehended significance, by the results of men’s study of the universe and its relations.
So is it also with that developed wealth of word and phrase by which intellectual and moral acts, conditions, and relations have come by degrees to be signified. All, as the historical study of language distinctly shows, has becn won through the transfer to an ideal use of words and phrascs which had before designated something physical and sensible. And the trans- fer was made in the usage of individuals and communities who saw a resemblance or analogy between the physical act and the mental, and who were ingenious enough to make an application of material already familiar to new and needed uses. Take as examples one or two of the terms we have just been employing: application is a ‘bending to,’ a physical adaptation of one line or surface to another ; trunsfer means ‘carry across ;’ zwnte/lectual comes, by an intricate series of changes, from a verb signifying ‘pick among.’ What agency other than that of the speakers of language has been at work here? We are ourselves all the time repeating the same pro- cesses in lively phrase. Citrcumvent and get around are but one
On Schleicher’s Views of Language. 41
metaphor, in an older and younger form; comprehend and understand are often familiarly replaced by the nearly equiva- lent modern phrases grasp or get hold of and get to the bottom (or into the heart) of, the figurative use of which is certainly a human product.
Once more, that large and conspicuous class of changes by which certain words are reduced from fulness and independ- ence of meaning to the value of connectives, signs of form and relation, equivalents of grammatical terminations, is of the same origin. We trace, for example, the history of have, from the timo when it signified possession only, to that when it has become in a part of its uses a,mere sign of completed action, an “auxiliary ’’ forming a “ perfect tense’’ (as in I have sat); and we find no trace of any alterative agency save a slowly changing usage, through which the speakers of Eng- lish (as of sundry other modern languages), without being conscious of what they were doing, or working reflectively toward an anticipated end, have converted the one thing into the other. So with of, which, from being in Anglo-Saxon time a full preposition, the same both in form and meaning with off, has now grown into a kind of detached and prefixed geni- tive ending. So, again, with tv, once a preposition governing a verbal noun, now an arbitrary “ sign of the infinitive,’ and even convertible and converted in childish and colloquial phrase into a representative of that verbal form (thus: “ will you do it? no, I don’t want to”). Ihave taken as examples some of the latest cases of this change, because, while not less fairly and fully illustrative than any which might be taken from other periods of linguistic growth, they are more di- rectly intelligible in their process. We say sometimes that such words change themselves in people’s mouths, without the knowledge of their speakers ; but we know, at the same time, that we are only talking figuratively, in the same way in which we might say that a fashion changes itself, or a law, or a popular opinion.
My illustrations of this immense and varied department of linguistic growth are scanty, but I think that they ought to be sufficient for their purpose. If there is in the whole de-
42 W. D. Whitney,
partment anything of a kind essentially different from them, or calling into action other forces than they imply, it has at any rate entirely escaped my quest. Nor am I aware that any student of language has ever attempted to point out any- thing inconsistent with them. Such alterations are all the time going on in our own speech without any question as to whence they proceed ; and the burden of proof evidently rests upon those who claim that in other times they have involved forces of a different character.
A hardly less extensive department is that which includes changes in the forms of words, alteration of their uttered sub- stance—phonetic decay, as it is sometimes loosely called. from the prevailing direction of the movement. I may be briefer in my notice and illustration of this, inasmuch as all authorities are virtually agreed in their attribution of its phenomena to a single prevailing cause—naimely, a disposition to economy of effort in utterance. This disposition, felt in human minds and directing the operations of human organs of speech, it is, which in all languages abbreviates long words, wears off end ings, gets rid of harsh combinations by assimilation, dissimi- lation, omission, insertion, compensation, and all the other figures of phonology, changes the tone of vowels and the place and mode of articulation of consonants, brings new alphabetic sounds into existence and lets old ones go into desuetude— and so on, through the whole vast list of modes of phonetic change. The ways in which the tendency works itself out are indefinitely various, depending upon the variety of human circumstances and human habits, as well as upon preferences and caprices which come up in a community in a manner often strange and unaccountable, though never justly awaken- ing the suspicion of an agency apart from and independ- ent of man. Every word which any one of us has learned to utter he has the power to utter always completely, if he will take the pains; but the same carelessness and haste which bring about the vulgarism pro’able and the colloquialism cajy’n, which make us say bus for omnibus and cab for cabriolet, _ tend to transmute gradually the whole aspect of our speech.
On Schleicher’s Views of Language. 43
When we learn German, we are conscious of a little special effort in pronouncing Anecht ; and the same feeling, in a less conscious form, converted the almost identical enzht of the Anglo-Saxons into our knight. The laws of phonetic muta- tion in speech are, in part the laws of the physical relations of articulate sounds ; but only in part, for else the phonetic his- tory of all related tongues would be essentially the same: the other great and indeterminable factor in the process is the will of men, in the forms of choice, willingness or aversion to articu- lating effort, sense for proportion and euphony, conservative tendency or its opposite, and other the like. And this, again, acts under the influence of all the inducements and motives, external and internal, which direct human action in other respects also. There is just as much and just as little that is arbitrary in the action of men on the form of language as in their action on any other of the elements which go to make up the sum of their culture. |
There is another form of mental inertia which leads to changes in the constitution of words. Something of exertion is involved in the learning and remembering of apparently irregular forms, like went from go, or brought from bring, or worse from bad, or feet from foot. If the great majority of past tenses in English are made by adding ed. of comparatives by er, of plurals by 8, there is economy of mental effort in making these usages universal, and saying goed, bringed, bad- der, foots. These particular alterations, it is true, being in very familiar and frequent words, sound strange and shocking to us; yet their like have borne no insignificant part in the reduction of English to its present shape ; and that their root has been in the mind and will of man admits of no denial or question.
If we thus need to call in the aid of no extra-human agen- cies in order to account for the changes of words, in respect either of meaning or of form, how is it with the production of new words and forms? This ought to be, if anything, the distinctively characteristic part of the growth of language, which should bring to light whatever of mysterious forces there may be involved in it. If names are given to things by
44 W. D. Whitney,
speaking men, then the will of men has at least something to do with the determination of language; if, on the contrary, names are given, always or ever, otherwise than by speaking men, then we ought to be able to catch the power in the act, and to analyze and describe it, and see whether it be like that which is exhibited in the growth of animal organisms. Now, in the first place, every one will have to acknowledge that men do sometimes give names to things. The father names his son, the author his book, the discoverer his isle, or bay, or plant, or animal, the inventor his machine or appli- cation of force, the scientist his stratum or epoch, the meta- physician his generalization—and so on, through an immense series of objects of thought and knowledge. Much of this, to be sure, does not gain universal usc, does not get into the very heart of the popular speech ; but that is perhaps because the essentials of popular speech were produced, not after a different fashion, but a long time ago. Parts of it, as circum- stances determine, do make their way into familiar and every day use, becoming as thoroughly English as any words that “came in with the Conqueror,” or even with his freebooting predecessors, the Angles and Saxons. Again, it must be con fessed that these are for the most part not productions of words wholly new, but adaptations or borrowings of ele- ments already existing in this or in other tongues. Yet this also is a matter of subordinate consequence. To the great majority of the men who are to use them, the words telegraph, dahlia, petroleum, miocene, with all their kith and kin, are pre- cisely the same as if they were forged brand-new out of the nomeuclator’s brain. And in the occasional instances in which such new fabrications are made, they answer the same purpose, and just as well, as the others. It is the easier and the customary way to apply already existing material to new uses in the extension of language; men will sooner assent to and adopt your name if it be of that kind; but their assent and adoption is all that is needed to make language of it, from whatever source it may come. We have already exam- ined, and referred without hesitation to human agency, the process by which appellations for new ideas are chiefly won—
On Schleicher’s Views of Language. 45
namely, by changing and adapting an old name to fit them. What is accomplished otherwise than in this method is in part by taking in consciously words out of other tongues. Thus, certain animals, or plants, or products, or peculiar instruments, or strange institutions, are brought within our sphere of knowledge in connettion with the names which they have borne where they were before at home, and we go on to call them by the same names; our English language coming by such means to include scattered elements from languages all over the globe. Or, what is of much more importance, there is some foreign tongue, to the stores of which customary re- sort is had when anything new requires to be expressed. Such a source of new expression to the English is the Latin, and, in a less degree, the Greek. No one, I believe, perplexes himself as to what may be the recondite organic affinity be- tween English and the classical tongues, whereby, when a new term is wanted, a Latin vocable presents itself, and is seized and put to use. The act of choice involved in the process. the determination by the will of man, is clearand undeniable ; all that the philologist attempts respecting the matter is to set forth the historical causes which have rendered possible and recommended our resort to these subsidiary sources. And when it is considered to what an enormous extent we have drawn upon the classical tongues, the dogma that men’s will has nothing to do with determining language gaina by this - alone a very doubtful aspect. But farther, still another pait of the new names called for in the uses of language is obtained by combining elements already existing In the language it- self, by making new compounds, or new derivatives with the aid of such formative elements, prefixes and suffixes, as the lan- guage hasin living use. In English, to be sure, this method of production is of minor importance, since the habit of com- position and abundant and varied derivation has become deadened with us. But English differs here only in degree from languages like the German, Latin, Greek, and Sanskrit. We do make compounds still, either loose ones, like ink bot- tle, steam-whistle, rail-fence, or closer, like inkstand, steam- boat, railroad ; and it has probably never entered into any
46 W. D. Whitney, one’s mind to doubt that such were actually made by us, and that the parts composing them did not grow together by any inherent force, separate from the determining action of the will of English speakers. And if this is the case with our compounds, it cannot be otherwise with the more abundant and various compounds of the other tongues to which we have referred. If any one be bold enough to maintain the con- trary, he may be challenged to bring forward his proof, and to instance an example of a word of which the constituent members have combined by an internal organic attraction. In this conclusion, however, is involved another, yet more important and far-reaching. On looking back into the his- tory of our family of languages, we find that the combination of independent elements to form new words has been a process of the widest range and most conspicuous consequence. Not only have names been thus made, but grammatical forms also ; the whole structure of inflective speech has had no other ori- gin. Every formative element, whether prefix or suffix, was once an independent vocable, which first entered into compo- sition with another vocable, and then, by a succession of changes of form and of meaning (changes which have been shown above to be due to human action alone), gradually ar- rived at its final shape and office. This can be proved by clear and acceptable evidence respecting so many formative elements, modern and ancient, that the argument by analogy from these to the rest is of a forde which cannot be resisted. The -ful and -less by which we make adjectives, the -ly which forms adverbs, the -d of the past tense in our ‘“‘ regular ’’ verbs, the m of am, the -th or -s of loveth or loves, arc all demon- strably the relics of independent words; and if these (along with many others which might be instanced), then, by fair inference, all the rest. The grammatical apparatus of those languages whose history we best understand is esseutially of the same kind with the -ful of helpful, and to whatever force we attribute the production of the latter we must attribute that of the former also. There are, it is true, left alive a few representatives of the antediluvian period of linguistic science, who hold that endings exuded from roots and themes by
On Schleicher’s Views of Language. 47
some indufinable force, having no analogy with anything that appears in language now-a-days; and such may, without ap- preciable damage to their reputation either for consistency or for insight, maintain the independent organic existence of language ; but all adherents of the prevailing modern school of historical philology, the school in which Schleicher him- self is one of the leading masters, accept an explanation of structural growth which not only admits but demands the will of man as a determining force.
We will give our attention to but one other mode of change in language, namely the loss of words and phlirases, their obsolescence and final disappearance. This doubtless pre- sents analogies with the wasting of tissues in organized bodies. But it really means and is nothing save that communities who have formerly used certain words come to use them more and more rarely, and finally cease to use them altogether. When we look for reasons, we seek them in the grounds of human action, and only there: the thing which this vocable designated has gone out of use and so out of mind, and there has been no farther occasion for its name to appear in men’s mouths; for this other, new expressions have chanced to arise and win acceptance, crowding this out of employment, which is existence ; for yet another, no explanation, perhaps, can be given save the unaccountable, but human, caprices of popular favor and disfavor. Forms are lost, too, by the opera- tion of phonetic decay, which destroys their distinctive signs, and so brings about their abandonment and oblivion; cases and genders, persons and moods, as our language more than others abundantly testifies, can go in this way; but they can go in no other. The same force which makes can un- make also, and nothing else can do it.
We have thus seen, or seemed to see, that words are neither made, nor altered in form or meaning, nor lost, ex- cept by the action of men; whence it would also follow that that congeries of changes which makes up the so-called growth or life of language is produced solely by human action ; and that, since human action depends on human will, languages,
48 W. D. Whitney,
instead of being undeterminable by the will of man, are de- terminable by that will, and by nothing else. And the strangest thing about it all is that I have made no asser- tion respecting matters of detail, and have instanced no case in illustration, which would not probably have been ac- cepted by Professor Schleicher and those who hold with him. So far as I am aware, no believer in language as a natural organism has ever professed or attempted to put his _ finger on this, that, or the other item in language as impos- sible to human agency, and exhibiting the peculiar organic force in action. Schleicher himself, certainly, abundantly admits in detail that which he denies in the totality. All the parts are as we have described them; only the whole is something entirely different. The parts are white, but put them together and they are black; every factor is positive, but the sum is negative! Passing strange indeed it is that the utter illogicalness of such a conclusion escapes these peo- ple’s notice. As we have already seen, that by which a cer- tain dialect differs from its ancestor, nearer or more remote, is not an indivisible whole; it is a mass of particulars, somo of them isolated, others hanging together in classes; and each of these particulars or classes has its own time, place, occa- sion, origin, and effects; their cumulative sum makes up the general result. Now it is easy to throw a group of objects, by distance and perspective, into such apparent shape as shall obscure or conceal their true character and mutual re- lations. Look at a village only a little way off upon the plain, and its houses are fluug together into a mass; trees grow out from their roofs; a cloud rests on the summit of - the church spire; the mountains behind are lower than the house-tops. If you refuse to judge appearances there ex- hibited by those of the similar village in the midst of which you stand, you may arrive at any the most ungrounded and- absurd views respecting them. Soin language: if you insist on standing aloof from the items of linguistic change and massing them together, if you will not estimate the remoter facts by the nearer, you will never attain a true comprehen- sion of them. And this is just what Schleicher has done in
On Schleicher’s Views .of Language. . 49
the essay of which we are treating. He rejects the genuine scientific method, which is to study thoroughly the phe- nomena which fall under immediate observation, with the forces they involve, and to reason cautiously back from these into the obscurer distance, always making due allowance for change of circumstances, but never needlessly postulating a new force. There is not a vestige of scientific character in his fundamental dogma; it is worthy only of the mythologic stage of linguistic study, when men were accustomed to veil plain facts in obscure and fantastic phraseology, and to assume quasi-personal causes for effects which are really due to the secondary workings of obvious and every-day agencies.
If the argument presented above, as to the presence of the human will asa factor in the growth of language, be found well- grounded and acceptable, then the question of the ‘‘ fixed laws”’ alleged to govern that growth is also virtually settled, and does not require detailed discussion. What we call “laws’’ are traceable everywhere, in the action of individuals and of com- munities, in the progress of human culture and human history, as well as in the changes of physical nature; the term is used, to be sure, in more than one sense, as designating generaliza- tions and inferred causations of quite diverse character ; but for that very reason a close examination is necessary in each particular case where the government of law is asserted, that we may avoid the gross, though too common, blunder of con- founding the various orders of law, and identifying their re- sults. An egg goes into the hatching-room and comes out a chicken ; a bale of cotton goes into the factory and comes out — a piece of cloth ; there is a palpable analogy between the two cases so far; and there are, beyond all question, laws in plenty, even physical laws, concerned in producing the latter result, as well as the former ;. but we do not therefore decline to peep inside the factory door, and satisfy ourselves with assuming that the cloth is a purely physical product, and an organism, because the chicken is so. Yet this, in my opinion, is pre- cisely what Schleicher has done. A very little unprejudiced and common-sense research applied to language suffices to show us that the laws under which its so-called life goes on are
50 WW. D. Whitney,
essentially different from those which determine the develop- ment of living organisms, animal or vegetable; they are simply modes of human action. Every law of speech has its foundation and reason in the users of speech—in their mental operations, their capacities, their wants and preferences, their physical structure, their circumstances, natural or historical, and their habits, the accumulated and concreted effects of all the rest. There is not less of linguistic mythology in setting up the government of language by law than the absence in it
of human action as a reason why it should be regarded as an _ organism.
It would be great ¢ cause for rejoicing if this mythologic mode of treating the facts of language were confined to a single scholar, or a single school. But it does, in truth, character- ize no small part of the current linguistic philosophy—even, or especially, in Germany, and among those who most affect profundity. Many an able and acute scholar seems minded to indemnify himself for dry and tedious grubbings among the roots and forms of comparative philology by the most airy ventures in the way of constructing Spanish castles of lin- guistic science.
Langunges, then, far from being natural organisms, are the gradually elaborated products of the application by human beings of means to ends, of the devising of signs by which conceptions may be communicated and the operations of thought carried on. They are a constituent part of the hardly won substance of human civilization. They are neces- sary results of human endowments and dispositions, and also highly characteristic results; yet only results, and not the sole characteristic ones, of man’s peculiar powers. Every human being, if endowed with the ordinary gifts of humanity, is put in possession, as part of his training, of a language, as he is of all the other elements of the civilization into the midst of which he happens to be born, and the acquisition of which makes him a developed man, instead of a mere crude savage, a being little higher than the highest of the other animals. If we are to give language a name which shall bring out its essential character most distinctly and sharply, and even in
On Schleicher’s Views of Language. 517
defiance of those who would make of it an organism, we shall call it an INSTITUTION, one of the institutions that make up human culture. The term,‘ probably enough, offends the prejudices of not a few; yet it is well chosen and correctly applied, and involves not a particle of derogation to the high dignity and infinite importance. of human speech.
The study, moreover, which takes for its object languages, their varieties, structure, and laws of growth, is not a natural science, any more than is the study of civilization at large, or of any of its other constituents, of architecture, of jurispru- dence, of history. Its many and striking analogies with the physical sciences cover a central diversity ; ; its essential method is historical.
Of course, its foundation being withdrawn, Schleicher’s whole argument in support of Darwinism falls to the ground, and there remains merely an interesting, and, if rightly used, instructive analogy between the two classes of facts and phe- nomena compared—one which Lyell (in his Antiquity of Man, chap. xxill.), with a soberness of judgment strangely in con- trast with the over-rash zeal of the German scholar, was con- tent to set forth as an analogy only. Darwinism must stand or fall by its own merits; it cannot be bolstered up by lin- guistic science. .
The second of the two pamphlets which I have undertaken to criticise is entitled “‘On the Importance of Language for the Natural History of Man.’”* It was published a couple of years later than the other, to which it endeavors to fill the office of a defense and support. Some persons, namely, hav- ing raised objections to the unsupported assumption there made, that languages are real concrete organisms, having a material existence, the second essay is intended to supply the lacking demonstration of that doctrine. Let us see how the demonstration is conducted.
The author begins with pointing out that the characteristic mode of activity of any organ—as, for example, of the stomach, the brain, the muscles—is now generally acknowledged to
* Ueber die Bedeutung der Sprache fur die Naturgeschichte des Menschen. Von August Schleicher. Weimar, 1865. 12mo. pp. 29.
52 W. D. Whitney,
depend upon the material constitution of that organ. So the locomotion of different animals, even the peculiar gait of in- dividual men, is conditioned by the structure of their organs of motion. The same is the case with language. This is the “audible symptom of the activity of a complex of material relations in the formation of the brain and of the organs of speech, with their nerves, bones, muscles, etc.”” The material differences of structure on which the differences of language in different individuals depend have never been anatomically demonstrated, and they may even prove forever too subtile for demonstration ; but that does not show that they are not real. What light is to the sun, that audible sound is tu these efficient peculiarities of organization ; it manifests them; and it may, in a philosophical sense, be said to be identical with them. Hence, languages have au independent material exist- ence, and the objections brought against their treatment as such are to be deemed and taken as set aside!
I solemnly affirm that this is, so far as I am able to make it, a faithful abstract of Schleicher’s argument ; and I refer incredulous readers to his text for its verification.
The most hasty examination of it cannot but make clear, in the first place, that the author, whether aware of it or not, has completely shifted his ground. A ngtural organism, which has grown and developed by inherent powers of its own, and under fixed laws, through a succession of ages, is one thing ; a symptom or inanifestation of a structural difference, which, speaking philosophically, may be said to be that differ- ence itself, in the same sense (rather a Pickwickian one, surely) in which light is the sun, is another and a very differ- ent thing: one is a being, the other is a function; one is an actor, or at least an agency, the other is an act or effect. All the inferences, for Darwinism and everything else, which Pro- fessor Schleicher founded on his former doctrine, are virtu- ally abandoned ; you cannot make the history of a function prove the transmutability of animal and vegetable species. The only feature, so far as I can discover, which the two doc- trines have in common is their denial of the agency of the human will: voluntary action is ruled out, on the one hand,
On Schleicher’s Views of Language. 53
because language is an organism, growing and developing by its own internal forces; on the other hand, because it is the necessary effect of real physical peculiarities of structure. This, then, is the point to which our attention has still to be particularly directed.
We have first to notice that it is not the uttered and audi- ble part or side of speech alone that Schleicher has in view. He does not intend simply that, constituted as we are, we must produce the articulated sounds, the alphabetic elements, which we actually produce, and no others. For this by itself would never lead to unity of speech in a community or race. Out of our alphabet alone, without importing a click, or a guttural, or atone, from other tongues, we might build up a language which should be as unlike our own as any that is spoken upon the face of the earth. No; his doctrine, as evinced by the whole course of his reasonings, is plainly this : the reason why I, for example, say hat, instead of hut, or cha- peau, or causia, or any other of the thousand words which peo- ple in various parts of the world use or have used to designate their head-coverings, is that my brain and my organs of speech arc so constituted and connected that Aat is to me the natural and necessary sign of this particular conception—and so with all the other signs that make up my language. Truly a most astounding doctrine! There are, I believe, few writers on language who would have the hardihood to maintain it. Hardly one would fail to acknowledge that, whatever natural internal connection there may have been in the initial stage of language between sound and sense, there is, at least, none now ; that the English-speaking child learns to call a hata hat, and could have learned to call it a Aut or chapeau—as, indeed, he often does, earlier or later; which of the names he acquires betg a matter of entire indifference to him until he has acquired one, and become so accustomed to it that it seems to him the “ natural’? name for his tile, and he can only by an effort change his habit and come to call it by any other name. Or, generalizing this—for what is true of this one sign is true of every other of which our language is com- posed—while each human being has the capacity of speech,
54 W. D. Whitney,
none is directed by nature to speak any one language rather than any other; the infant, of whatever race, acquires the lan- guage of those who are about him, or sometimes more than one, and could have acquired any other equally well; but the older he grows, the more the language he has acquired becomes to him that habit which is justly called a ‘ second nature,” and the harder it is for him to lay it aside for an- other, or add another to it. These are, it appears to me, clear and undeniable truths; there is neither mystery nor doubt about them; and their importance is so fundamental that he who overlooks or denies them cannot fail to make shipwreck of his whole linguistic philosophy.
Our view of the acquisition of language is not in the least at variance with modern scientific theories of cerebral struc- ture and action. There may be in the physical constitution of my brain something that makes me say hat; there may be atomic equivalents and atomic connections determining every item of my speech and all its combinations and uses; but it is a secondary or acquired something, a peculiarity effected by external causes, not inherent and self-determining. It is analogous with all the knowledge, the memories, the prefer- ences, the habits, the special aptitudes, which my experience and opportunities, working on a general and specific basis of capacities, have produced in me. That I choose to wear a hat at all, that I prefer one of a certain size and color, that I take my hat off when I meet a friend, that I remember the hats I have worn and where and when I got them, that I know how many I possess at this moment and where they are —all this depends, if you will, on infinitesimal peculiarities in the present structure of my brain; and it is all of the same kind with my capacity and habit of using the word hat. This is a trivial example; but it is not less instructive and decisive of the points involved than the most dignified one that could have been selected.
Again, our view does not make against the theory of the transmission to a certain degree of the effects of culture in the forin of higher inherited capacity. Among a certain number of persons born into such circumstances that they acquire
On Schletcher’s Views of Language. 55
English as their “ mother-tongue,”’ one may possess by de- scent a genius upon which even English, with all its force and beauty, imposes a laming constraint; while, on the other hand, and much more probably, there will be others whose meaner powers would be more in harmony with some lower form of speech, as Chinese or Malay. So it is everywhere ; if men were divided and languaged according to the kinship of their endowments, the present boundaries of races would be entirely broken up, and every community on earth would become a Babel. As things are, every man learns that lan- guage which circumstances place within his reach, whatever it may be, and works out and exhibits his higher or lower endowment inside of it; in his management and use of it. Even the humblest language that exists is so far beyond the capacity of even tle ablest human being to produce unaided, that its acquisition raises him to a plane of power indefi- nitely higher than he could ever have attained if left to grow up speechless. All that he can have reason to regret is that circumstances should not have been still more favorable to him, and enabled him to work out the whole force which it was in him to develop. And what is thus true of language is true of culture in general, in its other elements not less than in the linguistic.
Professor Schleicher has noticed, or has had his attention called to, the objection to his theory of language which is in- volved in the power to learn other languages than one’s mother-tongue; and he endeavors to set it aside—after the following fashion. First, pushing further a comparison al- ready made, he says that a man can also learn to go on all fours, or to walk on his hands, while nevertheless no one can doubt that we have a natural gait as men, conditioned by our bodily structure. But it must be evident at a glance that this comparison, at any rate, does not run on all fours. To make it other than helplessly lame, we ought to see that a human being if brought up by quadrupeds would move natu- rally on hands and feet together ; if by birds, would fly ; if by fishes, would swim ; in each case, without ever feeling a dis- position to walk erect upon his feet. For he who has never
56 W. D. Whitney,
learned any language but English, of whatever parentage he may really be, is undistinguishable from an Englishman, and never exhibits the slightest tendency. to relapse into the ances- tral dialect. But Professor Schleicher goes on to argue the matter upon other grounds. Again ignoring the question as to how a person obtains his * mother-tongue”’ at the outset, he raises a doubt as to whether any one ever really acquires in a complete manner any other language ; and, granting even that that be done, he suggests, as the very simple explanation, that such a one becomes in fact a different man from what he was ; another constitution of brain and organs of speech is substituted for, or added to, his natural one. Further, he continues, even if (which is not to be conceded) a person be- comes thus at the same time an Englishman, a Frenchman, and a German, it is still to be observed that these are related languages—in a certain sense, species of the same genus. But it is not at all to be credited that the same man can be master at ouce of wholly diverse tongues, like German and Chinese, or Arabic and Hottentot, any more than that he can walk easily and comfortably both on two feet and on all fours. Now it is an easy way of disposing of an adverse argument to discredit the facts on which it is founded; but we are con- vinced that what Schleicher refuses to believe is an undenia- ble truth: children of European parents do learn, where cir- cumstances favor it, those outlandish tongues along with their own, as readily and surely as those of the most nearly related European nations; they do not perceive or feel the differ- ence between a related and a non-related tongue; that is dis- coverable only by a process of reflection and learned compari- son of which no young child is capable. Instances of persons learning at once languages like German and Chinese are merely less frequent than the others, and for the simple rea- son that circumstances do not so often bring them about. When one has once schooled his thoughts to one form of ex- pression, it is true, the difficulty of acquiring a second will be partly proportioned to the resemblance or diversity between the latter and the former: but in this there is nothing strange or peculiar, nor does it in the least favor Schleicher’s theory.
On Schleicher’s Views of Language. 5T
One might just as well say of a person who has mastered a musical instrument, as the flute, that he makes it his mode of musical expression because the minute constitution of his brain and of his blowing and fingering apparatus render it a necessity to him; that he never acquires an equal mastery over any other instrument, or, if he does so, it is only in vir- tue of his becoming so far another being; that he may at the utmost become able to play kindred instruments, like the clarionet and bassoon ; but that the violin and the piano are entirely beyond his reach — proceeding then to argue that the musical notes of the flutist, as they reflect and represent pe- culiarities of his organism otherwise unmanifested, are them- selves material existences; and that the development of modern flute melody from the first rude tones of the ancient pipes exhibits the essential characteristics of organic life, and proves the truth of the Darwinian theory! I say it in all se- - Tlousness, such an argument would be precisely as good as that which Professor Schleicher has constructed, and which is one of the most striking examples I have ever seen of the way in which a man of high merit and worthy achievement in one department of a subject can in another deny the most funda- mental principles, be blind to the plainest truths, and employ a mode of reasoning in which there is neither logic nor com- mon sense. |
The subsidiary statements and reasonings of these two pamphlets partake fully of the unsoundness of their main ar- gument. Thus, in the immediate sequel of what we have just been considering, the author declares that speech is the sole exclusive characteristic of man, and that any given anthro- poid ape who should be able to speak would be called by us a man, [and a brother,] however unlike a human being he might be in other respects. As to this last assertion, it is so easy to speculate where the test of fact can never be applied, that I will not take the trouble to contradict it, although my own conviction is strongly against it, and I cannot but doubt whether Schleicher himself would have proved equal to fra- ternizing with his fellow-man if the case had been realized. But certainly, speech is so far from being man’s sole distinc-
58 W. D. Whitney,
tive quality that it is not a quality at all, in our author’s sense ; it is only a possession. The capacity of forming and acquir- ing speech is a quality, and one among the many which con- stitute the higher endowment of man ; but let the child of the most gifted family of the most highly cultivated race grow up untaught, in solitude, and he will no more employ a language than he will build a temple, paint a picture, or construct a locomotive. Not all the boasted development of the race will enable a single individual, if thrown upon his own unaided resources, to speak ; because speech, like the other elements of civilization, does not go down by inheritance, but by the process of teaching and learning.
It is not true, then, as our author argues later, that lin- guistic science leads us to the conclusion that man developed out of lower forms of animal life because language has been of slow devclopment, and without language man would not be man. The rise of language had nothing to do with the growth of man out of an apish stock, but only with his rise outof savagery and barbarism. Its non-acquisition by a given individual cuts off, not his human nature, physical or intel- lectual, but his human culture; it puts him back into a con- dition from which he would at once begin to advauce by slow degrees to that of a speaking man, as his remote ancestors had already done before him. Man was man before the de- velopment of speech began ; he did not become man through and by means of it.
In connection with this, Schleicher brings forward again a dogma which he has repeatedly laid down elsewhere with -great positiveness and confidence: namely, that ‘it is abgo- lutely impossible to carry back all languages to one and the same original language ;”’ that there must necessarily have been at least as many original languages as there are now- existing families of language. This is entirely wrong, and even a complete non sequitur from the premises which he him- self accepts. For he holds, with the historical philologists in
general, that all languages had the same morphological form at the outset; that is to say, that they began in the condition of bare roots, designating the simplest and most obvious
On NSchleicher’s Views of Language. 59
physical conceptions. He doubtless holds, also (I do not find a specific statement upon the subject, but it is an obvious and necessary inference from his expressed views), as others do, that it is not possible to point ont with certainty the precise roots and conceptions with which the different families of langnage began ; they are too much disguised and overlaid by the changes and additions of later linguistic growth to admit of being distinctly traced. Where, then, is the impos- sibility that the same roots should have served as basis of de- velopment to more than one family of languages? The question of probability we may discuss in any given instance as much as we please, but the assumption of impossibility is ruled out by the very nature of the case. To make this as- sumption, as Schleicher does, on the mere ground of the great unlikeness between the developed families, is quite illogical : for if languages starting even with the same completely de- veloped structure can come to be as unlike as are English, Welsh, and Hindi, for example, there is absolutely no amount or degree of dissimilarity which might not arise between tongues which had in common only their first rude elements. This seems a truth so incontestable that its denial is one of the strangest points in Schleicher’s linguistic créed, one that be- trays most tellingly the character of that creed, as made up of prejudices rather than of cautious and well-founded deduc- tions.
If there is another point in the creed entitled to contest the palm of unreasonableness with this one, it is our author’s view of language as an infallible test of race, and the only firm basis for a classification of mankind. ‘ How inconstaut,”’ he exclaims, “are the form of skull and other so-called distinc- tions of race! Language, on the other hand, is always a com- pletely constant characteristic.” And he goes on to point out that a German (we will say instead, an Englishman) may well enough chance to rival in wooliness and prognathism the _ most pronounced negro-head, while nevertheless he will never speak naturally (von Hause aus) a negro language. To ex- hibit the preposterousness of this claim, we have only to invert it, and say that it may well enough happen now and then
60 . W. D. Whitney,
that a person of African blood should rival in complexion, hair and Caucasian cast of features a descendant of purest Puritan stock or of the first families of Virginia, while nevertheless he will never, never speak as his mother-tongue the English lan- guage. I fancy that some of us have chanced upon facts not en- tirely consistent with that statement. Ishould like to see some adherent of Schleicher’s opinions going around in our Ameri- can community with an English grammar and dictionary, de- termining by the evidence of language to what race its various constituents belong. It would not be difficult in almost any American village to set up before him for examination a row of human beings who should show unmistakable traces of African, Milesian, Scottish, and German, as well as English, descent; and yet every mother’s son of them should speak English as his mother-tongue, and should not know a word of any other language under the sun. Aud our author’s imag- ined wooly and prognathic German, or any other German, would only need to be brought up from infancy in an African kraal, in order to speak African as naturally (von Kraale aus) as the child whose ancestors had lived for ten thousand years on the Aarrov. It is nothing short of gross judicial blindness that can make one overlook the infinite number of facts like these which the history of languages presents, and their bear- ing, and set up the mere accident, as we may fairly call it, of one’s mother tongue as the sole and sufficient test of race. One’s “ mother-tongue ”’ is determined simply by one’s teach- ers; and it is only because one’s teachers are usually one’s parents and a community akin in race with them that lan- guage becomes an indication, a prima-facie evidence, of race. On the broad scale, it is to a considerable extent a trustworthy evidence ; and its contributions to ethnology are of extraordi- nary and unsurpassed value ; but its degree of force in any individual case is to be measured by the degree of probability, determinable in part on other grounds than linguistic, that the given comuunity is one of descent and not of agglomeration or mixture. :
Another fallacy of Professor Schleicher’s—one, however, which stands in a more logical connection with his general
On Schletcher’s Views of Language. 61
theory of language—is his assumption that the primary differ- ences of language are geographical: that is to say, that forms of speech grew up in the outset resembling one another in the ratio of their proximity and of the accordance of the surrounding physical conditions. There is no good reason for holding any such doctrine ; it falls to the ground, at any rate, with the doctrine of the necessary physical origin of lan- guage, and is not unavoidably involved even in that. Not physical causes, but historical, determine language : dwellers in the same plain speak different tongues, without the slight- est tendency toward unification, save as the effect of com- munication and mixture; dwellers in the plain and on the mountains, in the interior and by the sea, in icy, temperate, and torrid climes, speak the same or nearly related speech, because it comes down to them by tradition through the sepa- rated representatives of a single community. Schleicher says farther that “in the later life of language, among men who live under essentially similar conditions, the language also changes itself uniformly, or spontaneously and in correspond- ing manner in all individuals who speak that language :’”’ thus ignoring the fact that only individual action tends to diversify language, and only communication to keep it uniform, and once more explaining as the result of physical forces phe- nomena which are in truth ascribable to human action, and to that alone.
In drawing his second pamphlet to a close, our author re- fers again to a very peculiar theory of his, more fully set forth elsewhere (in the introductory part of his Deutsche Sprache), that language-making and historical activity necessarily be- long to different and successive periods in the life of a race or nation, the former absorbing the whole national force while it is in progress, and rendering the latter impossible. A com- munity lies perdu while it is developing its speech (not learn- ing to talk simply, but working the language up to its highest point of synthetic structure), and then steps confidently forward to play its part in the drama of general history. This is so palpable a fancy, and a fancy only, that we nced lose no time over its confutation; we may simply notice that it in-
62 W. D. Whitney,
volves a most peculiar conception of language-making, since this really goes on as_long as the race lives, and cannot be shown to exhaust more nervous force in synthesis than in analysis ; a most peculiar conception of history, as if there were no history without record and publicity ; and a most: pe- culiar understanding of the circumstances which by their concurrence operate to bring a race forward into couspicu- ousness, or to make it take a part in those interworkings whose result is the higher civilization of the more gifted and favored races.
Finally, Professor Schleicher winds up with a bit of theory in pure natural history, which docs not precisely concern us as philologists, but yet is too characteristic to pass over, and which I accordingly give in his own words: “It is in the highest degree probable that not all organisms which entcred upon the road toward becoming man have worked their way up to the formation of language. <A part of them were left behind in their development, did not enter upon the second stage of development, but fell under a law of retrogression, and, as is the case with all such deteriorations, of gradual decay. The remains of these beings, who continued speechless, dete- riorated, and did not arrive at the condition of becoming hu- man, lie before us in the anthropoid apes’! This looks like Darwinism reversed: the apes do not so much represent a con- dition out of which man has arisen as that into which crea- tures that might have been men have fallen, through simple neglect of learning to talk! If we accept the doctrine, we cannot but be impressed with the grandeur of the work in which we, as a Philological Association, are bearing our hum- ble part. By encouraging and promoting, to the extent of our associated vapacity, the maintenance and progress of lan- guage, we perhaps contribute to preserve our own remote pos- terity and the whole human race from sinking to the condi- tion of the gorilla and the chimpanzee !
These peculiar and indefensible views of Schleicher appear more or less in all his later works which have occasion to deal with general questions of language. Thus for example, in the introductory part of his Deutsche Sprache (already more than
On Schleicher’s Views of Language. 63
once referred to), they make so much of a figure as to render that work, interesting and suggestive as it is, a most unsafe one to put in the hands of persons not qualified to use it in an independent critical spirit. But in the two pamphlets which we have been considering, they are presented almost pure and simple; there is hardly room beside them for the acute- ness of the comparative philologist to appear ; while we are, of course, able to pick out here and there a remark ora paragraph which sustains the reputation of the author, yet, as a whole, the essays are utterly unworthy of him, and can only be read with pain and regret by those who admire him and respect his memory. From the beginning to the end, in foundation and superstructure, they are unsound, illogical, and untrue, and must hurt the cause of science just so far as they are read ‘and accepted. I had suppposed that, in the bare and over- strained quality of their errors, they would carry everywhere their own refutation with them; but facts show that this is not so; there are still incautious sciolists by whom every error that has a great name attached to it is liable to be re- ceived as pure truth, and who are even especially attracted by good hearty paradoxes. These two papers have been trans- lated into French as the first and inaugural fascicle of a ** Phi- lological Collection,” or international series of important essays in philology ; and even so sound and careful a philolo- gist as M. Bréal has been misled into giving the inauspicious beginuing an implied sanction by letting his name appear alone upon the title-page, as author of the Introduction.* And the former of the two has been done into English and pub- lished in London by a Dr. Bikkers, who in his preface lauds it to the skies, as containing Cwith the sole exception of the dogma of the necessary diversity of primitive languages, which he rejects) only such doctrines as are to be taken for
* Collection Philologique. Recueil de Travaux originaux ou traduits relatifs a la Philologie et a U Histoire Littéraire avec un avant propos de M. Michel Breal. Pre- mer Fascicule. La théorie de Darwin. —De Uimportance du Langage pour U Histoire naturelle de Homme, par A. Schleicher. Paris, 1868. 8vo. pp. vi.31. M. Bréal’s preface is of but a page or two, and in it he indicates—thouyh, in my opinion, in a manner muc' less distinct and decided than the casc demandcd—his at least partial non-acceptance of Schleicher’s views.
c
64 | W. D. Whitney,
the established truths, the ‘* axioms,” of modern linguistic science (axioms they indeed are, in one respect: namely, that they are incapable of demonstration).* It was the falling in by chance with Dr. Bikkers’s version, a few weeks since, in @ library where it could only do unmixed harm, that led me to draw out and present these strictures.—| Views which might seem to be self-refuting require to be elaborately argued down when they are in danger of winning currency and ac- ceptance ; especially if they have to do, like these, with prin- ciples of fundamental importance. And reverence for the name and works of a truly great man should not lead us to cover up or treat with indulgence his errors, when they are sought to be propagated under the shield of his reputation, and tend, if accepted, to cast the science of language back into a chaos as deep as that from which it has lately begun to emerge.
* Darwinism tested by the Science of Lanquage. ‘Translated from the German of Professor August Schleicher, with preface and additional notes, by Dr. Alex- ander V. W. Bikkers. London, 1869. 12mo. pp. 70.
+ I had given the substance of them before a local society several years ago, on
the first appearance of the second essay, but had no intention of making them more publicly.
IV.—On English Vowel Quantity in the Thirteenth Century and in the Nineteenth.
By JAMES HADLEY, PROFESSOR OF GREEK IN YALE COLLEGE.
It is a well-known fact that the Modern Greek has lost the system of vowel quantity which belonged to the ancient lan- guage: xopun hatr and xopn village are prouounced alike by the Modern Greek; nor is it otherwise with Avcec thou wilt louse and Adac a loosing. In like manner the Romanic languages have lost the system of vowel quantity which belonged to the ancient Latin. Thus the Italian and Spanish mano, rosa, French main, rose are alike in vowel sound with Italian and Spanish vano, prosa, French vain, prose, though the former come from Latin mdnus, résa, and the latter from vdnus, prosa. It is an interesting question whether our own lan- guage has had in this respect the same experience as the Modern Greek and the Romanic tongues; whether we have wholly lost those distinctions of vowel quantity which un- doubtedly belonged to the Anglo-Saxon. |
It is sometimes said that we have no proper distinction of long and short quantity in English: all our vowels are alike in quantity; they are all equally susceptible of prolongation and abridgment; or rather, any vowel may be so prolonged as to excced the ordinary quantity of any other: thus fl/ may be so prolonged in utterance as to take more time than is ordinarily given to feed or file. But this must have been the case also with Greek and Latin vowels: their absolute time was not fixed, but variable. One speaker must have talked faster or slower than another ; the same speaker must have talked faster or slower at one time than at another ; even on the same occasion:one sentence or clause must have been ut. tered faster or slower than another ; and thus it could hardly fail to happen that a short vowel would sometimes have a longer sound than the ordinary or average quantity of a long vowel. But the average quantity of a short vowel was less than the average for a long one; or, with the same general
66 J. Hadley,
rapidity of utterance, the quantity of long vowels exceeded the quantity of short. I say simply, the quantity of long vowels eaceeded, was greater than, the quantity of short; not that the former was just double the latter. It may be pre- sumed that in ordinary spoken utterance there was not any so exact relation ; with the same tem, the long might equal 14 of the short, or 18 of the short, or 14 of the short, as well as just 2; and probably the average actual ratio was rather less than 2 to 1; the average long would occupy rather less than twice the time of the average short. The fixing of 2 to 1 as the precise numerical relation was probably the work of rhythmopeeia, or of rhythmopoia and melopeia together. When longs and shorts were combined in rhythmic composi- tion, and especially when a musical accompaniment was added, the longs and shorts must have a definite and fixed ratio to each other; and the. ratio of two to one was the most simple, obvious and convenient.
Now as regards our English syllables, it is certain that we have nearly or quite lost the feeling of length by position. By this 1 mean that we do not recognize a short vowel fol- lowed by two consonants as having any distinct relation to a long vowel followed by one consonant. We can perceive, of course, that there is more sound in fist than in fit, and more sound in fight than in fit; but we do not recognize any special relation of quantity between fst and fight. But in reference to vowel sounds, it should seem that our case is not essentially different from that of the ancient Greeks and Romans. There are certain sounds which, with the same tempo, the same gen- eral rapidity of utterance, we recognize as occupying more time than others ; we thus recognize the former as long, the Jatter as short. It is true, the English short vowels differ somewhat in quality of soynd,—that is, in position of tlie speech organs,—from the long vowels which most nearly re- semble them, the short being a little more open than the cor- responding long: there is a more open sound in fil than in feel, in full than in fool. in fell than in fai/,etc. But this does not affect the relation of quantity. It is clear that we have long vowel sounds in file and foul, in feel and fool, in facl and
On English Vowel Quantity. 67
foal, and in fall; aud that we have short vowel sounds in fill and full, in fell, in doll, in dally, in dull. Pronounce the two series in succession ; on the one hand, file, foul, feel, fool, fatl, foal, fall; on the other, fill, full, fell, doll, dally, dull:—the difference of quantity is manifest and unmistakable.
If then we have long and short vowels clearly distinguished from each other, it becomes a matter of interest to inquire whether the distinction is recent or ancient; whether, and how far, the vowels now sounded long or sounded short were sounded long or short respectively in early periods of the lan- guage. The question here raised is not whether our vowels have kept the same sounds, but whether they have kept the same quantities. It matters not that long @ has passed from
-its original sound (of ah) to that heard in fame ; long e, from the sound in prey to that in key; long 7, from the sound in pique to that in pike ; long 0, from the sound in bone to that in boon; long u, from the sound in prude to that in proud. These changes have taken place to a very great extent, but they do not affect the question now before us: the old long vowel remains long still. So if the short vowels i, %, é, (in full, full, fell,) have become a little more open; if on the other hand @ has generally become closer (as in dally), approaching the sound of & (in dell); if the old 6 has disappeared alto- gether, passing into a sound more open that of a in fall, and often described as its corresponding short (thus fall, folly) ; if %, and sometimes other short vowels, have sunk into the ob- scure and undistinguished sound heard in dull, done ;—these changes do not concern us here, as the old short vowel still remains short.
In the inquiry whether, and how far, we can trace back the present long and short quantities of English vowels, we find our best guide and help in a single (and singular) production of early English literature. I refer to the Ormulum, so called from the name of its author, the monk Orm, or Ormin. who wrote in the eastern part of England, some time in the thir- teenth century, or fully six hundred years ago. A scries of metrical homilies on the successive daily lessons of the church service, its interest is philological much more than literary.
63 | J. Hadley,
In the only manuscript from which it is known to us,—not improbably the only manuscript of it that was ever written,— we find on the part of the writer a careful and systematic re- gard to vowel quantity. He has a spelling of his own, to which he adheres with much consistency ; and in this spelling the most peculiar and conspicuous feature is the doubling of every consonaut which follows a short vowel. There was a tendency to this in the general English orthography even at that early period ; what is remarkable in this author is that he consciously carried out this tendency as a uniform and uni- versal rule. Thus he writes it with a double ¢, 7f with a dou- ble f, hundred with a double n and a double d, lasteth with a double s and a double sign for th. So much is he attached to this spelling, in spite of its strange and whimsical appearance, that in a preliminary address (seemingly very little needed) to the future copyist, he insists upon a careful conformity to it: Annd tatt he loke wel tatt he An bocstaff write twi33¢ss, E33wher ther itt uppo thiss boc Iss writenn o thatt wise. Loke he well thatt het write swa, Forr he ne ma33 nohht elless
Onn Ennglissh writenn rihht te word Thatt wité he wel to sothe.
Which may be modernized thus, in the same measure, all but the unaccented ending of the even lines:
And that he look [full] well that he A letter write twice [over],
Wherever it upon this book Is written on that wise.
Look he well that he write it so For he (ne) may not else
In English write aright the word, That wit he well to sooth.
Thus then the spelling of the Ormulum enables us to say, in the case of every vowel followed by a consonant in the same word, whether the author pronounced it—or, at least supposed that: he pronounced it—long or short. In order now to arrive at some general conclusions as to the persist- ence of vowel quantity in English, I have looked through the
On English Vowel Quantity. 69
vocabulary of the Ormulum, as presented with much fulness and clearness in the second volume of White’s edition (Oxford, 1852), and have noted the words which survive in the Eng- lish of to-day, including some few which, though lost out of the common language, are still generally known from their use in literary works of the modern period. I propose to state the general results of this comparison between the quantities of words as indicated in the Ormulum, and the quantities of the same words as heard in modern English. It should be said that the vocabulary of the Ormulum is not large. The work is nearly as long as the Odyssey ; but the number of _ different words used in it is much smaller than in the Greek poem. There is a great sameness—a wearisome sameness — in the contents of the book: the leading facts, principles, and precepts of the Christian system are repeated over and over again with little variety of expression. Still there are words enough to give a fair idea of the relation between the vowel quantities of Saxon English six hundred years ago and those of Saxon English at the present time. I say “ Sazon Eng- lish,” because few Latin words (aside from those already taken into the Anglo-Saxon), and fewer French words, are to be found in the Ormulum. Its language is as purely Teutonic as the modern German.
The most general statement suggested by the extended com- parison of which I have been speaking is this: that in the great majority of cases the vowels which had a long sound six hundred years ago are long now; those which had a short sound then are short now. And if the exceptions are pretty numerous, —if in a good many cases the long sound of the Ormulum corresponds to a modern short, or the short sound of the Ormulum to a modern long,— most of these excep- tional cases depend on a few obvious conditions, on clearly marked euphonic influences and tendencies, so that cases of capricious variation, of variation without apparent principle, are comparatively few. It will probably be most instructive, if we consider first (and indeed chiefly) these euphonic con- ditions and tendencies which have led to differences of quan- tity between the language of six hundred vears ago and that of to-day.
70 J. Hadley,
In the first place, then, let us look at the ‘cases where the loss of a consonant sound has occasioned the lengthening of a _ short vowel before it; asin alms, O. allmess, buy, O. biggenn, and the like. Such changes can hardly be considered as vio- lations of the old system of quantities. If there is here an alteration, an increase, of vowel quantity, it only takes the place of a consonant quantity withdrawn. It is simply that the time before occupied by a vowel and consonant is now oc- cupied by the vowel prolonged. This is sometimes described as an absorption of the consonant by the vowcl, sometimes as a vocalization of the consonant. Neither of these expressions gives a distinct idea of the nature of the change. Of course, in every instance of the kind there must have been a time of transition, when the consonant was beginning to be omitted, when the very speakers who omitted it were perfectly aware of its existence, and perhaps generally pronounced it, but oc- casionally let it drop with a lengthening of the preceding vowel. Now this consciousness of a consonant with a claim to be pronounced is an important element in the phenomenon. The speaker who does not really pronounce it, does not feel that he can omit it altogether; he does not feel that he is altogether omitting it. To his own feeling he gives ita kind of recognition. He perhaps brings the organs of speech into some sort of approach toward the position required for pronouncing the consonant, so that the preceding vowel passes into a sound more or less modified, which does duty for the consonant. If this modification continues to be made, then the resulting long vowel-sound will not be a mere simple prolongation of the preceding short, but something different, perhaps a diphthong. Yet it may very well happen that in this guasi-pronunciation of the consonant, the approach made by the organs to the position for that consonant will grow more and more slight, and the sound produced will differ less and less from a mere continuation of the preceding vowel; until] finally — and perhaps very soon — it comes to be just that and nothing else, and the consonant is replaced, as its claim for utterance is felt to be satisfied, by a simple addition of quan-
tity to the preceding vowel. But whatever may be thought
On English Vowel Quantity. T1
_as to tho rationale of the process, it is one of which we find numerous instances in comparing the Semi-Saxon of the Or- mulum with mcdern English. Thus where /, followed by an- other consonant, has been suppressed in utterance, though still retained in writing, the short vowel before it has become long in
O. allmess (alms), A. S. d/messe, Lat. eleemosyna.
callf (calf), A. S. cealf.
follc (folk, people), A. S. folc.
hallf (half, behalf), A. S. healf.
ille (each, every, Sc. ilk), A. S. dle, ele, ylc. sallfe (salve, ointment), A. S. sealf.
sallme (psalin), A. S. sealm, Lat. psalmus.
In should (O. shollde), would (O. wollde and wolde), we have a short vowel sound; but we may see from the ou that the vowel was first lengthened (show’d, woud, with ow as in youth) ; though afterward it became short again, by a new and independent change, similar to that by which good and stood have received their present short pronunciation. The same change has occurred also in could, which never had an t actually sounded. The Ormulun, like the Anglo-Saxon, has cathe, in which the th became d, and the vowel was after- ward shortened. As people were accustomed to write a silent l in should, would, and regarded could asa word of similar character, they put a silent 7? into that also. There would have been more propriety in the insertion of a silent »; for this letter belongs to the root as secn in can. It is a feature of the Anglo-Saxon in its earliest known form that it drops n before th or s, and lengthens the preceding vowel: as gés (goose) fur gans, s6dh (sooth) for sunth (which means being, existing, and is identical with Lat. -sens, in praesens, -sentis); and so ciéidhe for cunthe (could.)
Again, where g has been suppressed in utterance, the short vowel before it has become long in O. biggenn (to buy), abiggenn (to aby, pay for), A. S. bycgan, Abycgan.
leqgenn (to lay), A. S. leeyan. seyyenn (to say), A. S. secgan.
But in most cases of this kind the g appears in the Ormulum
12 J. Hadley,
softened into the consonant y-sonnd (3), which after the short vowel is written double (33): thus O. da33 (day), pl. dazhess, da33ess, A. S. diz, pl. dagas. dri33¢ (dry), A. S. dryge. e33lenn (to ail), A. S. eglan. e33therr (either), A. S. agdher. fa3zerr (fair), A. 8. fager. fle33! (flail), Germ. flegel, Lat. flagellum. ge33nenn (to gain), ga3henn (gain), O. N. gagn (advantage). gez3ulike (aptly, cf. ungainly), O. N. gegn (apt, clever), A. S. ungdgne (of no effect ). , la33 (lay), from lin (to lie), A. S. lég, from licgan. le33 (impv. lay), le33de (laid), from leqgenn, A. S lege, lagde, from lecyan. ma33 (he may), A. S. mdg. ma33 (may, maid), A. S. meg (femina, virgo). ma33denn (maiden), A. S. mégden. na33/enn (to nail), A. 8. ndglian. re33n (rain), A. S. regn. 8€33 (impo. say), se33'h (saith), se33¢e (said), from seggenn, A. S. sege, segdh, sidqde, from secgan. inn-se33/ess (seals), A. S. sigel, insegel. twe33¢nn (twain), A. S. twegen. twi33ess (twice), also twizess, A. S. twiga (i 2). thri33¢8s (thrice), also thrizess, A. S. thriga (14). wajzzn (waygon, wain), A. S. wign, wen. we332 (way), awe33, awe33¢ (away), A. S. weg, dweg. In 633therr,ma33 (maid ),and perhaps in tw233e88 and thri33e82, a long vowel of the Anglo-Saxon is found shortened (in the last two, however, not uniformly) in the Ormulum. This shortening may perhaps be explained as the consequence of an effort to make the feeble 3 more fully audible. The vowel may have been passed over lightly in order that a greater force of utterance might be brought to bear on the weak con- sonant following it, so as give this a distinct enunciation. lt would be perfectly natural, too, that the speakers of the language should become at length weary of this effort required for the weak consonant; and that they should then allow the consonant to be replaced by a mere continuance of the pre- eeding vowel, which would thus recover its primitive long quantity. | In the word master (O. ma33stre, A. S. mdgestre, mdgstre, Lat. magister) we du not lengthen the vowel: here the Scot- tish maister shows the truer (that 1s, the more analogical) pro-
N
On English Vowel Quantity. (63)
nunciation. In saith and said, the lengthened vowel sound which once belonged to them is still indicated by the az with which they are written.
The 33 of the Ormulum is not always to be traced'to an original g. In some instances it seems to have arisen from the diphthong et in the Old Norse, the language of the so- called Danes who came as invaders and settlers into eastern England: the vocabulary of the Ormulum shows evident marks of a Norse influence. Thus the plural pronouns ¢the33 (they), the33re (their), the3m (them) are not to be explained from A. S. thé, théra, thdm, but from O. N. their, theirra, theim ; — re338enn (to raise), not from A. S. résian, but from O. N. reisa ; — he3lenn (to hail, salute), not from A. S. Adi, but from the corresponding O. N. hezll (sanus, salvus), which, like E. hail, was often used in salutations. So a33 (aye, always, ever) is perhaps to be explained from O. N. ai, et, ey, which corres- pond to A. 8S. dwa; while a33 (may) may be a mere com- pound of ne and a33. The genitive Ke338eress, usually Kase- ress, from Kasere (Cesar, Emperor), might be accounted for in the same way; but for the 33 in 5e33sannz (bezants, coined _ in Byzantium), and in the proper name £33n0c (Enoch), we have no explanation to offer. |
Yet again, where a consonant 4 has been suppressed in ut- terance, the short vowel before it has become long in many words. I say “a consonant 4,” for the Anglo-Saxon h, where it stands at the end of a syllable or is followed by ¢ or th (dh), must be regarded as a true consonant. Thus in
O. bohhte (bought), from biggenn (to buy), A. S. bohte, from bycgan. brihht (bright), A. S. beorht, byrht, bryht. brohhte (brought), from bringenn (to bring), A. S. brohte, from bringan. cnihhtess (servants, soldiers, cf. knight), A. S. entht. dohhterr (daughter), A. S. déhtor. drohh (drew), also droh, from drazhenn (to draw), A. S. drég, from dragan. druhhthe (drought), A. S. drugadh. duhhti3 (virtuous, cf. doughty), A. S. dyhtig. ehhte (eight), ehhtennde (eighth), A. S. eahta, eahtodha, O. N. dttundi. Sehh, fe, (revenue, money, cf. fee,) A. S. feoh, fed. Sihhtenn (to fight), A. S. feohtan. Suihht (flight), A. S. flyht. bi-kahht (caught), also bikaechedd (catched) ; of doubtful origin. lihht (light, toe A. S. ledhe (likt).
T4 J. Hadley,
O. lthht (light, lux), A. S. ledht (ight). mahht, mthht (might), mihhte (he might), A. S.. meaht, mht, mihte. nahht, nihht (night), A. S. neaht, niht. nohht (nought, not), A. S. ndht, ndwiht, from ne and éwiht. ohht (aught), A. S. aht, contracted from dwiht. plihht (danger, state, cf. plight), A. S. pliht. rthht (right), A. S. riht. sahh (saw), from seon, sen (to see), A. S. seah, from seén ; but see p. 93. sthhthe (sight, appearance), innsihht (insight, knowledge), A. S. gesihdh. sohhte (sought), from sekenn (to seek), A. S. sdhte, from sécan. tahhte (taught), from techenn (to teach), A. S. tehte, from teecan. thohh (though), A. S. thedh. thohhte (he thought), from thennkenn (to think), A. 8. thohte, from thencan. thuhhte (seemed, cf. methought), from thinnkenn (to seem), A. S. thuhte, from thyncan. wehhte (weight), A. S. wiht, gewiht, from wegan (to weigh). wihht (being, person, cf. wight), A. S. wiht. wrihhte (maker, worker, cf. wright}, A. S. wyrhta. wrohkhte (wrought), from wirrkenn (to work), A. S. worhte, wrohte, from wyrann.
In some of these words (dohkhterr, drohh, likht, lihht, nohht, ohht, sohhte, tahhie, thohh), we find a shortening of the Anglo- Saxon long vowel, similar to that just noticed in e33therr, ma33, etc., and explainable in the manner then proposed ; though the combination of consonants (At), which in most of them follows the vowel, may have had something to do with its change of quantity.
Lastly, in a number of words which in the Ormulum have the consonant w repeated, showing that the vowel before it was then sounded short, this consonant is lost in English (or, at least, has no consonant power), and the vowel sound is long. Such are
O. chewowenn (to chew), A. S. cedwan. clawwess (hoofs, cf. claw), A. S. cla, dawu, pl. clawe. cnewwe (knee), pl. cnewwess, cnes, A. 8. cneow, cned. deewwenn (to bedew), from daw (dew), A. S. dedwian, from dedw. Sowwerr, fowwre (four), fowwerrti3 (forty), A. 8. fedwer, fedwertig. strawwenn (to strow, straw), A. S. stredwian, stredwian. throwwinnge (throe, suffermg), A. S. thréwwng. trewwess (trees), also treos, tres, sing. treo, A. S. treow, t 6. trowwe (true), trowwenn (to trow), trowwthe (truth), A. S. trcdéve, trcé6 can,
tredwdh
In most of these words we see an Anglo-Saxon long vowel shortened before the weak w, as we have already seen it before hand3. In nowwharr (nowhere), the shortening — which may
On English Vowel Quantity. 15
be compared to that of no in English nothing — really takes place before A, the true order of the sounds being that repre- sented in the Anglo-Saxon orthography, nédhwer. The words owwtherr, nowwtherr, (A.S. dwdher, ndwdher,) have the sense of either, neither ; but these last connect themselves with A.S. egdher. Theform nowwt (cattle, Sc. nowt) is to be explained from O. N. naut, while English neat corresponds to A. S. nedt. In these words, where the ww precedes a consonant, its sound can hardly have differed very much from that of the vowel w. Indeed the Latin aw is represented. by aww in clawwstremann (cloister-man, monk) from Lat. claustrum,%nd in the proper name Sannt Awwstin (Saint Austin, Augustine), though the Emperor (Kaserr-king) Augustus appears as Augusstuss.
We have now reviewed all the cases where a consonant, which in the time of the Ormulum was heard after a short vowel, is lost in modern English ; and we have seen that in all but a very few (should, would, saith, said, master, them, not), the pre- ceding vowel is long in our present pronunciation: even for these few, there is evidence that the most have been pro- nounced with long vowel sound, though that in more recent times has become short.
The next important point we have to consider is the effect of a weak r on the preceding short vowel. By weak r I mean to designate that peculiar sound of the letter which it has as- sumed in our present English, wherever it stands before a consonant or at the end of a word, as in far, farm, for, form, fur, firm. This is evidently weaker than the sound of r in farrow, forest, borough, merit, spirit, etc. According to Mr. Ellis, Ben Jonson, in his English Grammar of 1640, is the earliest writer who gives any sign of having recognized this distinction between a stronger and weaker r in the pronuncia- tion of our language. Perhaps the phenomenon itself, the weakening of r where it is final or followed by a consonant, may not be much older than that time. In the dialectic pro- nunciation of the Irish, which has in many points preserved the older English sounds, it has not yet established itself. This weak r is most easily produced after the neutral vowel,
76 J. Hadley,
so called, which is heard as short in cub, cud, and as long in curb, curd. Except after the sounds of ak and au (as in far, for), this w sound is always heard before a weak r: thus it comes in, as a brief yet perceptible element, after the proper vowel in fire, flour, fear, four, fare. When the preceding vowel was a short # or e, this has been overpowered by and merged in the following w sound. Thus fir and her are not distinguished in pronunciation from fur and Hur. The short u itself becomes long when this consonant follows it; as we see in comparing burgh with burrow, where the first has a long sound bef®re weak r, the second a short one before common r. The point with which we are now concerned is this, that the weak r, whether heard at the end of a word or before another consonant, is always preceded by a long vowel sound ; if the preceding vowel sound was originally short, it has become long. It is altogether improbable that in the time of the Ormulum r in this position had begun to assume its present weak sound ; but it had begun to affect the quantity of a preceding vowel. We find quite a number of words in which a short vowel of the Anglo-Saxon, standing before r in this position, had be- come long in the Ormulum, as in recent English. O. erd (place, region, Sc. airt), A. S. eard.
ern (earn, eagle), A. S. earn.
bern (children, Sc. bairns), also barrness, A. S. bearn.
bernenn (to burn), A. S. beornan.
birde (lineage, cf. birth), A. S. gebyrd (birth, lineage).
bord (board, table), A. 'S. bord.
cherl (young man, cf. churl), A. S. ceorl.
corn (corn), A. S. corn.
eorless (earls), A. S. eorl.
eorthe, erthe (earth), A. S. eordhe.
forth (forth), also forrth, A. 8. fordh.
kirrke-gerd (church-yard), A. S. geard.
hird (company, family, cf. herd), A. S. heord.
hirde (herd, shepherd), A. S. hirde.
hord (hoard), A. 8S. hord.
leornenn, lernenn (to learn), but lerrnde (learned), A. 8. leornian.
skarn (scorn), O. Fr. escorne.
stirne (stern, fierce), A. S. styrne.
swerd (sword), A. S. sweord.
word (word), A. 8. word.
On English Vowel Quantily. tT
O. 3eornenn, 3ernenn (to yearn), but zerrnde (yearned), A. S. geornian, from georn (desirous, O. 3eorne, 3erne).
It is not unlikely that some of these words, though length- ened in the Ormulum, may have retained their primitive short quantity in the prevailing English, until by the weakening of the r at a much later time they became long. It is certain that in most cases where an original short vowel has been lengthened before r, the change is not so old as the Ormulum. Thus in almost every instance where we have the sound of ah or au before a weak r, the word, if found in the Ormulum, shows a short vowel. Here belong
O. arrt (art), arrn (they are), A. S. eart, O. N. eru (sunt). arrctoss (north, cf. arctic), Gr. and Lat. arctos (bear, north). arrke (ark), A. S. eurc, Lat. arca. arrmess (arms, brachia), A. S. earm. barrli3 (barley), A. S. bere (Sc. dear). berrme (barm, leaven), A. S. beorma. berrne (barn), A. S. bere-drn, berern, bern.
Jeorr, ferr (far), A. S. feor. forr (for, prep. and conj.), A. S. for. JSorrme (former), A. S. forma. harrd (hard), A. S. heard. heorrte, herrte (heart), A. S. heorte. herrberrzhe (lodging, cf. harbor), A. S. hereberga. herrenenn (to hearken), A. S. hyrcnian. herrfessttid (harvest-time), A. S. hdrfest. horrs (horse), A. 8. hors. karrte (cart, chariot), A. S. crdt. marrch (month of March), Lat. and A. S. Martius. marrtirrdom (martyrdom), A. S. martyrdom, Lat. martyr. merrke (mark), A. S. mearc. norrth (north), A. S. nordh. orr (or), from oththr, otherr, A. S. édher. patriarrke (patriarch), Lat. patriarcha. scorrcnedd (scorched), O. ¥r. escorcher (to excoriate). sharrp (sharp), A. S. scearp. shorrt (short), A. S. sceort. sperrd (closed, barred, cf. spar), unnsperrenn (to unclose), A. S. sparrnian. starrc (firm, cf. stark), A. S. stearc. steorrne, sterrne (star), A. S. steorra, O. N. stiarna. thorrness (thorns), A. S. thorn. |" thweorrt, thwerrt (with ut; throughout, cf. thwart), A. S. thweorh. warrm (warm), A. S. wearm. werre (worse, Sc. waur), A. S. weor (evil). werrpenn (to cast, cf. to warp), A. S. weorpan. wharrfenn (to turn, cf. wharf), A. S. hweorfan. jerrde (rod, cf. yard), A. S. geard (virga).
7 J.. Hadley,
Somewhat less numerous are the cases in which the short vowel of the Ormulum corresponds to any other vowel sound than those of a4 and au Cfar and for) in modern English.
QO. barrness (children, Sc. bairns), also bern, A. S. bearn.
to-bresstenn (to burst, Sc. brust), A. S. berstan.
currsenn (to curse), A. S. cursian.
darr (dare), durrste (durst), A. S. dear, dorste.
Jerrs (verse), A. S. fers, Lat. versus.
Sirrst (first), A. S. fyrst.
Jirrthrenn (to assist, cf. to further), A. S. fyrdherian.
forrtherr (further), A. 8. furdhor.
girrdell (girdle), A. 8. gyrdel.
hirrtenn (to hurt), Dutch and M. H. Germ. hurten (to dash against). irre (ire), A. S. yrre.
kirrke (church), A. S. cyrice, Gr. xvptaxh. kirrtell (kirtle), A. S. cyrtel. mirrthrenn (to murder), A. S. myrdhrian. myrrha, myrra, myrre (myrrh), Lat. myrrha.
serrfenn (to serve), O. Fr. servir, Lat. servire.
skerrenn (to scare), O. N. skirra (to drive away).
, thirrst (thirst), A. S. thurst. thridde (third), thritti3 (thirty), A. 8. thridda, thrittig or thritig. turrnenn (to turn), A. S. tyrnan.
turrtle (turtle-dove), A. 8. turtle, Lat. turtur.
warr (aware), A. S. wdr. weorrc, werre (work), wirrkenn (to work), A. 8. weorc, wyrcan. werrse (worse), werrst (worst), wirrsenn (worsening), A. S. wyrsa, wyrst, wyr-
sian.
wurrm (worm), A. S. wyrm, weorm.
wurrth (worth, adj.), wurrthshipe, wurrshipe (worship), A. 8. weordh, weordhscipe. wurrthenn (to become, be, cf. woe worth the day), A. 8. weordhan.
But r is not the only consonant which has had this effect of lengthening the vowel before it. We find it produced also by 1, a liquid and a lingual like the r. Before / at the end of a word or followed by another consonant, a vowel originally short has often become long. Cases of this kind, in which the Ormulum still retains the short vowel, are the following :
O. all (all), allswa, allse (also), allmasst (almost), A. S. eal, ealswé, ealmeest. allderrmann (chief, ruler, cf. alderman), A. S. ealdorman.
allterr (altar), Lat. altare.
bulltedd (bred, bread from bolted flour), O. Fr. bulter, Wuter, M. H. Germ. biuteln.
fallenn (to fall, fallen), A. S. feallan, feallen.
fallse (false), A. S. fals, Lat. falsus.
galle (gall), A. S. gealla.
On English Vowel Quantity. 79
O. hallp (holp), holipenn (holpen), from hellpenn, A. S. healp, holpen, from helpan. halite (halt, lame), A. S. healt. pall (cloth, cf. pall), A. S. pall, pell, Lat. pallium. sallt (salt), A. 8. sealt. shulldre (shoulders), A. S. sculdre, pi. of sculdor. stall (stall), A. S. steal. walless (walls), grunndwall (ground-wall, foundation), A. 8. weal.
This lengthening of a vowel before / had already com- menced in the time of the Ormulum; and indeed in most of the instances found in the Ormulum, where a vowel origin- ally short is followed by the combination /d, the vowel appears as long in the Ormulum itself. Thus in
O. ald (old), but elldre (elder, older) and allderrmann, A. S. eald, yldre. bald (bold), beoldenn, beldenn (to embolden), A. 8. beald, bealdian, byldan. child (child), but chilldre (children), A. S. cild, cildru.
Jaldess (sheep-folds), A. S. gefeald.
Jedd (field), A. 8. feld.
gold (gold), gildene (golden), A. S. gold, gylden.
haldenn (to hold, holden), A. S. healdan, healden.
kald (cold), A. 8. ceald.
milde (mild), but millce (mildness, mercy), A. S. milde, milds or miltse saldenn (they sold), from sellenn (to sell), A. S. sealdon, from sellan. shildenn (to shield), A. §S. scildan.
talde (he told), from tellenn (to tell), A. S. tealde, from tellan. weldenn (to govern, cf. to wield), A. S. wealdan.
wilde (wild), A. S. wild.
jeldenn (to yield), A. 8. gieldan, gyldan.
Indeed, the Ormulum sometimes lengthens a short vowel before 7, where the modern English has it short. Thus in wel, also welle (well, A. S. well and wella, fons); and in wel, also well (well, A. S. wel, bene). The variation of quantity, which the Ormulum shows in the last of these words, is seen, continued to the present day, in Scottish weel, compared with English well. Further, before /d a short vowel is lengthened in the following:
O. cwaldenn (they quelled, killed), from cwellenn, A. S. cwealdon, from cwellan. dwalde (he dwelt), from dwellenn, A. S. dwealde, from dwellan (to hinder, delay). elde (old age, ef. eld), A. S. yldcdd, eld. oferrgildedd (gilded over), A. 8. ofergylded. seldenn (seldom), A. S. seldan. In scaldess (minstrels, scalds, from O. N. skdld-r, poet), and heold, held (he held, A. S. hedld), the Ormulum preserves -
an original long sound, which has become short in English.
80 | J. Hadley,
But the first of these is variously pronounced, as scdlds and as scaulds.
Before the liquid m followed by the mute 5, as before the similar combination /d, a short vowel is sometimes length- ened in the Ormulum. Thus in five words, three of which have a long vowel in English, while two preserve the earlier
short:
O. camb (comb), A. S.camb. . climbenn (to climb), A. S. climban. wambe (womb, belly), A. 8. wamb, womb. dumb (dumb), A. 8. dumb. lamb (lamb), but pl. lammbre, A. §. lamb, pl. lambru.
In like manner, before the liquid # followed by the mute d, a short vowel has hecome long in very many words. In &- hinndenn (behind, A. S. behindan), and hinnderrling (degene- rate, retrograde in character, A. S. hinderling), connected with EK. hinder, the Ormulum still retains the short sound; as it does before nt in funnt (font, A.S. font, cf. E. fount), munnt (mount, A. 8. munt), sannt (saint, A.S. sanct), where the vowel has become long in English. But the instances are far more numerous in which a vowel before nd is already lengthened in the Ormulum. Thus in
O. bindenn (to bind), bundenn (bound), A. S. bindan, bunden.
blind (blind), blendenn (to blind), A. S. blind, Wlendan.
Jindenn (to find), fundenn (found), A. S. findan, funden.
grindenn (to grind), A. 8. grindan.
grund (ground), but grunndwall (foundation), A. S. grund, grundweal.
hund (hound), A. S. hund.
kinde (nature, kind, kindred), A. S. gecynd.
minde (mind, memory ), but minndiznesse (memory), A. S. gemynd.
sund (sound, integer), A. S. sund.
-windenn (to wind), in attwindenn (to escape), wundenn ( wound), but winnde- clut and windeclut (winding-clout, swaddling cloth); A. 8. windan, wunden.
wunde (wound, vulnus), A. S. wund.
In the preterit singular, the Ormulum has band, fand, wand, like the Anglo-Saxon, but with short a lengthened; these forms, however, are not represented in English, where the vowel of bound, found, wound, comes from the plural forms, bundenn, fundenn, etc., A. S. bundon, fundon, etc. The change of vowel quantity before nd (as before /d and mb) is carried further in the Ormulum than in the modern English, being extended to a number of wordgin which it failed to establish
On English Vowel Quantity. 81
itself, so that the original short vowel is heard in their present pronunciation. Thus in
O. band (band), A. S. bend. ende (end, vulgar eend), A. S. ende. hand (hand ), oftener hannd, but with added -e always hande, A. S. hand. land (land), A. S. land. sand (sand), A. S. sand. sendenn (to send), but sennde (he sent), A. S. sendan, sende. shendenn (to shend, disgrace), A. S. scendan. strande (strand, bank), A. 8. strand. sunderr-run (private communing, cf. sunder, asunder), A. S. sundor. wand (rod, wand), O. N. vand-r, vénd-r, Goth. vandus. wendenn (to wend, turn, go), but wennde (he turned, went), A. S. wendan, wende.
The word freond, frend (friend, A.S. freénd) does not be- long to this series; it came with long vowel quantity from the Anglo-Saxon into the Ormulum, and passed thus into the older English, as we see from the spelling with te, which it has in common with the opposite, but strictly analogous, fiend (OQ. fend, A.S. feénd). |
It is a very curious fact that a lengthening of the vowel before ng, similar to that before nd, is frequent in the Ormu- lum, although unknown to modern English, in which the vowel before ng is always short. Examples are
O. gang (journey, cf. gangway), but ganngenn (to go), A. S. gang, gangan. genge (company, cf. gang), A. S. genge. king (king), A. 8. cyning, cyng. lang (long), bilenge (belonging to), A. S. lang, gelenge ; but lannge (long, diu), lenngre (longer), A. S. lange, lengra, leng. langenn (to long after), forrlangedd (desirous), A. 8. langian. mang, amang and amanng (among), A. S. dmang, gemang. * — ringenn (to ring), A. S. hringan. singenn (to sing), sungenn (they sung), sang (song), A. §. singan, sungon, sang. springenn (to spring ), sprang (sprang ), sprungenn (sprung ), offspring (offspring ), A. S. springan, sprang, sprungen, ofspring. stingenn (to sting), stungenn (stung), A. S. stirgan, stungen. strang (strong ), strengenn (to strengthen), but strenncthe (strength), A. S. strang, gestrangian, strengdhu. swingenn (to scourge, cf. to swing, swinge), A. S. swingan.
* The Norse grammarians recognize in that language a similar lengthening of primitive short vowels (a, o, u, 1), wben followed by ng (or nk ; also of a, o, u, when followed by Jf, lg, lk, lm, lp, ls): thus ldng-r (long), springa (to spring ), t@aga (tongue), veeng-r or veng-r (wing). It seems, however, to be questionable, whether, or how fur, this change belonged to the old language. See Heyne, Kurze Gramm. der aligerm. Dive 2d ed., p. 82.
82 J. Hadley,
O. thing (thing), A. S. thing.
bithrungenn (oppressed, cf. throng, i. e. press of people), A. S. bithrungen, from thringan, gethrang; but O. threnngdenn (they thronged), A. S. threngdon, from threngan.
thwang (thong), A. S. thwang.
tunge (tongue), A. S. tunge.
wengess (wings), O. N. veng-r.
wrang (wrong), A. S. wrang.
zung (young), rarely 3unnq, comp. 3unngre (younger), A. S. geong, gyngra.
In heng (hung, A.S. héng ; but henngde [hanged], A. S. hang- ode), the vowel was already long in Anglo-Saxon. Whether this extension to xg of the euphonic analogy which obtains for nad ever gained much currency in the language, may well be doubted. It seems certain that it cannot have prevailed at the time (probably in the fifteenth century) when the old long sounds of ¢ and u (as in pique, prude) began to pass into the diphthongal sounds heard in pike and proud ; for in that case, instead of saying king, thing, téngue, sting, as we now do, we should probably be saying king, thing, loung, soung, with the same vowel utterance as in kind, sound. It nay be observed, however, that in the most recent English there is a noticeable tendency to lengthen somewhat the short sound of o before ng, so that long, song are apt to be pronounced with much the same vowel sound as or, nor, for.
If now we have found in combinations such as /d, md, nd, ng, where tlre first letter is a liquid and the second its cognate sonant mute, a certain tendency to protract the quantity ofa preceding short vowel, it must be remarked that the ordinary tendency of a combination of consonants is in the opposite di- rection, — not to the lengthening of a preceding short, but to the shortening of a preceding long. The speaker slights the vowel in order to concentrate his energy of utterance on the following consonants, which thus massed together present some difficulty of enunciation. It is as in the Greek zeréo- TEPOC (poorer), for MTEVNOTEPOG, t. €. TEVHT-TEDOC, from zévnc, révyr-oc. Of this change — a long vowel shortened on account of two or more consonants following it — numerous examples are found in the language of the Ormulum, when compared with the Anglo-Saxon. Thus |
On English Vowel Quantity. 83
O. asskenn (to ask), A. S. dscian. blosstme (blossom), A. S. bléstma. chappmenn (chapmen, merchants), A. 8. cedpmen ; cf. O. chepinngbothe (market- booth). clennlike (cleanly), clennsenn (to cleanse), from clene (clean), A. S. clenlice, clensian, from clene. dredde (he dreaded), forrdredd (alarmed), from dradenn, dredenn (to dread), A. 8. dréd, drwden, from dredan; in O. as in E., the verb has passed into the weak conj. errnde (errand), A. S. erende. fedde (he fed), from fedenn (to feed), A. S. fédde, from fédan. Sifftis (fifty), from Jif (five), A. 8. fiftig, from fif. fosstrenn (to foster), fossterrfuderr (foster-father), A. 8. fésterian, foster fider. .goddspell (gospel) belongs here, if the A. S. word is g6 ‘spel (good tidings, = stayyéXoy) ; but this is now generally believed to be godspel (God’s word). hallzhenn (saints, cf. Hallow-een), haillzhenn (to hallow), A. 8. hdlige or hdlge, halgian. hiddenn (they hid), hidd (hid), from hidenn (to hide), A. S. hyddon, hyded, from hydan. keppte (he kept), from kepenn (to Leep), A. S. cépte, from cépan. lasstenn (tu last), also lastenn, A. S. lestan.' ledde (he led), ledd (led), from ledenn (to lead), A. S. ledde, leaded, from ledan. mosste (might, cf. must), A. S. mdste. nesst (nearest, next), also nest (Sc. niest), A. S. néhst. redd (read, part.), A. S. réded, from rédan. shadde (he parted), shadd (parted), from shedenn (to part, cf. to shed), A. 8. sceéd, scedden, from sceddan: of weak conj. in O. and E. siththenn (sithence, since), A. S. sidh tham, stdhdhan, perhaps sidhdhan. sleppte (he slept), from slepenn (to sleep), A. S. slép, from slepan: of weak conj. in O. and E. soffie (soft), A. 8. séfie. spredd (spread, part.), A. S. sprceded, from spredan. thratte (he rebuked, cf. threatened), A. S. thredtede, from thredtian. wepptenn (they wept), from wepenn (to weep), A. S. wedpon, from wépan: of weak conj. in O. and E. wimmann (woman), also wifmann, A. S. wifman, wimman, perhaps wimman. wissdom (wisdom), also wisdom, A. S. wisdom.
Probably the short vowel in wraththe (wrath), as well as in laththe (loathing, enmity), and kiththe scen in kiththeli3 (fa- miliarly, cf. kith), may be accounted for in this way, from the addition of a suffix the (A. S. dh, dhu) —the same as in strenncthe (strength), trowwthe (truth), from strang (strong), trowwe (true) — to the adjectives wrath (wréth, A. S. wrddh), lath (loathsome, hateful, A. S. lé@dh), and cuth (known, couth, A.S.ecd@dh): compare A. S. ledhdhu (offence), cdhdhu (home, household). It is possible that in these words, as well as in dredde, fedde, hiddenn, ledde, shadde, siththenn, wimmann, the
84 J. Hadley,
first vowel may have been pronounced long, the following con- sonant being written double only because it was sounded twice, first in the stem, and again in the suffix.” It is also possible, or even probable, that in some cases the change from long to short, now under consideration, may have taken place already in the Anglo-Saxon ;¢ but this makes no difference, either in the reality of the change, or in the cause from which it arose. :
In the list of words just given, the short sound which ap- pears in the Ormulum is maintained in moder. English. The case is otherwise with those which follow:
O. allmasst (almost), though the simple word is nearly always mast (most), A. S.
mast.
demmd (judged), from dem«nn (to judge, cf. doom, deem), A. S. démed, from déman.
derre (dearer), from deore, dere (dear), A. 8. deérra, from deére.
derrlinng (darling), A. S. dedrling.
dunnwarrd (downward), from dun (down), A. 8. ddunweard, from din.
hehhre (higher), also hehre, from heh (high), A. 8. hedhra, from hedh.
herrde (he heard), herrd (heard), from herenn (to hear), A. S. hyrde, hyred, from
hgran.
* Such a supposition must, however, be regarded as improbable for these words on account of the short quantity which they-have in English ; and especially im- probable for the preterits in -dde, on account of the corresponding participles dredd, hidd, ledd, ctc., in which a really double pronunciation of the dis hardly to be thought of.
What is here recognized as possible — that a vowel before a doubled consonant may have been long, the consonant being written twice because actually twice sounded — must be admitted also for lutte (he louted), as well as the comparatives derre (dearer) and nerre (nearer), mentioned in the next paragraph; and, perhaps with still stronger reason, for the words clennesse (clean-ness), fiffald (fivefold), leffull (belief-full, believing). As we have skillles (skill-less, ignorant), sefennnahht (seven-night, week), unnned (un-need, without constraint), sunderrrun (sunder-roun, private communing), forrraht (perverted, Germ. ver-riickt), it seems not unlikely that the r, n, and f would have been written thrice in derre, nerre, clannesse, fiffald, le full, if their first vowel had been short in sound; but the spelling of fidlike ( full-like, fully), stilliz (stilly), idelle33c = idellnesse (idleness), drunnkennesse (drunk- en-ness), unnitt (useless, Germ. un-niitz), forrswundennesse (remissness), orrath ax well as orrrath (inops consilii, O. N. dr-rddh), warns us not to lay too much stress on this considcration.
t The same posikibility is not to be overlooked in other cascs where the vowel quantity of the Ormalum differs from what must have been the primitive quantity in Anglo-Saxon. This is particularly true as to that lengthening of vowels in open syllables wich is soon to be considered : the change could hardly have gone so far in the language of the Urmulan, if it had not made a beginning in Anglo- Saxon times. 7
On English Vowel Quantity. 85
O. laffdi3 (lady), A. S. hlafdige ; cf. O. laferrd (lord), A. 8. hldford.
liccness (likeness), from lic (like), A. S. gelécnes, from gelic, but O. onnlicnesse (likeness, image).
lutte (he bowed, louted), from lutenn (to lout), A. S. ledt, from litan: of weak conj. in O. and E.
nerre (nearer ), from ner (nearly), O. N. nerri, from ner; cf. O. ner (nearer), from neh (nigh), A. S. nedr, from nedh.
oththr, orr, from otherr (all meaning or), A. S. ddher = dwdher (either): oththr, perhaps, by a confusion of A. S. odhdhe (or) with ddher.
Thurrsda33 (Thursday), O. N. Thérs-dag-r, A. S. Thunres-dég.
wennde (he weened ), from wenenn (to ween, think), A. 8. wénde, from wénan.
wesste (waste, desert, adj. and subst.), A. S. weéste.
From a continued working out of the same tendency, the English has a short vowel before two or more consonants in some words where the Ormulum shows the original long vowel quantity :
O. adle (disease, cf. addle), A. S. ddl.
breost, brest (breast), A. S. bredst.
brethre (brethren), A. S. brodhru.
buhsumm (pliant, compliant, cf. buxom), from A. S. bigan (to bend, bow). to-clef (he cleft, also clave, clove), A. S. cledf, from cleéfan (to cleave ). dost (dost, usual pron. dist), also dosst, from don (to do), A. S. dést, from dén. fifte (fifth), fiftende (fifteenth), A. S. fifta, fifleddha, O. N. fimtdndi. freond, frend (friend), A. 8. freénd.
gom (care, heed, cf. gumption), A. S. gedm.
hese (command, hest), A. S. hes.
heold, held (he held), from haldenn (to hold), A. 8. hedld, from healdan. monethth (month), A. S. ménadh.
naness, in forr the naness (for the nonce), from ceness (once, Sc. aines), A. S. cene.
, In the following words also the shortening may be explained on the same principle, since the vowel which is written before
their final liquid is little, if at all, represented in their actual pronunciation:
O. efre (ever), nefre (never), A. S. afre, nafre. becnenn (to beckon), A S. bécnan, bedcnian. bosemm (bosom, often pron. with long oo sound), A. S. bésm. brotherr (brother), pl. brethre (brethren), A. 8. brédhor, brédhru. moderr (mother), A. S. médor.
otherr (other), A. S. édher. wrepenn (weapon), A. S. waepen.
The cases which we have been considering show an ac-_ cented long vowel shortened in a close syllable, where it is
86 J. Hadley,
separated by more than one audible consonant sound from the vowel of the following syllable. We have next to notice a change which is the converse of this, —a change which has cut much deeper into the integrity of the old system of quan- tities — the lengthening of an accented short vowel in an open syllable (generally a penult), that is, when separated by only one consonant sound from the vowel of the syllable which fol- lows. This change has been carried to a very great extent in the modern German: geben (to give), nieder (down, cf. E. nether), tragen (to draw), nehmen (to take), are examples taken up at random outofan immense multitude. According to Schleicher, this change of quantity belongs to, and is a prominent feature in, the transition from Middle to Modern High German, which was made in the fifteenth century. In England the change must have commenced its progress carlier, as we find it carried very far in the language of the Ormulum, which belongs to the thirteenth century. We give first the instances in which the lengthened vowel seen in the Ormulum became so established in English usage as to remain long in the pronunciation of to-day. Thus O. -ale (ale), in bridale (bride-feast), A. S. ealu, brjdealu. In E. bridal it has be- come short again. azhe (awe), but also e33¢ (fear), A. S. ege; cf. O. N. eqja (to strike with fear or awe).
bakenn (to bake), A. S. bacan.
bede (prayer, cf. bead, bead-roll, beads-man), A. S. gebed.
berenn (to bear), borenn (born), A. S. beran, boren.
bidell (crier, messenger, cf. beadle), A. S. bydel
brasene (brazen), from brass (brass), A. S. brdsen, from brds.
brekenn (to break), A. S. brecan.
bridledd (bridled), with 1 from bridell (? not found in O.), A. S. bridel.
bule (bole, tree-stem), in bulaxe (axe, hatchet, cf. pole-axe), O. N. bolézxi, from
bol-r.
care (care), A. S. cearu, caru.
chari3 (mournful, anxious, ef. chary), A. S. ceariy, from cearu
chosenn (chosen), from chesenn (to choose), A. S. coren, from cedsan.
clofenn (cloven), A. S. clofen, from cleéfun (to cleave).
cnapess (boy’s, cf. knave), A. S. cnape, cnafa (boy).
cnedenn (to knead), A. S. cnedan.
dekenn (Levite, deacon), A.S. diacon, Lat. diaconus.
dale (dale), A. S. dal: original quantity preserved in E. dell.
dragzhenn (to draw, drawn), A. 8S. dragun, dragen.
kirrkedure (church-door), A. S. duru (door).
On English Vowel Quantity. 87
O. efenn (equal, even), but pl. effne, vb. effnenn, A. S. efen, efenian.
ele (oil), A. S. ele, O. Fr. oile, oille, Lat. oleum.
etenn (to eat), but impv. ett (eat), A. 8. etan, et.
Jaderr ( father), A. S. fader.
farenn (to go, fare), but impv. farr, A. S. faran, far.
biforenn (before), also biforr, A. S. beforan.
bifrorenn (frozen), A. S. froren, from jredsan (to freeze).
gate (way, cf. gait), O. N. gata (way), A. S. geat (gate).
qreefess (ditches, cf. grave), A. S. graf. -
-qume (man), in bridgume (bridegroom), A. S. guma, brydquma.
hatenn (to hate), hete (hate), A. S. hatian, hete.
hefenn (to raise, heave), hofenn (hove, hoven), A. 8. hebban, hofen.
hire (her), A. S. hire.
hizhenn (to hasten, cf. to hie), A. S. higtan.
hope (hope), A. S. hopa.
ifell (evil), A. S. yfel.
kechell (cake), O. N. kaka.
ladenn (to draw out, cf. to lade water, also ladle), A. S. hladan.
late (late), but lattre (latter), lattst (last), A. S. lat, latra, latost.
lashe (law), A. S. lagu.
Sorrlorenn (lost, cf. forlorn), from forrlesenn (to lose), A. 8. forloren, from _for- leésan. |
makenn (to make, Sc. mak), but impv. macc, A. S. macian, maca.
mele (meal, flour), A. S. melu.
mete (Meat, food), A. S. mete.
efennmete (commensurate), from mett (ineasure, cf. mete), A. S. gemet: O. metelike (meetly), A. S. gemetlice.
nakedd (naked), A. S. nacod.
name (name), but nemmnenn (to name), A. S. name, nemnan.
binethenn (beneath), but niththrenn (to lower, cf. nether), A. S. beneodhan, nidh- erian.
nizhenn (nine), nizhennde (ninth), A. S. nigon, nigodha, O. N. niundi (ninth).
oferr (over), but also offr, A. S. ofer.
openn (open), but oppnenn (to open), A. S. open, openian.
rezhellboc (rule-book), A. S. regol, Lat. regula.
sake (quarrel, cf. sake), A. S. sacu.
forrsakenn (to forsake, forsaken), A. S. forsacan, forsacen.
same (same), A. S. same (pariter), O. N. sam-r (idem).
sezhenn, sene (seen), A. S. sewen, segen, sén, from seén (to see).
shame (shame), but shamm/asst (shamefaced), A. S. sceamu, sceamsast.
shapenn (to form, create, cf. to shape), A. S. sceppan.
skathenn (to harm, to scdthe), A. S. sceadhan.
slazenn (slain), from slan (to slay), A. S. slagen, from sledn.
smeredd (anointed, cf. smeured), A. S. smyred, from smyrian, cf. smeoru (oint- ment).
spekenn (to speak), A. S. sprecan, and specan,
stélenn (to steal), A. S. stelan.
stirenn (to stir, move, Sc. steer), A. S. styrian.
swerenn (to swear), A. S. swerian.
88 J. Hadley,
O. anndswere (answer), A. 8. andswaru. takenn (to take, taken, Sc. tak), but impv. tacc, A. S. tacan, tacen, tac. tale (reckoning, number, cf. tule), A. S. talu. tholenn (.0 suffer, Sc. thole), A. S. tholian. wakenn (to wake, watch), A. S. wacan.
waterr (water), but wattrenn (to water), A. S. wiiter, wiiterian. weorelld, werelld (world), but gen. vvcoridese werridess, A. S. weoruld, woruld, world.
wrekenn (to wreak), wrache (wreak, revenge), A. S. wrecan, wracu. wuke (week), A. S. wucu. gate (gate), also gate, A. S. geut.
It isremarkable that this change is carried toa much greater extent in the Semi-Saxon of the Ormulum than it is in modern English. It should seem that there must have been a reaction early established, which set limits to the tendency, and main- tained the short vowel in many words where it had begun to be lengthened. Instances of this kind—where an accented short vowel in an open syllable is lengthened in the Ormulum, but the same vowel is found short in English (mostly, indeed, in monosyllables with final consonant sound) — are the fol- lowing :
O. abufenn (above), = bufenn, from A. S. 4, be, and ufan. °
beodenn, bedenn (bidden), A. S. boden (commanded), beden (cntreated). [The A. S. verbs biddan (to entreat) and beédan (to command) are pretty much confounded in O.]
bisscopess, pl. of bisscopp (bishop), A. S. bisceop; bisscopess had a secondary accent on the o.
byte (bit, morsel), A. S, bite.
bodi3z (body), A. 8. bodig.
bule (bull), Dutch bul, O. N. boli.
clepedd (called, cf. clept, yclept), from clepenn, A. S. cleopod, from cleopian.
ctide (cud), A. S. cud.
cumenn (to come, also as part.), but impv. comm, cumm, A. S. cuman, cumen, cum.
cwike, pl. of ewice (living, quick), A. S. ewic.
dide (did), from don (to do), A. S. dyde, from dén.
drake (dragon), A. S. draca, Lat. draco.
drifenn (driven), from drifenn (to drive), A. §. drifen, from drifan.
Jretenn (to fret, trans.), A. S. fretan (to eat up).
glade, pl. of gladd (glad), gladenn (to gladden), A. S. gléd, gladian.
godess, gen. of godd (god), but pl. goddess, A. S. god.
gresess (grasses), Sing. in gresshoppe (grasshopper), A.S. grds, gars, gdrshoppa.
hafenn, but habbenn (to have), hafesst, but haffst (hast), haffde (had), A. S habban, hdfst, hafde.
hefiz (heavy), A. 8. hefig.
heofenn (heaven) in comp., but as sep. word heoffne, heffne, A. S. heofon.
On English Vowel Quantity. | " 89
O. hiderr (hither), A. S. hider. hise, pl. of hiss (kis), A. S. his. huni3 (honey), A. S. hunig. kide (kid), O. N. kid. kiness, kine (comm. kinness, kinne), gen. and pl. of kinn (kin, kind), A. S.cyn. _lifethth (liveth), from libbenn (to live), A. 8. lifadh, libban. dimess (limbs), sing. not in O., A. 8. lim, pl. leomu, imu. litell (little), pl. little, A. S. lytel. lokenn (shut in, cf. locked), A. S. locen, from lican. lotess, pi. of lott (lot), A. S. hlot, pl. Alotu. lufe (love), lufenn (to love), but luffsumm (pleasant), A. S. lufu, lufian, lufsum. mani3, mani (many), A. 8. manig. mikell (great, many, mickle, much), but pl. miccle, A. S. micel. mineteress (money-changers, cf. minter, mint), A. 8. myretere, from mynet (money), Lat. moneta. muneclif (monk-tife), A. §. munec (monk), Lat. monachus. naru (narrow), but pl. narrwe, A. S. nearu. nile (nill he, i. e. will not), but nillt (wilt not), A. S. nelle, nelt. ofne (oven), dat. of ofenn (2 not in O.), A. S. ofen. peninng (penny), A. S. pending, pening, penig. rathe (quickly, cf. rathe, rather), A. S. hradhe, hradhor, from hrdédh (quick). risenn (risen), from risenn (to rise), A. 8. risen, from résan. rotenn (to rot), A. S. rotian. Saterrda33 (Saturday), A. 8. Saternddg, Lat. Saturni dies. seofenn, sefenn (seven), but also seoffne, seffne; seofenntiz (seventy): A. S. seo- Jon, hundseofontig. shetenn (to shut up), A. S. scyttan. sikerr (sure, Sc. sicker), O. Sax. sikor, O. H. Germ. sihhur, Lat. securus. stne, rare for sinne (sin), A. S. syn. skatheles (unharmed, scdthless), A. S. sceadha (harmer). stafess, pl. of staff (letter, cf. staff, old pl. staves), A. S. staf (staff, letter). stede (place, cf. stead), A. S. stede. | stekenn (to confine, cf. to stick, remain fast), A. S. stician. stoke (stock), dat. of stocc (? not in O.), A. S. stoc. sume, pl. of summ (some), A. S. sum. sumerr (summer), A. 8. sumor. sune (son), A. S. sunu. Sunennda33 (Sunday), from sunne, rarely stine (sun), A. S. sunne, Sunnanddg. thiderr (thither), A. 8. thider. Sullthrif un (complete, cf. thriven), O. N. thrifinn, from thrifask {to thrive). thripell (triple), from Lat. triplex, Fr. triple, confused with A. S. thri; cf. prov. Eng. thribble. tredenn (to tread, trodden), A. S. tredan, treden. whiderrwarrd (whitherward), A. S. hwider. widewe, comm. widdwe (widow), A. S. widwe, wydewe. wilenn (to will), but willt (wilt), A. S. willan, wilt. witenn (to know, cf. to wit, O. E. to weet), but impv. witt, A. S. witan, wit. writenn (written), from writenn (to write), A. S. writen, from writan: cf. O. writess, pl. of writt (writ), A. 8. writ.
90 J. Hadley,
O. wude (wood), A. S. wudu. wunedd (wont), from wunenn (to accustom), A. S. wuna, gewuna (custom). getenn (to get), ett (gets), bizetenn (gotten), A. S. begitan, begiten. 3ifenn (to give, given), also written with g, but impv. 3/7, A. S. gufan, gifen, gif That this change—the lengthening of an accented short vowel in an open syllable — was still in progress at the time of the Ormulum, so that the usage in respect to it was then unsettled and fluctuating, is apparent from indications in the book itself. A number of the words given in the last two lists have here and there a mark of short quantity written over the vowel, as if the writer, having first given it as long with only a single consonant after it, was afterwards inclined to recall his judgment, to set it down as short, and therefore drew a curve line over it, this being an easier way than doubling the consonant by interlineation. What makes this explanation more probable is the fact that, while there are more than forty distinct words which in one place or another have this short mark over them, it occurs in almost every case over an ac- cented vowel in an open syllable. That there was a special vacillation on the part of the writer as to the quantity of such vowels, seems a natural, if it is not a necessary, inference from this fact. Thus berenn (to bear) is once at least written with a mark of short quantity over the accented vowel; and the same is true of bede (bede), dale (dale), hatenn (to hate), hete (hate), ladenn (to lade), late (late), mele (meal), mete (meat), name (name), stelen (to steal), takenn (to take), tale (tale) ; also bite (bit), cude (cud), kine (hin), lifethth (hiveth), sine (sin), stede (stead), thrifenn (thriven), wilenn (to will), witenn (to wit), writenn (written).
Under the broad euphonic analogies and tendencies which have now been described come all but a comparatively small number of the cases in which the modern English quantity differs from that in the Ormulum. There remain, however, some few changes which are not altogether of an isolated char- acter. <A long vowel of the Anglo-Saxon and the Ormulum has in a good many instances been shortened before a final mute. This is especially the case with the old long o before a final k-sound. The long quantity of that vowel was indicated
On English Vowel Quantity. y1
in early English by doubling the o: thus 6éc was written book. The gound afterwards changed to that which we hear in spook, spool; and still later was shortened to its present pronunciation. Instances of this kind are
O. boc* (book), A. S. bec. croc (hook or crook, device), O. N. krok-r. lokenn (to look), A. S. locian. Jorrsoc ( forsook), from forrsakenn (to forsake), A. S. forséc, from forsacan. toc (took), from takenn (to take), A. S. téc, from tacan.
Occasionally other long vowels have become sliort before a k-sound, as in
O. brukenn (to use, enjoy, cf. to brook), A. S. brican. Sic (fig), in fietre ( fig-tree), A. S. fic, Lat. ficus. seoc, sec (sick), A. 8. seédc. struc (passed, ef. struck, O. E. strook), A. S. strdc, from strican. wic (dwelling, strect, cf. Swanwick, Greenwich), A. S. wic.
The same change of quantity has taken place not unfre- quently before d, seldomer before ¢: thus
O. blod (blood), A. S. bled. Jlod (flood), A. S. fled. god (yood), A. S. géd. stud (stood), from stanndenn (to stand), A. S. stéd, from standan. wod (mad, Sc. wud), A. S. wed. bred (bread), A. S. bredd. died (dead), A. 8. dedd. dreedenn, dredenn (to dread), A. 8. dredan. hafedd (head), A. 8. hed fod. shedenn (to part, cf. to shed), A. S. sceddan. shredenn (to shred, pare, cf. Sc. screed), A. S. screddian. but, comm. butt (but, except), A. S. bitan. fot (foot), A. S. fot. hat (hot), A. S. hat. , letenn and létenn (to let, allow, also as part. let), pf. let (he let), A. S. latan, leten, pf. lét. swat (sweat), A. S. swt. wete (drink), from weet (not in O., wet, Sc. weet), A. S. weet. wat, also wait (wot), from witenn (to know, cf. to wit), A.S. wdt, from witan.
In bedethth or biddethth (biddeth), forrbedethth (forbiddeth),
the long form comes from A. 8. beédan (to command), the short one from A.S. biddan (to entreat) : in db:ddenn (to com-
* In this word, and in several others, where a long vowel has become short be- fore a final consonant, the Scotch retains the earlier long quantity: thus buik, bruik, bluid, gude, ete.
92 J. Hadley,
mand, to entreat) of the Ormulum, the forms of the two verbs are very much confounded.
The few cases in which the difference of quantity between the Ormulum and the modern English is not to be explained from principles already set forth, will be found, so far as I have noted them, in the following list :
O. amen (amen), Gr. ’aphy.
an (an, one), rarely ann; nan (none) ; onnan, anan, also anann (anon); A. S. dn, ndn. Eng. alone, atone, only, and Sc. ane, nune, preserve the original long quantity.
ani3 (any), A. S. enig.
beon, ben (to be, been), A. S bedn (to be): the long sound of been is still some- times heard.
cariteth (charity), Fr. charité, Prov. caritat, Lat. caritas.
chele, also chéle (chill, subs.), A. S. céle.
clath (clothing, cf. cloth, but pl. clothes), A. S. clddh.
cuthe (could), A.S. cédhe.
def (deaf, cf. Sc. deave, to deafen), A. S. dedf.
deth (death), A. S. deddh.
deofell, defell (devil, Sc. deevil), A. S. deéfol, Lat. diabolus.
doth (doth, usual pron. diith), from don (to do), A. S. dédh, from don.
flcesh (flesh), A. S. flesc.
gluterrnesse (gluttony), O. Fr. gloutonnte (from glouton, Lat. glato(n), from gli tire): in gluterrnesse English affixes are attached to the root (glut) of the Latin and French words.
gyn (art, device, cf. gin), shortened from O. Fr. engin, but perhaps confused with a derivative of O. N. ginna (to deceive).
inoh (enough, also enow), A. S. gendg, gendh.
bikeechedd (catched), also bikahht (caught) ; of uncertain origin.
profete, prophete (prophet), Lat. prophéta.
publicaness (publicans), Lat. piblicani.
redi3 (ready), also redelike, A. S. rede, redlic.
riche (rich), A. 8. rice.
sari3 (sorry), but sare (sorely), A. S. sdrig, from sér (sore).
selili3 (happily), A. S. selig, geselig (happy, Sc. seely), whence E. silly.
shephirde (shepherd), from shep (sheep), A. S. scedp, scedplirde.
shunenn (to shun), A. S. sciinian.
tene (ten), rarely tenn, tende (tenth), A. S. tn, tén, teddha, 0. N. tiundi (tenth). The old long quantity is preserved in the compounds thirteen, thirteenth, etc., and in Sc. teinds (tithes).
onnxeness (against, again), onnzen (again, against), A. S. ongedn (against).
jet (yet), strangely lengthened in O., A. S. git, get, yiet, gyt.
drunncnenn (to drown, trans.), in form = A. S. druncnian (to get one drunk), cf, O. N. drukna (to be drowned).
ennoell (angst), A. S. engel, dngel, Lat. angelus.
On English Vowel Quantity. 93
O. flumm (river, cf. flume), O. Fr. flum, Lat. flumen.
funnt (font, cf. fount), A. S. font, Lat. fons, font-is.
irrene (of iron, ferreus), from irenn (iron), A. S. tren (ferrum and ferreus) : rr in trrene perhaps an oversight.
munnt (mount), A. S. munt, Lat. mons, mont-is.
sannt (saint), A. S. sanct, Lat. sanctus.
sleckenn,® slekkenn (to slake), A. S. sleac (slack), gesleccan (to slacken), O. N. slakkva (to slake).
thurrh (through), A. S. thurh, thuruh, (through, thorough).
waccnenn (to woken, trans. and intrans.), A. S. wécnan: lengthened in E. under influence of to wake, O. wakenn, A. S. wacan.
whamm (whom), from wha (who), A. S. Awam, from hwa.
jgocc (yoke), A. 8. geoc, gioc.
In the case of been, could, deaf, death, enough, ready, again, against, the spelling shows that they came into English with the long quantity which they had in the Ormulum. The preterits barr (bare, bore, A. S. bdr), bat (bit, A. 5. bat), brace (brake, broke, A.S. brde), comm (came, A. S. cwam, com), cwathth (quéth, A.S. cwddh), sahh (saw, A. S. seah). space (spake, spoke, A. S. sprdc), 3aff (gave, A. S. geaf), have not been placed in the foreguing lists, because the English forms, though used in both numbers, correspond apparently to the plurals of the A. S. (beron, biton, breecon, ewdmon or cémon, cweedon, s@won or segon, sprecon, gedfon) and the O. (berenn, comenn, sexhenn, spekenn, 3efenn): in et (he ate, A. 8S. dt) with long e like pl. etenn (they ate, A. 5S. wton), the same ex- tension of the plural quantity to the singular appears even in the Ormulum. The ow in the English preterits bound, found, wound, — A. S. 1, 3 sing. band, fand, wand, p). bundon, fund- on, wundon; QO. sing. band, fand, wand, pl. bundenn, fundenn, —-is to be explained in the same way.f
We have not yet attended to the suffixes of inflection and derivation ; but for these only a few words will be necessary. The inflectional endings are all short in the Ormulum: there is reason, indeed, to believe that such as were originally long
* The digraph ck, in the Ormulum, is equivalent to cc or kk, and marks the vowel before it as short. At the end of a word, or before a consonant, cc is alone used : if a vowel follows in the same word, ck or kk takes its place.
t For several of these preterits the Scottish dialect has forms — such as brak, cui, spak, fand, etc. — which correspond to those here given from the Ormulum.
94 J. Hadley,
had become short during the Anglo-Saxon period. Thus the Ormulum has . -ess in the gen. sing.: as flashess kinde (flesh’s kindred, A. S. flesces gecynd). -ess in the plural: as lachess (leeches, A. S. kecas, earlier lecds or lacids). -err in the compar. forrtherr (further, A. S. furdhor), from forth (forth, A. S. fordh): in the compar. of adjectives the O. has -re (A. S. -ra, -re), as Sulre ( fowler, A. S. filra). In ¥E., -er has become long by change of r. -esst in the superl : as deresst (dearest, A. S. dedrost, d+érest). -esst in the 2d person of verbs: as heresst (hearest, A. S. hyrest). -ethth in the 3d person of verbs: as lokethth (looketh, A. S. lécadh). enn in the past part.: as haldenn (holden, A. S. healden). -edd in the past part.: as wundedd (wounded, A. S. wundod).
The -de of the weak preterit is usually added directly to the stem, as dredde (he dreaded). The present participle termi- nates in -ennde (A. S. -ende), but is rare in the O., the only instances where it is a proper participle being bernennde (burning), dwallkennde (misleading), glowennde (glowing), and stinnennde (stinking). The suffix -inng is very frequent, but always forms a verbal substantive; while -ung, which in Anglo-Saxon is more uscd for this purpose, is in O. confined to the word reowwsunng (rueing, repentance) and two or three others. Suffixes of this kind, found both in O. and in E., are
-ell (A. S. -el) : as girrdell (girdle).
ene (A. S. -en) : as brasene (brazen).
ere (A. S. -ere): as mineteress (minters, money-changers) ; very rare in QO.; another instance perhaps in forrle3errnesse (fornication).
-inng (A. S. -ing, -ung): as biginninny (beginning), laferrdinngess (lordings).
-issh (A. S. -isc): a» shepisshe (sheeplike, sheepish).
-iz3 (A. S. -ig) : as modi3 (moody).
-linng (A. S. ting) : as derrlinng (darling).
-nesse (A. S. -nes) : as godnesse (guodness), wittness (witness).
-stere (A. S. -stre) : only in huccsteress (huckster’s).
In final -ene and -ere, the first ¢, short in A. S., is length- ened by the open syllable: the English -er is of course long. Final -23 is long in consequence of the partial vocalization of the g (compare the effect of a weak 7), while the correspond- ing suffix -y of the later language, discarding the semivowel, has returned to the short quantity of A. S:-ig. The same changes appear in the numerals twennti3, thritti3, fowwerrti3, etc. (twenty, thirty, forty, etc., A. 8. twentig, thrittig, fedwer- tig, etc.). Butin the suffix -/3 (our -ly) the vowel was origin-
On English Vowel Quantity. 95
ally long, the words which contain it being compounds of the adj. like (O. lic, A. 8S. gelic). Suffixes of this kind (really words in composition) are the following, found both in the Ormulum and in English:
-dom (A. §. -dém) : as horedom (whoredom).
-fald (A. §S. -feald) : as threfald (threefold).
-fasst (A. S. -fast): as stedefasst (steadfast); in shamm/fasst (shamefuced) the form has been changed by mistaken popular etymology.
full (A. S. -ful): as sinnfull (sinful).
-had (A. S. -hdd) : a8 ma33dennhad (maidenhood, maidenhead).
dac (A. S. Idec): only in weddlac (wedlocky.
-les (A. S. -leds): as childles (childless) ; rarcly ess, as endeless (endless).
-li¢, -like or -liz (A. S. lic): as eorthlic, eorthlike, eorthliz (earthly).
-mann (A. S. -man): as allderrmann (alderman).
-shipe (A. S. -scipe): as wurrthshipe and wurrshipe (worship).
-summ (A. §. -sum): as halsumm (wholesome).
-warrd (A. S. -weard): as affterrwarrd (afterward).
In -fald and -shipe, the Ormulum has lengthened an Anglo- Saxon short vowel, from the influence of /d in the first case, and of an open syllable in the second ; as to the last, the Eng- lish agrees with the Anglo-Saxon. The short vowel of -warrd has been lengthened in English by the weak 7; while the long vowel lias been shortened in -dum, -had, -lac, -les, two of which end with mute sounds. In -les, the change had commenced in the thirteenth century. The Ormulum has in many words a suffix -le33c —as seen in godle33c, = godnesse (goodness) — which corresponds to -leik-r (= A. 8. -lée), a frequent suffix in the Old Norse. Peculiar cases are rihhtwis (righteous, A. 8. rthtwis), and stallwurrthli3 (stoutly, cf. stalwart, A. S. stilweordh).
It has seemed desirable to add here a list of all the words, not already noticed, which appear in the Ormulum with the same vowel quantity that they have in modern English. The cases noted in the foregoing lists as showing such an agree-, ment consist of words in which the earlier quantity has been changed under .cuphonic tendencies which had commenced their course before the time of the Ormulum. But in nearly all the words which follow, the primitive quantity has re- mained unchanged from Anglo-Saxon times to the present day.
96 | J. Hadley.
In this list, as in those before it, where the same root is found with different suffixes, only one form has been given, unless there were special reasons for adding others.”
O. er (ere), A. S. ar: superl. eresst (erst) in allreresst, A. S. erest. ere (ear, auris), A. S. edre. est (east), A. S. edst. , ath (easy), unnethe (with difficulty), cf. O. E. eath, uneath, A. S. eddhe. affterr (after), bafftenn (behind, cf. abaft), A. 8. gfter, baftan. uh (1 owe), A.S. GA. ahnenn (to obtain, conquer, cf. to own), A. S. dgnian (to possess, acquire). allfa (alpha), Gr. &dga. alls (as), from allswa, A. S. ealswd. amm (am), A. S. eom. aand (und), rarely and, A. S. and. anngrenn (to anger), O. N. dugr (vexation), dngra (to vex), see note on p. 81. anntecrist (antichrist), Lat. antichristus. appell (apple), A. 8. dpl, dppel. ar (early, Sc. air), sce er. asse (ass), A. S. assa. asskess (ashes), A. S. asce. athess (oaths), A. S. adh. att (at), A. S. dt. axret (axe), A. S. ax. azenn (own), A. S. dgen; from a3henn (to possess, cf. to owe), A. S. dgan. bacc, bacch (back), A. 8. bde. tem (beam), in sunnebem (sunbeam), A. S. bedm (tree, column). bere (bier), A. 8. ber. beetenn (beaten), A. S. bedten. bannkess (banks), A. S. banc (mound). bapptisstess (baptist’s), Lat. baptista. bathe (both), A. S. bd twd. bathth (bath), A. S. bddh.
* Monosyllables which end in vowels are not included in this list, nor in any of those already given, for the reason that the peculiar orthography of the Ormulum fails to indicate their quantity. There can be little doubt, however, that they were regularly long, those which were originally short —as i, also icc (Z), he (he), fra (from, cf. fro), etc. — having been already lengthened. In modern English, all accented final vowels are long: if the articles a and the are short, it is because they have given up their separate accent, and become incorporated as proclitics _ with the words that follow them. In the Old Norse, too, a root-vowel is regu-
larly long when it stands at the end of a word, so that short final vowels appear only in unaccented syllables of inflection or derivation.
t The writer of the Ormulum does not repeat the double consonant x. He writes aze (are), fox (for), ore (or), sexe (six), waxenn (to war), though in all these the preceding vowel must have been short. He does, however, write warzrenn in at least one instance; and it is quite likely that in the preterit wex (A. S. wedz) the vowel was long.
é
On English Vowel Quantity. 97
O. bedd (bed), A. S. bed. belle (bell), A. S. belle. bennche (bench), bennkedd (benched), A. S. benc. bettre (better), betist (best), A. S. betera, betst. biddenn (to bid), A. S. beédan; see p. 91. abidenn (to abide), abad (abode), A. S. &bidan, &'dd. biggenn (to dwell), unnbiggedd (anbuilt on), cf. Sc. to big (to build), biggie (built), O. N. byggja (to build, dwell), A. S. b@7an. birrzenn (to bury), A. S. byrgan. bisscopp (bishop), A. 8. bisceop, Lat. episcopus ; see p. 88. bitenn (to bite), A. S. bitan. bitterr (bitter), A. S. biter. blretenn (to bleat), A. S. blatan. toblawenn (bloated, cf. to blow), A. S. bldwan (to blow). ° blettcenn (to bless), blettsedd (blessed), A. S. bletsian, gebletsod. dlisse (bliss), A. S. bis. ithe (blithe), A. S. blidhe. dlome (bloom), O. S. bémo, O. N. blém, A. 8. bléwan (to blow, bloom). dlunnt (blunt), O. N. blunda (to wink, drowse, be dull). bone (boon), A. S. bén, O. N. bén. bote (amend, cf. boot), A. S. dét. bothe (booth), O. N. bitdh. bozhess (boughs), A. S. bé7, béh. brad (broad), A. S. bréd. brass (brass), A. S. bras. brennd (burned, cf. brand), A. S. brenned, brand. breress (briars), A. §. brér, brer. brid (bride), A. S. bryd. brinngenn (to bring), A. S. bringan. cobrisenn (to bruise), A. S. brysan, tébrysan. buce (goat, buck), A. S. bucca. bure (abode, cf. bower), A. S. bir. oarrh (town, borough), A. S. burh, burg. butt (but), also but, A. S. dian. buttenn (to but), Fr. bouter. buzhenn (to bow), A. S. biugan. cann (he can), A. 8. can. casstell (castle), A. S. castel, Lat. castellum. chaff (chaff), A.S. ceaf. chepinngbothe (market-booth), cf. E. cheap, A. 8. cedp (bargain, price). chesenn (to choose), ches (chose), A. S. cedsan, ceds. chesstre (city, town, cf. -chester in names of place), A. S. ceaster, Lat. castrum. chilldre (children), pl. of child, A. S. cildru, from cid. toclaf (he clave, clove), A. S. cled/, from cledfan (to cleave). clathess (clothes), pl. of clath, A. S. clddh. clene (clean), A. 8. clene. clippenn (to clip), O. N. Kippa (to shave off). cludess (hills, cf. cloud), A. S. cléd (rock, hiil). clutess (clothes, cf. clout), A. S. clu (clout). 13
98
0.
ed. Hadley,
cnawenn (to know), cneow, cnew (knew), A. S. cr Awan, cnedw.
cnelenn (to kneel), A. S. cnedwian.
enif (knife), A. S. cnif.
crafft (craft, science), A. S. créft.
eribbe (crib), A. S. crib, cryb.
crisstene (christian), crisstnenn (to christen), from Crist (Christ), A. S. eristen, cristnian.
crummess (crumbs), A. S. crume.
crune (crown), O. Fr. corone, Lat. coréna.
cullfre (cudver, dove), A. S. culfre, culufre.
cumnenn (to know, be able, ef. cunning), A. S. cunnar.
cuppess (cups), A. S. cupp, cuppa.
cuth (known, Se. couth), unncuth (unknown, uncouth), A. S. cudh, unendh.
ewellenn ( 0 quell, kill), A. 8. cwellan.
cwen (queen), A. S. cwén.
ewennkenn (to quench), A. S. dewencan.
cwice (living, quick), but with -e ewike, A. S. cwic.
deel, dale, del (part, deal), dalenn (to share, deal), A. S. dal, delan.
dew (dew), A. S. dedw.
dede (deed), A. S. ded.
dellfenn (to bury, cf. delve), A. 8. delfan (to delve).
dellta (delta), Gr. dédra.
demenn (to judge, cf. to deem), A. S. déman.
deop, dep (deep), A. S. dedp.
deor, der (animal, cf. deer), A. S. dedr.
deore, dere (dear), A. S. deore.
dejenn (to die), O. N. deyja.
dinnt (blow, cf. dint), A. S. dynt.
dippenn (to dip),-A. S. dyppan.
dom (doom, judgment), A. S. dém.
don (to do), A. S. dén.
drem (sound, connected with dream), A. S. dredm (sound).
dreort3, dreri3 (sad, dreary), A. S. dredrig.
drezhenn (to suffer, Sc. to dree), A. S. dredgan.
drifenn (to drive), draf (drove), A. S. drifan, drdf.
drinnkenn (to drink), drannc, drunnkenn, A. S. drincan, dranc, druncen.
dun (subs., down, hill), dun (adv., down), A. S. dtin, dime.
dusst (dust), A. S. dust.
dwellesn (to dwell, delay), A. S. dwellan (to hinder, delay).
ec (also, eke), A. S. ede, éc
efenn (evening, even), A. S. wfen.
effisone, effisoness (eftsoons), A. S. eft and séna.
egge (edge), eggenn (to egg on), A. S. ecg, ecgan, O. N. egg, egqja.
ekenn (to eke, increase), A. S. écan.
elldre (elder), A. S. yldra, from eald (old).
elless (else), A. S. elles.
ezhe (eye), pl. ehne, ezhne (eyne, Sc. een), A. S. edge, edgan.
Jierenn (to make one fear), feredd (afeard, afraid), A. S. dferan, dfcred.
Jawe (sew), A. S. fed, =fedwe.
On English Vowel Quantity. 99
O. fanngenn = fon (to take, cf. fang), A. S. fon, part. fangen. Jasst ( fast, firm), fusste ( fast, quickly), A. S. fast. Jfasste ( fast, jejunium), A. S. fasten. fatt (fut, vat), in reclefatt (incense-vessel, censer), A. S. fat. Jecchenn (to fetch), A. S. feccan. fedenn (to feed), A S. fédan. "fell (skin, fell), A. S. fel, fell. | fell (he fell), A. S. fedl, from feallan (to fall). Vowel shortened in O. and E. fend (enemy, fiend), A. S. fednd. Jeorthe, ferthe ( fourth), A. S. fedrdha. . ° Jesstnenn (to fix, fasten), A. S. fdstuian. ' Jet (feet), from fot (foot), A. S. fet, from fot. _ fezenn (to join, cf. to fay), A. S. fézan. Jif (five), A. S. fi - filenn (to file, defile), A. 8. fylan, dfylan, from ful ( foul). Jilenn (to fill), A. 8. fyllan, from ful (full). Jir (fire), A. S. fir. Jisskess ( fishes), A. S. fise, pl. fiscas. Slerd (mockery, cf. fleer, subs.), O. N. ferdh (deceit). Jleon, flen (to flec), A. S. fleén. Jletenn (to flow, float, cf. to fleet), A. S. fledtan. SJlezhenn (to fly), fleh (flew), A. S. fledgan, fledh. JSlittenn (to remove, ef. to flit), O. N. fletta (to strip). Jloce ( flock), A. S. floc. flor ( floor), A. S. fiér. Jlowenn (to flow), A. S. fléwan. fode ( food), A. S. feda. folighenn (to follow), A. S. folgian. fox ( for), A. S. for. Jrawarrd (away from, cf. froward), A. S. framweard. JSremmde (strange, Sc. frem), A. S. fremde, from prep. fram (from). JSressh (active, fresh), A. S. fersc, O. Fr. fres, fem. fresche. Jresst (period of time, Sc. frest, cf. to frist), A. S. fyrst. Srosst ( frost), A. S. forst. Sul (foul), A. S. fal. full ( full), A. S. ful, full. fulluhht (baptism), A. S. fullukt, from fulvian, fullian, whence probably E. to full (cloth). gaddrenn (to gather), togeddre (together), A. S. gad(o)rian, tégdd(e)re. gan (to go, gone), A. S. gd.a (to go). ganngenn (to go, Sc. to gang), A. S. gangan. gast (spirit, ghost), A. S. gdst. gat (goat), A. S. gat. gatt (he got), in bigatt (obtained), A. S. begeat, from begitan (to get). gazhenn (gain), O. N. gagn (advantage). gessthus (quest-house), A. S. gdst. gillt (guilt), A. S. gylt. biginnenn, -gann, -gunnenn (begin, -gan, -gun), A. Ss. on-ginnan, -gan, -gunnen. gladd (glad), but with -¢ glade, A. S. gldd.
100 7 J. Hadley,
O. gledess (gleeds, glowing coals), A. S. gléd. glowennde (glowing), A. 8. gléwan (to glow). godd (god), but godess {god’s), A. S. god. goddspell (gospel), A.S. godspel, see p. 83. grat (great), A. S. gredt. gredi3 (greedy), A. S. gradig. grene (green), A. S. gréne. gresshoppe (grasshopper), A. S. gaérshoppa. gretenn (to greet), A. S. grétan. grimm (grim, stern), A. S. grim. gripenn (to gripe), in bigripenn (to rebuke), A. S. gripan (to gripe). grissliz (hideous, grisly), A. S. grysitc. habbenn (to have), also hafeun, huffde (he had), A.S. habhan, hdfde. heelenn (10 heal), A. S. helan. hep (heap), A. S. hedp. her (hair), A.S. her, hér. hete (heat), A. S. heeto. hethenn (heathen), A. S. hedhen. hewenn (to hew), A. S. hedwan. hal (whole), A. S. hdl. hali3 (holy), A. S. halig. ham (home), A. S. hdm. — hannd (hund), also hand, A. S. hand. heh (high), A. S. hedh. helle (hell), A. S. hel, helle. hellfe (handle, helve), A. 8. helf, hielfa. hellpenn (to help), A. S. helpan. henngde (he hanged), A. S. hangian, hongian (to hang). heoffne, heffre (heaven), A. S. heofon. her, here (here), A. S. hér. herenn (to hear), A. 8. hyran. hew (appearance, cf. hue), A. S. hiw (Ato ?). hidenn (to hide), A. S. hydan. hill (hill), A. S. hyll. himm (him), A. 8. him. hiss (his), A. S. his. hof (hove, heaved), from hefenn, A. 8. héf, from hebban. bihofethth (behoveth), A. 8. behdfadh. horedom (whoredom), from A.S. hére (whore). huccsteress (huckster’s), Dan. héker, M. H. Germ. hocke, hucker. hunngerr (hunger), A. S. hungor. hunndredd (hundred), A. S. hundred. hunnte (hunter), A. S. hunta. hus (house), A. S. his. husell (eucharist, housel), A. S. hisel. icchenn (to move oneself, cf. to hitch 2). idell (idle), A. S. édel. if = 31f (if), A. S. gif ille (evil, ill), A. 8. yfel, O. N. ill-r.
On English Vowel Quantity. 101
O. illke (same, cf. Sc. ‘of that ilk’), A. S. ilca. inn (in), A. S. in. inn, inne (inn, lodging), A. S. tn, inne. inoh (enow, enough), A. S. gendg, gendh. trenn (iron), A. S. iren. iss (is), A. S. is. itt (it), A. S. Ait. kanndellmesse (Candlemas), A. S. candel-mdase, Lat. candela. kanunnkess (of canonic person, monk’s), Lat. canonicus. kaserr (Caesar, emperor), A. S. cdsere. kelenn (to make cool), A. S, célan (to make cool), célian (to grow cool). kemmpe (Sc. kemp, champion), A. S. cempa. kene (keen, bold), A. S. céne. kepenn (to keep), A. S. cépan, cypan. kinn (kin, kind), A. S. cyn. kinndlenn (to kindle), O. N. kynda. lacchenn (to catch, cf. latch), A. S. ldccan. lade (leading, way, cf. lode), A. S. lad. leche (leech), A. S. lace. lye (belief), A. 8. geledfa. lefess (leaves), A. S. led f. len (reward, cf. loan), A. S. ledn. lepenn (to leap), A. S. hledpan. laste (least), irreg. lengthened, A. S. ldsest, ldst, superl. of lyte/. lewedd (lay, cf. lewd), A. S. lkewed. laf (loaf), A. S. hléf. laferrd (lord), A. 8. hldford. lah (low), superl. lazhesst, O. N. [47-r. lahhzenn (to laugh), also lahg3henn, A. S. hlehhan. lammbre (lambs), pl. of lamb, A. S. lambru, from lamb. lannge (adv. long), from adj. lang, A. S. lange, from lang. bilappedd (lapped, enclosed), A. S. lappa (lap, border). lare (lore), A. S. lar. lasse (less), A. S. léssa, compar. of lytel. lath (louthsome, hateful), A. S. lédh. lattre (latter), lattst (last), from late, A. S. ldtra, ldtst, from lat. lawe (heap, hill, cf. Houns-low, Braddaw), A. S. hlew. ledenn (to lead), A. S. ladan. lefe (leave, permission), A. S. ledf. lefenn (to believe), A. S. gelgfan. bilefenn (to remain), lefethth (r cmaineth, cf. to leave), A. S. lain (to leave). lefftenn (to lift), from A. S. lyjt (air). lenenn (to lend, loan), A. 8. lenan. lenntenn (Lent, spring-time), A. S. lencten (spring). leo, le, gen. leoness, leness, (lion, lion’s,) A. S. led or leén, Lat. leo. leof, lef (dear, cf. lief), A. S. ledf. leosenn, lesenn (to loose), A. S. lysan. Sorrlesenn (to lose), A. S. forkésan. lettenn (to let, hinder), A. S. lettan.
102 J. Hadley,
O. lezhenn (to lie, mentiri), A. S. led7an.
lhude (loudly), A. S hlide, from hliéd (loud).
libbenn (to live), A. S. libban.
lic (like), A. S. gelic
lic, lich (body, corpse, cf. lyke-wake), A. S. /ic.
lif (life), bilife (quickly, belive), A. S. lif:
iif (air, cf. Se. lift), A. S.-ly/t.
likenn (to like, please), A. S. lécian (to please).
lim (lime), A. 8S. lim.
lin (to lie, jacere), A. S. licgan, but lidh (for ligdh).
lisste (it pleased, cf. to list), A. S. lyste, from. lystan.
lisstenn (to list, listen), A. S. hlystan.
little, def. and pl. of litell (little), A. S. lytel.
lithe (gentle, cf. lithe), A. S. lidhe.
loft, in o lofft (aloft), A. S. on lyfte (in air, on high), O. N. @ lopti.
lott (lot), A. S. Alot.
lozhe (flame, Sc. low), A. S. lég, O. N. logi.
luffsumm (lovesome, pleasant), from lufe (lore), A. S. lufsum, from lufe.
lusst (lust), A. S. lust.
lutenn (to bow, Se lout), A. S. liétan.
macche (mate, consort, cf. match), A. S. gemdcca: and in same sense make (make, obs.), A. S. gemaca.
meeless (meals), A. S. mel.
menelike (common, cf. mean), A. S. gemane.
mann (man), menn (men), A. S. man, men.
mare, mar (more), mest, mast (most), A. S. mdra, mest.
mede (meed, reward), A. S. méd.
melltenn (to melt), A. S. meltan.
menenn (to moan), A. S, menan.
menenn (to mean), A. S. menan,
meoc, mec (meek), O. N. miuk-r (lenis).
messe (mass), A. S. mdsse, Lat. missa.
miccle, pl. of makell (mickle, much), AS. micel.
middell (middle), A. S. middel.
mile (mile), A. S. mél.
mille (milk), A. S. mile, meolc.
min, mi (mine, my), A. S. min.
minnstre (minster), A. S. mynster, Lat. monasterium.
misstenn (they missed), A. S. missan (to miss).
mod (mood), A. 8. mdd.
mone (moon), A. S. mona.
muth (mouth), A. S. midh.
narrwe, def. and pl. of naru (narrow), A. S. nearu.
ned (need), A. S. nyd, néd,
neddre (adder, for nadder), A. S. nddre.
nedle (needle), A. S. nedl.
neh (nigh), A. S. hedh, néh.
neow, new (new), A. S. niwe, neowe, with short vowel.
ner (nearly), O. N. ner; also ner (nearer), A. S. nedr, comp. of nedh.
nett (net), A. S. net.
On English Vowel Quantity. 103
O. niththrenn (to lower, ef. nether), A. S. nidherian, from nidher.
off (of), A. S. of.
offrenn (to offer, sacrifice), A. S. offrian, Lat. offerre.
offie (oft), A. S. oft.
ollfenntess (gen., camel’s), A. S. olfend (camel), from Gr. édégas (elephant).
onn (on), A. S. on.
oxe (oz), A. S. ora.
pappe (nipple, pap), Lat. papilla.
paradis (paradise), Lat. paradisus.
pin (pain, Sc. pine), pinenn (to pain, cf. to pine), A.S. pin, pinian.
pitt (pit), in hellepitt (hell-pit), A. S. pyt (puteus, fovea).
ploh (plough), Dutch ploeg, O. N. plog-r.
posstell (apostle), A. S. apostol, Lat. apostolus.
preost, prest (priest), A. S. predst, Lat. presbyter.
racchess (scenting-hounds, cf. rach), A. S. rdce.
red (counsel, Sc. rede), radenn (to counsel, Sc. rede), A. S. red, redan.
refenn (to rob, reave, bereave), A. S. redfian.
ramm (ram), A. S. ram.
rann (ran), A. S. ran, from rinnan (to run).
rannc (rank, luxuriant), A. S. rane.
rap (rope), A. S. rép.
reccnenn (to reckon), A. S. recenian.
recless (incense, cf. reek), also reccless, A. S. récels, from récan, redcan (to reek).
redenn (to read), A. S. rédan.
reddenn (to rid), A. S. hreddan.
rekkenn, reckenn (to reck), reckelws (reckless), A. S. récan, receleds. Vowel shortened in O. and E. .
reowenn (to rue, repent), birewenn, A. S. hreowan.
resste (rest), A. S. rest, rdst.
rhof, rof (roof), A. S. hréf.
ridenn (to ride), A. S. rédan.
rime (number, metre, cf. rhyme), A. S. rim (number).
rippenn (to tear, spoil, cf. to rip), A. S. ryppan (or rypan 2).
risenn (to rise), ras (rose), A. S. risan, rds.
rode (rood, cross), A. S. réd.
rote (root), O. N. rét, A. 8. wrdtan (to root).
ruhh (rough), A. 8. hreéh, hreéw: shortened in O. and E., see p. 74.
rum (room), A. S. ram.
run (secret converse, cf. roun and rune), A. S. rin.
saccles (guiltless, cf. sackless), A. S. sacleds.
sete (seat), A. S. set (insidiae).
sdre (sorely), A. 8. sdre, from sdr (sore).
sawenn (to sow), A. S. sdwan.
sawle (soul), A. S. sdwel, sd.
scrennkenn (to supplant, make shrink), A. S. dscrencan, from scrincan (to shrink).
scribe (scribe), Lat. scriba.
sed (seed), A. S. sed.
sekenn (to seek), A. S, sécun.
104 J. Hadley,
O. sellenn (to sell), A. S. sellan, syllan (to give). sellf (self), A. S. self, sylf. | semenn (to seem), A. S. séman (to make agreeable). sennde (he sent), from sendenn, A. 8. sende, from sendan. bisennkenn (to sink, trans.), A. §. sencan, from sincan. seoffne, seffne, also se(o)fenn (seven), A.S. seofon. Seon, sen (to see), A. S. sedn. serrzhe (sorrow), A. S. sorg. settenn (to set), A. S. settan. settledd (seated, settled), A. S. setlan (to settle). sexe (six), sextene (sixteen), sexti3 (sixty), A. S. six, sixtyne, sixtig. shefess (sheaves), A. S. scedf. shethe (sheath), A. S. scedh, sceddh. shewenn (to show), A. S. scedwian (to view). shall (shall), shollde (should), A. S. sceal, sceolde. shannkess (legs, shanks), A. 8. scanc, scanca. shene (sheen, adj.), A. S. scyne, scéne. shep (sheep), A. S. scedp, scép. shifftedenn (they divided, cf. to shift), A. 8. sciftan (to divide). shinenn (to shine), shan (shone), A. S. scinan, scdn. shir (sheer, pure), A. 8. sctr. shothwang (shoe-latchet), A. S. sced, scé (shoe). shrifenn (to shrive), shriffte (shrift), A. S. scrtfan, scrif. shrud (clothing, cf. shroud), A. S. scréd. sibb (kin, Sc. sib), A. S. sid. sikenn (to sigh), A. S. sican. sillferr (silver), A. S. seolfor, silfor. sinne (sin), A. S. syn. sinnkenn (to sink), sunnkenn (sunken), A. S. sincun, suncen. siththenn (sithence, since), A. 8. sidhdhan (stdhdhan ?). sittenn (to sit), satt (sat), A. S. sittan, sdt. skill (skill), O. N. skil (distinction, distinct notion). skinn (skin), A. S. scinn, O. N. skinn. ° slap (sleep), slepenn (to sleep), A. S. slap, slepan. slan (to slay), sloh (slew), A. S. sledn, sléh. slaw (slow), A. 8. sidw. sleh (cunning, sly), O. N. sleg-r. sloth (track, cf. sleuth-hound), O. N. siédh. smacc (savor, smack), A. S. smdc. smec (smoke), A. 8. sméc, from smeécan. smethe (smooth), A. S. smédhe. smitenn (to smite), A. S. smitan. sone (soon), A. S. séna. soth (sooth), A. S. sédh. speche (speech), A. S. sprec, spec. sped (speed), A. S. spéd. spell (utterance, cf. spell), spelidrenn (to spell), A. 8. spel, apellian. steep (steep), A. S. stedp. staff (letter, cf. staf’), A. S. s#ta/f.
On English Vowel Quantity. 105
O. stallwurrthliz (stoutly, cf. stélwart), A. S. stdlweordh.
stan (stone), A. S. stdn.
stanndenn (to stand), A. S. standan.
steorenn, sterenn (to direct, steer), A. S. stedran.
stikkess (sticks), A. 8. stycce, sticce (piece).
stille, still (still, quiet), A. S. stille.
stinnch (stench), A. S. stinc.
stinnkenn, stannc, stunnkenn, (to stink, stank, stunk,) A. 8. stincan, stanc, stuncen.
stinntenn (to cease, cf. to stint), A. S. stintan (to blunt).
strem (stream), in waterrstrem, A. S. stredm.
strete (street), A. S. street, Lat. strata (via).
strenncthe (strength), from strang (strong), A. S. strengdhu, from strang.
streon, stren (race, cf. strain), stre(o)nenn (to beget), A. S. streénan (procreare).
stunnt (stupid, cf. stunt), A. S. stunt (blunt, stupid), from stixtan.
suhhghenn (to sigh, cf. sough), origin uncertain.
summ (some), A. S. sum.
sunne (sun), rarely stine, A. S. sunne.
sur (sour), A. S. sir.
susstress (sisters), A. S. sweostor, swustor.
suth (south), A. S. sudh.
swelltenn (to dic, cf. to swelt, swelter), A. S. sweltan.
swet (sweet), A. S. swete.
swift (swift), A. S. swift.
swille (such), A. S. swylc, swile.
swin (swine), A. S. swin.
swinne (labor, O. E. swink), A. S. swinc.
swollzhenn (to swallow), A. S. swelgan.
tacchenn (to teach), A. S. tecan.
tamenn (to generate, to teem), A. 8S. tyman, téman, from tedm { progeny).
tress (tears), A. S. tedr.
takenn (token), tacnenn (to betoken), A. S. tdcen, tdenian.
tel‘enn (to tell), A. S. tellan.
temmple (temple), A. 8. tempel, Lat. templum.
temmpredd (tempered), A. §. temprian (temperare).
tende (tenth, cf. -teenth in thirteenth, etc.), A. S. teédha, O. N. thundi, Frisic tianda: cf. also Sc. teinds (tithes).
tene (ten, cf. -teen in thirteen, etc.), A. S. tyn, tén.
tene (injury, vexation, O. E. teen), A. S. teén, teéne.
ther (there), A. S. ther.
thann (than), A. S. thonne.
thanne, thann (then, when), A. S. thonne, thénne.
thannkenn (to thank), A. S. thancian.
thatt (that, conj.), A. 8. that. ,
thennkenn (to think), A. 8. thencan, thencean.
theos, thes (thighs), A. S. thedh, thes.
thin, thi (thine, thy), A. S. thin.
thinnkethth (secms, cf. methinks), A. S. thync(e)an, thinc(eJan.
thiss (this), A. S. this.
thrapenn (to chide, Sc. threap), A. S. thredpian.
14
106 J. Hadley,
O. threfald (threefold), A. 8. thrtfeald.
threnngdenn (they thronged), A. S. threngan, from thringan.
thi esshenn (to thresh), A. S. therscan.
thrifenn (to thrive), thraf (throve), O. N. thrifask, threifsk. thusennde (thousand), A. 8. t/.tsend.
thuss (thus), A. S. thus.
tid (time, cf. tide, even-tide), A. S. tid.
till (till), O. N. til.
time (time), A. S. téma.
timmbrenn (to build, cf. timber), A. S. timbran, from timber. tithennde (tidings), from A. S. tidian (to betide, befall).
towarrd (toward), A. S. téweard.
tradd (he trod), from tredenn (to tread), A. S. tré 1, from tredan. trapp (trap), A. 8. treppe.
tun (town), A. S. tun.
twafald (twofold), from twa (two), A. S. twd.
twellf’ (twelve), A. 8. twelf.
bitwenenn (between), A. S. betwednan.
twennti3 (twenty), A. §. twentig.
twinne (twin), A. S. getwinne.
unn- (un-, negative prefix), A. S. un-. .
unnderr (under), A. 8. under.
upp, uppe (up), A. S. up; O. upponn (upon), A. 8. uppan.
ure (our), A. S. ure.
uss (us), A. S. ds; shortened in O. and E.
ut, ute (out), A. S. vt, tte; O. abutenn (about), A. S. bitan. utterlike (outwardly, cf. utterly), A. 8. % or, uttor (outer, utter). wac (weak), A. S. wdc.
wede (clothing, cf. weed), A. S. wed, wade.
wannt (wanting), O. N. van-t, neut. of van-r, A. 8. wana (defect): hence O.
wanntethth (wanteth).
wass (was), werenn (were), A. S. wis, weron.
wasshenn (to wash), A. 8. wascan.
wawenn (woes), pl. of wa, A. S. wdwan, pl. of wdéwe, from wd (adv. and in-
terj. wo).
waxenn (to war, grow), also warxenn, A. 8. weaxan.
wecche (watching), A. S. wéiece.
weddenn (to wed), A. 8. weddian.
well (adv., well), also wel (Sc. weel), A. S. wel (bene).
welle, dat. of wel (a well), A. S. well and wella (fons).
wenenn (to ween, think), A. S. wénan.
wennchell (child, cf. wench), A. S. wincel (proles), wencle (ancilla). _ wennde (he went), from wendenn (to go, wend), A. 8. wende, from wendan. wepenn (to weep), A. S. wépan.
wesst (west), A. S. west.
wher (where), A. S. hwer.
wheete (wheat), A. 8. hwete.
whanne, whann (when), rarely whane, A. S. hwonne, hwdnne. whatt (what), A. S. Aswdt. |
whellp (whelp), A. 8. hwelp.
On English Vowel Quantity. 107
O. wheol, whel (wheel), A. S. hwedl.
wheththr (whether), A. S. hwddher.
whil (while), whilumm (whilom), A. S. hwill, dat. pl. hwtlum.
whille (which, Sc. whilk), A. S. hwyle.
wicche crafftess (witchcrafts), A. S. wicce (witch).
wicke, wikke (wicked), probably connected with A. 8. wiccian (to bewitch).
wid (wide), A. 8. wid.
widdwe (widow), also widewe, A. S. widwe, wydewe.
wif (woman, wife), A. §. wif.
wiless (wiles), A. S. wil.
wille (will, subst.), A. S willa.
willt (thou wilt), wollde (would), from wilenn, A. S. wilt, wolde, from willan.
win (wine), A. 8. win.
winndwenn (to winnow), A. S. windwian.
winnenn (to toil, win), wann (he won), wunnenn (won), A. 8. winnan, wan, wun- nen. -
winnterr (winter), A. S. winter.
wis (wise, prudent), A. S. wis.
wise (wise, manner), A. 8. wise.
wiss (certainly, cf. J wis, O. E. ywiss), A. 8. gewis.
wisste (he wist, knew), from witenn (to wit, O. E. to weet), A. S. wiste, from witan.
withth (with, prep.), A. S. widh (against, towards).
withthess (withes), A. 8. widhig (willow), O. N. widhi-r; vowel shortened in QO. and E.
witt (knowledge, wit), A. S. wit; also O. witt (know thou), A. S. wit.
wittness (witness), A. S. witnes (witn s ?).
wokenn (they woke, watched), from wakenn, A. S. woécon, from wacan.
wrath (wrdth, adj.), A. S. wrddh: for wraththe (wrdth), see p. 83.
wrecche (wretch), A. S.w deca, wrecca.
wrejzenn, wrezhenn (to accuse, bewray), A. S. wrégan.
writenn (to write), wrat (wrote), A. S. wrttan, wrdt.
writt (writ), A. S. writ.
wulle (wool), A. S. wull.
wunnderr (wonder), A. S. wundor.
3ellp (boast, cf. yelp), A. 8. gilp, gelp.
ger (year), A. S. 4 dr.
gett (gets), from 3etenn, A. 8. git, from gitan.
3H (give), impv. of 3ifenn, A. 8. gif, from gifan.
ff (if), also iff, A. S. gif.
gol ( Yule, Christmas), A. 8. gedl
jonnd (yonder), bizonndenn (beyond), A. 8. geond.
gunngre (younger), from 3ung, rarely 3unng, A. S. gyngra, from geong.
jure (your), A. S. edwer.
juw (you), A. 8. dw.
V.—Anglo-Sazon and Early English Pronunciation. By FRANCIS A. MARCH,
PROFESSOR OF THE ENGLISH LANGUAGE AND COMPARATIVE PHILOLOGY IN LAFAYETTE COLLEGE.
The Anglo-Saxons, Old Saxons, and Icelanders mark off their poetry into verses by the rhythmic repetition of letters. The most common form of it is alliteration, the repetition of the same initial sound in the first accented syllables of certain words.
The Common Narrative verse in Anglo-Saxon consists of two sections, and in a perfect verse two syllables of the first section and one of the second begin with an alliterating sound. If this be a consonant sound, it must be the same in all the syllables ; following consonants of a combination need not be repeated, though sc, sp, and st usually go together.
It would seem then that alliteration may furnish some evi- dence of the pronunciation of the consonants in Anglo-Saxon and Early English, first, as to what consonants are sounded alike; secondly, as to the order in which the consonants of any combination were uttered.
Mr. A. J. Ellis in his great work on Early English pronun- ciation gives the sounds of all the letters in Anglo-Saxon and English, and wishes to bring together all the evidence con- nected with the subject, but neither he nor any of his critics, so far as I have seen, have directed their attcntion to allitera- tion. There are several combinations as to which the evi- dence from this source seems to be decisive against his repre- sentation of their sound.
1. The initial combinations Al, hn, hr, hw, lie believes to have been pronounced lh, nh, rh, wh. But these are frequent alliterating letters all through the Anglo-Saxon poetry, and regularly alliterate with A alone, or with other combinations beginning with 4. I will cite a single cxample from each of the three great sources, Beowulf, Cedmon, and Cynewulf.
Hl. Weé purh holdne hige hildford pinne. Bedw. 267.
and té6 heofnum up hiledre rérdon. Cad. 1675. hleatre behworfen, ah in hele ceafl. Andreas, 1705.
On Anglo-Saxon and Early English Pronunciation. 109
Hn. heord-weordunge hndhrun rince. Bedw. 952. Gyld mé mid hyldo pet ic pé hnedw ne wes. Ceed. 2823.
Hr. habban on healse hring-weordunge. Bedw. 3017. habhan wolde, und rh é6fgefo:. Cred. 1360. hide for hergum: hrefn weorces gefeah. Elene, 110.
Hw. efter heleda hryre hwate Scyldingés? Bedw. 2052. heft mid hringa gespanne. Hwearfhim eft nider. Ceed. 762. hdlig under hrusan, pe gé hwile nu. Elene, 625.
Such is the regular alliteration. There are a few lines in which Aw may alliterate with w. Thus in Psalms, lxxxii. 10, ‘the three words wegnes : hweél :: windes occur, and in Psalms, cvi. 28, windes : hweodu ::. weordad, where hw may be the second alliteration in the first section; but it is to be noted that a very large number, more than a third of all the verses in Anglo-Saxon poetry, have but one alliterating word in the first section, and that these verses are therefore com- plete without counting the hw as a rime.
There are also a few lines having both w and A as possible alliterations to hw, where w is in the foot which oftenest rimes:
hwate Scyldingds ; gewdt him hdm ponon. Bedw. 1601.
In the following verse there is no alliteration except be- tween word and hwile, or very light particles:
Ne mdgon gé pa word gesédan pe gé hwile nu on unriht. Elene, 582.
It would be easy to emend the few lines of this kind; but they are not numerous enough to throw doubt on the cur- rent pronunciation, and I am inclined to believe that some of them do really represent an exceptional pronunciation of the poet or scribe who gave the poem its last touches. There are other indications that in some dialects 4 was going silent. A few verses are found where a word beginning with a vowel alliterates with one in A followed by a vowel :—
se pisne Ar _—sihider onsende. Andreas, 1606. ealdum ceorle hondslyht giofan. Beodw. 2972.
See also more examples cited by Heyne, Beowulf, page 102, note on line 2930, none of which however are as sure, as might be wished, as, indeed, is the case with those here cited. On the whole, the handful of irregularitivs in the riming of hw and kh serve to assure us that the poets really followed the sound as they heard it. |
110 F. A. March,
Mr. Ellis also holds that in the initial combinations of 2, the w is secondary ; that wi, wr were / and r labialized as in French loz, roz. But w is the alliterating letter :—
wépnum gewurdad. pd pdr wlonc heled. Bedw.331. éd pet for wlence and for wonhygdum. Ced. 1678. wigspéd wid wrddum purh pet wlitige tred. Elene, 165. Wederd peéden wrece leorné:le. Bedw. 2336.
on }-ds wer-peéde wraccan laste. Crd. 2822
and pd wig-preece on gewritu setton. Elene, 658.
To this evidence from Anglo-Saxon may be added that from Old Saxon and Old Norse; all these combinations are com- mon in the Heliand, and those beginning with 4 are common in the Edda, and the alliteration is the same as in Anglo- Saxon. Single examples will suffice. From the Heliand (ed. Heyne, 1866) :—
endi mid hluttrun trewun. Ward thé the hélago gést. 291.
holdan hérron, hnég imu tegegnes. 2419.
an thea hétun hel hriwig-méde. 4448.
Ak thea moétun hwerban an that himiles lioht. 1920.
umbi thiu word an gewinne: stédun wlanka man. 3928.
theréd wdr-sagoné word thém wrekkiun sagda. 631. From the Edda (ed. Liining, 1859) :—
Hliods bid ek allar helgar kindir. Voluspa, 1. er vil hérmug tvau haigum at rinum. Gudrun, 3. 4. Valdi henni Herfodr hringa ok men. V6 uspn. 23. Hvern létu peir héfdi skemra. Hmiskvida, 15.
When we consider how important a part alliteration plays in the history of these tongues, how old and deep its work is, it seems necessary to admit that these peoples must have heard the alliterating sounds as first and most emphatic in their combinations, and that for all literary purposes, at least, transliterations of the old texts should preserve the priority of the rimming letters.
In the Early English alliterative poems, as fast as these combinations change their pronunciation, the words with - which they alliterate change. The 4 of Ad and hr is no longer written, and lord, of course, alliterates with 7, ring with r. WA alliterates with w:—
And pere mizte pow wite, if pow wolt which pei been alle. Vision of P. P. ii. 44. And with him to wonye with wo while God is in heuene. ii. 106.
On Anglo-Saxon and Early English Pronunciation. 111
I should expect the Anglo-Saxon alliteration of Aw to be preserved in Old English of the northern dialect. In the latest alliterative poems we still find wl, wr, riming with w:
Lo, suche a wrakful wo for wlatsum dele3. Deluge, 307. That I mizht worchen his wille pat wrouzht me to man. Vision of P. P. i. 82.
This brings wr down to Chaucer’s time : He was a wel good wright, a carpenter, Cant. Tales, Prologue. 614.
A similar examination of Anglo-Saxon and Early English alliteration assures us of the pronunciation of cn = kn, gn, and other initial combinations, about which however there is, I believe, no difference of opinion.
There is some question as to the sound of c, se. g, p, d, and the rune called wen, which the Germans generally transliter- ate by v, and the English by w.
It has been thought by some that ce was sounded like ch, that in ceaster (English -chester), for example, from Latin castrum, the ce is the Celtic notation for ch. But the allitera- tion shows that the poets did not discriminate it, nor c before t, y, from the regular c =k:
-peet sid ceaster hider on pds cneorisse. Andreas, 207. of Caldéa_ censtre dlédde. Crd. 2200. his cymé kalend ceorlum and eorlum. Menolog. 31.
and so often. But in Piers Ploughman such words are spelt with ch and alliterate only with each other or French ch :
For in charnel atte chirche cherles ben yuel to knowe. vi. 50. And pat chaude or plus chaud _for chillyng of here mawe. vi. 313.
while ¢ before e, t, y, in French words alliterates with s.
Ac panne swore Symonye and cyuile bothe. ii. 168.
Se when followed by a parasitic e is by many pronounced sh. But in Anglo-Saxon poetry all sc’s are alike :
Swd scridende gesceapum hweorfad. Traveler, 135. scearp scyldwija gescdd witan. Bedw. 288.
while in Early English such words alliterate only with each other or with 8:
I shope me in shroudes as I a shepe were. Vision of P. P., Prol. 2. Mowe be siker pat her soule shall wende to heuene. i. 130.
112 F. A. March.
There are some words in which ge represents ¢consonant (English y) of other languages: geong has corresponding words in Sanskrit, Latin, Gothic, German, beginning with the sound y; in English it is young and in Anglo-Saxon it is sometimes spelt tng. But this word, like other similar words alliterates freely with all words beginning with g:
geong : geardum :: God. Beow. 13. gedgode :: gledwést. Cad. p. 221, 1.
And so other words, foreign proper names especially, even
retaining their spelling with ¢; lacobes :: géde. Psalms, Ixxxvi. 1. lordane :: gréne. Ceed. 1921.
In Early English a large number of words with the.y-sound initial came in from the Norman and alliterated with each other. The Anglo-Saxon stragglers varied awhile, but finally went over tothe Normans. It seems probable therefore that g in these words, as certainly in others, was intended to rep- resent some palatal sound, either the same as g in go, or so like it as to have not been clearly discriminated from it.
It has been generally believed that d represents the sonant th = dh which we have in thine, other, smooth, and the rune p, the surd in thin, loveth. But it seems to be admitted that so - far as actual use in the manuscripts is concerned, they are mere calligraphic variations, like the two forms of Greek sigma; and we find that the words which in English begin with dh alliterate freely with those which begin with th: pine (thine) alliterates with ge-peahtunge (thinking), Psalms, liv. 13; pene (that), with purstige (thirsty), Ps. Ixi. 4; ponan (thence) with Zhrdcia (Thrace), Meters, 26, 22 ; and the like.
The regular shifting is from surds to sonants. Wherever the English has surd ¢# the presumption is that the Anglo- Saxon also had it; but the English sonant is a natural weak- ening of a surd, and is of no grect weight against the evidence of alliteration.
A surd and sonant do not combine in the same syllable. We may be sure that the ending of the third person singular of the indicative present of syncopated forms of surd roots is surd: pincp (thinketh) and like words, end in a surd &h.
On Anglo-Saxon and Early English Pronunciation. 1138
There are other euphonic laws which indicate that th rather than dh was the common sound in Anglo-Saxon, final as well as initial. (See March’s Comparative Grammar of the Anglo- Saxon, § 194.)
There is no Latin word in Anglo-Saxon poetry beginning with vorw. The general law of shifting is in favor of the sound of English w rather than the German, since the German w is a weakening of the English. The sonant continuous labial represented in Anglo-Saxon by f between two vowels, and which has changed to v in English, probably resembled the present German w == bh, an unstable transition sound from b to v. |
Alliteration throws little light on the vowels. The poets prefer to have riming vowels differ. There seems little room for any wide divergence from the sounds accepted in Germany and America, and set forth with such accumulation of proofs by Mr. Ellis. The laws of change in vowel sounds make it doubtful whether the pure a sound ever obtained any general currency among’ people of Saxon traditions in words where we find it already weakened in Anglo-Saxon. The Normans perhaps did not adopt in their speech, and did not try to ex- press in their spelling those modifications of the @ sound which the Anglo-Saxons distinguished by @, ea, e, 0, @; but we know that these very tendencies to change a towards e on the one hand and o on the other have been active and modi- fied a large’ part of the language, and it is likely that the progress has on the whole gone on ever since the Anglo-Saxon times, and that grammars and dictionaries which represent the Roman a as in general use in such words in the reign of Elizabeth, speak rather for the court and the college than for the folks at home.
15
VI.—Some Notes on Ellis’s Early English Pronunctatton.* By CHARLES ASTOR BRISTED
One volume of this book was noticed in the North Ameri- can Review rather more than a year ago, since which time two of the remaining three volumes have appeared.
Our first feeling in connection with the work is one of ex- cusable national pride. It is, partially at least, a publication of the Chaucer Society, and one of the principal supports of the Chaucer Society is Professor Child of Harvard, who had already so well merited of the republic of letters by bringing about the Percy Reprint. And it is gratifying to find our countryman’s deserts so freely and gratefully acknowledged in England. The current report of the Chaucer Society says: ‘The active help of Professor F. J. Child, of Harvard, has been continued ; to him we owe nearly all our fresh mem- bers. The enlightenment of the managers of public libraries in the United States contrasts favorably with the disregard of our work by the librarians, or rather library-committees of Great Britain and Ireland. The society has twenty public libraries subscribing to it in the United States, against eight in England, one in Ireland, and none in Scotland.”’
I think we ought to rejoice in these things and to give them full prominence without being deterred by any bugbear of mutual admiration or self-praise. The American philologist has an up-hill task. He is much less encouraged by both the general public and individuals than the professors of the posi- tive and mechanical sciences are. Ile finds fewer like-minded men. He needs all the support he can get.
On the very first page of the first chapter of Mr. Ellis’s
* On Early English Pronunciation, with especial reference to Shakspere and Chau- cer, containing an investigation of the correspondence of writing with speech in England, from the Anglo-Saxon period to the present day, preceded by a systematic notation of all spoken sounds by means of the ordinary printing type. Including a re-arrangement of Professor F. J. Child’s Memoirs on the Languaye of Chaucer and Gower, and reprints of the rare tracts by Salesbury on English, 1547, and Welsh, 1567, and by Barcley on French, 1521. By ALEXANDER J. Exxis, F.R.S.,F.S. A., Fellow of the Cambridge Philosophical Society, formerly Scholar of Trinity College, Cambridge, B. A. 1837, &c., drc., §rc.
On Ellis’s Early English Pronunciation. 115
book we find a statement which deserves wide circulation, as it helps to correct some erroneous ideas which have lately come into vogue. ‘‘ We speak,” says Mr. E., “‘ of the ‘dead’ languages of Greece and Rome, unconscious that our own English of a few years back has become as dead to us.”’ This is perfectly true, and true in a wider sense than our author uses it, because he had only occasion, for his purpose, to em- ploy it with reference to pronunciation and idiom. We have often heard it asserted within a few years, and especially by what I may call the phonetic interest, that the difference not merely between present and early English, but between the former and Anglo-Saxon, is almost entirely and solely a differ- ence of orthography. Of which popular error I hardly think it necessary to go into a deliberate and detailed confutation before this body. No one with even a moderate knowledge of the subject, having before his eyes Havelok the Dane and King Henry II.’s Proclamation (to go no further back than the 13th century), could seriously and honestly maintain such a proposition.
But before examining the body of the work we must care- fully scrutinize the introduction, containing the palaeotype, the author’s system of notation and literation for scientific pur- poses.
It is called palaeotype because the old (Roman) types are used in it, without accents or diacritical marks, though some of the letters are turned or reversed. The elements called letters (though some of them are combinations of as many as five letters) reach the high number of 270, besides which there are nearly fifty tonic signs, used as suffixes. These ele- ments are intended to express all known sounds, not only in the Indo-Germanic but also in the Asiatic and African Jan- guages.
While giving due praise to the remarkable industry and learning displayed in this system, we cannot avoid regarding it as somewhat cumbrous and overdone. The letters and com- binations are not in all cases happily chosen. @ for our ng has an odd and repulsive effect and ¢ for the Spanish d (which is either d@ or dh flat, or th sharp, according to position, and
116 oo C". A. Bristed,
certainly ‘needed’ no separate sign,) may very well mislead. But the most puzzling is the.ae diphtheng. Mr. Ellis uses this, so far as: | can make him out,* to express the shortest or very nearly the shortest possible: a,,something shorter than any or- dinary continental form of the letter, and doubles. it (@@) to express a prolongation of this, sound, not so long or full as the alphaf in words like gar. Now, in spite of. its occurrence as.an Anglo Saxon alpha, the diphthong @ is apt to he associ- ated in our minds with the-sounds of (English) @ and é, rather than with any modification of alpha. (Thirty years ago, some of the professors and tutors of Yale College, and prob- ably of some other New England colleges, did most barbar- ously shorten this diphthong where it. occurred in Latin ante- penults, but they made it é not d.) Mr. Hllis’s use of @ has confused in one place the writer of the notice in the North Amer- tcan, who takes his double diphthong, for (English) d@; and I ence thought it had deceived Mr. Ellis himself, in writing pa- kaeo-typically the Scotch word plaid, till I found afterward that he really supposed the word to be pronounced pladd—a singu- lar slip for one of his dialectic knowledgs and experience.t But these are trifles; the main objection is to the whole theory of the scheme, ‘which is like an attempt to exhaust the inexhaustible. Take, as the first case, the varieties of alpha from the shortest and thinnest (say English @) to the long- est and broadest (say German d@) which passes gradually into (English) aw. Between @ and awa delicate musical ear might detect an almost infinite number of shades, especially when the modifying power of following consonants is taken into ac- count.§ Mr. Ellis himself says, that “if we descended into
* He gives ae illustration man, cat, sad. But the vowel of cat is surely not the alpha of man / it is rather the first syllable of manners.
t Iuse alpha as a convenient term througnont for the continental or non-Eng: lish. power of a. ° _ See Scott’s rhymes, laid, plaid, maid, On 1 a subsequent re- reading of the book, the conviction has been forced upon me, that Mr: Ellis is sometimes con- futed by his own notation and that he has brought tegether under his «, ea, gounds as different as ordinary English & and a. __
§ RX, according as more or less strongly sounded, has a marked effect on a pre ceding vowel. Most of us are familiar with the Irish pronunciation of horse. ‘Many, perhaps most persons would express this in English letters by harse+ it is
On Ellis’s Early English Pronuniaction. 117
every. minute shade:of spoken:sound; thé variety would be so interminable that all definite character would be lost’’ (ch. i, p- 19), but if he does not grasp at these tnterminable varieties, he strives to admit as many of them: as possible, and im par- ticular to make out for each language.its own distinct set of vowel sounds, thus increasing (as it seems to me) the already sufficiently great difficulty of : ‘acquiring foreign. tongues.”
' My own idea, I confess, is the reverse of this. In the in- terest of practical philology, I always try to'bring together as many sounds of different languages as I ean, making small account of the inadequate and often erroneous rules given in grammars and dictionaries. We may often find an equivalent which, though perhaps not scientifically accurate, is sufficient for practical purposes. Thus our.own é may not strictly cor-
generally thus printed by those who undertake. to represent the Irish patois. Others might compare it to the proper name Horace with the a elided, and might write it hor’ce or horrse, which last is the combination I should adopt to explain it to a foreigner who understood English. But maay persons, as I have said, hear the vowel as d. ‘The Bath pronunciation mentioned by Mr. Ellis seems akin to the Irish. He was asked for a piece of card as he supposed, when the ap- plicant really wanted a cord. It is to be regretted that he does not state whether the people of Bath pronounce the r more strongly than Englishmen generally ; if they do, it explains the modification of the vowel., Last winter a lady who is an excellent linguist was reading or declaiming French verse to a small circle. I noticed that she gave a peculiar sound to e before r, e. g. instead of perdu she said pardu, or something so near pardu' that I could not express it otherwise either in speech or writing. When her attention was called to the peculiarity (of which she was of course unaware), she suggested what appeured to be the true reason of it, namely, that in endeavoring to avoid the common Anglo-Saxon fault of not sounding the continental r fully, she laid a little extra stress on it, and thus modified the preceding vowel. On the other hand, the suppression of r also modifies the vowel before it. Vide New England gal and gaa for girl, and young- NewYork /ést for first. My German gardener, calls horses, hésses. ‘These are cases of the vowel’s being thinned. In other well-known instances, the labializa- tion or complete elision.of the r broadens it.
- #In the distinction, elsewhere noticed, which Mr. Ellis so often and strongly
msists on, between our we and the French out, our 7 and ou and the: continental gi and au, he has no lack of support; nevertheless I cannot agree with him as to the amount, importance, and nature of the differences. It seems to me that the diphthongal sounds are the same, .that is, have the same elements; but we pronounce them more quickly and compactly, .it being i in accordance. with the genius of our language to compress and condense diphthongs.
118 C. A. Bristed,
respond to the French é and its general equivalent at diph- thong ; but it is so near that we may consider it an equiva- lent. Our numeral ten does not absolutely reproduce the name of the distinguished French critic, but it is so near that no Frenchman would misapprehend it, and by adopting this pronunciation we avoid the common Anglo-Saxon error of calling him Zane, a sound which one of his countrymen might possibly misunderstand. The name of the French poet and novelist who died a few years ago, may not be exactly our word merry with the accent reversed, but it is very nearly so, and by calling it mer-ry’ we avoid the danger of pronouncing it like our proper noun Mary with the accent reversed.
But Mr. Ellis’s minute subdivision is not confined to the vowels. He has an odd theory of labialized letters and every initial consonant followed by the digamma sound he considers a distinct sound, not merely g, but dw, fw, pw, rw, &e. The writer of the notice in the Worth American has expressed his dissent from this conceit, but a word or two more may be said of it. Most of these new elements are formed by taking to pieces the French diphthong oz (Eng. waa or waw), in what I must call a most arbitrary manner. Take roz for in- stance: what ground is there for connecting the digammatic sound with the initial consonant? None whatever. It be- longs to the following vowel. Take away the initial r, and you still have the sound waw, as in ote or the first syllable of oiseau. Take away thet or the o, and replace the r as in roseau, ristble, and what becomes of the digamma? Mr. Ellis makes out a great difference between our we and the French out, which I consider altogether imaginary, so far as the hearer is concerned, which is the main point. For here, to parody the Eastern mot, “the speaker is one and the hearer is another,” though we cannot continue the quotation and say, in all cases, that “there is no harm done.’’* Mr. Ellis’s view is what I should call too subjective. He is guided by the manner in which the sound is formed, the position of the
* I once saw an Englishman’s pronunciation of ou: expressed in French comic writing by wi. But this must have been a mistake, the French diphthong ut being one of the most difficult sounds to an Anglo-Saxon.
On Ellis’s Early English Pronunciation. 119
vocal organs. But what practically concerns the philologist in such cases is the effect rather than the cause. The Ger- man or the Charleston w may be formed by a position of the organs different from that which produces the ordinary v. I can only say that it sounds to me and to most persons like »v, and I do not see the good of calling or writing it bh.
Apart from these general objections, I think special excep- tions should be taken to two particular points in the litera- tion, namely the notation of t and ow, and of ch. The first ele- ment of the dipthongal sign #, and the diphthong ow (or ow) he expresses by a turned e; but what sound he intends is doubtful. In the equivalent within brackets he makes it @ or a slight modification of d, to which there is no particular objection; but the example which he gives is the indistinct short vowel of du¢; and he illustrates a modification of the sound by the French e (muet) and eu. What analogy, not to say identity, is there between these indistinct sounds, (which we may roughly call z) and @? It may be remarked here that Mr. Ellis, with what I consider his usual superfluity of distinction, finds two if not three different vowel sounds in the pronoun J, the noun eye, and the affirmative ay.
Our ch’ he represents by tsh, as the French and German grammars usually do. I am convinced that this -notation, ex- cept for final ch, is erroneous, and that the old English ortho- epists, wrong in 80 many respects, were right when they gave é: or ty as the continental representative of the sound.
If we begin a word with ¢sh, this combination has a tend- ency to form a syllable by itself; thus T-s-h-a-r-l-e-s would naturally be not Charles, but Tisharles, a dissyllable. If I were trying to give a Frenchman an idea of our ch in chair, I should refer him tothe French proper name Thiers (the name which so many people absurdly call Tears); it is not the exact equivalent of Thiers, but it comes very near.
The proper notation of medial ch may be explained by these three combinations of English words:
hurt shoe, hurt you. her chew (chew pronounced like first syllable of choosing).
120 sO. A. Bristed,
I think every. one will.admit that the:second of these com binations is nearer to the third than the first is, and that it slides into: the third imore readily than the first does.
Final eh alone is-accurately represented by ésh.
The same observations apply mutatis mutandis to the nota- tion of our g soft andj.” 6: :
Some of Mr. Ellis’ foreign transliterations, as hinted in the palaeotypic scheme and expanded in the body of the book, strike me as very odd. He considers French ¢ the representa- tive of our %,-and would transliterate pity in French by pété. One is curious to know how he ‘would transhiterate petty. I have already expressed my conviction (founded on a ten years’ residence in Paris,) that our’é is the proper represen: tative of the French é and its equivalent (in most positions) ai. Thus the French word for wing (atle) is virtually our letter U.
He says that the Spanish ‘guttural 2, or in | modern orthog- raphy 7, becomes eh (that is our sh) in French. Evidently Mr. Ellis never drank a Kéres Koblere in Paris. The jota becomes *& in French ; comparative philology supplies many analogies to this change.
In Italian he has actually transposed the o aperto, and o chiuso making the former o and -the latter aw. He also supposes the Italian ex to be.a diphthong like the Spanish ew, forgetting that-there are no diphthongs in Italian and that Europa is as much a quadrisyllable as paura is a trisyllable.+
ee — ee a
* These views called forth, as I ‘had anticipated, strong opposition. Professor Schlegel went so far as to deny any resemblance between the initial ti of the word tiens and our initial ch, even after I had shown him in Mr. Van Name's paper on the Creole Dialects that some of the West-Indian negroes had made out of tens bon a verb chamber or chomber “ to hold.”
Mr. Ellis himself, while maintaining that the two sounds are “ quite distinct,” admits that the passage from one to the other is “ very short and swift ”’ (p. 205),
Practivally, I put the case thus: When a given sound does not regularly exist in a language, but may be approximately expressed by two combinations of let- ters, one unfamiliar, the other familiar, by which should we transliterate it? And I answer without hesitation, the latter. Hven when the letter power exists but is not common in the connection, I prefer to use a more familiar combination, e. g. to transliterate keg in French by quégue.
t If any exception to this rule can be found, it is o¢ which sometimes comes very near to our own of (and in the most common words, such 4s vo!) much nearer indeed than Mr. Ellisswould probably admit.
On Ellis’s Early English Pronunciation. 121
The awful Dutch guttural sch he admits as a soft sound, on Dutch authority (one would suppose the Dutch easily pleased in the matter of soft sounds,) apparently because in some dia- lects it is pronounced like shr.
If 1 am blamed for dwelling so long on this introductory subject of letter-notation, I can only reply that the palaeotype is a main thread of the work, running all through it, and con- stantly attracting attention, and I shall therefore take the liberty of going on still further in the same direction, and noting certain points where Mr. Ellis is in the main right, but not so clear or full as he might be, and where his obser- vations suggest additional remarks.
In reference to the peculiar Welsh 72, he says that it is not the least like thl or shl, which sound all Englishmen give it. He is right in saying all Englishmen, for though Shakespeare wrote Fluellen for Llewellyn, the th and f are so closely con- nected, that Shakespeare’s orthography cannot be justly con- sidered an exception. How all Englishmen come to make this mistake he does not attempt to show, nor is his explana- tion of the true sound as intelligible as it is elaborate. It seems to be ¢l preceded by a click or cluck, which modifica- tion must be sometimes inaudible, since (as he tells us) the word dlan, when shouted from a distance, sounds like ¢lan.
The French Wl mouillé (ll preceded by ¢) is merely a longz or y. Whatever may have been the case once, there is now no sound of / remaining in it. This, Mr. Ellis admits in the third volume, though he does not always write in the other volumes as if he were quite sure of it. If any one doubts the fact he may be convinced by these two circumstances, that French comic writers use the lJ mouillé to express an exagger- ated and affected pronunciation of 7, e, g, patrille for patrie ; secondly, that half educated persons frequently substitute it for y in writing; thus a porter will spell the participle of avoir, aillant instead of ayant.
The Spanish liquid @J seems to be in a transition state. When medial, as in Caballo, the J is certainly sounded ; when initial, as in llamar it is frequently dropped. The Italian gi always, I perleres in good Italian, preserves the J.
6
e
122 O. A. Bristed,
I will conclude my remarks on this part of the subject with some observations on the Spanish d. Mr. Ellis, on the au- thority of some recent Spanish grammarians, gives to this letter, whether initial, medial, or final, the sound of ¢h flat. His example is deidad, pronounced dheidhadh. The old grammars (1 mean those in vogue forty years ago) give three different sounds to the Spanish d, according as it is, first, int- tial or following a consonant; secondly, medial after a vowel ; thirdly, final. In the first case the pronunciation is the same as that of the English letter, in the second our ¢A flat, in the third our th sharp. So that each d in deidad (deidhath) would have a different sound. Now which of these rules is right, or are they both right according to their respective dates, and has the Spanish language undergone what we may call a certain amount of orthoepic degradation during the present century? The result of my inquiries surprised me, and will probably surprise some of my hearers. On the one hand I have not been able to satisfy myself that initial Span- ish d is anywhere pronounced differently from our own d, or that final Spanish d ever has the sound of th flat, so that the old grammars may be considered more trustworthy, so far, than Mr. Ellis’ more recent Spanish authority.* But on the other hand I discovered startling traces of orthoepic degrada- tion. That the lower classes in South America drop the final dand slur over the medial, so as almost or quite to elide it, need not surprise us any more than their pronunciation of z and ¢ soft as 8, or the apparent negroism prevalent in Cuba, of substituting a vocalized r for the strongly trilled final r, e.g. amaw (or something very like it) for amar. But I was greatly astonished to find the same elisions and others equally marked, in the Spanish capital. My son who has just returned from Madrid, where he spent several months (having passed the early winter in other parts of Spain,) tells me that per- sons of the best society drop the final and eliminate the me- dial d. No one but a foreigner says usteth, the natives all say
* Is it the most recent? He relies on Cube v Soler, 1851. I have seen guides published within the last ten years which insist on the pronunciation of the d everywhere like English d. But this, even if theoretically right must be practi- cally wrong. Vide infra.
On Ellis’s Early English Pronunciation. 123
“st€é. But what more surprised me was a slurring of medial 7, the very thing which is the usual reproach against Anglo- Saxons in speaking the Romance tongues. And I was even still more struck with the elision of medial s before a conso- nant, especially before ¢, precisely similar to what has hap- pened in French. The fashionable pronunciation of Madrid seemed to my son studiously effeminate, a deliberate weak- ening or dropping of the strong consonants. Mr. Hay, the well-known author, tells me that the Castilians drop the final s of their particles, e. g. for greater they say, not mas grande, but ma grande. He was almost certain that he had heard the orator Castelar say lo reyes for los reyes.* In the observa- tions of these two gentlemen, who though liberally educated, and good practical modern linguists, are not professional philologists or even, strictly speaking, scholars, we cannot look for scientific accuracy, and must expect some discrepan- cies. Thus, my son thought that the apocope of s was con- fined to the lower classes, and Mr. Hay had not clearly noticed either the apocope or the syncope of d.
We are now at length able to proceed to the main body of the work. The general results of Mr. Ellis’ elaborate investigations are, | believe, tolerably well known to scholars, and with two, or at most three exceptions, generally acquiesced in. That our
“Castelar is, I understand, an Andalusian, so that his pronunciation would be no more test of pure Spanish usage than Patrick Henry’s would have been of pure English. In a subsequent conversation my son explained the case thus:
The Andalusian dialect tends to drop final letters, even r. [So that the Cuban amaw is not a negroism, as seemed probable.} It also drops medial s; e. g. azul celete for azul celeste. ‘This provincial pronunciation has largely invaded Madrid. The Spanish teachers say this is owing to the Cubans, who are mostly of Anda- lusian origin and have brought back their pronunciation to the capital. [Rather a roundabout route! but not absolutely impossible when we consider the extent of sectional feeling and consequent separation between the Spanish provinces, and the fact that Cuba is a common ground to them.]
In respect to d, while it is sounded clearly and like our own letter as an initial und in the -1po terminations, [this peculiarity Mr. Ellis has correctly noted,| it is dropped entirely from the -apo terminations, but without compromising the dis- syllable or the separate sound of the vowels. Thus prado is pra-o (or, as Mr. Ellis would writcit pra.o). D medial befor: r is th flat (dh); thus the d of podrir is the d of the Irishman’s dress. But some exquisites drop even this d, and say Mari for Madrid! This pronunciation, however, is generally condemned and ridiculed as an affectation.
124 C. A. Bristed,
peculiar vowel powers are of modern date; that our ancestors pronounced @ as alpha, (which of the modifications of alpha is a question of secondary importance,) ¢ as d, 7 as ee, €a@ as @, Ou as oo, and o7 or oy as 00 ee, —these points may be con- sidered pretty well established. There is some doubt as to the date of the change from alpha to English @. Mr. Ellis considers it not yet to have taken place in Shakespeare's time ; others are of opinion (basing their conclusion partly on the same data,) that it had then taken place, or at any rate that the vowel was then in a transition state. As to the early pronunciation of ai diphthong, the difference of opinion is strongly marked. Mr. Ellis considered it to have been what for practical purposes we may call our i. Mr. Payne, no mean authority on such matters, attributes to the early diph- thong its present sound. The French diphthong naturally follows the English, as they rhyme in early verse, macaronic or other. Mr. Ellis, with most laudable fairness and accu- racy, has quoted Mr. Payne’s ipsissima verba at full length, so that the reader may compare both their views at lei- sure. Every reader must make his choice. I incline to Mr. Payne. Some of the instances not adduced by him, and on which Mr. Ellis dwells rather strongly, seem to me to con- firm the views of the former gentleman rather than of the latter; e.g. the remarkable political macaronics in the Auch- inlech MS. To show how obvious the medial rhyme is, I will read eight lines of it. [We must remark beforehand: 1st, that den in the first line is for Pon, and the verbs fere and defere are spelled with an e, not with az diphthong; 2d, that the present French nasals were much weaker then, or hardly existed, so that we must rhyme French and English ent; 3d, that crey in the 5th line is for crows; 4th, that our 7 was then ee, so that fire was feer, and shire (as it still is to some extent in England,) sheer. ]
“«Len peut fere et defere, ceo fait il trop souent, It nis nouper wel ne faire, perfore Engelond is shent. Nostre prince d’ Engletere par le consail de sa gent, At Westminster after pe feire maden a gret parlement.
On Ellis’s Early English Pronunciation. 125
La chartre fet de cyre ieo lenteink et bicn le crey
Tt was holde to neih pe fire and is molten al awey.
Ore ne say mes que dire, tout i va a Tripolay,
Hundred, chapitle, court an shire, all hit gop a deuel wey.”
These alternate French and English lines seem on the face
of them to rhyme in their quatrains or double-pairs not only finally but medially; to pronounce with Mr. Ellis “it nis nouther wel ne fire” in the second line, breaks up the rhyme.* I think also that Mr. Ellis’ explanation of the connection be- tween the diphthongs ai and of in French (given it is true in another writing, but not contradicted in this book, and about which we shall have more to say,) tellsagainst him here. Mr. Payne considers that in Norman and Early English, at, ay, ei, ey, and even their reversed forms, ta, te, had the sound of a, as well as ea and ae; in short he holds to what we may call an etacism in old French.t |
* The rhyme “ lasteth ay,” ‘“ miserere mei” (Ellis, p. 447), I take to be a make- shift rhyme of a monosyllable with a dissyllable, such as share and layer would be now.
t The sound of é and its general equivalent ai dipthong in modern French I presume to be of late introduction ; and this is apparently assumed by both dis- putants.
If we can trust De Saint Lien — who styled himself Gentilhomme Bourbonnais, Anglicised his name into Holyband and published The French Littleton in 1609,— the diphthong must then have been in the transition state. He gives no fewer than three different sounds to aior ay. First, that of é in the final syllable of the future indicative, in ay (present of avoir) and scais (present of savoir) Secondly, ‘the first personof the first perfect [what is now usually called the preterit definite] is sounded as it is wrjtten.”” Whatever this may mean, the distinction is really cu- rious. It is intelligible enough that aimerais (conditional) = atmere, while aimerai (future) = aimeré: but on what principle could the final syllables of the preterit and the future (ainat, aimerai,) be differently pronounced? Thirdly, in all other cases, ‘‘sound it as gaye [gay] gaping,” that is evidently our 4.
A writer in Zhe Academy has pointed out the etymological difference of é and é, the latter representing a Latin @ while the former represents a Latin e. But hia inference that these two e’s have exchanged their original sounds is very question- able. I cannot bring myself to believe that mére, pére were ever pronounced like the first syllables of Méry, périr, or of our own merry, perry.
There is another point connected with the English diphthong a: which ought not to be passed over; so I shall introduce it here, though it comes in somewhat clumsily. The old standard dictionaries (Walker, &c.) give five sounds of English a, the four which are familiar and obvious, and a fifth, “‘ the sound of az in fair.”” No difference exists at present between fare and fatr in the pronunciation of Englishmen or Americans, and this fifth sound of ais either an anachronism or a mere fantasy Which is it? When I ventilated this question twenty years ago, Professor Halde- man threw out a suggestion that the English of the last century gave a French sound
126 os C. A. Bristed,
But the most disputed of Mr. Ellis’ points is the French sound of «, which is attributed to our w, both long and short, until after Shakespeare’s time. If he is correct, the change that has occurred is most extraordinary ; the French uv is now not only so foreign to the Anglo-Saxons that they cannot (without much effort and instruction and practice) pronounce it, but they have not a conception of what it is like; and at this day there are dozens of poetasters in this country, and not a few in the mother country who imagine (judging from their practice) that French words Jike perdu and plus rhyme with English words like you and through; they might almost as well rhyme them to bee and flee. Mr. Ellis, however, would say that this assertion is too broad, the sound does exist in Anglo-Saxendom, to wit, in Scotland and in some of the English counties. I am by no means sure that this Low-
to ai, pronouncing fair, for instance, like the first syllable of ferry. Lam not aware of his rcasons for this hypothesis, nor can I find any direct reference to such a pro- nunciation in Mr. Ellis; but there is a circumstance which bears on this point indirectly though it can hardly be said to confirm Prof. Haldeman’s view. In the last century cheerful and chair full were sufficiently alike in sound to give rise to a bad conundrum noticed by Mr. Ellis “ Why is a fat man a merry man?” “ Be- cause he is a cheer-ful.” Now our first impulse is to refer this to the pronunciation of chair as cheer, which we know was then current to some extent in England, as indeed it is to this day in South Carolina and possibly in some other Southern States. I am unable to satisfy myself how far this pronunciation extended, whether it included all words in which az is followed by r, whether it included any in which ai is not followed by r, but there arc indications in the literature of the day that so far as it did prevail, it was the result of a reaction against a Hibernian attempt to revise the archaism of 4 for ea diphthong, which is a well-known Hi- bernicism to the present day. But another mode of explaining the conundrum occurs to me. The pronunciation cherfiud was known in 1780 and is still known in England, and indeed is conformable to the analogy and genius of the lan- guage, which shortens in compounds the long vowel of the simple, know, know!- edge ; vine, vinyard; Whitby from white, and even Whitsunday from white — if that be the true derivation of the word which, however, is doubtful. Now we know that final r in English is utterly destructive of the sound of 2 changing it into the indistinct %, whether the final syllable be or be not accented. Is it not then possible that Englishmen of the last century in trying to give the at some such [theoreti- cal] sound as Prof. Haldeman supposes, may have run it into % It is entirely odd to suppose our great grandfathers talking of a fur day and sitting on a chur: yet it is equally strange to suppose them sitting on a cherr in a ferr day; and if we reject both these, I see no third alternative except to pronounce this fifth sound of a altogether imaginary, and the old lexicographers under a delusion when they nvented it.
On Ellis’s Early English Pronunciation. 127
land Scotch ui is the French u; but allowing it to be identi- cal, 1 do not admit the inference; for the Lowland Scotch was a very peculiar dialect of English, resembling, indeed, that of the border counties, but very distinct from the midland and Southern varieties. As to the Devon, Mr. Ellis confesses (note on pp. 635-6, chap. vii) that the sound in this dialect is vague and indeterminate, sometimes verging on 00, more frequently on his oe oe, that is the French ew long as in jetne ‘‘ fasting,’ or the long sound of that indistinct vowel which we usually consider as %.* The evidence of the 16th and 17th century orthoepists 1s very conflicting, as we might expect, considering the frequent confusions and errors which occur in our so much more enlightened days. It is certainly annoying and perplexing fo find two men of the same locality, educa- tion and social position, contradicting each other point blank, as in the case of Wallis and Wilkins. The critic of the Worth American thinks that in such a strait we should prefer the writer who states that a foreign sound is strange and difficult to his own people; but it would hardly be safe to adopt this rule broadly and without limitation. However, the very ut- most weight which we can assign to Bullokar, Wallis, and the others who identified the sound of «in French and English, is to let them offset the contrary statements of Wilkins, Eron- dele, and the others who discriminate between the sounds. We are then thrown back upon the general probabilities of the case, and these are surely in favor of the present sound. For my own part, I doubt very much if the French w was ever fairly naturalized in England, even in the French specch of the upper classes. We all remember how Chaucer describes his prioress as speaking French
“ful faire and fetisly After the scole of Stratford at the Bowe For French of Paris was to hire unknowe,”—
showing that in his time there was a French French, and an
English French. The French w is and always was a strange sound to the Spaniard, the Italian, the Portuguese, the Dutch-
* It would be well to obtain the opinion of some French orthoepists as to the identity or non-identity of these Scotch and provincial sounds with the French u.
128 C. A. Bristed,
man, yes, and we may add the German, in spite of the theo- retical and grammar-inculcated correspondence of the German ue diphthong with the French vowel.
There are three passages in Mr. Ellis’ work which have particularly attracted my attention.
1st, As to the numeral one. It doubtless surprised most of us when our attention was first called to the fact that the di- gammatic prefix to this word is of comparatively recent date. Mr. Ellis is unable to fix the exact time of its introduction, but it was not before the 17th century. I have a conjecture (I do not pretend to call it anything more) that while passing from own into wun, the word took the intermediate form, un. Let us examine what may be called a crucial passage in Love’s Labor Lost: ‘‘ Master Person, quasi perse-on, and if one should be perst, which is the one?” with our modern pronunciation this is unintelligible, except the vulgarism of perse for pterce, still common in some parts of America. Mr. Ellis pronounces: ‘‘ If own should be pirst which is the own 7’ But suppose we throw out the words “is the” and read, “Tf un should be pirst which un?” then the clause becomes modern rustic English, for English rustics say un for one to this day.”
2d. As to the English prosthesis and apheresis of the let- ter h. He begins by speaking of the words in which initial A is properly mute. These he reduces to five, heir, honest, honor, hostler, hour (and of course their derivatives). Humble and humor he leaves doubtful. This may surprise some of us. | believe no Americans say humble, and very few say Aumor. We do not suppose it good English to say so. Dickens evi- dently sounded the 2 of Aumble, (see Uriah Heep in David Copperfield,) but Dickens, after all, was not a man of the high- est education. It is certainly singular to. find an American aphaeresis contrasted with an English quasz-prosthesis. I must say that on this point my own recollections do not agree with Mr. Ellis’ experience.
* They also use it for zt, Arm, them - — a sort of indefinite pronoun like French on, but only in the objective case.
My conjecture on the intermediate stage of un may perhaps be confirmed or re- futed by reference to the rhymes of the period.
On Ellts’s Karly English Pronunciation. 129
The main question, however, relates to the English actual prosthesis, or prefix of the A where it is not written, conjointly with the equally common apheresis. Respecting these pecu- liarities, the popular American conception is erroneous. There is a current idea (made the most of by would-be dramatists and comic writers) that all Englishmen except those of the very highest class (and even many of those) always prefix an initial h, where there is none, and omit it where it is written ; and that the same persons are guilty of both the prosthesis and the apheresis. This is a grave error. The result of my own experience and observation I sum up thus: In London, no man above the rank of a servant or a small tradesman says a hass or a hangel. Thackeray and Trollope may be quoted against this assertion. But Ido not put implicit faith in nov- elists as philological authorities, and though Thackeray is sometimes very happy in transliteration, he is not uniformly so, e. g., when he writes cage instead of cadge to express car- riage in the flunkey dialect. Let us, however, to avoid ap- pearance of hair-splitting, draw the line a little higher, so as to include the upper order of tradesmen ; there the prosthesis must stop. But the apheresis extends much higher. Gen- tlemen in the best society sometimes talk of ’owse and ’ome: they are not conscious of saying so, any more than a Massachusetts gentleman is conscious of calling a certain familiar article of dress his cot or his caught instead of his coat. But when we go into the western and northwestern counties, and what we may broadly call the manufacturing districts, the prosthesis extends much higher up into society. It is common in what the English call the upper-middle classes, and among men whom we should call merchants. The apheresis, on the other hand, is not so common there, in the best society, though on this point I feel less certainty. The difference is conformable to what we might expect; as we go northward the aspirate becomes more prevalent, and (so to speak) more respectable. Perhaps my conclusion is an imperfect generalization, but Mr. Ellis bears me out in the first part of it, at least. He says, (chap. iii., p. 221,) “In the practice even of the most es- teemed speakers, ham in names of places has no aspirate, ez-
17
130 C. A. Bristed,
hibil, exhibition, lose h, and his, him, her, when unaccented drop their kh. It is extremely common in London to say a tome for at home.”* The general impression given by his remarks (which are not so full or clear as could be wished) is that the prosthesis denotes a lower grade of society than the apheresis. At any rate it is certainly wrong to assign both, indiscriminately, to all or nearly all. Englishmen, as the aver- age American is apt to do.
The entire absence of this erroneous apheresis and pros- thesis of A from all varieties of American dialect has often struck English tourists. There is, however, one Americanism —though not solely an Americanism—connected with h. It is the aspiration of the initial wh. Good English practice en- tirely omits the A here, making, for instance, no distinction between the noun witch and the relative which. The only Lon- doner (if, indeed, he could properly be called such) whom I ever heard aspirate the initial wk was Prof. Jolin Grote. To aspirate the wh is Scotch, Irish, American. Thackeray, to express it in the speech of his O’Mulligans and Costigans, reverses the letters, writing hwat for what. It is probable, however, that the north-county men, who have some of the Scotch peculiarities, aspirate the initial wh, though I cannot recollect any of my acquaintances who did, e. g. Tom Taylor, who is a Cumberland man, certainly did not.
Before taking leave of the letter , let us say something about it in its opposite position—at the end of a word. Mr. Ellis shows how in Shakespeare’s time the final 4 after o had the same diacritic power as final e mute; the name written moth in Love’s Labor Lost was pronounced mote ; the page is not called moth as a fluttering flilbertigibbet lad, but mote as a mere speck or mite of a boy. Mr. Grant White wished to improve on this, and give the same power to medial h, which seems tobe an error The medizval Latin orthographies show that this diacritic power of final 4 was once very general. It is still retained in German. The most natural way to a German of expressing the sounds dte, dte,is eth, oth. An English acquaintance of mine, named Mate, who had reported
* His paleotype is perhaps % tome. I am not sure of it.
On Ellis’s Early English Pronunciation. 131
himself viva voce to his landlord at Baden, was much disgusted at appearing next day in the watering-place gazette as Mr. Meth.
8d. As to the diphthong o, the different powers and changes of this diphthong in different languages, were discussed by Mr. Ellis, in a paper read before the Philotogical Society, in 1868, of which paper a portion only is embodied in the present work. (I may be allowed to mention here as an illustration of the difficulties with which the American student has to contend, that after failing to find this volume of the Society’s transactions in the public libraries of more than one city, and equally failing to find any bookseller who would undertake to import it for me, 1 was compelled to write to the author him- self.)
In a paper which I had the honor to present at the first meeting of this association, I suggested that the first element of the Greek diphthong o may originally have had a digam- matic power, which was afterwards lost. This is Mr. Ellis’ view. There isa continental theory—I am unable to say who first put it into shape; it may be found in the transactions of our Oriental Society—which bears on this question. The theory is as follows: the diphthongs a: o differ from the other diphthongs in being more simple sounds, and in having some of the qualities of short vowels, e. g. when final they throw back the accent to the ante-penult. -: is the equivalent of ¢, (just as at in French is of é) and oof». The epithet pidor applied to the two vowels does not mean, as generally sup- posed, smooth, unaspirated, (which indeed involves a contra- diction when applied to v;) it means simple, as opposed to the compound or diphthongal equivalent. This hypothesis is at least ingenious. It must assume, however, the non-existence of any digammatic element in m, or at any rate, its disappear- ance before the names Epsilon, Upsilon, were given to those letters.”
The connection of o with ov, in the apparent relation of short and long, which Mr. Ellis insists on, may help to prove that
* We must remember also that omega in certain cases throws bnck the accent to the ante-penult.
132 C. A. Bristed,
o and w were different vowel powers; I do not think it neces- sary as a support to the existence (at some time) of a digam. matic clement in o.
The Roman OE certainly looks like an equivalent of the Greek OI and partly invented for that purpose. But there is a very prevalent opinion among scholars that it was only a conven- tional representative, not an equivalent, and that it had the sound of English @. This conclusion they draw, not mercly that oe has become e in the Romance languages (for that only from the fact proves that the diphthong had this sound in later Latin, which no one ever doubted,) but also from the occasional interchange of ae and oe in Latin, even in words like coelum which are at least cognate tothe Greek.* They are not, how- ever, agreed among themselves as to details, and especially as to the explanation of pairs like moenia, munta, poena, punire. Some, (our friend, Prof. Hadley, is of the number) do not ad- mit that oe ever had any sound but [Engl.] @. Some, like Corssen, allow an original sound which had something of u in it.— Some think that the o was at first digammatic. Among these last is Mr. Ellis, who holds that as o was originally [Engl.] wee, so oe was originally [Engl.] wd. This is cer- tainly an easier way of accounting) for the above mentioned pairs, than to suppose that the Romans were at first undecided whether to express a by the vowel w or their new diphthong oe, and so split the difference by taking wu as its representative in some words and oe in others. Jt also explains the later sound of ve as simple e by the dropping of the prefix. But the whole subject is involved in difficulty, as I have shown in the paper already referred to, where I really understated the case by omitting to take account of this very difference be-
* Mr. H. A. J. Munro, however, condemns coelum as a late coinage like sylva.
+ This was Corssen’s first opinion. He afterwards came to the conclusion that the old Roman oi split into @ and u according as the second or first element of the diphthong prevailed in popular pronunciation. He also rejects the close connec- tion of the Greek diphthong with the Latin vi of words like vinum.
{ Written in paleotype, ue ue. Perhaps Mr. Ellis would hardly allow his u to be digammatic, but it comes to this practically, if the combinations are to be con- sidered diphthongs at all.
§ By the singular but not impossible absorption of the second element.
On Ellis’s Early English Pronunciation. 133
tween the Greek diphthong and its Roman representative, which is, perhaps, the most puzzling of our confusions. I must confess that I cannot see my way clearly through it. At an earlier period, when the two languages were farther remov- ed, the Romans seem to have had no difficulty in expressing the o by vi—doubtless pronounced wee. At a later period, the sound of the o is so strange to them that they are obliged to make a diphthong partly on purpose to represent it—and after all, the representation is only conventional ; the foreign sound is not correctly renlered. If oe was merely [Engl.] 4, what was the use of inventing it when they already had e and ae with the same sound ?
Two possibilities suggest themselves :
The first is, a variation of the Roman e towards 1, of which there are some apparent signs; but the absence of anything like Iotacism in the Romance tongues seems to negative this supposition at once. I therefore dismiss it, though somewhat unwillingly.
The other is as follows: We may suppose, that whatever the original sound of the second element in «, and whether it ever had, or, having had, still retained a digammatic prefix, it had come to be no longer a pure ee but contained some ad- mixture of u. (A digammatic prefix might well pass into such a shade of uv.) Then the sound would be foreign and difficult to the Romans, who seem not to have had any false sounds in their language, and were greatly puzzled by the v. Although they made a special sign for the Greek diphthong, their clumsy attempts at pronouncing it might soon degener- ate into simple [Engl.] d. Much as the Germans, though they have made ¢hetr oe a conventional respresentative of the French ew and wu, stumble fearfully at these French sounds, generally rendering them by e, sometimes even by Alpha. A German valet de place, at Munich, once told me that Lola Montes had marshes (avait des marais) in a certain street, meaning that she had lived (avazt demeuré) in it.
Professor Drisler’s opinion is somewhat peculiar: He thinks that in the pairs like poena, punire, the u represents the oe, not by substitution or omission, but by simple contraction, oe becoming w, just as in Greck oe becomes ov.
1384 C. A. Bristed,
‘On the whole, I must reiterate my inability to arrive at any satisfactory solution of all our difficulties. The only two pos- itive conclusions at which I have succeeded in arriving are both negative (pardon the Hibernicism ;) I am confident that the classical sound of « was not,as in modern Greek, a simple t (Engl. ee,) and that the first sound of oe was not as in later Latin a simple e. ,
‘We now pass to the diphthong in French. The interchange of o¢ with ui in certain proper nouns and adjectives (as Fran- gais, frangais, or Frangois frangots ; aimais, aimatt, or armors, aimot!)-up to a comparatively recent date* is a puzzle at first sight, for there seems little connection between the sounds. Mr. Ellis considers that the second element of the diphthong was originally a French e, and that ot thus passed into az by dropping what I should call the digammatic element, thus Frengwais, Francais. This explanation is plausible and probable. Moé, toé, (that is in English letters nearly mweh, tweh), have always been provincialisms for mot, tot. The Nor- man use of a? (or its equivalents) for ot is of ancient date. At one time the sound of az had encroached largely on oi. Sait for soit and fraid for frotd were good seventeenth-century French (p. 184). One of the curiosities of the confusion is that the names of the French letters, afterwards bé, cé, dé, &c., and now be, ce, de, were, in the sixteenth century, boy, coy, doy, &c.
The present sound of the diphthong, Mr. Ellis considers to be waa rather than waw. He has abundant French authority for ‘this ; yet I have heard the latter in Paris at least as often as the former. It is safest to consider the pronunciation as fluctuating between the two sounds, much as the English pro- nunciation of vase fluctuates between vaaz and vauz.t
We finally arrive at English. Mr. Ellis has shown pretty conclusively that the original English pronunciation of 02 was
* There is at least one living writer, Couture, the painter, who still-ases oi for ai in hia verbs.
tI believe all Americans pronounce vase to rhyme with case. Moore. (an Irishman) rhymes it with grace, habitually. Sotheby (an Englishman) rhymes it with draws.
On Ellis’s Early English Pronunciation. 135
[Engl.] voee. Our ancestors pronounced boy as an. American pronounces “ buoy,” booee. (In regard to this word buoy I had always supposed our pronunciation a pure Americanism, but Mr: Ellis says that it is also that of all nautical men in England. In ordinary English society the word is pronoune- ed exactly like boy.
In its passage to the present sound, the diphthong (in the 17th century) came near taking that of 7; but this pronun- ciation was condemned, and has passed into a vulgarism and a Hibernicism. It is curious to note, as showing how com- - pletely such things arc matters of fashion and convention, that exactly the reverse has happened in German with the ez, the pronunciation of which as oz, though theoretically correct, is practically considered vulgar.
I am inclined to dwell somewhat on this interchange of 01 and i, because it seems to me to afford additional proof of Mr. Ellis’ error in considering the first element of these diphthongs to be the indistinct @, rather than alpha in one and [Engl.] au in the other. As alpha broadens into au, so ¢ broadens into oi; the change is clearly accounted for. According to Mr. Ellis’ notation, it is less easy to understand or explain.
Mr. Ellis considers that our o7 is a peculiar sound, not ex- isting in the other European languages. Sed guaere.
Loose and desultory as these remarks on the diphthong ot and its representatives may seem, there is, I think, a leading idea to be gathered from them, whether we consider the first element as 6, [Engl.] 00, or [Engl.] au, the second as z [ce] or e [d,] there appears in the four languages which we exam- ined a certain similarity of progress; the first element has originally a digammatic force; this digamma afterwards dis- appears, and the second element is modified. We cannot be so sure of this in Latin as in the other languages, but even there, we have at least a probability. In French the digamma still remains except in the few proper nouns, adjectives, and inflections, where oz has given way to at.
I have already trespassed too long on your patience, but it is hard to part from Mr. Ellis without taking some notice of his labors and arguments in favor of orthographic reform.
136 C. A. Bristed,
Most of the recommenders of phonetic systems have been met with ridicule, and it is not unjust to add that many of them have deserved nothing better. Ignorance, indolence, love of notoriety, that blind radicalism which, with about as much wit as a mischievous monkey, attacks whatever zs, merely be- cause it is,—such are the inspirations of the majority of them. Some are so ignorant of Greek as to call themselves phonic which, if it meant anything would mean murderous—and they certainly do murder common sense and logic; some so ignor- ant of English that they do not know by what rule or on what principle the a of mate is long. From such persons Mr. Ellis is removed tote coelo. He is a man of wide and deep knowl- edge, a scholar, a mathematician (sixth wrangler at Cam- bridge,) apparently a musician; he has labored many years at his particular specialty, and his very mistakes are such as no ignorant man could make. Whatever he says should re- ceive the most respectful consideration. He sees the great inconveniences and practical absurdities of the present system (all of which are exaggerated here), particularly the subor- dination of authors to printers. If an Englishman feels this, how much more must an American, above all a New Yorker, in the majority of the printing-houses of whose city the Web- sterian cacograpby reigns supreme. All the faults of the ex- isting orthography are set forth well and truly and without exaggeration. But Mr. Ellis soon makes an admission which is indirectly fatal to the proposed reform. He says (Chap. vi., pp. 624, sqqg.), that there can be no absolute standard of pronunciation, and this position he illustrates at length. I was able to add another illustration within fifteen minutes of reading the chapter. After giving the last-century pronun- ciation of sewer (drain) which was shore, he expresses the present sound by.a monosyllable which most persons would be apt to pronounce like sure, I had scarcely read this when there called on me two English gentlemen of about the same age, education, social and official position. ‘“‘ How do you pronounce this word?”? I asked. The reply of one was a monosyllable very like Mr. Ellis’, the reply of the other a dis- syllable very like the pronunciation of most Americans. Sev-
On Ellis’s Early English Pronunciation. 137
eral of Mr. Ellis’ own pronunciations would not find universal acceptance. As an orthoepist, he sees the great advantage which a scientifically-framed phonetic system would have in giving the pronunciation of each writer. We might almost be able to get at the statistics of dialetic peculiarities and to tell how many Massachusetts men call a coat a cot, and how many call it a caught. But we must remember that it is not the sole use of written language to inform the people of one section or country how the people of another section or coun- try prenounce. By giving all local and even individual pecu liarities a species of legitimate sanction in serious writing, we debase the standard of pronunciation still more, while we break up the standard of orthography altogether. The attempt to create a perfect system of spelling would inevitably end in the confusion and destruction of all system. When Mr. Ellis, and some fellow-reformers, started a newspaper on the im- proved plan, they came to grief on the very title. It was Phonetic News ; some pronounced the second word nvoz, others nyooz; of course each pronunciation required a different nota- tion. This anecdote, reported in perfect good faith by the author, strikes me as an excellent practical reductio ad absur- dum. . If I have spoken freely of many passages and statements in this book, it is not from want of admiration for the work, which is a monument, not only of industrious, but of intelli- gent learning, a treatise of exceeding interest and value. I am sure we all deeply regret that the author’s impaired health has delayed the completion of it hy the appearance of the last volume.
18
VII._—On Algonkin Names for Man. By J. HAMMOND TRUMBULL.
The Indian speaker never generalized. He loved figurative expressions, but however frequent his metaphors or dark his allegory might be he used no word of which the literal mean- ing was not exactly limited. His language supplied him with specific names for all known objects, qualities, and rela- tions, and its marvelous possibilities of synthesis enabled him to frame new terms as often as new distinctions were required. It grew by progressive differentiation,—from genera to spe- cies, from species to varieties and individual peculiarities. There is not, perhaps, in the Indian mind — certainly not in the structure of Indian languages — absolute incapacity for generalization, but the scrupulous avoidance of it as a defect, whether in thought or speech, is a characteristic of the race.
Though the Algonkin languages are poor in general names, yet we find in all of them certain elements of synthesis which may be regarded — from one point of view, or another, — as rudiments, or as vestiges, of such names. These are not used as independent words, but in composition they take the place of the ground-word or principal root — their denotation being limited or directed by the attributive prefixed. Such, for ° example, is the (Chip.) terminal -abo (after a vowel, — wabo ; Abnaki -a"ba, -wa"bw,) denoting “ drink,” found in many spe- cific names, but never without a prefix: asin Chip. wiasa- abo (meat-drink) broth, ishkoté-mabo (fire-drink) whiskey or other ardent spirit, mashkiki-wabo (herb-drink) liquid medi- cine, totosh-abo (breast-drink) milk, ete. In a few inst:nces, such a generic expression which in one dialect is inseparable, in others has attained—or has not yet lost — independent position as a «specific name. In the Massachusetts language, -min, denoting ‘‘small-fruit’’ (berry, nut, or grain), does not appear to have been uscd without an attributive, e. g. wuttahe- min (Chip. od¢éimin) heart-fruit, a strawberry, weno-min twine- fruit, a grape, wompi-min white-fruit, a chestnut, etc.: in the Delaware and in some western Algonkin languages, -mtn is similarly employed in composition, but is also used inde-
On Algonkin Names for Man. 139
pendently as the name of a single species — the bilberry or huckleberry.
Other grammatical devices by which the deficiency of these languages in general names is compensated need not here be pointed out. That such a deficiency exists is indisputable, yet it has been often disregarded in the selection of words for comparison of different languages and dialects. No one has recognized more clearly than did Mr. Gallatin “the extreme precision of the Indian languages,” and their poverty in “ ge- neric designations or words,’”* but of the first twenty English words in his “‘ Comparative Vocabulary of fifty-three tribes” Cin Trans. Am. Antig. Society, vol. i., pp. 307 and after,) fif- teen are relative and general names not one of which can be accurately translated by a single word in any Indian language. Every Algonkin dialect has names for an “ elder brother,” a ‘younger brother,’ a “ twin-brother,” a ‘son of the same father,”’ and a “son of the same mother,” and has moreover two forms of some or all of these names, one used exclusively by men, the other by women. But in no dialect can there be found the precise equivalent of the English “ brother,” in its largest denotation. The nearest approximation to it is, perhaps, by a term which, in some languages, designates ‘‘ one of the other sex, born of the same parents ;’’ spoken by a woman, this word means “ brother,’ — by a man, “sister.”
The names by which May has been designated, by different tribes, or, more accurately, which most nearly correspond to the English appellative in its two meanings, “ an individual of the human race” (homo), and “one possessing in a high degree the distinctive qualities of manhood ”’ (vir), have occa- sioned much perplexity to vocabulary makers. Mr. Hale, in a note to his Vocabularies of North America (Zrans. Am. Ethnol. Society, ii. 74), remarks that “in general, there was no means of ascertaining with precision the existence of this distinction.”” He has, however, nearly indicated its true character by the suggestion that, in vocabularies, the term ‘answering to vir will usually be found under man or hus-
* Transactions of the Am. Ethnological Society, vol. ii., p. cxxxi.
140 J. H. Trumbull,
band,” and the “ term answering to homo, under ‘ Indian, na- tive.’ The truth is, it is as impossible to find an Indian equivalent for homo as for man. By resorting to the Latin, we only halve the difficulty, not remove it. There is not in any American language any single name applicable alike to the red man and the white, to native and foreigner, to ally and enemy, to chief and counselor and to prisoner and slave, and in its largest sense common to both sexes. For vir a term nearly correspondent may be found in every. dialect — though seldom, if ever, as a primary word ; but homo is un- translatable by an Indian.
In Algonkin languages—and the same probably is true of all others spoken by North American nations — we have three classes of names for Man, into the composition of which enter three or more different roots. These are indicated, not very clearly, by Roger Williams, in the introduction to his Key into the Language of America (16438):
‘‘T cannot observe that they ever had (before the coming of the English, French or Dutch amongst them) any names to difference themselves from strangers, for they knew none; but two sorts of names they had, and have, amongst them- selves. First, general, belonging to all natives, as Ninnuock, Ninnimissinniwock, Eniskeetumpaiiwog, which signifies men, folk, or people. Secondly, particular names, peculiar to the several nations of them amongst themselves, as Nanhiggané- uck, Massachuséuck,”’ etc.
Of the three “ general ’’ names, the second, xinni-missinnu- wock, is formed from missin (with indef. suffix, missin-nin,) a derivative of mtss: great, much (multus), and comprehends all homines who are not virt, corresponding etymologically and in its denotation to the Greek oi roAdoi, or the Latin mul- titudo. It was a general name for tributaries, captives, slaves, — that is, for all mankind, the speaker’s nation and its allies excepted. The prefix ninni-, however, limits it to inferiors of the speaker’s own race, as will presently be shown. Eliot employs missinnin for ‘man’ (homo) in Gen. vi. 7, and in the plural, misstnninnuog, for “ people,” Exod. xxiv. 2, 3, Deut. iv. 838, etc. In Jonah, i. 8, howaé missinnin ken? ‘of what
On Algonkin Names for Man. 141
people art thou ?” would convey to an Indian the meaning of ‘‘what kind of slave (or, whose servant) art thou ?”’
In Williams’s other names, ninnu-ock (ninnu-og, Eliot) and eniskeetompau-wog, both plurals, we find two roots common to all Algonkin languages. They vary in pronunciation (and in the phonetic notation employed by different writers), one as nin, nen, enin, aren, len, illin, etc.; the other as omp, a"bé, abé, dpé, dp, etc. These two roots are combined in the Abnaki aren-a"bé and the Delaware len-dpé, and the former is repeated, as a prefix, in the Delaware tribe-name lennt lendpe.
Mr. Heckewelder, who received with unquestioning faith the legends of his chosen people, the Delawares, and was convinced that theirs was the parent stock from which all Algonkin nations were derived, found in this tribe-name new evidence of their high antiquity and purity of race. Lenni lendpé, he says (History of the Indian Nations, p. 25), ‘signifies original people, a race of human beings who are the same that they were in the beginning, unchanged and unmized.”’ As to the analysis of the name, he is not quite clear. Zenno, he says, signifies ‘a man;” in the names of quadrupeds, “‘ a male.” Lendpe signifies man—‘“in a more extended sense,’ —and “in the name of the Lenni Lenapé, it signifies people, hut the word lenni which precedes it has a different signification and means original, and some- times common, plain, pure, unmixed.” ‘Under this general description [and very general it is, certainly,] the Indians comprehend all that they believe to have been first created in the order of things.” (Corresp. with Duponceau, pp. 368, 412.) Mr. Cass, in the North American Review for January, 1826, remarked the ‘“‘ confusion in Heckewelder’s ideas of the name in question,” and offered another — and a worse — translation of it. DLenee, he says, ‘ generally and properly means ‘ male’,”’ and ‘“‘ the true meaning of lenaupé is ‘common’.’’ He was as far as was Mr. Heckewelder from detecting the connection between lenno ‘“‘ man” and a word meaning “ original, com- mon, plain,’’ ete.
To discover the primary signification of each of the two roots found in len-dpé, we will look first to the Massachusetts
142 J. H. Trumbull,
language, where the materials for etymological research are more abundant and, generally, more trustworthy than in the Delaware.
Every savage believes in the superiority of his own tribe and nation to all others. He and his are the real men: the rest are servants, tributaries, missinninnuog. Whatever is greater than himself passes out of his order of being and be- comes to him manitou ‘preternatural.’ The Illinois, says Marquette, call themselves Zhe Men, ‘comme si les autres sauvages au prés d’eux ne passoient que pour des bestes.” This conviction of personal and tribal excellence stamps itself on every savage language. In some of the North American tongues its traces are very plainly marked. Notwithstanding the want of a substantive verb, ‘‘] am ” is a constant element of Algonkin grammar. The demonstratives and relatives which in Indo-European languages appear to have been de- rived from the primitive pronoun of the third person are in the Massachusetts and other eastern Algonkin dialects mani- festly related to the pronoun of the first person. The Indian conception of man was as one ‘ like’ himself. Men of his own nation were “ such as I,”’ nostrates, and his was the “original,’” ‘‘common,” normal, type of humanity.
The (Mass.) pronoun of the first person singular is nén ; as 2 prefix, n’; plural ntn-awun. The demonstrative of inanimate objects is nz ; of animate beings, xd (noh, Eliot) ; of place, na there. The distributive ‘some,’ ‘any,’ ‘of the kind of,’ is ’nnt or unni. Resemblance or identity was expressed by ni-unnt (neane, Eliot) such as this, or nan same; ni-nan the same thing, nénan the same person; ’nnth (Eliot), ani CR. W.), it is so, or itis the same ;" nana (nanwe Eliot) common, usual, i. e. ‘such as’ ours, or ‘of our kind,’ hence, ‘ native,’ ‘indigenous.’ Kliot wrote nanwe missinninnuog ‘“ common people,” Mark xii. 37 (= ninnimissinniwock of Roger Wil-
* Comp. Chip. in-, int-, prefixed to verbs, “to signify a certain way or manner in which something is done or used,” etc.; e. g. 1N-dbi he so looks ; od ’1Nabaman he so sees him; 1nagode “it hangs so; nind 1nawa “T resemble him; ini1dé “it is cooked in a certain manner” (so); rn1gini ‘“‘he is so large;’” INo “‘it is so”
Baraga.
On Algonkin Names for Man. 143
liams, before quoted), and nanwe wut-epistle-um Jude for “ the general epistle”’ etc.*
’"Nnin-u (enin), pl. ninnuog, which Roger Williams gives as one of the “ general names belonging to all natives” and “ sig- nifying men,” was occasionally used by Eliot in the plural and, with an attributive prefixed, in the singular, for ‘ man,’ ‘men ;’ but the Indians restricted its denotation to men Iltke’ themselves, of the common: or native type, of the speaker’s kind (though not necessarily of his tribe or nation). It is opposed to penwwi (Abn. pirwi) strange, foreign, of another kind.
In other Algonkin dialects, the Massachusetts ’ninnu or enin-u becomes (Abnaki) arent, (Quinnippiac) ren, (Dela- ware) lenno, (Illinois) illini, (Cree) ethinu, etc., — meaning
* Schleicher (Vergl. Grammatik, 2te Aufl. p. 642) considers the root of the lst sing. pronoun in Indo-European languages — ma ‘I,’ ‘ me,’ — identical with the verbal root ma ‘to think,’ ‘to measure,’ and with the ma in Sansk. ma-nu, Goth. ma-n ‘the thinker,’ ‘man’: for since “we must not ascribe to the primitive language the abstract conception of the Ego,—what,” he asks, “should ‘I’ be, originally, but ‘man’?” The likeness of the corresponding roots in Algonkin languages is as noticeable — and the probability of their original indentity is at least us great—as in the Indo-European. Compare Chip. nin I, me, 1N-/ 80, such, ININ-t man, and nind’ 1InEN-dam (intrans.) I think, suppose, it seems to me, I am (so) minded, nind’ INEN-dda (trans. jnanimate) I think of it, think it (so), mind’ INEN-ima (trans. anim.) I think of him, think him (so). But I do not believe that the Indian —of Asia or America— waited fur the demonstration “ cogito, ergo sum,” as a necessary preliminary to sclf-assertion or to the vocal designa- tion of his fellow-savage. Without rising to “the abstract conception of the Ego,” he in some way discovered and expressed the distinction between ‘this, — me’ and ‘that, — my Like,’— alter ego. His mental states and activities, — his likes and dislikes, opinions, regards, emotions, — how he was affected by an exter- nal object, what he thought of it, how he estimated or measured it, —he would naturally express by ‘it is so to me” (though perhaps not so to another); ‘‘Z so regard, feel, esteem, believe, think it.” Of the same object, onc might say nin mino-f.Ndan I well-think it, it to me is good; another, nin jing-EN-dan I hate it, it to me is odious; toward the same individual and with reference to the same act one would express his emotion by nin-nishk-EN-tma I am angry minded at him, an- other by nin bdp-ineEN im I am laughter-minded, joyful; what one remembers (mikwéndin = mikoa-en-ddn finds in thought), another forgets (wanéndan = wani-EN-dan misses in thought, or bdn-en-ddn ceases thinking of).
_ In Chip. tnéndam (= Del. eléndam, Abn. erérdam, Mass. unantam), only én rep- resents the root: -dam is the grammatical formative, and the prefixed in- is the the adverbial ‘so, ‘in such manner,’ which is dropped when the verb receives any other prefix — as in minwendam, nishi&niman, etc.
144 J. H. Trumbull,
always, a ‘common man,’ of the speaker’s kin or kind. Used as an adjective, the Mass. nanwi, Abn. arent, Del. lennz, de- notes the ‘common,’ ‘usual’ or ‘ native,’ as distinguished from penwwi, Abn. ptrai, the ‘strange,’ ‘foreign,’ of ‘ other kind’: e.g. Abn. arent wdama" common or native tobacco, aren-a"dwé he speaks Abnaki (comp. pirw-a"dwé he speaks a foreign language) ; Del. /en-achpoan, common (i. e. Indian) bread, /enna-meek common fish (the sucker, found in almost all streams), den-chum common or Indian dog, (distinguished from the species introduced by Europeans), etc. Zeisberger translates ‘‘ Lennape, an Indian; Linni lenape, Indians of the same nation.” -
In lendépé, we have this adjective in synthesis with an in- separable generic. Heckewelder (Corresp. with Duponceau, 411) says that the termination ap or ape “ belongs to animals walking in an erect posture; hence, denape man.’ It is found in all pure Algonkin languages (Mass. -omp, Abn. -a"bé, Penobscot -ombé, Chip. ¢@bé, etc), but nowhere as an inde- pendent word. Asa generic suffix it denotes ‘an adult male.’ With a demonstrative prefixed (n’) it designates ‘ the male,’ or as an adjective, simply, ‘male.’ The primary meaning of the root may have been nearly that which Heckewelder sug- gests. It appears in the Mass. ompa-, Chip. omi-, a prefix to verbs of lifting, raising, erecting, etc.: e. g. Mass. ompa- ndéeu “he lifts himself up,” from a stooping position, John viii. 7; Chip. omBinan “he lifts or raises it up,” oMBdbate “ the smoke ascends,’ OMBashin “the bread rises,’ etc. (comp. Abnaki aBds: a standing tree); as an adjective, in Mass. noMPads male, noMPoshim male beast, pish nomPaiyeuw kah squaiyeuw there-shall male-be and female-be (Gen. vi. 19) ; in Chip. nasé male, nin-naBEm “‘ my husband ”’ (Baraga), etc. The dependence of the Indian warrior and hunter on his bow is expressed in its designation as ‘belonging to the adult male,” and by transferring it from the class of inanimate (‘ignoble ’) objects to the animate or ‘ noble’: Mass. ohtomp, Abnaki ’ta"BI, Powhatan attawPp or auAtaB.
Len-dpé (= Abnaki aren-a"bé, mod. Penobscot aln-ombe,)
On Algonkin Names for Man. 145
denotes “a common adult male,” i. e. an Indian man; lenno len-dpé, an Indian of our tribe or nation, and consequently, vir, ‘“*a man of men.” The roots, len and a"p, correspond more nearly to nostras and mas than to homo-and vir ; but the former is as exclusively masculine as the latter, and can- not be prefixed to a feminine noun-generic.
Recurring now to Roger Williams’s division of names into ‘‘ general, belonging to all natives,” and “ particular, peculiar to the several nations amongst themselves,’ we will first .trace these two principal roots, under their dialectic modifi- cations, through the several Algonkin languages, and after- wards notice some of the names for men of inferior race, — for enemies, strangers, and foreigners, — into the composition of which neither of these two roots may enter.
l. Man of the ‘common’ or ‘ native’ type; of the speaker’s kin or kind; nostras. Root, ’NEN, ’REN, "LEN, — from an ear- lier in ? with a demonstrative prefix, or reduplication. As an adjectival, it denotes ‘common,’ ‘indigenous,’ sometimes ‘mere.’ Formed as a verb, ‘to be a man (like ourselves),’ hence, in many dialects, ‘ to dave.’
Old Algonkin (Nipissing), inini: nin-ininyu “Iamaman.” Howse.
Chippeway, inini, pl. inini-wak. | Ottawa, anini.
Massachusetts, -inin, pl. -ininnuog ; ninnu ‘“ male,’’ Eliot (Mark x. 6.)
Narragansett, ’nnin, inin; pl. ninnuog. R. W.
Menomini, tnin. Potawatomi, nini (Lykins), entn & (P. Jones).
Saki (Sauk), nenni. Maximil. Musquaki (Foxes), nini.
Montagnais, trini-ou [he is] man; triniou-in “life.” Le Jeune, 1634.
Abnaki (Kennebec), aren-t; as adjective ‘simple,’ ‘ plain,’ ‘ mere.’
Quinnippiac, ren (pl. renewak) man. Peirson, 1658.
New Sweden, “rhenus, Mann : renappi, Menniskis. ” Campanius.
Delaware, lenno, pl. lennowak. Zeisb. "
Shawnee, dint “man,” lindwat “ Indian ;” lindwai-wt “he lives.”” Howse.
delnoieh “Indian.” Whipple.
Illinois, :ini.
Miami, elaniah (Volney), ahlanuah (Barton).
Micmac, el’nu, l’naot (Maill.), al’nu (Howse).
Montagnais of Labrador, il’no.
Cree, ethtn'u “man, an Indian,” Howse; Western Cree, Aiyenu, Maximilian. {[Comp. 4st pers. pronoun nitha I, and net’étin I do so, I so act. Howse remarks that “the th is so softly uttered that a nice ear only can detect it,” and, among the western Crees, it ‘‘is lost in the 3 or y ; nitha becomes niya [= niia], ethinu is fyinu. The western Crees call themselves
19
146 J. H. Trumbull.
Néayaég which Dr. Hayden translates: “those who speak the same tongue.” Of Néhethéwuk, the equivalent in the dialect of the Hudson Bay Crees, Howse makes “exact beings, or people,” and Sir John Richardson, ‘exact or complete men.”’}
‘Shyenne, ita’ni (adj. male, of man) ; eta/nio “ people.” Hayden.
Atsina, nithun'a. Comp. nathan't-nita “to live.” Hayden.
Arapaho, inen!, pl. inen'a. Comp. tninek'tina “to be alive.” ? Blackfeet, ninnow, nenow. Howse. Hayden has nin'’a “chief,” but for Ia- dian, ni-i!-tsa-ta!-pi. Comp. nistu’a [= Cree nita] I, nitsinan mine. Powhatan. The generic name appears in such compounds as Strachey’s rawERUNNUwA “an old man.” For ** man,” John Smith has nemarough (by a misprint, probably, for nematough), and Strachey, ntmatewh. This is the equivalent of nemat (Strachey; and so in the Massachusetts dia- lect,) “my brother,” my mate, with the verbal formative (= Mass. ne- mat-ou he is my brother, or mate).
Nanticoke, thn, iin, “Indian.” Wohacki, for “man” in Gallatin’s vocabu- lary, means ‘his body,’ ‘himself,’ = Mass. wukhoghi (Eliot), Narrag. wuhhoek (R. W.).
The characteristic of the pronominal root is constant throughout. The prefixed demonstrative, or reduplication, varies, with changes of dialect, as », /, r, and (rarely) y; is lost in strong aspiration of the following vowel; becomes a soft, scarcely audible th in the speech of the eastern Crees and the Atsinas of the northwest, and among the Shyennes is represented by ¢.
Without intending to follow the Algonkin name beyond the presumed limits of the Algonkin group, I may be permit- ted to allude to the fact that the Crees and Atsjnas are neigh- bors of Athapascan tribes, suggesting the possibility of rela- tionship between the Cree ithi'nu, Atsina nithun’a, and Shyenne tta’ni,—and the Chepewyan dinnie, Takulli tennt, Umkwa tiiné, Navajo dennat’ and Apache n’de, all having the same meaning, “man, native.” The likeness of the east- Algonkin ’nnin-u, inin-i, to Labrador-Eskimo innuk, pl. innuit, man, innu-wok alive, una he, this, ingna the same, etc., 1s not less noticeable.
II. An adult male: designated by the inseparable noun- generic -a"be, -d¢p, -omp, or other dialectic modification of the root a"B. With n’ demonstrative prefixed it designates ‘ the male,’ or as an adjectival, simply, ‘ male.’ With a pronomi nal prefix, it may mean ‘ husband, —e. g. Chip. ne-nabem [the final m is possessive,] my husband ; but Baraga in his
On, Algonkin Names for Man. 147
Otchipwe Dictionary, marks this use of the word as “ unpo- lite.” With the prefix ’nen (ren, len,) it denotes a ‘common man,’ 1. e. an Indian adult male. With other attributives, it forms class-names and tribe-names.
1. With the demonstrative prefix, designating ‘ the male’ ; and, with the possessive suffix (-om, -em, -m), ‘ husband’ :—
Algonkin and Chippeway, ndbé ; ni-ndbem [my male,] my husband.
Ottawa, ndpé; ni-nd-bom.
Menom., naupe-om, ndpium, “husband.” Dr. James.
Potawat., nawbam ; nin-nawbam “my husband.”
Miami, [ndpem husband,] nenapéma my husband. Volney.
Illinois, nampeheman husband. Gallatin.
Montagnais, napiou “ man,’ _ndpen “husband ” (naapen. Gabriel).
Naskapi (Scoffie), naabouh “ man,” naahpen “ husband.”
Abnaki, na"bé-, prefixed to names of male animals.
Massachusetts, nomp ads male, a male; nomp-oshim, nomp-oshimwus a malé quadruped ; nompai-yeuo [he is] male, Gen. vi. 19. Eliot. Roger Wil- liams does not use omp- or nomp- as a prefix ; but for a male beast has enewdshim, = Del. lenno-wechum (Zeisberger),
Cree, ndpayoo man, pl. napeywuk: ne-nabem ‘“‘my husband.” Howse.
Nanticoke, nddp. Heckw. Pamptico, nuppin “ Indian.”’? Lawson.
2. With the prefix ’nen, (ren, /en,) ‘ common,’ ‘ native,’ ‘ of our kind ;’ designating an Indian adult male:
Abnaki (Kennebec), aren-a"bé “homo ” Rasles.
(Penobscot), alnombé ; alnambay. Vetromile. - Delaware (N. Sweden), renappi, Campanius — who has, incorrectly, piri renappi for ‘‘ strangers.” (Unami) len-dpé “an Indian,” pl. len'dpewak. Zcisb.
Mississauga, linip? [‘‘linneep.” Barton.]
3. With other attributives, forming class and tribe names:
Mass. Mitomp, Narrag. nitép, Abn. nida"bé, Del. (N. Swed.) nitappi, Powhatan netab, pl. netapewh (Strachey), nitoppu (J. Smith),—the familiar “ netop” of the early colonists, sometimes translated: “ brother,’’ but by Roger Williams, more accurately, “ friend,’”’—denotes a brother by adoption or affinity, one who is regarded as a brother; literally, ‘man of my family,’ or ‘my kinsman.’ The prefix (Mass. nit-) may be translated ‘of the family, ‘domestic’; as in Chip. néta “my brother-in-law”? (Baraga), Mass. and Narr ag. nitassu (netassu, El.) a domestic animal.
Mass. Ket'omp (kehtomp, El.) chief r man ; from keetti (kehte,
El.) chief, greatest.
148 J. H. Trumbull,
Mugwomp great man, captain; from mogki great, powerful.
Kinomp (Abn. Kina’bé “ homme courageux, brave, géné- reux,” Rasles), a “‘ brave”; Eliot uses it for “ captain ”’ in John xviii. 12, where Mayhew (1709) substitutes mukquomp = mugwomp ; Micmac keenap “ warrior, hero,” Rand.
Nonk'omp young-man ; literally, light or slender man, from: nonk'2, levis.
Pinomp (penomp, El.; Del. piuldpé “a big boy,” Zeisb.) a new (i.e. a chaste) man: from pinu (Del. pilt-, Chip. 5272-,) new, strange, unused, chaste. Perhaps the most curious mis- take in Eliot’s version of the Bible is the use of penomp for “ virgin,” e.g. in Gen. xxiv. 16, Isaiah vii. 14, 1 Kings, i. 2, and Matt. xxv. 1, where the parable is of the ten penompaog, i.e. chaste young men. With the Indians chastity was a mas- culine virtue, and it is easy to see how Eliot’s interpreter, misunderstanding his question, gave him nescius vir for nescia virt.
Delaware Kigdpe (Zeisb.), Abn. kiga"bé, a young unmar- ried nian, is in those dialects the equivalent of Mass. pinomp. The corresponding feminine appellation in the Delaware (Unami) is-kikochque, Zeisb., Ottawa gigang “ virgin, maid,” Baraga. Blackfoot astt’-api = Del. kigdpe.
Abnaki (Kennebec) seé'na"bé, modern Penobscot senomdt, Mass. sannup (Wood, 1634), was the common designation of an Indian man, in the vigor of manhood, married, or master of his lodge. Rasles translates it by “vir.” The signification of the prefix is not quite clear. The word is not found in Eliot or Mayhew, but was much used by the English colon- ists, who understood “sannup and squaw ” to mean “ Indian man and woman.” Possibly, the former name is a contrac- tion of anisina’bé—which, in other Algonkin dialects, has nearly the same meaning, but is not found in the Massachu- setts of Eliot or the Abnaki of Rasles.
Old Algonkin (Nipissing) alisinapé, Lahontan; mod. Alg. and Chip. antshin-abe (“Indian”) Baraga; Ottawa nishan- dba; Potawat. nishinapé ; Penobscot wlisenombi “good man.” The same prefix, with zrint (= inini) as the generic, is found in Montagnai’s artchi-irini-ouak (pl.), men (Le Jeune, 1634),
On Algonkin Names for Man. 149
modern arrishirini (Vetromile). Compare Blkf. ntitsatdpt ‘Indian’? (Hayden). The prefix signifies ‘good,’ ‘ well- doing’; Mass. wunnesu, Del. wulisso, Alg. and Chip. onijishi “he is fair, beautiful, fine, good.”’ (Bar.)
Narrag. enisketomp contr. ’skitomp [skeetomp, R. W.] “man”; Quinnip. eansketambe “an Indian,” wusketambaug (pl.) “ men,” “ people ”; contr. sketambaugh, Peirson ; Mass. wosketomp, Eliot, Mayhew, and Cotton, for “ man”; Etche- min oskitap, uskidab, “man,” n’oskitapaim “my husband”’ (comp. uskitch-inu “Indian man,” Vetromile, = ouskejin, Barrett, and Micmac uskiginu “Indian man,” Vetromile) ; Naskapi (Scoffie) of Labrador, nashkapou [he is] Indian (Ga- briel), naskupi and “ nasquapee,” which, says Mr. Hind (£z- ploration of Labrador, ii. 96), they translate, “ people stand- ing upright.” The prefix appears to be the equivalent of Chip. onishk-, in onishka “he rises, stands erect,” particip. wenishk-ad “ one who stands erect,”’ etc. — repeating and em- phasizing the meaning of the generic -omp, -dp.* In May- hew’s version of John’s gospel (1709), unashketomp, pl. -paog (not found in Eliot,) is used for “officers,” ch. vii. 46, xvii. 12, 22, and in Wood’s vocabulary, Mass. (1634), sasketupe is translated “a great man.”
Blackfoot mata'pi “ man” (Hayden) belongs to this class. The prefix may be from ma’ts: “ brave.”’ The generic affix, for “ male,”’ is found also in asi’taPi “ young man,’ sako'taPi “boy,” nii/tsataPi ‘ Indian,” and in the names of Blackfeet bands, e. g. A’petuPt “ Blood people,” Mumitupio (pl.) * Fish Indians,” etc. Comp. BIkf. etapr ‘‘ to live”? (Hayden).
The Micmac designation of an adult male is peculiar. Gal_ latin’s vocabulary gives (from Maillard) Micm. tchinem “ man (vir): tchenem-emool *‘ husband ;” [em is possessive, and -ool is dn affix of the 2d and 38d person sing., ‘thy’ or ‘her’. ] Rand’s vocabulary (in Schoolcraft) has n’cheenum-ocom “ my husband ”; wobatka-cheenum “ white man” (but this last is probably white man’s Micmac, of modern formation). Only. in nilhetop “ my friend ’’ (Gal.) = Abn. nita"be, Mass. nitop,
* Comp. Del. LENni LeENape, and (in an other group of languages) the Pawnee tribe-name Cha!-hiksi cha/hiks “‘men of men.
150 J. H. Trumbull,
do I find the generic suffix for‘ male’ which is common to all Algonkin languages. The tribe-name—the true vir— corresponding to the Alg. and Chip. anishinabé, Del. lenno- lenape, Naskapi nasquami etc., does not appear in the Micmac vocabularies. The etymology of tchinem is obscure. It may be a dialectic corruption of Abn. seé'na"be (Mass. “sannup’’), with the loss of the p by the nasalization of the preceding vowel.
III. Man inferior in degree or kind; not ‘of us’ or ‘such as we are’:
Mass. missininnuog, Narr. missinnuwock, “folk, people” (R. W.),—if of the speaker’s nation, ninni-missinnu-wock,— has been previously noticed. Literally, “the many,” o¢ roAAoi : Abnaki mesairwak “ ils sont plusieurs” (Rasles). From the same root, Mass. musst (and redupl. mdmussi) wholly, of the whole; Narr. mitsst-su “the whole of him”; Del. messisu ; Abn. messtooi *“ fout entier’’; Chip. misi, misiwe, “‘ every where,” ‘ all,”
Mass. penwi ( venbwe El.) strange, novel, different —whenee penovwot stranger, foreigner; pl. penmwohtedog strangers, is used by Eliot for “the heathen,” Ezek, xxxvi. 3, 4, and else- where, and for “gentiles.” Abn. pirwi-arena"bé “ homme étranger” (Rasles), pirt ‘de nouveau,” = Del. pili, Chip. bini, ete. The Chip. maiag- has nearly the same meaning, —‘ foreign, strange, changed” (Bar.); maiag-anishinabé “a strange Indian from another tribe; in Scriptural lan- guage, pagan, gentile,” matag-isi “ he is a foreigner ” (Bar.).
Mass. howdn, auwon; pl. howanig, somebodies, any-bodies, or interrogatively, who is this? who are these ? (Narr. awdiin “there is somebody,” awdiin ew6? “who is he?’ R. W.) As an adjective, howaé any, some kind of. Abn. aovenni, Micm. wen, Del. auween; Cree, owena who? pl. owéneki; ow'eukk some one, any one; Chip. awénen who? Hence one of the designations of Englishmen by the Indians of New England,—usually written awannux or owanuz; Narrag. ‘“« awaunagus-suck Euglish-men, . . . as much as to say, These strangers” (R. W.), Pequot waunnuksuk (Stiles). Abn. amennots ‘Frenchman ”’ (Rasles) has the same etymology.
On Algonkin Names for Man. 151
IV. Nations of different language, enemies, and Europeans, were usually designated by a verb or participle in the ani- mate-plural, without affixing @ noun-generic. The principal tribe of the lroquois, for example, was called by the Algon- kins of New England ** Mohowaug-suck or Mauqudu-og, canni- bals or men-eaters”’—as Roger Williams explains (Key, p 16)—* from mého to eat." Eliot writes this verb, mawhau‘ he eats what lives (or an animate object) ; noh moohhukque “he that eateth me,” John vi. 57; mawhauqua-og they who eat what lives, etc. Hence, the name “ Mohawks” adopted by the English, and the Dutch Mahakuaas, contracted to Ma- quas.* (Comp. Abn. ne-m@ha"’ak mégawak “I eat the Iro- quois,” Rasles.) The French and northern Algonkins may have derived the same name, * Maquas,” from Alg. makwa a bear,— Ganniagwari, the national name of the Mohawks, sig- nifying ‘“‘a she bear”; but it is nearly certain that to the a dians and English of New England, the “ Mohawks ”
‘* Mauquauogs ”’ were, by name, * cannibals.”
A Mohican tribe in eastern Connecticut received from their enemies (Narragansetts and Niantics) the name of Paqua- tauog, or Pequtt6og (R. W.), destroyers, ravagers, and passed into history as “ Pequots,” only a small band, which had de- serted the main tribe, retaining the national name of AlvA- hekanneuk (Wolves) corrupted by the English to ‘ Mohe- gans.”’
The “Eskimos” bear an Algonkin nickname which describes them as “eaters of raw flesh’; Cree eskwa-mwayw, Abnaki eski-moha", he raw-eats (animal food).
The name Algonkin—Algoumequin and Algonquin of the French—has been extended over a great family of nations and languages. “The Algonquin was the mother tongue of those who greeted the colonists of Raleigh at Roanoke, of those who welcomed the Pilgrims to Plymouth. It was heard from the Bay of Gaspe to the valley of the Des Moines; from Cape Fear, and, it may be, from the Savannah, to the land of the Esquimaux ; from the Cumberland River of Kentucky to
*“ The AMauquawogs:or Mohowawogs, which signifies men-eaters.” R. W. in Letter to Winthrop, 4 Sfuss. Hist. Collections, vi. 239.
152 J. H. Trumbull,
the southern branch of the Missinipi.”* Yet the origin of the name has, I believe, never been pointed out, and scarcely two authors agree in fixing the locality of the tribe to which it originally belonged.f Mr. Gallatin (Synopsis of the Indian Tribes, p. 24) found it “difficult to ascertain whether this name did belong to any particular tribe, or was used only as a generic appellation.” Etymology removes the difficulty. An Algonkin was, eo nomine, removed from all “ local habi- tation.” No tribe ever called itself or was known to neigh- boring tribes by the name. It was not even a “ generic appellation,” until the French and English adopted it 4s such. We first meet with the “ Algoumequins” in Champlain’s narrative of his voyage to Canada in 1603 (Les Sauvages, cte. repr. Quebec, 1870, pp. 6, 8, 9). He was in company with M. du Pont-gravé and had as interpreters two Indians of some Algonkin-speaking tribe—probably Montagnez from Tadoussac,— whom Pont-gravé had carried to France on his return from a former voyage to the St. Lawrence. At Pointe de Saint Matthieu (now Pointe aux Allouettes) at the mouth of the Saguenay, opposite Tadoussac, they found a war-party of Indians “of three nations, the Estechemins [Etchemins], A/goumequins, and Montagnez,” returning from a successful expedition against the Iroquois. The Montagnez were already at home, on the north bank of the St. Lawrence ; the Etchemins and their country were well known to the French, but the *“* Algoumequins”’ were new acquaintances. Their name—or what Champlain understood to be such— must have been learned from themselves or their allies, and must belong to one of the dialects which we call Algonkin.t The 7 is clearly an interpolation, for it does not belong to
* Bancroft’s History of the United States, iii. 237.
+t See, in Shea’s Charlevoix, vol. ii., p. 8 (note 3), a collation of the princi- pal authorities — exhibiting, as the editor remarks, “most remarkable differen- ces of opinion ” on this point.
¢ The learned author of Etudes Philologiques sur quelques Langues sauvages, in a later work (Jugement Erroné de M. Ernest Renan, etc. 2me éd. Montréal, 1869) which did not come in my way until after this paper was presented, derives the name “ Algonquin” from the Huron, a dialect of the Iroquois. The Harons and the Algonquins were allies, he remarks: the former, impatiently awaiting the
On Algonkin Names for Man. 153
the Montagnais, Etchemin, or any other Algonkin language at that time known to the French. The termination -in, or as it was afterwards occasionally written -ain, is that of the French adjective (as in Mexigquain, or -cain), but it perhaps represents, as in some other tribe-names of French adoption (e. g. Champ- lain’s Quenongebin, Ochataguins, Otaguottouemin, etc.), an original -inin ‘man,’ or its plural. In Champlain’s later pub- lications ( Voyages, editions of 1619 and 1632, and the Map,) he writes ‘ Algommekins”’ for ‘* Algoumequins.”? We have then as the base of the name, A’goumek or A’gommek—and recognize an equivalent of the Virginian Accomac, the Nar- ragansett Acawmen-cake “jand on the other side” or acdw- muck (R. W. Key, pp. 3, 4), Mass. o*gkomuk and ogkomuk (Eliot), Abn. a"’gma"mek “en dela,” “ au-deld,” and Aga’- meno'ki ‘ France” (Rasles), Cree akémik (Howse), Chip. agéming (Baraga). Among the Montagnais at Tadoussac, or by the Etchemins of l’Acadie,—anywhere, indeed, east of the Ottawa River,—the original Algonkins would have naturally been designated by their eastern confederates as men from “the other side,” from “the beyond-river country.’ The, editor of the Quebec reprint of Champlain’s voyages, in a note to the account of the first meeting with the “Algoume- quins ” in 1603 (Les Sauvages, p. 9), suggests, unconsciously, the derivation of the name, by the remark that they lived on the Ottawa River “ et au-dela.’”*
ree
coming of their friends, used to ask one another, 1AKO-KEN? which is Huron for ‘ Est-on arrive?” And /ako-ken, at first “une sorte d’appel militaire,” came to be the recognized designation of a tribe and nation, and finally was corrupted to Algonquin! That the “ Algoumequins”” whom Champlain met on the lower St. Lawrence in 1603, years before he visited the country of the Hurons or promoted the Algonkin-Huron allinnce, made themselves known to him by a name bor- rowed from an “sppel militaire” in a foreign language — and which required an interrogation mark to give it mcaning — is, to say the least, improbable. With- out raising the question whether even French ingenuity could extract “ Algon- — quins” from “ Jako ken?’’ —is not such a derivation of a tribe-name as absurd as the worst of the etymological blunders of Schoolcraft and Duponceau which the author of Ktudes Philologiques has so glecfully exposed ?
* The “ Algoumequins ”’ encountered at Tadoussac in 1603, appear to have be- longed to the tribe which afterwards became known to the French as Kiche- sipiriniuek (i. e. Gireat-river men) and “ Sauvages de l’Isle.”” These occupied the Tle des Allouottes (as it is now called) in Ottawa River —the “Great River of
154 | J. H. Trumbull,
In the Jesuit /telations, the name changes from “ Algomme- kins”? to ‘ Algonquaines’’ and, finally, ‘“‘ Algonquins.” This change was perhaps effected by the influence of the Huron name for the same tribes. The Hurons, who spoke a dialect of the Iroquois, designated their “ Algommekin”’ allies as ‘© Aquannaque,” i.e. “of a different language,” “ foreigners ”’ (Sagard). ‘“ Our Hurons ’’—writes Father Lallemant in the Relation for 1641 (Quebec edition, p. 72),—* call the Neutral Nation Attiwandaronk,* that is to say, ‘ people of a slightly different language’; as for the tribes which speak languages | which they (the Hurons) cannot at all understand, they call them Akwanake [ = Aquannake of Sagard], of whatever na- tion they may be, that is to say ‘strangers.’”? The Huron name became more familiar to the French than that by which the tribes on Ottawa River had first been called,—these tribes, when at home, could not properly be designated as * from the other side,”’—and there was sufficient resemblance between a gwa'mek and a kwanake to make the transition from Algomme- kin to Algonquin easy. |
The Chippeways call the modern Algonkins, Odishkwa- gamig ‘ Lake-enders,’ from ishkwa at the end of, and gami lake (literally, water). Mr. Schoolcraft gives a translation and analysis of this name—of which he seems to have re- garded “Algonquin” as a corruption or the equivalent.
The eastern tribes gavé, as we have seen, the same name to countries of Kurope as to the region between the Ottawa River and the great lakes: Narr. acawmen-éake, Abn. aga’- menw'ki, Chip. ag@ming, ‘land on the other side’ or ‘over the water.’ To the Nipissings and the Montagnez, the French and English were “ Algonkins.”
The French in Canada were called sometimes Awennats- ak‘ somebodies’ ( = Narr. awaunagussuck “ these strangers,” see p. 150, ante); but were usually distinguished as ‘ Wood en-boats’—Alg. Mittigouchiouek CLahontan, who translates,
a
the Algommekins ‘of Champlain's later voyages and Map of 1632. Perhaps the appellation Achesipirinivsck was originally given to all the tribes and bands liv- ing on or near the “ypreat river” (hitchi-sipi), to distinguish them from the ‘“‘small-lake men ” (Nipissirinimek) dwelling near Lake Nipissing.
* Whence probably the modern .Adirondack.
On Algonkin Names for Man. 155
inaccurately, “‘ constructeurs devaisseaux ”’), Chip. wemitigoji- wag (Baraga); Cree Wem’slégoso-ak.
The English in New England were specifically described as ‘* Coat-wearing ” (Narr. Wautaconduog, R. W.), but soon re- ceived the appellation by which Anglo-Americans, and. since the separation of the colonies from Great Britain, the in- habitants of the United States, have been designated by all — northern tribes,— Big Knives.” ‘They call Englishmen Chdugquaqu ock, that is, Knife-men,” from chauquog knife (R. W.). In other dialects, different names for ‘knife’ are em- ployed: e. g. Alg. and Chip. mokoman, whence .Chip. Kitchi- mokoman “an American ’”’ (Baraga), and Aitcht-mokoman. aki [great-knife-land,] the United States ; Cree, Ketsimohkoman. Del. Mechan-schican, ’Chanschican (Heckw.), Miami Mitchi- malsd (Volney), Blackfoot Omakstod, and Arikara Nehstkuss all have the same meaning, though formed from different roots. .
The Alg. Aganesha, Chip. Jaganash and Saganash, Cree Agathasu, Hakaiahsu, Miami Ayd@ldchima (Volney), and pro- bably Yengees—by double corruption, “ Yankee,’’—represent Algonkin imitations, more or less successful, of “ English,” “ Anglais” or ‘ces Anglaises.”
There are Algonkin names for “ whites”? and “ blacks,” but these are without any generic.affix to restrict their ap- plication to ‘men’: e. g. Chip. Waidbishkiwed “‘a white man ora white woman ” (Baraga), a participle (subjunctive) from wabishkiwi to be whitish, pale,—and Miami Oudbkiloketa “white skin” (Volney): Chip. Adaketewiias and Cree Kis- kitowtias, ** black flesh ’’; etc.
V. For Woman there are names corresponding nearly to femina, mulier, and uror. The first—which has been angli- cized from east-Algonkin dialects, as “‘ squaw,” — as a gen- eric suffix denotes one ‘of woman-kind,’ as a prefix signifies ‘female,’ without restriction to the human species. Eliot did not employ it independently for “‘woman.”’ In Gen. vi. 19, he wrote ptsh nompai-yeu-w kah squai-yeu-m “ they (animals) shall be male and female,” but in Gen. v. 7, wosketomp kah mittamwossis-soh ukkezheiih “male and female (aman and wo-
156 J. H. Trumbull.
7
man) created he them.”” Witha suffix denoting ‘ living crea- ture, ‘ animal,’ — squd-as (contracted by R. Williams to sgudws) is ‘a female,’ without distinction of age or condition. So, sgudshim (squd-oshim) a female quadruped, Abn. skwés- sem, Del. ochquéchum. It has the place of a noun-generic in the Mass. nunk-squé young woman ; Narr. keegsquaw virgin (R. W.); sonkisqud, contr. sonsq’ (and Narr. saunks, R. W.) mistress, sachem-squaw ; etc.
Though this general name is found either as an indepen- dent word or as an element of synthesis in every Algonkin language, it is not easily traced through the published vocab- ularies, in which it is often confounded with or represented by names for multer and uzor. It does not appear under ‘*Woman,” in the Micmac, Etchemin, Abnaki, Massachu- setts, Mohican, or Miami vocabularies given by Mr. Gallatin, but it occurs in some of these under “ Girl” or “ Wife.”’
Old Algonkin, ickoue, Lahontan. Chip. ikwé, Baraga, ecquoy, Long.
Ottawa, akwe, Bar., ekwa, Tanner. Potawatomi, dkwé, ukgquah, Gal., ogué, qué Lykins.
Delaware, ocHque-u woman, Zeisb.; H’quelt woman, quai !tchitz (dimin.) girl. Whipple. New Swed. aqeo; as a suffix, -éque. Campanius.
Nanticoke, achquahzke ; suffixed, in pecnquaH girl. Gallatin.
Shawnee, equiwa, dimin. squithetha girl, Johnston; s’squawowdh, dimin. s’squaw the e thah girl, Whipple.
Powhatan, -usqua, in wironausqua “woman queen”; dimin. usquaseins “ girl.” Strachey. [For “woman” Strachey has cucheneppo, cutsseneppo; J. Smith crenepo ; of which I can make nothing.
Mohican, -esgua: in peesquasoo girl, Edwards; peesqudthuh, Jenks.
Mass. and Narr. squa-, squée, female ; squads, E)., squdws, R. W., a female.
Abnaki (Kennebec) skové- (prefixed) and insep. -skaé, female; as in na”k- skooé ‘“‘ fille,” koossthoo-skooé “‘vierge.’’ Penobscot, kosiuskwe virgin, Vetromile.
Etchemin, -sque. Pelsquasis girl, Kellogg ; noksqus-ak “girls” [young women], Barrett
Micmac, -shquei [-chkoei, Maill.] insep. generic ; contractde, -ishk. [Na"zkwé, narkwe, Vetromile, the equivalent of Abn. na"k-skooé, Mass. nunksqud, young woman, has been improperly used by some of the Catholic mis- sionarics (I find it also in Mr. Rand’s vocabulary—as noksow) for “ vir- gin.” In Vetromile’s ‘Indian Good Book,” Naxwhet Mali stands, in the Creed, Rosary, ete: for “the Virgin Mary.” The prefix (Abn. na”k-, Mass. nunk-, nonk-, Chip. ndng-) means ‘light’ (levis); in this connection, ‘ not full grown’: comp. Mass. nunkomp young man. Naz- whet (Vetr.) is the participle of na"rhwa.
Montagnais, schqguow woman; dimin. sguastsh girl: comp. tishquah [his]
On. Algonkin Names for Man. == ~ 107
wife. Gabriel. [TY essarawi- and the participle tessarawit used for “‘ vir- gin”’ in the Montagn. Prayers, Creed, Confiteor, etc., in Vectromile’s Indian Good Book, (ec. g. Mari eiamitsh tessarawit Maria semper virgo,)
_ to an Indian denotes — like Eliot’s penomp —a chaste male. It becomes feminine only by suffixing the generic -shyua. Comp.Alg. and Chip. “‘nintéssanaw I am in a virginal state (a male speaking),’’ participle taies- sanawid, and “nin tessanAkKWeEw I ama virgin (a female speaking),”’ ptep. taiessanakwid. Baraga.
Naskupi (Skoffie), schow woman, squash girl, teshquouet wife. Gabriel.
Cree, tskway! oo [she is] woman, of woman kind. Howse.
Blackfoot, ski-, a fem. prefix to names of animals: but ahi/ma woman, pl.
akiks; aki'kuen girl. Hayden.
In the far-off Arapoho isi! woman (and as fem. prefix), and the correspend- ing Atsina (Falls Indian) ith’a and ithe'i, we nearly lose trace of the harsh guttural ocHyue of the Delawares and the Alg. ikwe.
For mulier we find in different Algonkin languages at least three names:
(1.) Abn. phainem [p’hainem], Rasles; mod. Penobscot, phanem (“ sanoba ala phanem man or woman,” Ozunkh.) ; Mohican p’ghainoom, Jenks.
(2.) Micm. epit, pl. epttgik, Maill. ; abit woman, aibitis girl, n’t’aibit-em [my] wife, Rand.
(3.) Mass. mittamwossis, contr. mittamwus El. (muttumwus Mass. Psalter), used both for mulier (Gen. ii. 22; iii. 2,) and uzor (Gen. ii. 24, 25; iii. 8; Ephes. v. 22); Narr. mittémus woman, wife, R. W.; Miami metaimsah, Schooler. Vocab. ; Chip. mindimoté old woman; mindimotémish, always pre- ceded by a possessive pronoun, “ wife, [my] bad old woman.” Baraga. The affix -ish is derogatory, but is not always to be translated by ‘bad.’ ‘ My poor old wife’ is better — if, as is not certain, Baraga’s analysis of the word be correct. The Abnaki mana’-dagwéssw, which Rasles gives (with p’hainem) for “femme,” is probably an equivalent of Chip. mindimoié- mish. Comp. Powhatan utumpsets, tumpsis, old woman (Strachey) ; Menom. metamo woman (Schoolcraft); Shyenne matum'ha (Hayden).
The names for wxor need not be considered, in this con- nection. For “my wife” the Indian ugually said “ my wo- man,” and in the second and third person the feminine gen- eric (-sque, -kwé) suffixed to a man’s name or title designated his wife: e. g. Chip. ogima chief, ogimdkwe the chief’s wife.
158 J. H. Trumbull,
The principal results of the analysis which has been at- tempted in this paper may be briefly recapitulated, as follows:
1. There is no Algonkin name for Man ( = homo) com- mon to both sexes and to all varieties of the human species.
2. The name of largest denotation is one which designates Man as a being of the speaker’s race and language, his like, of his kind or kin. .
3. This name (Alg. znzn-z, Mass. enin-u) is related to the pronoun of the first person (Alg. nin, Mass. nin, prefixed, n’), to the demonstratives animate and inanimate, to various words expressing likeness, relation or identity ; when used as an adjective, it distinguishes the common, usual, and native, from the strange, unusual, or foreign; and it is the theme of a verb meaning ‘to live’ i. e. ‘to be a man,’ to be such as other men. The root of this name, if not identical with, is not distinguishable from the root of verbs meaning ‘to think,’ ‘to be minded.’
4. Only the second n of the name belongs to the root (IN or IN). This is constant in all pure Algonkin languages. The prefixed demonstrative (or reduplication) varies in dif-— ferent dialects as in-, en-, ar-, el-, eth-, et-, etc.
5. Names for Man — wr are formed by prefixing attribu- tives to the inseparable noun-generic (-a"B, -AP, -OMP) de- noting an adult male. With a prefixed demonstrative, this generic forms the adjective na"be, ndbé, nompé, ‘male’: with the adjectival tnin- ( —aren-, len-, etc.) it designates, as in Del. len-épé, ‘a common male,’ i.e. an Indian man: with other attributives, it forms class, tribe, and specific names, e. g. Alg. antshin-abé, Abn. seé'n-a"bé, Mass. wosket-omp.
6. Inferiors, enemies, and Indians speaking a different language, were designated as “ slaves,” “captives,” “ stran- gers,” or merely ‘“ somebodies”’; collectively, as “the many,” oiwoAdoi. Names given to Kuropeans and to foreign tribes were sometimes formed from inanimate nouns, e. g. “« Wooden Boats,” for Frenchmen; “ Big Knives,” for Anglo-Ameri- cans ; sometimes from verbs or participles animate, as ‘“ Eat- ers of raw flesh,’ for the Eskimos; “They who eat what lives, or is alive,” for the Iroquois ; “* The Clothed ”’ or “ Coat-
On Algonkin Names for Man. 159
wearers,” for Europeans. Kindred and friendly tribes were often designated by their geographical position : the Mipissirini- wek, (Nipissings) and other tribes between the Ottawa and the lakes were, to the Montagnez, A"gowmek “on the other side,” the Indians of Maine were “of the east land ” (Abnaki), to western Algonkins.
7. For Woman, some modification of the root of Chip. 7xws, Mass. esqua, ‘ femina,’ is found in every Algonkin language, as an inseparable generic if not as an independent name. It is the common appellation of both multer and wzor, and its diminutive, of pwella: but there are distinct names for mulier and uzor in every language, as there are also for juvencula and virgo, though Eliot does not appear to have discovered in the Massachusetts dialect either of the last two, and one has often been mistaken for the other in the compilation of vocabu- laries and by translators.
VIII.—On Some Forms of Conditional Sentences in Latin, Greek, and Sanskrit. |
By JAMES B. GREENOUGH, TUTOR IN LATIN IN HARVARD COLLEGE.
In the treatment of the Protasis and Apodosis or Hypo- thetical Period the so-called General Supposition, since it was first distinguished by Professor Goodwin, has, among Ameri- can scholars at least, been generally recognized in Greek, but in Latin it has not commonly been admitted to exist except as a Grecism.
It is useful for a logical treatment of the subject in Latin, as well as for the purposes of comparative syntax —a branch of the science of language that has not as yet been much at- tended to — to determine whether this form of protasis and apodosig is an heirloom of the Indo-European family, or only a special development in the Greek, which was afterwards borrowed by the Latin.
Now the nature of this form is this: the protasis or con-
160 J. B. Greenough,
dition does not refer to a single act or event, in present or past time, upon which the conclusion is founded. This form it is the province of the indicative in both clauses to express. Nor does the protasis refer to a future act or event upon the hapnening of which the apodosis will be true. This form is expressed either by the optative in both clauses, or by the subjunctive in the first with the future indicative in the second. But what the protasis does refer to, is the indefinite or repeated happening of an act or event in all time (or at any past time), upon every occurrence of which the apodosis becomes (or became) true. The formula for this kind of hy- pothetical sentence is: ‘‘If ever he does (or did) this, it is ‘(Cor was) in all-such cases well,’”? — and this idea is clearly distinguishable from “If he is now doing this, it is well,” as well as from “If he should do (or does) this, it would (or will) be well.” And in Greek this distinction is carefully maintained. The first is ’Eay rovro Tparri) kadoc exer. The sec- ond is 'Ei rotro xparre cade Exe. The third is "Eay rovro xparry kada@e te. To use Prof. Goodwin’s examples:
"lly wore Saopog ixnrat aot rd yépac roku peilov, Ll. i. 66. If ever (that is, in all cases in which) a division comes, your prize is always greater. -
"Re ree avreiroe evSuc reSvixee. Thuc. viii. 66. If any one re- fused (in all cases of refusal) he was put to death.
So also with all indefinite relatives and relative particles :
"Hyix’ &y 8 oixor yéerwrrat, Cpdow obk avacyera. Arist. Pax. 1179; "Ore ikw rov detvov yévouvro rodAvt abrov azéAecrov. Xen. Anab. ii. 6.12. The protasis is expressed by the subjunctive or optative, according to the time of the condition, but the apodosis is some form of the indicative expressing a general truth either in the present or past. Occasionally however the protasis also is ex- pressed by the indicative, like the particular condition pres- ent or past, a fact which is significant as throwing light on the Latin usage. Now in Latin, at least in the monuments of literature, this form of construction is not ordinarily dis- tinguished from the particular condition in present ‘or past time, so that the exceptional Greek is the usual Latin form. There are however some traces of this construction in Latin
On some Forms of Conditional Sentences. 161
which agree with the Greek though from their rarity they have usually been considered as imitations.
It has long been noticed that the second person singular of an indefinite subject has a kind of affinity, so to speak, for the subjunctive mood. It is obvious that there is no magic about this second person which should change the mood of the verb when with any other person it would be indicative; the dif- ference must be in the nature of the thought expressed. Upon examination these cases are found to be of two kinds. They are always either protases or apodoses. The protases of this kind correspond for the most part to two forms of apo- doses. Their apodoses are either in the subjunctive or some other form referring to the future, or in the present indicative used to express a general truth. In the first case they are clearly cases of future protasis and apodosis of the ordinary form: e. g. In qua (amicitia) nisi, ut dicitur, apertum pectus videas, tuumque ostendas, nihil fidum, nihil exploratum habeas. Cic. Laelius, xxvii. 97. So also, Ubi enim istum invenias quit honorem amici anteponat suo? Ibid. xvii. 64,—where the pro- tasis is omitted,an exceedingly common usage. So, Quod st eti- am possts quidvis deferre ad alterum, videndum est tamen quid ille possit sustinere. Ibid. xx. 713, — where the participle in dus takes the place of a future form. Nowin the second case this construction has never to my knowledge been analyzed, but it seems to me unmistakably the same as the Greek general sup- position. Take for instance, Vita humana prope uti ferrum est, st exerceas conteritur. Cato. Carmen de Moribus, cited by Aulus Gellius, xi. 2 (Jordan, Religuiae Catonis, p. 83). Here we have the subjunctive in protasis referring to any one of a series of acts, with the indicative in the apodosis expressing a general truth, precisely parallel with éav rotro rparrns cadic Eyer, which is the regular construction. So, Virtutem necessarto gloria etiamsi tu id non agas, consequitur, Cic. Tusc. Disp. i. 38. So also with relatives and relative particles: Bonus segnior fit ubi negligas, Sall. Jug. 31; Cum animum ab istis imaginibus ad veritatem traduzerts nihil relinquitur, Cic. Tusc. Disp. 55.
One can hardly realize the frequency of this construction
2
162 J. B. Greenough,
unless his attention is particularly called to it. In fact it is almost universal in writers of all periods. In order therefore that the frequency as well as the nature of the construction may be understood, examples are given at some length:
Istaec ubi periclum facias aculeata sunt, Those things when you try them are full of stings. Pl. Bacch. v.63; At nune ‘st attingas eum manu, paedagogo puer tabula disrumpit caput, But now if you lay your hand on him the boy breaks lis mas- ter’s head with a slate. Ibid. v. 440; Quom patrem adeas postulatum puero sic dicit pater. Ibid.; Quos quom censeas esse amicos reperiuntur falsi falsimoniis. Ibid. v. 540; Jn re mala animo si bono utare adjuvat. Capt. v. 202; Nam doli non doli sunt, nisi astu colas; set malum maximum, si td palam pro- venit (observe the third person in the indicative, in precisely the same construction). Ibid. v. 221 ; Consimile’ st quom stertas, quasi sorbeas, Mil. Gl. v. 820; De magnis divitiis, siquid demas plus fitan minus? Trin. v. 348 ; Mage st exigere occuptas duarum rerum exoritur optio. Ibid. v. 1052; Stultus et sine gra- tia’s. Tibi recte facere? quando quod facias perit. You're a fool and thankless fellow. Do rightly by you, when what one does is lost upon you! Aul. v. 335.
Uxor, si cesses, aut te amare cogital, aut etc. (see whole passage). Ter. Ad. v.32; Ille quem beneficio adjungas, ex animo facit. Ibid. v.72; At tamen ‘ubi fides’ si roges, nil pudent. And. v. 637.
Tantum remanet, quod virtute et recte factis consecutus sts. Cic. Cat. Maj. xix. 69; Quae (exercitationes virtutum) cum diu multumque vizxeris mirificos efferunt fructus. Ibid. iii. 9; At memoria minuitur, credo, (parenthetical) nist eam erer- ceas aut 8. sis natura tardior. Ibid. vii. 21; Quae (utilitas) tamen ipsa efflorescit ex amicitia etiamsi tu minus secutus sis. Lael. xxvii. 100. Sed haee adjuvant inoratore . . 8i quid persequare acrius, ut invitus et coactus facere videare. De Orat. ii. 43; Hie quantum fit mali, sinon levem testem lacseris ? Ibid. ii. 74; Ista discuntur facile si et tantum sumas quantum opus sit, etc. Ibid. iil. 23; Plures ineuntur gratiae st uno tempore dicas pro pluribus. Brut. 57, ad fin.; [ Conformatio] sententiarum perma- net, quibuscunque verbis uti velis. de Orat. il. 52.
On some Forms of Conditional Sentences. 168
Nam si velis quod nondum vetitum est, limeas ne vetere; at si prohibita impune transcenderis, neque metus ultra neque pu- dor est (where in the first sentence we have the second person in both protasis and apodosis, and hence the subjunctive ; in the second the apodosis has the third person, and hence the indicative). Tac. Ann. iii. 54.
Plerique, quae delicta reprehenderis malevolentia et invidia putant. Sal. Cat. 3.
That this construction is not imitated from the Greck 18 sufficiently proved, it seems to me, by its universality at all periods of the language, and secondly, by the fact that in the same sentence the second person is expressed by the sub- junctive but the third by the indicative, a distinction which is entirely foreign to the Greek. Obviously if the earliest writers were copying from the Greek they would have used the subjunctive throughout, as the Greek does. Several of the examples from Plautus are of this kind: e. g. Jn mala uxore atque inimico siquid sumas, sumptus est ; tn bono hospiti alque amico quaestus est quod sumitur. Plaut. Miles Gio., v. 673 ; Volup est quod agas, 81 id procedit lepide. Ibid. v. 947 ; Nimia’ st voluptas, st diu afueris domo domum ubi redieris, si tibi nulla aegritudo animo obviam’st. Stich., v. 524. See also Tac. Ann. iii. 54, and Plaut. Capt., v. 221, cited above.
There are a few cases where the same construction i is used in the first and third persons: e. g.
Nihil profictunt mercatores nist admodum mentiantur. Cic. de Off.i. 42; Luxuria vero cum omni aetate turpis, tum senec- tuti foedissima est. Sin autem libidinum etiam intemperantia accesserit, duplex malum est. Cic. de Off. 1. 34; Neque alitur si factat ullam inter suos habet auctoritatem. Caes. B. G. vi. 11 ; — see also, Caes. B. C. ii. 24, Q. Curt. vi. 5.
* It will be observed that all the cases thus far have been in present or general time, corresponding to the Greek sub- junctive and expressed with the present tense in the apodosis.
But it has long been noticed that the same construction was used exceptionally, especially by later writers — not to my knowledge by Cf€ero,— of past time, with the imperfect in apodosis, corresponding to the Greek optative with the im-
164 J. B. Greenough,
perfect. It is this construction which has been considered a Grecism. S% quis a domino prehenderetur consensu militum eripiebatur. Caes. B. C. iii. 110.
The cases are numerous but as the construction is familiar the passages are only referred to :—
Caes. B. C. ii. 41; Liv. ii. 27, iii. 11, xxxiv. 38, iii. 19, 1. 32, viili.8 ; Q. Curt. iv. 30, ad fin.; Nepos, xvii. 8, xx. 4, xxv. 2, xvii. 1, xviii. 3; Sallust. Jug. 58; Tac. Amn. vi. 30.
This same construction also occurs in Sanskrit, that is, with the protasis in the subjunctive or optative and the apodosis in the indicative, expressing a general truth. The more com- mon construction however is with the subjunctive or optative in both, as in the first case of the second person in Latin, mentioned above.
For Sanskrit scholars I may cite Rig-Veda, i. 94. 15; i. 79. 2;i. 79. 3; Hitopadesa, 800.
Now we see this construction regularly used in Greek, very frequently found in Sanskrit, and in Latin universal in one form and occurring occasionally in other forms in writers of various dates. We are naturally led to conclude that this construction was in use more or less in Indo-European times and was received by each of the languages as a part of the original inheritance. It was afterwards developed by the Greek almost to the exclusion of other forms, so.that the indica- tive is only the exception, but gradually disused by the Latin except in one or two cases, to re-appear again occasionally in the later language, while the Sanskrit retained both construc- tions.
It is worth while perhaps to consider the origin of the con- struction. This is not difficult to see if we consider the ori- gin of the use of the subjunctive and optative in protasis generally. There can be no doubt that the original form of the hypothetical sentence (so far as the condition is future or general or contrary to fact) is represented by such expressions as “Tulle hane opinionum, luctum sustuleris” and ‘“ Naturam expellas furca, tamen usque recurret,” in which two codrdinate clauses appear, the first containing a command, and the sec
On some Forms of Conditional Sentences. 165
ond a consequence of the (implied) performance of that com- mand. Now it is obvious that such a formula may develop into two constructions, the one meaning ‘ Let one do so and so, (and) such will be the effect ;” the other stating the con- sequence more generally, as “‘ Let one do so and so, (and) such is (always) the effect.” These two correspond to the - particular and general protases. In Greek both forms are used in their proper place. In Latin the former is chosen for future protasis and many general protases, as it obviously makes little difference whether one says ‘‘such is the effect ” or such “ will be the effect,’’ when the statement is intended to be general.
In the case of the second person singular of an indefinite subject the language adopts both, but for the ordinary gene- — ral condition it takes another form, of different origin, in which the protasis is not a command at all but a statement of a fact which for the moment is assumed to be true. This form gives the indicative in both clauses, a construction which as we have seen occurs also exceptionally in Greek. The Sanskrit adopts most commonly the first form both for future and general conditions, but it also often uses the second with the indicative in the apodosis, like the Greek.
In regard to the origin of the protasis generally, see the able treatise ‘‘ Der Gebrauch des Conjunctivs und Optativs tm Sanskrit und Griechischen,” by B. Delbriick (Halle, 1871).
PROCEEDINGS.
- THIRD ANNUAL SESSION,
HELD AT NEW HAVEN, JULY, 1871.
AMERICAN PHILOLOGICAL ASSOCIATION.
New-Haven, Cr., July 25, 1871.
Tux Association assembled, agreeably to notification, in the Representatives’ Hall of the State-House, at three P.m., with the — President, Chancellor Crosby, in the chair.
The report of the Secretary was then read and adopted. In the report it was anhounced that the following persons had become members of the Association, according to the provisions of the constitution, during the course of the year.
Professor Frederic D. Allen, East-Tennegaee University, Knoxville, Tenn. ; Professor J. Graeff Barton, College of City of New-York ; Mr. F. S. Batchelder, Stafford, Ct.; Rev. J. H. Blakeley, Bordentown, N.J.; Miss Mary L. Booth, New-York; Mr. P. Born, Selinsgrove, Pa.; Mr. Elihu Burritt, New-Britain, Ct.; Mr. 8. M. Capron, Hartford, Ct. ; President Alexis Caswell, Brown University, Providence, R. I.; Rev. Dr. Talbot W. Chambers, New-York; Professor Elie Charlier, (Life-Member,) New-York ; Professor Elisée Charlier, New-York ; Professor Lyman Coleman, Lafayette College, Easton, Pa.; Pro- fessor Nelson E. Cobleigh, East-Tennessee Wesleyan University, Athens, Tenn. ; Rev. William B. Corbyn, Quincy, IJ].; Mr. A. Crittenden, Brooklyn, N. Y.; Professor William C. Crane, Baylor University, Independence, Mo. ; President E. S. Dulin, St. Stephen's College, Columbia, Mo.; Dr. F. Ebener, Baltimore, Md.; Professor William M. Fisher, Baylor University, Indepen- dence, Mo.; Professor Leon C. Field, Cazenovia, N. Y.; Mr. James B. Greenough, of Harvard University, Cambridge, Mass.; Mr. Edward M. Greenway, Jr., Baltimore, Md. ;, Professor “alvin 8. Harrington, Wesleyan University, Middletown, Ct.; President Lucian H. Hammond, Lebanon Valley College, Annville, Pa. ; Professor Willabe Haskell, Bucksport, Me. ; Professor B. J: Hawthorne, West-Tennessee College, Jackson, Tenn. ; Profes- sor H. W. Haynes, University of Vermont, Burlington, Vt.; Professor M. W. Hamphreys, Washington College, Lexington, Va.; Professor John T. Huntington, Trinity College, Hartford, Ct.; Professor William H. Jeffers, Wooster College, Mo.; Professor 8. A. Theo. Jobe, St. John’s College, Little Rock, Ark.; President Abiel A. Livermore, Unitarian Theological Seminary, Meadville, Pa.; Professor Ch. Louts Loos, Bethany College, Betha- ny, W. Va.; Professor J. H. McDaniels, Geneva, N. Y.; President George H. Magoun, Iowa College, Grinnell, Iowa; Professor Daniel Marvin, Jr., Ra- cine College, Racine, Wis.; President F. A. Muhlenberg, Muhlenberg Col- lege, Allentown, Pa. ; Mr. Bennett H. Nash, Boston, Mass. ; Mr. C. W. Naseau, Lawrenceville, N. J.; President Cyrus W. Nutt, Indiana University, Ind. ; Professor Edward North, Hamilton College, Clinton, N. Y.; Rev. E. G. Par- sons, Derry, N. H. ; Professor Oval Pirkey, Christian University, Canton, Mo. ;. President J. C. Pershing, Female College, Pittaburg, Pa.; Professor A. J. Quinche, University of Mississippi, Oxford, Miss. ; President Joseph Shea, St. John’s College, Fordham, N. Y.; Professor E. Snyder, Illinois Industrial University, Champaign, Il].; Professor Friedrich Stengel, Columbia College School of Mines, New-York ; Professor Daniel S. Talcott, Bangor Theological:
4 Proceedings of the
Seminary, Bangor, Me.; Professor J. Henry Thayer, Andover Theological Seminary, Andover, Mass.; Professor E. H. Twining, Minnesota University, St. Anthony, Minn. ; Professor William 8. Tyler, Amherst College, Amherst, Mass. ; President T. R. Vickroy, Lebanon Valley College, Annville, Pa. ; Professor Stephen 8. Young, Bowdoin College, Bangor, Me.
The Treasurer’s report was presented, read, and accepted. It showed the receipts and expenditures of the past year to have been as follows :
RECEIPTS. Balance on hand, July 26th, 1870... 0.2... ..cccccccccccscsccesonces O32 86 - Annual assessments paid. .......... eee eeccccsecceccuce eccccccse 245 00 Foes from mew members.......ccceeccccsccccsmcccccccccecccccess S10 O $587 86 EXPENDITURES.
Printing of Transactions, 1869-70, 500 copies............cccarecces $108 43 Printing, postage, stationery, and other expenses............... eee. 59 78 Total expenditure, 1870-71......... cece eect cans eesecees oo. - 8468 26 Balance on hand, July 25th, 1871.......cccccscccccccccesccecccucs 119 60 $587 86
The Treasurer, for the Committee of Publication, announced that the volume of Transactions for 1869-70, was printed and ready for delivery to members.
Upon motion, Professor T. R. Lounsbury and Professor W. L. Montague were appointed Assistant Secretaries.
The address of welcome to the Association to the hospitalities of New-Haven was given by Lieutenant-Governor Francis Way- land. A response was given in behalf of the Association by the President, Chancellor Crosby, of the University of New-York.
The Association then proceeded to its regular business. The first paper, upon “ Inaccuracies in Grote’s Narrative of the Re treat of the Ten Thousand,” by Professor Fisk P. Brewer, ot Chapel Hiil, N. C., was read, in the absence of the author, by Professor James Hadley. |
Professor Brewer's criticisms were confined to the interval between the time when Xenephon received command and the arrival at Trapezus. He showed, by a minute examination of passages, that many of the details given by Grote are inconsistent with, or at least not warranted by, the statements of Xenophon and Diodorus, the only authorities whom he quotes. Thus, where Xenophon is represented (vol. ix., p. 78, Harper’s ed.) as saying to his fellow-lochagi, <‘ The enemy will be upon.us at day-break.” The expression (p. 79) that Apollonides “ protested against it as tnsane” is hardly justified by Xenophon’s ¢Avapeiv. P. 80 speaks of four commanders as seized by Tissa- phernes, instead of five. On p. 88, we find “cavalry and bowmen,” in place of “bowmen and slingers ;’ and on p. 89, “four thousand horsemen and
American Philological Association. 5
darters,” where it should be “a thousand horsemen and about four thousand bowmen and slingers.” For “darts” and “ darters” on the Persian side, (pp. 88, 90,) no authority is found. The “galloped” of p. 92 is too strong for Xenophon’s rpoceAdoac. That the houses of the Karduchi were “ comfort- able,” (p.95,) is not proved by yaAxouact rapméAAote xareoxevacuévac, On the same page it is intimated, without warrant, that the Greeks watted before taking what was necessary for refreshment. On p. 96, Kleonymos and Basias are spoken of as “two distinguished men” among “ several” Greek warriors mortally wounded ; yet Xenophon does not say that any others were killed, and these are only described as xado re xdya8o, Instead of a repeated refusal by Cheirisophus to obey Xenophon’s entreaties, (p. 96,) there was really but a ‘single instance. The two marches, (p. 102,) from the residence of Tiribazus to the river Teleboas, should be changed to five. For the statement (p. 105) that the reliefs from Cheirisophus were “sent back to bring up exhausted soldiers who had been Jef? behind,” there is no evidence. The attendance of the native youths (p. 106) was not confined to the fatigued soldiers. The “nine days’ march” on p. 109 should be changed to twelve. The statements on p. 113, as to a certain soldier who had been a slave in Athens—that he was exported from home in his boyhood, and that he had escaped from slavery (with the suggestion as to the time and place of this event)—are not sustained by the language of Xenophon. |
The second paper, by the Rev. Stephen R. Riggs, missionary to the Dakotas, “Concerning Dakota Beliefs,” was read, in the absence of the author, by Hon. J. H. Trumbull.
This paper was not intended to cover the vast field of things of which their faith takes hold, but rather to select such as are most characteristic and such as are important enough to have made an impress on their language.
The Dakota names of the heavenly bodies were first noticed. Their family or generic name is Wi. The sun is the day Wi, the moon isthe night Wi, and the stars are the “ battle-axes” or “ war-clubs” of Wi, (us-chanhpi,) per- haps because they are regarded as a great war-party, marshaled under the great captain Ws. The morning star is the “ light-shooting star ;” the even- ing star is the “large star ;” the parallelogram of Orion is “the bearer” or “the bier ;” and the milky way is “the Spirit Road,” along which men’s spirits, they aay, pass to the great Hereafter. When the moon wanes, it is believed to have been gnawed by mice, and they say, wi-yashpapi, “ the moon is bitten off.” The sun is the real Wi; it “ appears” in the morning, “ goes down” or is “ cast into” some place at evening. To it the Dakotas pray, offer sacrifices, and dance the “sun dance.” They address it as “grandfather,” and the moon as “ grandmother.”
Thunder is “the Flyer,” Wa-kiy-an, and is represented as a great bird. Of lightning, “ god-manifestation” or the “ spirit coming-home,” they make no representation. Of the four quarters of the heavens, the east and the west are “the sun rising” and “the sun going-down;” the north and the south are Waziya and Itokaga, regarded as gods, ever in conflict and each in turn victorious. ;
Boulders are the “ «olid gods,” “hard wakan.” These they worship, paint-
6 Proceedings of the
ing them red, decorating them with swan’s down, and offering sacrifices. The boulder is toonkan, “ grandfather,” by preSminence.
Oonktehi, the great god of the waters, may be regarded as their chief di- vinity ; certainly one of their oldest gods. Tohim they attribute the making of the world, by bringing up earth-seed from the deep waters. The name Oon-kte-hi is net resolvable into its elements, and seems to belong to an early ‘stage of the formation of the language. The name of the Takooshkanshkaa “‘ god of motion,” is of more recent origin. He is the Jupiter of the Dakotas and the object of frequent worship.
Heyoka, the personification of contrariness, the grotesque, the anti-natural, is one of the old gods, not much worshiped now. He is the god of fable and romance, :
Wakan is an ancient word which represents the god-worship of the race. Every object of worship is-wakan, and is worshiped because it is wakan. Its compounds are manifestly of recent date; a gun is a “oaken iron,” a horse is “ wakan dog.” The idea of the “Great Wakan” (wakan-tanka) can not be anoldone, It is their designation of the white man’s God, and they have learned it from the white man. The “wakan dance” has been borrowed from other Indians, and ts not an old institution with the Dakotas. Mr. Riggs gives some accoant of this dance, and of the initiation to the secret society by which it is performed, and proceeds to speak of the Dakota belief as to the soul and its future state. Nagi means “shade” or “shadow,” as well as “soul” or “ spirit.” Of one who has breathed his last, they say, nagi iyaya, “ the spirit (or shadow) is gone.” They believe in the separate existence of the soul, and in a “house of spirits,” wanagitipi. Every thing, even the dumb boulder, has a spirit. The world is full of spirits, who cause all dis- ease and death. The conjurer works his cures by expelling or overcoming one spirit by another. To do this is pikiya, from a root piya, meaning “ to make over again,” “ to renew,” “to mend.”
Sacrifice is probably an old form of Dakota worship. Mr. Riggs has ob- served it offered most frequently to the “ painted stone,” toonkan. The offer- ing was sometimes a small dog, killed and painted red. He points out the apparent relation between toshna and wayushna, “to offer sacrifice,” and yushna, “to drop,” “to let slip,” “to miss,” “to make a mistake,” and be tween the words for “ labor” and “ sin.”
The observance of twohduse, or voluntary abstinence from something “sacred or forbidden”—the taboo of the Dakotas—is next mentioned. Then, their belief in omens or presentiments, (wohkdeche,) and in dreams, (shamna,) with their “ vision-seeking,” (Aamdepi,) by fasting, prayer, etc. The relation of what has been seen in a vision (hamdohdaks, “ declaring the vision”) must be given in waken language, a sacred dialect, whose words have a peculiar meaning. In this dialect, “man” is the “ two-legged being ;” a dog is “ the four-legged animal ;” s black bear is “the black wakan,” etc. Their songs are often composed in this wakan language.
The next paper was on the “ Imperfect Tenses of the Passive Voice in English,” by Fitz Edward Hall, Honorary D.C.L Oxford, of Marlesford, England.
This paper was presented and read by Professor Whitney, who prefaced it
American Phiological Association. 7
with a brief account of the author; the latter, though an American by birth and education, having spent his life so much abroad as to be less known than he should be to American scholars, He is a native of Troy, N. Y., and a graduate of Harvard, of the class of 1846. Having been shipwrecked on the coast of India, he was led to remain there, and to enter the British service, in which he held successively the positions of professor of Anglo-Saxon, Sanskrit, and Hindustani, inspector of schools for a province, and, after his return to England, librarian, for a time, of the India Office. In Sanskrit and Hindu- stani he has shown scholarship and done work not surpassed by those of any other living scholar of English birth. He has also always been an industrious and fruitful student of English, modern and ancient ; is one of the editors of the Early English Text Society’s series of publications; and has now in ‘hand an extended series of chapters on points in the history, grammar, and criticism of our language. It was at his own suggestion, Professor Whitney said, that Dr. Hall had prep.red and forwarded a paper for presentation at this meeting of the Association.
The subject discussed is the propriety of such locutions as is being built. Dr. Hall begins with quoting expressions of opinion, generally unfavorable and often violently denunciatory, respecting these locutions and their origina- tors, on tlie part of various recent authorities. He then inquires respecting the time of their appearance, They are not mentione | in Priestley’s grammar, (1772,) nor in Bretland’s extension of it, (1785 ;) but Skillern (1802) gives a complete paradigm of verbal conjugation on the plan they involve. They are found used by Southey in 1795, by Coleridge in 1797, and by Lamb, Landor, De Quincey, and others, in passages which the writer quotes and re- fers to in full. And this sort of phraseology is becoming more and more com- mon; the best English reviews, magazines, and journals are constantly marked by it; and some of the choicest of English writers employ it freely. After ts tn building had been corrupted to is a-building, and this had come to be felt as vulgar and abbreviated to ts building, a just avoidance of ambiguity led, in the latter part of the eighteenth century, to the creation of ts being built. There were two present participles in use, active and passive, namely, building and being built, and as an active imperfect or continuous tense had been formed by prefixing #s to the former, so now likewise a passive imper fect by prefixing the same auxiliary to the latter. The form is nat resolvable into ts being + built, as has been strenuously urged by objectors, but into is + being built. Overlooking this, men like G. P. Marsh and R. G. White have been misled into strangely exaggerated reprobation of the new phraseology. The strictly analogical relation of ts building and ts being built is so obvious that it can not have failed to suggest itself to many minds, though it has es- caped the notice of the authorities mentioned. The “ absurdity” of combining. és and being is wholly imaginary. If ts being built is wrong because Latin ens edificatus est is inadmissible, then ts building is alao wrong, because edt- ficans est would be bad Latin. If be and ezist are completely synonymous, then % existing ia as bad as is being. If ts being involves an absurd repetition, then sono stato and ist gewesen are also absurd. Mr. Marsh’s claim that con- sistency would demand, equally with is being built, its analogues will be being built and would have been being built, and their like, is not to be allowed ; we say, for example, preparedness, but not understoodness, designedly, but not
8 : Proceedings of the
acknowledgedly, a now too notorious fact, but not a never tc be sufficiently eze- erated monster ; practical usage having the right to decide how far it will go in a given direction of expression, where a compromise is to be made between desirable clearness and a felt awkwardness of phrase. To pronounce the locution “ unidiomatic” implies a wholly new definition of idiomatic, and as to “being opposed to the genius of the language,” that is a sounding phrase which has no philological value. The strength of those who decry the modes of speech here in question ‘consists mainly in thvir talent for calling hard names, and their opposition really proceeds from no higher motive than lite- rary conservatism and dislike of novelty.
The paper closes with a parallel between the new phrase is being built and the word its, which was new and shocking to the purists not very long ago; and it is made to appear that the latter involves worse violations of sound principle and analogy than the former.
Evrenina SEssron,
The annual address was delivered by the President of the As- sociation, Rev. Dr. Howard Crosby, Chancellor of the University of New-York.
After alluding to the progress of the Association in its work, and the favor- able prospects under which it enters the third year of its existence, Dr. Crosby spoke somewhat as follows:
“ Linguistica or philology may be considered either as a science or as a phi- losopliy. Under the first aspect we may gain some idea of its extent by think- ing of the vast number of languages which are to be investigated, not only those now spoken, but also many of which we have but the fossils. It touches here psychology and history, and enables us to know the unseen. A linguis- tic criticism is the source of all true commentary. By philology we can re- construct prehistoric man, and read the history of times before the Olympiads and Nabonassar. Languages are never lost. By this science, the original unity of the human race is already nearly proved. The philologist is also in part a physiologist and an anatomist, because he must study the organs of speech. He seems to be the centre of all science; he is the universal inter- preter ; therefore he can not be contracted or illiberal. He receives from all and bestows upon all. Again, philology as a philosophy speculates on the value of language to man, and its relation to his mind. These speculations are not to be confounded with the facts of the science. Man has worked out language for himself, according to his needs. Language has wrought its marvels; its triumphs are the triumphs of our race. But itself records its weakness by its constant use of negatives. Every profound thinker has found himself fettered by language. Hence disputes and mieunderstandings have arisen. Also in poetry, in devotion, in music, language is shown to be imperfect ; it can never be made sufficient for the whole realm of thought. Man, in his development, must have a nobler and fuller language than he has to-day. This may be in a new creation with spiritual bodies.”
The President, in conclusion, referred to the field of American languages as especially open to the researches of the Association, suggesting its divi gion into sections and the organization of local branches.
American Philological Association. 9
After the close of the address of the President, Professor Com- fort read a paper upon “The Order of Precedence in Study of the Ancient and the Modern Languages.”
After stating the present condition of the discussion with reference to the position of language in education, Professor Comfort proposed the following reorganization of our system of linguistic education :
The study of one living language should be commenced by pupils when ten or twelve years of age. As much time, or more, should be given to the study of this modern language, as is now given to Latin in the academy or preparatory school. The method will at first be conversational and practical. Rigid grammatical instruction will be given later in proportion to the growth and progress of the student.
Two years before the close of the academic course the study of a second living language will be commenced. Upon entering college, the proficiency of the pupil in these two languages will be nearly equal. These two modern languages will take the place of Latin and Greek in the studies which are required for.entrance to college.
During the whole of the freshman year, these two languages will be studied according to the most rigid philological method. During the re- mainder of the college course, at least one study at a time in other branches of science will be pursued from text-books in one or the other of these mo- dern languages.
The study of Latin will then be commenced in the sophomore, and that of Greek in the junior year. Both of these ancient classical Janguages will be pursued about two years.
It will then remain for philological faculties in post-collegiate universi- ties, or for professional schools, like the school of philology which is to be opened in connection with Yale College, to give that high training and cul- ture in all branches of linguistic science, which is in general so lacking in America. Associations of linguists, like the American Oriental Society and the American Philological Association, have alao a work to perform, in the promotion of linguistic science in America, which also is beyond the province of any school of instruction.
It is only through the influence of the (post-collegiate) universities, and of the various philological societies in Germany, that philology has, like the other sciences, attained such a high stage of development in that country.
Professor Comfort then gave at length the arguments in favor of this plan, claiming that it contains the natural order and method of the study of language, and that it offers great advantages over the present system, or con- flicting systems, to all classes of students: to those who shall finish only the academic course ; to those who complete the collegiate course ; to those who, after leaving college, shall study in professional or technical schools; and to those who shall become professional linguists.
The objections that are urged to this plan were passed in review. <A very respectable minority among the best philologists and educators in Europe, and quite a number in America, already favor this change. A number of the features in the plan proposed are adopted, and with eminent success, in some of the best colleges and other schools in Europe and America.
10 | Proceedings of the
MokRNING Srssron, Wennespay, Jury 26, 3871.
The first paper of thé morning was on “ English Vowel Quan tity in the Thirteenth Century, and in the Nineteenth,” by Professor James Hadley, of Yale College, New-Haven, Ct.
Professor Hadley remarked that the modern Greek and the Romance lan guages have lost the systems of vowel quantity which belonged to the ancient Greek and Latin ; and raised the question whether the same is true of Englishas compared with Anglo-Saxon. It is admitted that in English we have nearly lost the feeling of length by position, (where a short vowel stands before two or more consonants ;) if flet and fight seem each longer than fit, wa do not re- cognize fist as having to fight any definite relation of quantity. But as to vowel sounds, it can not reasonably be doubted that those in jfile, foul, feel, Sool, fal, foal, fall, occupy more time in average enunciation than those in Sil, full, fell, dol, dally, dull,
If, then, the present English has long vowels in some words, and short vowels in others, how far do these quantities agree with those which belonged to the same words in earlier periods of the language? In deciding this ques- tion, valuable help is to bé derived from the Ormulum, a series of metrical homilies on the daily lessons of the church service, composed by the monk Orm in the thirteenth century. In the only known manuscript, written per: haps by the author’s own hand, a consonant is regularly doubled when it-fol- lows a short vowel : thus, ¢, tf, hendred, are written, ttt, vif, hunndredd. We can see then what vowels were sounded short, and what long, by the writer of this work ; and by comparing them with present pronunciation, we can make out the nature and extent of the changes which have taken place since that time.
Sach a comparison shows that, in the great majority of cases, the syllables which then had long or short vowels, have the same quantity now; and that the exceptional cases, where the quantity has changéd, can mostly be referred to certain recognizable euphonic influences and tendencies. These euphonic causes of alteration in vowel quantity, it was the main object of the paper to set forth in their nature and working.
Thus, vowels have become long, since the thirteenth century, through the suppression of a following consonant, as in alms, buy, day, brought, etc. ; in light, four, etc., the original long quantity has been restored in this way. Vowels have been made long also by the effect of a weak 7 before a consonant or at the end of a word, as in for, dare, church. So, before the liquid J, as in all; and especially before /d, as in child ; and before the similar groups mb and nd, as in climb, kind. These changes before liquids had begun in the time of the Ormulum, which in some respects carries them further than the English ; it often lengthens a short vowel before ng, as in king, tunge, (tongue.) Before other combinations of consonants, a long vowel has been shortened, and was so in many cases in the Ormulum: thus in sleppie, (slept,) soffte, (soft).
But the most frequent change In English quantities has been caused by the tendency to lengthen an accented short vowel in a penultimate syllable, when separated by only one consonant from the vowel of the final syllable. This
American Phitological Association. 11
tendency, which has produced the long sound in evil, chosen, name, etc., is carried much further in the Ormulam, where heavy, risen, love, etc., have long vowels. It has also prevailed very extensively in the German.
Other euphonic tendencies to change of quantity were pointed out, whicb, however, have a more limited range of application. The paper closed with some remarks on suffixes, such as -dom, -hood, ly, which were long in the Or- maulum, but have become short in modern English. -
Upon motion of Dr. A. B. Hyde, the Executive Committee were requested to cause a catalogue of the members present to bé printed for distribution.
The next paper was entitled, “ Notes on A. J. Ellie's Early En- glish Pronunciation,” by Mr. . A, Bristed, of Lenox, Mass.
This book awakens a pardonable pride in us, when we consider that it is published by the Chaucer Society, of which our countryman, Professor Child, is so active a member. The work is a monument of industry, intelligence, and learning ; nevertheless, we must take exception to some things in it. The paleotype is too cumbrous, and makes hypercritical distinctions. The notation of several sounds, both English and foreign, is open to objection. Mr. Ellis has committed the error of transposing the two Italian 0's, (chtuso and aperto,) also that of making diphthongs in Italian.
An examination of the Spanish D leads to the conclusion that the Spanish language is undergoing an orthoepic degradation, by the syncope and apo- cope, not only of D, bat of 8, and even of R; and is passing through a stage similar to that which the French has already undergone.
With the majority of the old English sounds, as fixed by Mr. Ellis, no fault is to be found. The time, however, when long A received its present’ sound, most critics would put further back than the author has done. Mr. Paine’s views on the diphthong AI (both in old English and old French) are, on the whole, more plausible than those of Mr. Ellis. What was “the fifth sound of A, at in far,” given by the eighteenth century lexicographers ? Was it the short E of ferry (— French é,) or the indistinct short U? As to the French sound which Mr. Ellis assigns to the old-English U, we may well doubt if it was ever naturalized in English, even among the French-speaking population. In Chaucer’s time there was already a French French, and an English French ; his prioress spoke the latter.
Three subordinate points in the work specially attract our attention, 1st. The word one. Possibly, in passing from own to wun, it had an intermediate stage of un. The pun in Love’s Labor Lost, read with this pronunciation, be- comes modern rustic English. 2d. The (present) prosthesis and apheresis of H in English conversation. Mr. Ellis says truly that the former denotes & lower stage of society than the latter. Might he not have added that in the Manufacturing districts the prosthesis extends to more respectable classes than in the metropolis and southern counties? It is a common popular error in America to attribute the prosthesis and apheresis to the same classes and persons. Few Englishmen aspirate the H of initial W H; all Irishmen, Scotchmen,and Americansdo. 8d. As to the diphthong O01; an examination of it in English and French, compared with Greek o: and Latin O E, suggests
12 Proceedings of the
a probability that in all four languages the first element had originally a digammatic force, which was afterward dropped in three of them.
In regard to the reformation of our orthography, while Mr. Ellis states very fairly and forcibly the defects and inconveniences of our present mode, he makes an admission fatal to the proposed change. “There can be no ab- solute standard of pronunciation ;” therefore, there can be no fixed standard of phonetic orthography.
The next paper was upon “ Anglo-Saxon and Early English Pro- nunciation,” by Professor F. A. March, of La Fayette College, Easton, Pa,
’ The alliteration of Anglo-Saxon poetry gives a good indication as to what consonants were pronounced alike, and in what order the consonants of any combination were pronounced. The early English alliterative verses enable us to date approximately the changes of sound. An account was given of the alliteration of ¢, sc, g, h, j, th, and the combinations i, hn, hr, hw, wl, wr.
Certain laws of phonetic change sometimes give a clue to sounds where alliteration fails ; th as in thin, is not distinguished in the alliteration from th in thine ; but surds gradually weaken to sonante, sonants do not change to surds. Hence arises a presumption that words pronounced with surds in English had them in Anglo-Saxon.
Again, a surd and sonant do not combine in the same syllable. Hence, the th of the syncopated forms of verbs ending in a surd must have been surd : thinedh (thinketh) is incredible.
The Anglo-Saxons distinguish shades of vowel sounds which the later Nor- man English neglect. Words in a, for example, may exchange it for ae, ea, eando. The traditional pronunciation of the descendants of those who used these weakenings of a, has probably never been pure ain those words which are still pronounced weak.
The statements of grammarians who describe any vowel as having a single uniform sound in early English are hasty generalizations.
AFTERNOON SESSION.
The Association met at three p.m, the Vice-President, Professor W. W. Goodwin, of Harvard University, in the chair.
The first paper read was on “The so-called Attic Second De- clension,” by Professor F. D. Allen, of the University of East- Tennessee, Knoxville, Tenn.
The close connection of the nouns in o¢, o with those in o¢, ov was recog- nized by the ancient grammarians. No satisfactory exposition of the origin of this form of declension is found in modern grammatical works. . The words belonging to it have w at the end of the stem in place of the o of the common form. This w is, however, in no case original, that is, descended from a pre- Grecian 2. Examination of the individual words, with the aid of collateral forms and in the light of modern etymology, shows that the wo stems have arisen : -
American Philological Association. 13
1. From stems in ao (older afw or ajo) by interchange of quantity, as in ‘Arpeidew from ’Arpeidao: . thus Aeds, vedc, dvoyewr, from Aaoc, (?a/o¢,) vade, (vaoFo¢,) dvdyaov, (dvu-yaFjov.) Ktihner’s.view, that vo was first contracted into w, and an e arbitrarily prefixed, is singularly perverse,
2. From stems in ao (older aFo, ajo or aco) by contraction : diuves, deivue, dyjpwc, from dizvaoc, deivaog, (-ovaFo¢,) dyypaog, (dynpac-oc.)
8. From stems in wo (older ofo) by shifting of quantity: Kéws, yaAdwe, from Kaog, (xoFog=cavus,) yaAwog, (yadof or.)
4. From the same by contraction: Koe, ydAwr, Go, (CoFor,) xatpwc, (rarpofo¢g =patryus, cf. suus, from old Lat. sovos,) dAwe, (FadoFoc,) Aaywc from Aaywos.
The remaining words, to be similarly explained, although some of them can not be with such certainty analyzed, are: "Adwe or 'ADoug, (’ADoFor,) Mivus, (MivoFoc,) wepivewc, (wept-vaFoc,) “dvapewc, (-Kpafjog,) xparaiAewc, (-AaFoc,) Ews, (dFo¢,) d&tdxpewc, Tuvdapews, Boripewc, 'Augidpews, wAétwe, (tAejoc,) iAews, from tAaog, ow¢ from odor, xdAuc, Tere, Kéwe, and several others.
The inflection explains itself without difficulty. The accent of the nomi- native remains in all the cases.
But GAlws, ndtows, Mivwo may follow the third declension, and conversely fpws and zyrpec, commonly of the third declension, occur in forms of the Attic second declension. The line is nowhere to be strictly drawn; all these words are onein formation. The stem ends in ofo or of interchangeably. That the fuller form is the original one, the etymology of xurpw¢ shows, The second declension form is therefore the earlier in all these words. The other words declined like fpws, namely Tpoc, dusc, Sac have doubtless a like origin, (the accent of gen. pl. Tpdwy, ete., may be thus explained,) and this whole class are seen to be, as it were, estrays from the Attic second declension. The acc. sg. in wv, 80 common in the Att. 2d-decl., "Adw, Ka, etc., corresponding as it does to fpw, (contr. from f#pwa,) is to be regarded as a third declension form, refer. able to the heteroclisis just explained.
The word éuc¢, Epic #0¢, Doric Fos, is distinct from these. The Epic and Doric forms come from an sstem atcoc, (Sanscrit ushay.) The Attic éuc (2d decl.) is generally thought to be due to a mere blunder of the language, but I am inclined to recognize a genuine vowel-stem aico, to which the parallel form exists in Sanscrit uaha.
This form of declension is not entirely confined to Attic, but is more a favorite there than in other dialects, particularly in the less elevated style or diction.
The third paper was on “A Mode of Counting, said to have been used by the Wawenoc Indians of Maine,” by Hun. J. H.
Trumbull, of Hartford, Ct.
The late Dr. J. G. Kohl, of Bremen, author of a “ History of the Discovery of Maine,” published by the Maine Historical Society in 1869, mentions, as a possible “ reminiscence of the Northmen among the Indians of New-England,” the fact that “among the Wawenoc Indians near Pemaquid, certain nume- rales have been handed down by tradition, bearing a resemblance to the Ice- landic, which may have been derived by them in their barter with the north- ern strangers” who visited New-England in the eleventh century.
These Wawenoc numerals were first brought to notice by R. K. Sewall,
14 Proceedings of the
Eeq., of Wiscasset, in a communication to the Maine Historical Society, January, 1868. They were printed in the Estorical. Magazine for March, 1868, with a note from the Rev. Dr. Ballard, of Brunswick, Me., (since deceased,) who asked “ Whence did they come? Did Madoc bring them here in his semi-true, semi-fabulous voyage? or did Northmen leave them on the coast f”
The Wawenocs were a tribe or band of the Abnakis. They became extinct about 1750. Tradition affirms that they used these numerals, in their inter- course with the whites, early in the eighteenth century. Not one of the numerals, however, belongs to the Abnaki or to any other aboriginal language of New-England.
The writer was convinced they were not Icelandic. If, as he was inclined
to believe, they were of Welsh origin, he saw no reason for looking back to Madoc, or the twelfth century, for their introduction. After searching an- successfully all the English and Scottish provincial glossaries, he has lately come upon their track. First, he ascertained that the knowledge of these namerals was not confined to the Wawenoes or to Maine. Two friends in Hartford had learned “ the Indian way of counting”—manifestly of the same _ origin as the Wawenoc numerale—nearly fifty years ago, from their father, who resided in Massachusetts, and afterward in Windsor, Vt., (but never in Maine.) A lady in Western New-York had been taught the same way of counting by her mother, who used to live near the Narraganset Indians in Rhode Island. So, if the numerals were of Norse origin, the Northmen must have taken great pains to disseminate knowledge of their numerical sys tem among the “ Skrellings” of all Vinland, and the Skrellings must have had excellent memories, to preserve the strange sounds with so little corrup tion for seven or eight hundred years. A few months ago, light came from an unsuspected quarter. Mr. Alexander J. Ellis, of London, in the Transac- tions of the Philological Society, 1870, gave some specimens of English d'a lects written In “Glossic.” Among these was a method of scoring sheep, used in the dales of Yorkshire. The Yorkshire score was surprisingly like the Wawenoc and Narraganset numerals! That they came from the same source, there could be no question. In answer to a letter of inquiry, Mr, Ellis most obligingly communicated all the information he had been able to obtain respecting this mode of counting, and promises further investigation. The score was printed for the first time, probably, in Mr. Ellis’s paper on Palsotype, in 1867. He obtained it from a lady who learned it fifty years ago in Yorkshire. He has since received it, with some variants, from corres- pondents in Leeds, and elsewhere, and ascertained that it has been used within the memory of persons now living, in counting sheep. One correspondent thinks it was “ prevalent in the East-Riding of Yorkshire, and right through to Thirsk.” - Mr. Ellis agrees with the writer that the seore is, partly at least, of Welsh origin. Whether it was brought into New-England from Wales or from Yorkshire is uncertain. There were Yorkshiremen in almost every township before the middle of the seventeenth century, and a good many Welahmen have visited Maine since the time of Mateo. There is clearly no reason for assigning ite introdaction toa high antiquity.
These scores are to be regarded rather as tally-marks or counters than as true cardinal or ordinal numbers. They were employed in counting off by
American Philological Association, 15
fives, tens or twenties. Traces‘of some such systems may be found in many achool-boy rhymes for “ counting out.”
The fourth paper was on “ The Newly Discovered Relationship of the Tuteloes to the Dakotan Stock,” by Rev. Joseph Anderson, of Waterbury, Ct.
Mr. Anderson’s paper consisted largely of oxtracte from letters of the well- known philologiet, Horatio Hale, Esq., now residing at Clinton, in the Pro. vince of Ontario, Canada, giving an account of a visit to Nikungha, the last survivor of the tribe of the Tateloes, and reporting a digcovery made at that time. This venerable Indian, who has died since Mr. Hale’s visit, at the advanced age of a hundred and six years, or thereabout, resided on the Re- serve of the Six Nations, near Brantford. The Tuteloes, of whom he was the last representative of pure blood, had been Jooked upon by ethnologists as an Iroquois tribe, chiefly because holding a place in the Iroquois confederacy. But the list of words obtained by Mr. Hale from Nikungha showed: conclu. sively that the Tutelo language belonged net to the Iroquois but to the Dako. tan stock. Mr. Hale’s list embraced over two hundred words: of these there are none which appear to be related tothe Iroquois languages, while a consi- derable number of them bear a marked regemblance to the Dakota or the closely allied Omaha.
As introductory to the extracts from Mr. Hale’s letters, Mr. Anderson gave a rapid sketch of the fortunes of the Tuteloes, from the time of the earliest records, when they were situated in southern Virginia and northern North- Carolina, until their removal, as one of the nations of the Iroquois confederacy, to the Reserve in Canada. He presented, next, a comparative vocabulary of twenty-five Tutelo, Dakota, and Nottoway-Iroquois words, in the preparation of which he had been assisted by Rev. Stephen R. Riggs, author of the “ Dakota Grammar and Dictionary” published by the Smithsonian Institution ; and considered in conclusion the bearings of Mr. Hale’s discovery upon the whole subject of Indian migrations. One of the questions suggested by this newly-discovered relationship is, how to account for the separation of this single, isolated tribe from the extensive group of affiliated nations now situat- ed to the west of the Mississippi River, and its establishment, so far away from the others, on the Meherrin River in Virginia. This question was dis cussed with special reference, on the one hand, to Mr. Lewis H. Morgan’s hypo- thesis, (North American Review, January, 1870, p. 50,) that the course of the Dakotan migration was from the Rocky Mountains eastward, by the way of the Platte River and the Black Hills of Nebraska, to the head-waters of the Mississippi ; and, on the other hand, to the opinion, which has found favor with some of our American ethnologiste, that while other Indian stocks may have come from the north-west, and moved continuously eastward, the Dako- tan stock came from the east and has been moving westward. The opinion which seems to harmonize the ascertained facts most readily, although leav- ing undecided the line of movement of the Dakotan migration, is that of Mr. Hale, that “in former cimes the whole of what is now the central portion of the United States, from the Mississippi nearly to the Atlantic, was occupied - by Dakotan tribes, who have been eut up and gradually exterminated by the intrusive and more energetic Algonkius and Iroquois.”
16 Proceedings of the
Of the twenty-five words in the comparative vocabulary embodied in Mr. Anderson's paper, eleven are evidently of the Dakotan stock, and five others probably so. In view of the great divergence among the confessedly Dakotan dialects, and the strong tendency to dialectic variation in all the American languages, the number of words which are the same or nearly the saine in the two lista is surprisingly large. The following may serve as examples: “Two” in Tutelo is némp; in Dakota, ndnpa. In Nottoway, on . the other hand, it is dekanee. “Four” is tép in Tutelo; in Dakota, tépa ; but in Nottoway, hentag. ‘“ Seven” is sagoin in Tutelo, in Dakota, shakowing. “Nine” is sang in Tutelo; in Omaha, shanka. The Tutelo for “ father” is tat ; the Omaha, ndade ; the Tutelo for “ fire,” pétt ; the Dakota, péa ; the Tutelo for “ water,” mdant ; the Dakota, mini. Some of tlie less obvious resem- blances are equally suggestive to the comparative philologist.
In the evening there was no session, as the members of the As- sociation attended a brilliant reception which was given to them in the galleries of the Yale School of Art.
Mornina Session, THurspay, JuLy 27.
The Association met at nine 4.m., the President, Dr. Crosby, in the chair.
The following persons were announced as having been elected members of the Association in accordance with the provisions of the constitution:
Professor A. M. Black, Monmouth College, Monmouth, Ill.; Rev. Charles E. Brandt, Farmington, Ct. ; Rev. Dr. Horatio Q. Butterfield, New-York; Mr. H. L. Boltwood, Princeton, Ill.; Professor Franklin Carter, Williame- College, Williamstown, Mass. ; President William C. Cattell, La Fayette Col- lege, Easton, Pa. ; Professor Francis J. Child, Harvard University, Cambridge, Mass. ; Professor Edward B. Cole, Yale College, New-Haven, Ct.; Professor J.C. Daniels, Olivet College, Olivet, Mich. ; Professor George E. Day, Yale Col- lege, New-Haven, Ct.; Miss Mary C. Dickinson, Northampton, Mass. ; Pro fessor John B. Duncan, Washington College, Topeka, neas; Professor Evan W. Evans, Cornell University, Ithaca, N. Y.; Professor William C. Fow- ler, Durham, Ct.; Mr. A. W. Fowler, New-York ; Professor J. N. Fradenburg, Fredonia, N. Y.; Professor Danie] C. Gilman, Sheffield Scientific Institute, New Haven, Ct. ; Professor Joshua B. Garritt, Hanover College, Hanover, Ind. ; Pro- fessor William H. Green, Princeton Theological Seminary, Princeton, N. J. ; Professor Ephraim W. Gurney, Harvard University, Cambridge, Mass.; Mr. Horatio E. Hale, Clinton, Ontario, Canada ;‘ Professor James H. Hanson, Waterville, Me.; Professor Samuel Hart, Trinity College, Hartford, Ct. ; Professor J. H. Jewett, Olivet College, Olivet, Mich.; Miss Rebecca Lowrey, New-York ; Dr. Washington Matthews, Fort Buford, Dakota Territory ; Pro- fessor G. C. Merrill, Washington College, Topeka, Kansas; Professor John - L. Mills, Marietta College, Marietta, Ohio; Professor Edward D. Morrie, Lane Theological Seminary, Cincinnati, Ohio; Professor William M. Nevin, Lan- caster, Pa.; Mr. A. Parish, New-Haven, Ct.; Professor Lewis R. Packard, Yele College, New-Haven, Ct.; Mr. Tracy Peck, Bristol, Ct.; Professor D. L. Peck, University of Alabama, Tusca , Ala:; Rev. Dr. John Pike, Rowley, Mass.; President Noah Porter, Yale College, New-Haven, Ct. ; Rev. Stephen R. Riggs, Lake Traverse, Minn.; Professor Timothy H. Roberts, Whiiney’s Point, N. Y.; Professor William C. Russell, Cornell University
American Philological Association. 17
Ithaca, N. Y.; Mr. Eugene Schuyler, U. 8. Legation, St. Petersburg, Russia ; Mr. Wesley Sawyer, Boston, Maas.; Professor Charles A. Schlegel, Female Normal College, New-York; Professor John 8. Sewall, Bowdoin College, Brunswick, Me.; Professor T. K. Smith, Colby University, Waterville, Me. ; Professor Frank Snow, Kansas State University, Lawrence, Kansas; Professor Thomas A. Thacher, Yale College, New-Haven, Ct.; Dr. Joseph Thomas, Philadelphia; Professor Crawford H. Toy, Southern Baptist Theolo- gica] Seminary, Greenville, 8. C.; Professor Henry M. Tyler, Knox College, Galesburg, Ill.; President Milton Valentine, Pennsylvania College, Gettys- burg, Pa.;: Mr. M. Warren, Providence, R. I.; Professor Albert S. Wheeler, Florida, N.J.; Professor Alonzo Williams, Brown University, Providence, R. I.
The first paper of the morning was on “ Strictures on the Views of August Schleicher respecting the. Nature of Language and other related subjects,” by Professor W. D. Whitney, of Yale College, New-Haven, Ct.
Professor Whitney said that he had been led to prepare this paper by hav- ing fallen in with an English version of one of the essays to which it related, prefaced by an extravagant and undiscriminating laudation of its merits. When false views were presented and urged under the recommendation of highly considered names, it was necessary to take pains to refute them. He disclaimed any lack of respect for Schleicher, or of grateful appreciation of his many and great services to comparative philology, in criticising his lin- guistic philosophy.
The first essay spoken of was published in 1863, and bears the title “ The Darwinian Theory and the Science of Language.” It is an attempt to prove Darwinism true by the evidence of language ; because languages, like ani- mals or plants, are natural organisms, which grow according to fixed laws, and are not determinable by men’s will. This view of language the writer endeavored to prove false. He went through with all the particular modes in which a language comes to differ from another, its predecessor and ances- tor, and showed that in each of them human agency is concerned, and no other agency beside ; and that what was true of every part was, of course, true of the whole, theirsum ; and that this sum, and nothing else, constituted the so-called life of language. Of course, this being shown, the attempted proof of Darwinism falls away of itself.
The second essay was published a year later, and was a defense of its pre- deceasor. It is entitled “‘ Respecting the Importance of Language for the Na- tural History of Man.” In professing to support the dogma formerly laid down—that a language is a natural organism—it really changes ground en- tirely, claiming that a Janguage is the necessary result and expression of a specialty of physical organization in the person speaking it. This new doc- trine, it was shown, is equally erroneous with the other. It entirely ignores the fact that every child. learns its “ native language,’ and might as easily have learned any other. Schleicher’s attempts, partly to deny, partly to ex- plain away, the fact that one is able to learn other languages, in addition to that which he has learned first, or his “ native language,” are also a lamenta- ble failure. The subsidiary dogmas—that language is the sole exclusive cha- racteristic of man; that it is the sole reliable test of race; that there must necessarily have been many primitive languages, and, therefore, many origi- nal and independent races of men; that linguistic science leads us to the con.-
i) Proceedings of the
qlusion that men descended from the anthropoid apes; that language-making and historical activity can not coexist, but characterize successive periods in the Jife of a race; and so on—were one after another taken up, and their re- futation attempted. It was claimed by the writer that the two essays were, in their foundation and whole superstructure nsound and illogical to a de- gree rarely equaled.
The second paper was on “ The Origin of Language,” by Pro- fessor F, A. March, of La Fayette College, Easton, Pa.
The roots of speech are from two sources, man’s natural language, and the sounds made by external objects. An examination of parte of the Bible, Ho mer, Nala, The Hitopades’a, Beowulf, Kalewala, and Hottentot legends was reported as affording evidence to prove that the sonnds of external objects are not what men most think of and nanre, and so going to show that imitation of such sounds did not furnish any considerable part of the fundamental sounds of language.
The facts are such as we might anticipate from psychological and physio- logical considerations.
Instinctive speech consists of utterances and vocal gestures to express the most vital needs of animal man, to invite, repel, warn, woo ; to utter joy, pain, surprise, love, hate, weariness, and the like. These are what he would talk, about first and most; then would come objects which affect these feelings and wants, named as 80 affecting them.
Again, the life of a word is found in a permanent relation in a living man between certain states of his mind and certain nervous and muscular habit of his vocal organs ; a relation often hard to establish. In natural language it is established by instinct, and such sounds may be imitated by the slightest assent of the will. But the imitation of the sounds of external objects de- mands energetic effort of the imagination as well as of the muscular sense and of other powers, and it is to be classed in its beginnings rather with pan- tomime than speech proper, and is likely to be one of the rarest exercises of the mimetic power. |
The latest school of science inclines to give the firat man, as distinguished from the “ ape-like progenitor,” a large accumulation of inherited sound-signs. They therein agree, as to substance of linguistic fact, with the old belief that man was created with the divine endowment of language. There is no saltus in man’s history to be called an origin of language ; but the present lan- guages are proper growths from the natural language of the first free agenta. The articulations seem to be distributed among the signs partly by the law of the least effort, the easiest going to the most used roots. Most objects are named at firat as they affect man, the denotation being eked out by gesture, and the selection of objects to be named being directed by the eye; but after- ward a number of complex nanies are not built from roots, but made by whim as children make jargon names, or by onomatope ; and a few of these are incorporated into language. It would seem possible to accumulate data for determining with fair probability the extent of the original speech, mainly by the scientific study of the speeches of savages, and the vocal sounds of in- fants, deaf-mutes, idiots, and the lower animals.
American Philological Association. 19
The third paper was upon “An Old Latin Text-Book,” by Colonel T. W. Higginson, of Newport, R. I
The writer desired to present the study of languages from the literary and artistic, as others had presented it from the scientific stand-point. Beginning with a sketch of the delightful childish associations which, in his case, sur- round a certain old book, (Leverett’s New Latin Tutor,) he passed to a general assertion of the value of classical study to a boy as impressing him early with a sense of the beauty of words. Beneath all the excessive atten- tion paid to the classics in our colleges there was this merit, that the sense of beauty, as distinct from mere science, had thereby been kept alive. The boy, he maintained, was naturally classical in taste, rather than romantic; the growing youth loved German literature, but the boy preferred the more defi nite outlines of the classics: as Emerson said, “Every healthy boy is s Greek.” The current scientific tendency was to deprive English style of its beauty, and make it merely frank, manly, and direct; but the sense of art must be kept alive by the study of pure literature, and especially the models remaining to us from the Latin and Greek. [The entire paper has simce been published in the volume entitled AWantic Hesays, by this author. ]
The fourth paper was upon “Sign-Language as indicating the Law of Vocal and Written Language,” by President G. W. Samson, of Columbian College, Washington, D. C.
Sign-language, though specially the means of intercommunication between deaf-mutes, is the method to which children, and rude tribes with but the vocabulary of a child, resort to make their wants and sentiments known. In its elaborated form, seen in all the characters used in early written communi- cation, it is, and must be, composed of three elements, the mimic, thé tropic, and the phonetic. The communications of uniffStructed mutes, limited to material wants, are chiefly mimic, being a mere pointing to an object, as bread desired, and to the organ, as the mouth, which feels the want. This element is also a prolific fount in the elaborated modes of communication of instructed mutes; as it is in the kindred address to the eye in all hiero- glyphic and primitive alphabetic signs. The second element, the tropic, pro- vides for the expression of sentiment; by which mental convictions and emotions, as distinct from bodily wants and material operations, are commu- nicated. The third, or phonetic element, is the only possible means provided by the language of signs, whether those signa be the motions of the mutes or the characters drawn by the pen, to enable mutes to communicate proper as distinct from common names.
In the merely elementary education heretofore given to mutes, these three elements of sign-language have been found adequate after the generations of inventive skill employed to enlarge their number. It became a serious ques- tion with the first deaf-mute college, whether any power of invention would prove adequate to the task of multiplying signs which could communicate every form of conception and sentiment essential to the comprehension of the several departments of mathematical and metaphysical investigation re- quired in the stady of natural science, of mental, moral, and ssthetical phi- loseophy. This difficulty has led to the effort to train mutes to learn by the
20 : Proceedings of the
eye to copy vocal utterance ; and this attempt, as well as the enlarging of the field of sign-language, throws light upon the origin and growth of lan- guage.
The effort to train mutes to vocal utterance by mere observation of the play of the vocal organs in others, calls attention to the facts that the vowels re- quire but a very few, ten or twelve only, fixed positions of the vocal organs; while the consonants, only from twenty to twenty-five in number, are pro- duced by movements observed in four organs, the lips, teeth, tongue, and larynx. All these fixed positions and movements can be carefully noted by watching with the eye and by placing the hand on the throat. The force of the utterance necessary to make these pesitions and movements convey vocal sounds to those who hear, is learned by holding the hand of the mute before the mouth of the teacher. The elements of vocalization thus become so simplified and practically comprehended by the mute, that in due time he can follow the speaker and distinctly respond, as sure and confident in the use of his natural organs as the trained pianist is of the utterance that will come from tquching his artificial keys.
In the advance of invention in sign-language, it is found, as might be anti- cipated, that only simple ideas, or roots, can at first be formed ; that concep- tions of time, mode, quality, etc., enumerated in the categories of Aristotle and of Kant, must be conveyed in separate signs; while time permits the shortening, the combining, and finally the elaboration of grammatical pre- fixes and affixes to root-signs, so that case, mode, and tense assume the cha- racter found in the most polished tongues.
Thus in a new language forming in aggeration, the origin of language as ‘an invention, and its growth to maturity, may be scanned.
Upon motion, the President appointed Professor W. S. Tyler, Professor H. N. Day,’and Rev. J. Anderson, a committee to sc- lect the place for the next annual meeting.
Upon motion, the president appointed Rev. Dr. Hyde, Profes- sor S, S. Haldeman, and Rev. C. H. Brigham a committee to nominate officers for the ensuing year.
The president announced that the members of the Association were invited by the president of Yale College to visit the differ- ent buildings and cabinets of the college during the intermission between the morning and afternoon sessions.
AFTERNOON SESSION.
Dr. J. Thomas, of Philadelphia, and Dr. N. W. Benedict, ot ‘Rochester, read papers upon the “ Pronunciation of the Latin and Greek Languages.” The reading of these papers was followed -by a long discussion, but as the committee on pronunciation ot the ancient classical languages, which was: appointed last year, and to which the whole subject was referred, had made no further report, no additional action was taken by the Association.
American Philological Association. 21
EVENING SEssI0n.
The first paper was on “ The Celtic Elements in French,” by Professor A. H. Mixer, of Rochester University, N. Y.
Buffon said many years ago, “Le style, c’est ’homme.” - The linguistic science more recently has added with emphasis, The language; it is the na- tion. Every people, it is now claimed, may be analyzed by its speech. Lan- guage furnishes to civil history its background. It takes up the thread where history began, and traces it back to the beginning of the nation—the infancy of the race. It thus solves the most difficult problems of ethnology ; those pertaining to the beginnings of our primitive races. But more than this, the speech farnishes much of the very minutia of subsequent history, and all the more important changes and phases of the progressive national life, Hence the testimony of language and that of history must agree. Where these now seem discrepant, we must suppose the disagreement but © apparent, as the future investigations may prove.
This article proposes to examine the linguistic testimony as the nationality of the French Is the Frenchman essentially Celtic or Roman? Every stu- dent of the French language and peuple finds from the outset of his work this discrepancy, that the language.appears more completely Roman than the people does, or than the facts of history seem to justify. Are we mistaken in the facts of language or those of history? What are the historical proba- bilities ?
The French nation is the ieeue of three. successive waves of. migration. The Roman invasion, where history begins, found the country occupied by the Celts. These form the basis of all future national and linguistic growth: The picture of the ancient Celt, as drawn by Roman writers, strikingly cor- responds in every feature with that of the Frenchman of to-day: “In order to understand the history of the French nation,” says Heeren, “ it is necessary to consider it the issue of the Celtic race. It is thus only that we can explain this character which, in spite of the various intermixtures to which the Celtic population has been subjected, remains even to-day in the French such as it is delineated in Cesar.” “The Celts were not barbarians, but true heralds of civilization wherever they settled ; the equals of Saxons and Romans and Greeks, whether in physical beauty or intellectual vigor.”
With such testimony to the magnitude and character of this ancient peo- ple, can we doubt the necessarily powerful influence it exerted upon any na- tion with which it combined? Were the millions of Gauls absorbed and lost ? Was there not here, as In all such cases, a compromise, and will not the language, when properly analyzed, show this ?
All possible influence of the Celt in the formation of the French is included in contributions to the vocabulary, and changes effected in the other elements, chiefly the Latin. The contributions to the vocabulary were very few, for reasons readily apparent; but the changes wrought in the Latin elements ‘were numerous and great, including. both chanjyes in words and changes in syntax or the fundamental structure of the tongue. Perhaps the most strik- ing feature in the French is, that nearly all the words appear here ortho- graphically and phonetically shorter than in the language from which they
22 Proceedings of the
are derived. The phonetic decay is immense. Is not this remarkable feature, which even Schleicher thinks is due to some strong local influence, to be credited in a high degree to the Celt, whose hasty and impetuous tem- perament would naturally tend to bring about just this result? Itisa sig- nificant fact that all the simple sounds in Freneh are found in NeoCeltic Bretonne dialect, and also all those of the Bretonne with the single exception of the guttural ch are in the French. Several of these sounds were unknowy to the Latin. The nasa] sound, the most marked peculiarity of the French pronuaciation, appears due to the Celt. Indeed, so numerous and great have been the changes from this source as to lead us to conclude that in the most characteristic feautures of her phonetic system the French language i is not of Latin, but of Celtic birth.
The changes in the syntax have been equally numerous and radical. The fandamental law of the Latin was synthesis and dependence ; that of the French, analysis and independence. The Celt has broken the proud strac- ‘ture of the Roman into fragments. These fragments are used in the forma- tion of the new speech, but that the Celt is the artisan is seen throughout all. He has also caused to enter into the new tongue that simplicity, directness, elasticity, and vivacity—that spirit—that something which renders it surely more Gallic than Roman. The testimony of language is thus found to har- monize with that of history in leading us toward the conclusion that the French character is cosentially the offspring of the ancient Celt.
The second paper was entitled “Studies in Cymric Philolo- gy,” by Professor E. W. Evans, of Cornell University, Ithaca, N. Y.
This paper was devoted to the discussion of various questions relating to the early Welsh. .
After stating the consonant changes which, as shown by Zeuss, character- ize the transition from old to middle Welsh, (p, t, c, when not initial, to b, d, g; b and m to v; etc.,) Professor E. showed, further, that these changes, when the consonant is not followed by another, are regularly attended with the lengthening of preceding short vowels: thus, cat, battle, becomes cdd, epavol, colt, becomes ddatl ; etc.
Professor E. showed by examples that the verb-ending -it, 8d sing. pres. act. ind., (compare Irish -id and Latin -it,) which Zeuss, or his editor, finds pre- served only in one Welsh gloss, is really of frequent occurrence in the old Welsh pets.
As another result of his own researches, Professor E. showed that the -ator, -etor, -itor, of the old Welsh poets, heretofore treated as gerund and supine endings, are really passive endings of the finite verb—the equivalents of the Irish -ither, -ithir, and the Latin -atur, -etur, -itur.
He impugned Zeuss'’s etymology of the name Cymro, Cambrian, (cyn, con, and bro, regio,) by showing that the early derivatives of the name indicate eymra as its older form.
Among other subjects discussed by Professor E., was an old Welsh gloss of the Folium Luzemburgense, which Zeuse passes over as obscure,
American Philological Association. — 23
The next paper was on * Algonkin Names of Man,” by Hon. J. H. Trambull, of Hartford, Ct.
The Indian speaker aimed always at extreme precision. He never genera- lized. His vocabulary was poor in generic terms. It grew by progressive differentiations—from genera to species, from species to varieties and indi- viduals. There was not, perhaps, in the Indian mind, certainly not in his language, any inherent incapacity for generalization ; but he avoided it as ° a defect, whether in thought or speech.
A large proportion of the English words selected for the basis of compara- tive vocabularies of American languages are generic or class names. None of thege can be translated by a single word, in any Indian language. Every Algonkin language, has, for example, names for “elder brother,” “twin brother,” and “ younger brother,” for “son of the same father,” and “ eon of the same mother ;” but in no dialect is there a precise equivalent of the ap- pellative “ brother.”
The names given by various tribes to man, or rather the names’ which have been assumed to be the equivalents of the English word man, in its two senses, of “an individual of the humaa race,” (homo,) and “ one possessing in a high degree the distinctive qualities of manhood,” (vir,) have occasioned much perplexity to the vocabulary-makers. The truth is, that it is as impos- sible to find an Indian word with the precise meaning of the Latin homo as of the English “man.” By resorting to the Latin, the difficulty is halved merely, not escaped. No American language has any single name applicable alike to the red man and the white, to the speaker’s own tribe or nation and to his enemies or subjects, to young and to old, to chief and to councilor, prisoner and slave, and in its larger sense common to both sexes. Vir, it is true, finds in almost all dialects a correspondent term ; but homo is untrans- latable by an Indian.
Mr. Trumbull proceeded to analyze and discuss the meaning of the two classes of names for ‘“‘ man ” found in the Algonkin languages—both cf which are combined in the Abnaki aren-ambé, and the national name of the Dela- wares, lennitendpe. He pointed out the errors of Heckewelder, Cass, and Duponceau in the analysis of this Delawarename. The Massachusetts ninnu, Abnaki arent, Delaware lenné, are identical ; n, r, and 7 being interchangeable: in Algonkin dialects. Each means a man “ of the same kind” as the speaker, that is, an Indian—an “ origional,” or “common,” or “ normal ’’ man, ‘as op- posed to a “stranger,” or “ foreigner,” or one “of another language.” This contrast is preserved in the names [linois, (lenné or ilini, with the termi- nation given it by the French,) “ men of our kind,” and Peoria, from piroue, “strange,” “foreign,” which was a village of Indians speaking a strange language.
Omp, dpe, embe, ia a noun generic denoting an “adult male;” primarily, the Latin maa, not viv. With one or another prefix, it denotes a “ chief man,” “captein,” “ husband,” “brother,” etc. Its primary signification is “to stand upright ;” ¢ walking in an erect posture,” Heckewelder translates it.
Lenni-Lendpe means, “adult Indian men,” “ viri of our race ;” “men like ourselves” —of the “‘ common” or “ normal ” type,
This was illustrated by the analysis of tribe names in several languages of | the Algonkin stock.
q v
24 Proceedings of the
Morning Session, Fripay, Jury 28.
The committee upon nominating officers for the ensuing year reported the following nomipations, whjch were carried unani- mously :
President.—Professor W. W. Goodwin, Harvard University,
‘Cambridge, Mass.
Vice-Presidents.—Rev. Dr. A. C. Kendrick, Rochester Univer- sity, N. Y.; Rev. Dr. W. S. Tyler, Amherst College, Mass. |
Secretary and Curator.—Professor G. F. Comfort, New-York.
Treasurer.—Hon. J. Hammond Trumbull, Hartford, Ct.
Additional Members of the Executive Committee:—Rev. Dr. A. N. Arnold, Chicago University, Il]. ; Chancellor Howard Crosby, New-York University ; Professor James Hadley, Yale College, New-Haven, Ct.; President G. W. Samson, Columbian College, Washington, D.C.; Professor J. B. Sewall, Bowdoin College, Brunswick, Me.
Upon the motion of Dr. Samson, it was voted that the duties of the officers-elect should commence with the close of each annual session.
The report of committee for selecting the place for holding the next (fourth) annual meeting of the Association, fixing upon Pro- vidence, R. I., was unanimously adopted.
The first paper of the morning was on “The Chronology of some of the Events mentioned in Demosthenes on the Crown,” by Pro- fessor W. W. Goodwin, of Harvard University, Cambridge, Mass.
Professor Goodwin said, It is generally conceded by scholars that the de- crees, laws, and other documents included in the text of Demosthenes on the Crown are spurious; but the fact that Clinton’s Fastt Hellenici, which recognizes the documents in question as of historic authority, stil] remains the chief authority in Greek chronology, coupled with the recent repub- lication of a popular edition of Aeschines against Ctesiphon which follows Clinton’s chronology in general, makes it worth while to call attention to the subject. If these documents are taken as authority, it is absolutely impossi- ble to make any consistent chronology of the events which preceded the bat- tle of Chaeronea, and it will strike every one who examines the subject that not a single name of an archon, and not a single important date given in these documents, can be shown to be correct. But if we follow the plain statements found in the text of Aeschines and Demosthenes, nothing can be simpler than the whule order of events. Aeschines tells us that in the last month of the year of Charondas (338-7 B.c.) Demosthenes was appointed one of a commission of ten for restoring the walls of Athens. Now, the whole object of Aeschines in this part of his argument is to prove that the decree for crown ing Demosthenes was proposed during the latter’s year of office, It was
T(
American Philological Association. * 25
therefore proposed during the year which followed his election, that is, the year 337-6 B.o. Asthe crown was to be proclaimed at the Great Dionysia in the spring, the decree of Ctesiphon must have preceded that festival; and as the indictment must have been brought against the decree immediately after the proposal—that is, between its passage by the Senate and the first day on which it could be presented in the assembly—we must place the two events togeth- er, probably in the winter or early spring of 337-6 B.c. But Aeschines further states that his indictment preceded the death of Philip, that is, the summer of 336 B.c. But the documents above mentioned place the decree in October of the year of Euthycles, (who never was Archon at all,) and the indictment in March of the year of Charondas, 338-7 B.c. The same is true of the dates of the events which precede the battle of Chaeronea. Leaving out of account the spurious documents, we see that Aeschines made his famous speech at Delphi in the spring of 339 B.c. ; that Philip was chosen general of the Am- phictyons in the autumn of the same year; that immediately after his elec- tion Philip passed Thermopylae and seized Elatea. The panic at Athens and the embassy of Demosthenes immediately followed. The alliance with Thebes and the entrance of the Athenian army into Boeotia succeeded ; and the skirmishes between the allied forces and Philip’s invading army fell in _ the winter, one of these being called by Demosthenes “the winter battle.” The decisive battle of Chaeronea was fought in the following summer. The year of the delivery of the orations is shown, by a great variety of testimony, to have been 880 B.c,
The second paper was on “The Mode of expressing Number in certain Indian Languages,” by Mr. George Gibbs, of New. Haven, Ct.
Mr. Gibbs referred to a paper by him published in the Smithsonian Miscel-
laneous Collections, (160, Appendix B,) on the numeral systems of certain languages, showing that modifications, either by direct change or the intro- duction of particles, were then used in counting different objects. This pe- culiarity exists not only in very many American languages, as, for instance, the Algonquin, the Iroquois, the Selish, the Mexican, and others ; but among some of the South-Sea Islands, and even in the Yoruba, an African language. The idea appears to be in all cases descriptive. In some languages, it seems to correspond to the “ animate” or “inanimate” objects counted ; but in others, it has a vastly wider range, and the connection between them can not easily be traced. _ The most extraordinary development known is in the Cakeniquel or Gua- temalteco of Central America. The late Mr. Theodore Dwight, a member of the American Ethnological Society, had found in a rare work on that lan- guage, not leas than forty-five modifications of the simple or common nume- rals, of which it will be sufficient here to mention a few: In counting words, the syllable pak is added, cutting off the last letter of the simple numeral, thus, Aun, one, hupah, one word ; cay, two, capah, two words.
In the same way, other changes are made in counting by threads or strings ; by hands; by pairs; by closed hands, or fists; by drops; by fragments, splinters, crumbs, and swallows in drinking; by spoonfuls; in counting
/§ 11
26 Proceedings of the
timbers, poles, and fishes ; in counting provinces ; parties; globular objects, as eggs, loaves of bread, etc.
In the Cherokee language, according to the Rev. J. B. Jones, the same idea enters into the verb; as, for instance, “I take a long object ; I take a round object ; I take objects folded, or which can be laid one upon another ;” but it does not appear in the numerals themeel ves.
Further remarks were made upon a peculiarity in the Uniapa language, of one of the Micronesian Islands, in which it appears that a systematic ana- tomica] vocabulary existe ; also distinct names, not systematic, for geometrical figures, etc., and finally distinct numerals for different classes of objects.
Mr. Trumbull read some extracts from a letter lately received from Dr. Washington Matthews, U. 8. A., Post Surgeon at Fort Buford, D. T., accompanying specimens of a dictionary of the language of the Minitares or Gros Ventres, Indians living near the north branch of the Missouri, between the Mandans and the Yellow Stone River, This language, of which vocabularies have been published by Say, the Prince Wied-Neuwied, and Dr. F. V. Hayden, is of the Dakota stock, nearly related to (if not to be regarded as really a dialect of) that of.the Aubséroke or Crows. Dr. Matthews, who has had favorable opportunities for acquiring _a knowledge of the language, had nearly completed a dictionary of it, when his manuscript collections were destroyed by fire, at Fort Buford. He is now busily engaged in re-writing his work, in which he aims to exhibit the analysis of every conr pound word and to refer derivatives to their roots.
“I consider (he writes) that not only is the analytical method the best way of studying the aboriginal tongues, but it is the only way of studying them, that is, of giving to them an attention which we may dignify by the name of study. In converting an oral into a written language we can only secure a reliable orthography by a carefal analysis. I have found that In- dians, in conversation, will take the same liberties with their language, that Europeans do with theire—abbreviating, suppressing sounds, and slurring words together, where it suite thelr convenience. A word as it falls from the lips of a speaker is not always to be pat down with a certainty of correctness. In different connections, and in the mouths of different indjviduals, the same word will often sound differently, and the sandard word, as it should be writ- ten, can only be discovered by analysis. A person who endeavors to ‘ pick up’ a language from the Indians themselves, will so frequently hear phrases in- stead of words, and experiences such difficulty in obtaining the single word sign for any idea, that he wil] progress but slowly unless he attempts to ana- lyze from the beginning ; it is even difficult to fix a long compound word in the memory, unless the component parts of it have been discovered. So in the first steps to the acquisition of a language, as well as in the more tho- rough examination of its grammatical construction, he must be assisted by analysis.”
American Philological Association. 27
“In comparing one language with enother, or in endeavoring to drew cor- clusions as to origin from such comparison, analysis is indispensable, gud the erroms committed hy those who disregard it are aggravating,” ag Dr. Mat- thewa shows by examples from the vocabularies. “ Again, in languages to- tally different we occasionally find words of similar meaning, pronounced alike, or nearly alike, and only from analysis we can learn their dissimilarity.” Mr. Catlin, endeavoring to establish the Welsh origin of the Mandans, calls attention to the “striking similarity’ of the Mandan Maho penela, “the Great Spirit,” and the Welsh Mavwr penaethir,“ to act asa great chief—head or principal—sovereign orsupreme.” “The analysisof the Mandan name would have shown Mr. Catlin that it is formed from ma, ‘the, that, which,’ hopini, ‘medicine, mystery,’ etc., and te, (pronounced hie or hées, except in compounds,) ‘ great;’ that it should be written ma-hopinite,‘ the great mystery’ or ‘modicine;’ and that there is no possible connection between it and the Welsh mawr, penacthir, ‘ head-man ;’ moreover, the Mandans call their high- est deity or deities, no¢ ‘The Great Spirit,’ but ‘Chief of Life,’ ‘Maater of Life,’ (Omahank numakei,) or ‘ The Firat Man,’ (Numak mahhana,) as we may learn even from the first part of Mr, Catlin’s work—written before he started his hypothesis of Welsh origin.” ,
The next paper was on “ An Ancient Bulgarian Poem concern- ing Orpheus,” translated by Rev. C. F. Morse, for thirteen years a missionary of the A. B.C. F. M. in Bulgaria.
| It was presented, in the absence of the author, by Professor Whitney. The poem is said by Mr. Morse to have been brought to light by 8. I. Verco- vich, of Serres, and printed at Moscow, in 1867, by a Russian ethnographical society, with a prose translation and preface in Russian. Ite age is unknown; but it must have been composed since the Bulgarians received Christianity, and, judging by the dozen Turkish words contained in it, since the Turkish conquest of the country. Probdbly it is two or three centuries old, and pro- duced in Macedonia. The metre is somewhat irregular ; repetitions of the end of one line at the beginning of the next are a frequent feature. The translation given is quite literal, and not in verse.
The poem tells how Orpheus went to Arabia to win a royal bride, and finally succeeded in bringing her home, after overcoming various obstacles and enemies of a supernatural character. The whole action is completely, almost absurdly, supernatural, and borrows nothing whatever from Greek tradition or fable, except the name of the hero, and his skill upon the harp, and power to work wonders by its means.
Professor A. O. Kendrick, of Rochester University, N. Y., read & paper on the “ Pronunciation of the Greek Aspirates.”
He first gave an account of the origin of the aspirates, and then controverted the arguments of Professor George Curtius in favor of their being uttered like . the Sanskrit aspirates with a separate sound of the breathing. The paper ep. deavored to set aside Professor Curtius’s arguments drawn successively from the movableness of the breathing, or its easy disengagement from the body of the consonant, from the testimony of inseriptions, from the barbarian cor- ruptions in Aristophanes, and from the mode of transliterating the Greek aspi-
28 Proceedings of the
rates into Latin ; and also to show the insufficiency of his answer to Arendt’s objection to the dictum, founded on the difficulty or impossibility of pro- nouncing the frequently recurring combinations ¢d, 7d, with a separate ut- terance of the breathing. The purpose of the paper was negative rather than positive ; alming rather to weaken the force of the arguments for the aspirate theory, than to establish the opposite view.
' Mr. J. B. Greenough, of Harvard University, presented the next paper, showing that the “ General Supposition,” first dis- tinguished by Professor Goodwin, in Greek, which ‘is expressed by the subjunctive or optative in the protasis and the indicative of a general truth in the apodosis, was also found in Latin.
He argued that the construction was an heirloom of the family, and not, as is commonly supposed, developed separately in Greek.
This view he supported from the fact that the second person singular of an indefinite subject in a hypothetical sentence regularly takes this form in Latin in all periods of the language and in all kinds of writers, though the first and third persons take the more usual Latin form of the indicative in both clauses. As in Cato, Carmen de Moribus, Vita humana prope uti ferrum est, si exerceas conteritur. While there were hundreds of instances of this form in writers of all periods, not more than two or three cases of the indica- tive could be found. That this was not an imitation of the Greek, he argued from its universality, as well as from the fact that in the same sentence the subjunctive of the second person and the indicative of the other two appeared side ‘by side, whereas in Greek there was no such distinction ; but the modal forms were used in all persons alike in this construction. He also showed that the same usage occurred both in the Vedic and later Sanskrit. From these arguments he concluded that the construction belonged to the time of Indo-European unity, or, what practically amounts to the same thing, if it is not really the same, the time of Greeco-Italic Sanskrit unity.
' Hon. J. H. Trumbull presented, and gave.an abstract of, a paper comprising “ Contributions to the Comparative Grammar of the Algonkin Languages,” founded on twenty-five versions of the Lord’s Prayer, in nineteen languages and dialects of the Algonkin stock.
The writer had endeavored to give a literal English translation of each of these versions. Brief grammatical notes were added, and some of the most obvious peculiarities of the several dialects were pointed out. The difficulty of forming.an accurate judgment of the nearness of relationship between two Indian languages, by the comparison of vocabularies compiled or translations made by different persons and at different periods, was incidentally illustrated. The unlikeness of two versions does not prove the unlikeness of the two lan- - guages in which they are made. It may be attributable to the incompetence of the translator, or to the wide range in the selection of words more or less nearly equivalent to those of the European text which do not admit of exact transixtion into an American language. The word “ bread,” for example, in the fourth petition of the Lord’s Prayer, was translated by Eliot, in the lan- guage of Massachusetts by nummeetsuong[anon]ash, (Matthew 6 : 11,) and
American Philological Asscoiation. 29
petukgunneg, (Luke 11:8.) The former word (corrected by the insertion of two syllables which were omitted by the printer or transcriber) means “ our eatings” or “ meals ;” the latter is, literally, “something rounded,” and was the Indian name of a smal! round cake or loaf. In the Abnaki versions, the word selected means “ baked (or roasted) grain ;” in Zeisberger’s Delaware, it is, simply, “ something baked ;” in Edwards’s Muhhekan, tguogh, another way of writing the familiar “tuckahoe” of Virginia and the Middle States, from the same root as Eliot’s petukqunneg, and denoting “something round.” The Algonkin (Canada) and some Chippeway versions have a word by which the Indians distinguished a loaf of bread of European fashion, as “ something to be cut off from,” that is, to be sliced, or cut in pieces.
President Samson gave a short abstract of a paper, the read- ing of which he deferred for lack of time till the next session, upon the “Families of Languages as developed in the Mediterra- nean Civilization, and their Influence upon each other.”
The following resolutions were offered by Rev. C. H. Brigham, and were carried unanimously:
Resolved, That the members of the American Plilological Association thankfully acknowledge the lavish and graceful hospitality which they have found in the homes of the citizens of New-Haven, and from the officers ot Yale College. ’
Resolved, That the thanks of the Association be given to the local commit- tee of New-Haven, for their arduous and incessant labors to provide for the comfort and convenience of the members of the Association during the present session.
Resolved, That the thanks of the Association be given to the State Govern- ment of Connecticut for their kind permission to the Association to use the rooms of the Siate-House during the present session.
President Samson offered the following resolution, which, after considerable discussion, was referred to the Executive Committee ;
Resolved, That a committee be appointed by this Association to consider and report upon the expediency and practicability of securing a comprehen- sive analysis of the English language, as spoken and written in the Ameri- can States, which shall be in harmony with those prepared to aid students in other modern languages, now so generally spoken among us and required in our schools.
The same disposition was made, after some discussion, of the | following resolution, which was offered by Mr. Sawyer, of Bos- ton: ,
Resolved, That the Executive Committee be requested to consider and re- port to the Association at the next annual meeting, a plan for the systematic division of the proper work of the Association and for holding preliminary local meetings.
The minutes of the Association were read, and, after some re-
marks by the President, the Association adjourned at twelve m.
MEMBERS OF THE AMERICAN PHILOLOGICAL ASSOCIATION.
Frederick D. Allen, East-Tennessee University, Knoxville, Tenn.
William F. Allen, University of Wisconsin, Madison, Wis.
Joseph Anderson, Waterbury, Ct.
Martin B. Anderson, University of Rochester, Rochester, N. Y.
N. L. Andrews, Madison University, Hamilton, N. Y.
Albert N. Arnold, Baptist Theological Seminary, Chicago, D1.
John G. Barton, College of the City of New-York.
F. L. Batchelder, Stafford, Ct.
H. Louis Baugher, Pennsylvania College, Gettysburg, Pa.
E. L. Beadle, Poughkeepsie, N. Y.
Nehemiah W. Benedict, Free Academy, Rochester, N. Y.
Otis Bisbee, Poughkeepsie, N. Y.
James H. Blakely, Female College, Bordentown, N. J.
G. R. Bliss, Lewisburg University, Lewisburg, Pa.
Porter C. Bliss, Washington, D. C.
James R. Boise, University of Chicago, Chicago, II].
Alice R. Boise, Chicago, II.
Henry L. Boltwood, Princeton, III.
Mary L. Booth, 89 Madison avenue, New-York.
P. Born, Selingsgrove, Pa.
James P. Boyce, Southern Baptist Theological Seminary, Greeneville, 8, C.
Charles E. Brandt, Farmington, Ct.
H. C. G. Brandt, Clinton, N. Y.
_ Fisk P. Brewer, University of North-Carolina, Chapel Hill, N. C.
Charles H. Brigham, Ann Arbor, Mich.
Charles A. Bristed, Lenox, Mass.
Charles J. Buckingham, Poughkeepsie, N. Y.
R. J. W. Buckland, Rochester Theological Seminary, Rochester, N. Y,.
Elihu Burritt, New Britain, Ct.
Horatio Q. Butterfield, 62 Bible House, New-York.
Henry A. Buttz, Drew Theological Seminary, Madison, N. J.
Samuel W. Capron, High School, Hartford, Ct.
Franklin Carter, Williams College, Williamstown, Mass.
Alexis Caswell, Brown University, Providence, R. I.
William C. Cattell, La Fayette College, Easton, Pa.
Charles F. Chandler, School of Mines of Columbia College, New-York.
Talbot W. Chambers, New- York.
Elie Charlier, (Life-Member,) Institute for Young Men, 124 East Twenty. fourth street, New-York.
Elisée Charlier, Institute for Young Ladies, Madison Avenue and East Thirty-third street, New-York.
Francis J. Child, Harvard University, Cambridge, Mass.
American Philological Assuciation.
Henry F. Clark, Berea College, Berea, Ky.
Nelson E. Cobleigh, East-Tennessee Wesleyan University, Athens ‘Tenn. William P. Coddington, Syracuse University, Syracuse, N. Y Lyman Coleman, La Fayette College, Easton, Pa.
Henry M. Colton, Yale School, 1193 Broadway, New-York. George F. Comfort, (care of Harper & Brothers,) New-York Edward B. Coe, Yale College, New-Haven, Ct.
Williain B. Corbyn, Quincy, Ill. \ William C. Crane, Baylor University, Independence, Texas. A. Crittenden, Packer Collegiate Institute, Brooklyn, N. Y. Howard Crosby, University of New-York, (302 Second avenue, New-York.) Edward P. Crowell], Amherst College, Amherst, Mass.
Henry N. Day, New-Haven, Ct.
George E. Day, Yale College, New-Haven, Ct.
J.C. Daniels, Olivet College, Olivet, Mich.
H. A. Dearborn, Tuft’s College, Malden, Mass.
Elizabeth H. Denio, Albion, N. Y.
Schele De Vere, University of Virginia, Charlotte, Va.
Mary C. Dickinson, Northampton, Mass.
E. 8. Dulin, St. Stephen’s College, Columbia, Mo.
Benjamin W. Dwight, Clinton, N. Y
¥rederick Ebener, Baltimore, Md
Asher B. Evans, Lockport. N. Y.
Evan W. Evans, Cornell University, Ithaca, N. Y.
Leon C. Field, Cazenovia, N. Y. P
William M. Fisher, Independence, Mo,
J. B. Feuling, University of Wisconsin, Madison, Wis. Martha B. Flint, Monticello Academy, N. Y.
William C. Fowler, Durham, Ct.
A. W. Fowler, New-York.
Norman Fox, William Jewell College, Liberty, Mo
Simeon T. Frost, Amenia, (Seminary,) N. Y.
Helen M. French, Mount Holyoke Seminary, South-Hadley, Masa, Merrill E. Gates, Albany, (Academy,) N. Y.
Joshua B. Garritt, Hanover College, Hanover, 111.
Willis C. Gaylord, Rochester, N. Y.
Elizabeth L. Geiger, Burlington College, Burlington. Towa. George Gibbs, New-Haven, Ct. ‘
Daniel C. Gilinan, Sheffield Scientific School, New-Haven, Ct. Joseph H. Gilmore, University of Rochester, Rochester, N. Y. William M. Goodrich, Poughkeepsie, N. Y.
William W. Goodwin, Harvard University, Cambridge, Mass. James B. Greenough, Harvard University, Cambridge, Mass. Edward M. Greenway, Baltimore, Md.
Ephraim W. Gurney, Harvard University, Cambridge, Mass. James Hadley, Yale College, New-Haven, Ct.
S. S. Haldemann, University of Pennsylvania, Columbia, Pa, Horatio Hale, Clinton, Ontario, Canada.
American Philological Association,
Laecian H. Hammond, Lebanon Valley College, Annville, Pa. James H. Hanson, Classical Institute. Waterville, Maine. Albert Harkness, Brown University, Providence, R. L.
Calvin S. Harrington, Wesleyan University, Middletows, (2. Samuel Hart, Trinity College. Hartford, Cs
Willabe Haskell, Bucksport, Maine.
B. J. Hawthormme, West Tonnessee College. Jackson, Tenn.
H. W. Haynes, University of Vermont, Burlington, Vt.
John H. Hewitt, Olivet Vollege, Olivet, Mich.
8. D. Hilman, Dickinson College, Carlisle, Pu.
Thomas Wentworth Higginson, Newport, B. I.
Oscar Howea, Shartleff College, Alton, DIL
C. J. Hadson, Genesee College, Lima, N. Y.
John S. Huntington, Trinity College, Hartford, Ct.
Ammi B. Hyde, Alleghany College, Meadville, Pa.
Milton M. Humphreys, Washington College, Lexington, Va. William H. Jeffers, Wooster College, Wooster, Mo.
8. A. Theo. Jobe, St. John's College, Little Rock, Ark. Edwin E. Johnson, Trinity College, Hartfort, Ct.
Asahel C. Kendrick, University of Rochester, Rochester, N. Y. A. C. Kimber, Jr., St. Stephen’s College, Annansdale, N. Y. Louis Kistler, North-Western University, Evanston, I. William G. W. Lewis, Poughkeepsie, N. Y.
Abiel A. Livermore, Unitarian Theological Seminary, Meadville, Pa. Charles L. Loos, Bethany College, Bethany, W. Va.
Benson J. Lossing, Dover Plains, N. Y.
Thomas R. Lounspury, Yale College, New-Haven, Ct. Rebecca Lowry, 205 West Sixteenth street, New-York.
0. F. Lumry, Wheaton College, Wheaton, II.
William F. Lush, Rochester, N. Y.
Bela P. MacKeon, Cornell University, Ithaca, N. Y.
George F. Magoun, Iowa College, Grinnell, Iowa.
Joseph H. McDaniels, Geneva, N. Y.
George McMillan, Hillsdale College, Hillsdale, Mich. Edward H. Magill, Swathmore College, Philadelphia, Pa. Francis A. March, La Fayette College, Easton, Pa.
Dan Marvin, Jr., Racine College, Racine, Wis.
Washington Matthews, Fort Buford, Dakota Territory. Milton Meury, Cold Spring, N. Y.
Charles M. Mead, Andover Theological Seminary, Andover, Mass. W. J. Milne, Lockport, N. Y.
John L. Mills, Marietta College, Marietta, Ill.
A. H. Mixer, Rochester University, Rochester, N. Y. William L. Montague, Amherst College, Amherst, Mass. Charles D. Morris, Mohican Lake, N. Y.
Edward D. Morris, Lane Theological Seminary, Cincinnati, Oh‘o, F. A. Muhlenberg, Muhlenberg College, Allentown, Pa.
American Philologtcal Association.
James ki. Munson, 84 Park Row, New-York.
J.H. Myers, Milton, N. Y.
Bennett H. Nash, 62 Boylston street, Boston, Mass.
C. W. Nassau, Lawrenceville, N. J.
_ William M. Nevin, Franklin Marshall College, Lancaster, Pa. Edward North, Hamilton College, Clinton, N. Y.
Thomas H. North, St. Catharine’s, Canada.
F. W. A. Noetz, Muhlenberg College, Allentown, Pa.
Cyrus Nutt, Indiana University, Bloomington, Ind.
vohn C. Overhiser, New-York.
Lewis R. Packard, Yale College, New-Haven, Ct.
William A. Packard, College of New-Jersey, Princeton, N. J. BE. G. Parsons, Derry, N. H.
A. Parish, New-Haven, Ct.
J. Person, Belvidere, (Academy,) N. J.
John C. Pershing, Pittsburg Female College, Pittsburg, Pa,
D. L. Peck, University of Alabama, Tuscaloosa, Ala.
Tracy Peck, Bristol, Ct.
Oval Pirkey, Christian University, Canton, Mo.
Noah Porter, Yale College, New-Haven, Ct.
Samuel Porter, National College of Deaf-Mutes, Washington D.C E. L. Prentice, Poughkeepsie, N. Y.
A. J. Quinche, University of Mississippi, Oxford, Miss.
John H. Raymond, Vassar College, Poughkeepsie, N. Y. Stephen R. Rigys, Fort Wadsworth, Dakota Territory. Timothy H. Roberts, Whitney’s Point, N. Y.
Lewis L. Rogers, Indiana Asbury University, Greencastle, Ind. Ezekiel G. Robinson, Rochester University, Rochester, N. Y. Edwin R. Ruggles, Dartmouth College, Hanover, N, H. William C. Russell, Cornell University, Ithaca, N. Y.
George T. Rider, Poughkeepsie, N. Y.
George W. Samson, Rutgers Female College, New-York. Wesley C. Sawyer, Boston, Mass,
A. J. Schem, 114 Nassau street, New-York.
Charles A. Schlegel, Female Normal College, New-York. Henry Schliemann, Paris, France.
Eugene Schuyler, United States Legation, St. Petersburg, Russia John S. Sewall, Bowdoin College, Brunswick, Maine.
Jotham B. Sewall, Bowdoin College, Brunswick, Maine. Joseph Shea, St. John’s College, Fordham, N. Y.
Hiram W. Sibley, Rochester, N. Y.
A. B. Smith, Poughkeepsie, N. Y.
§. K. Smith, Colby University, Waterville, Maine.’
E. Snyder, Illinois Industrial University, Champaign, IIL
J. W. Stearns, University of Chicago, Chicago, Il.
William Steffin, St. John’s College, Annapolis, Md.
Frederick Stengel, School of Mines of Columbia College, New-York. William A. Stevens, Denison University, Granville, Ohio.
American Philological Association.
Daniel S. Talcott, Bangor Theological Seminary, Bangor, Maine. E. A. Tanner, Illinois College, Jacksonville, 1).
J. Henry Thayer, Andover Theological Seminary, Andover, Mass. Thomas A. Thacher, Yale College, New Haven, Ct.
Joseph Thomas, 116 North Eleventh street, Philadelphia, Pa. Crawford H. Toy, Southern Baptist Theological Seminary, Greenville, 8. C J. Hammond Trumbull, Hartford, Ct.
Henry M. Tyler, Knox College, Galesburg, Il.
William 8. Tyler, Amherst College, Amherst, Maas.
E. H. Twining, University of Minnesota, St. Anthony, Minn. Milton Valentine, Pennsylvania College, Gettysburg, Pa.
John C. Van Benschoten, Wesleyan University, Middletown, Ct. Addison Van Name, Yale College, New-Haven, Ct.
Thomas R. Vickroy, Lebanon Valley College, Annville, Pa.
Julia E. Ward, Mount Holyoke Seminary, South-Hadley, Mass. Horatio G@. Warner, Rochester, N. Y.
M. Warren, Providence, R. I.
J. B. Weston, Antioch College, Yellow Springs, Ohio.
Mrs. A. E. Weston, Antioch College, Yellow Springs, Ohio. Albert 8S. Wheeler, Florida, N. J.
Nehemiah White, St. Lawrence University, Canton, N. Y. William D. Whitney, Yale College, New-Haven, Ct.
Alonzo Williams, Brown University, Providence, R. I.
Edwin H. Wilson, Rochester University, Rochester, N. Y. | Alexander Winckll, University of Michigan, Ann Arbor, Mich. George W. Winslow, Classical Institute, Evanston, Ill. . Stephen J. Young, Bowdoin College, Brunswick, Maine,
=~
OFFICEHRES.
1871-72.
PRESIDENT, WILLIAM W. GOODWIN.
VICE PRESIDENTS, ASAHEL C. KENDRICK, WILLIAM §. TYLER.
SECRETARY AND CURATOR, GEORGE F. COMFORT.
TREASURER, J. HAMMOND TRUMBULL.
EXECUTIVE COMMITTER, The officers above named and— ALBERT N. ARNOLD, JAMES HADLEY,
HOWARD CROSBY, GEORGE W. SAMSON, JOTHAM B. SEWALL.
U.C. BERKEL
_—
This voiuiix 2 ou with funding trow the National Endowment for the Humanities, 1990.