Vedic Reduplication of Nouns and Adjectives 9781463221744

Citation preview

Vedic Reduplication of Nouns and Adjectives

Analecta Gorgiana

336 Series Editor George Anton Kiraz

Analecta Gorgiana is a collection of long essays and short monographs which are consistently cited by modern scholars but previously difficult to find because of their original appearance in obscure publications. Carefully selected by a team of scholars based on their relevance to modern scholarship, these essays can now be fully utilized by scholars and proudly owned by libraries.

Vedic Reduplication of Nouns and Adjectives

Edward Hopkins

  2009

Gorgias Press LLC, 180 Centennial Ave., Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright © 2009 by Gorgias Press LLC Originally published in All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC. 2009

‫ܛ‬



ISBN 978-1-60724-590-2

ISSN 1935-6854

Extract from The American Journal of Philology 14 (1893)

Printed in the United States of America

AMERICAN

JOURNAL OF PHILOLOGY VoL. XIV,

I.

WHOLE

No. 53·

I.-VEDIC REDUPLICATION OF NOUNS AND ADJECTIVES. The object of this paper is to point out the noteworthy parallelism to be seen in the reduplication of nominals and in that of verbs in the Vedic language. As the verb has present, perfect, aoristic, and secondary (intensive, desiderative) reduplication, so the noun or adjective may appear sporadically reduplicated in such wise as to suggest present, perfect, aoristic, or secondary reduplication. These terms, although manifestly incongruous in connection with nominals, I shall employ for the present, in order more clearly to indicate the mutual resemblance of the several formations. The conditions of verbal reduplication are plainly given in the grammars. In the present a long radical is shortened, r is replaced by i, and in some cases a radical a may be represented by the same palatal vowel. Characteristic of aoristic reduplication is the preference for a heavy syllable (adiidu[iam), sometimes at the expense of the radical (avi'l/a{am from Vii§), with the change of radical a orr to z (dftjanat) when the radical is light, without it when the radical is heavy (adadak.~am). The present reduplication of a with i is, on the other hand, unlengthened; thus, presentjigiiti, t![i(hati; aorist ajijanat.' 1 It is, however, extremely rare to find roots having at the same time reduplicated aorists and presents. Only four of these, so far as I know, show an aoristic reduplication that is the lengthened form of the present, viz. in RV. bihhiyfzt bibhayat, pipar?i ptparat, yuyoti (Jliyot?), with one more in RV.-AV.: titir- (RV. pte.), atitaras, AV. The few roots remaining that have a mutual but more differentiated reduplication are mamatsi dmimadanta, vavartti avi1Jffat, juhuras jihvams (later jih-). These are all in RV. Add one case of equivalence: su,pati, RV, with asu~avus, AB. All other roots reduplicated in the

AMERICAN JOURNAL OF PHILOLOGY.

2

The reduplicated aorist has then affinity with the present and with the perfect (pluperfect adadakEJam), and some of its supposed forms are to be interpreted rather as perfects, if their meaning be employed as an aid. 1 In one case an approach to perfect form is seen in the employment of the -e (middle) as 3d person, att!ape, yet with aoristic reduplication. The reduplication of desideratives has this in common with the temporal reduplication just noticed, that a and r are reduplicated with i, while occasionally the reduplicating stem-vowel is made long. Desiderative forms are chiefly found in the present system in the early language. The reduplication of the intensive, on the other hand, is more closely connected with that of the perfect. In distinction from the present and aoristic forms, radical a, r are reduplicated in the perfect with a, not with i. In agreement with intensive reduplication this vowel is often long (dadhfira), though generally short. Intensives, which belong, as a primary class, to the early and die out in the later language, are reduplicated in three ways: x) simple form, a with a, i with e, u with o; 2) middle form, the whole root is reduplicated or the root in weakened form; 3) strong form, the root is reduplicated, with a connecting vowel added. Examples are: 1) vavad, cekit, fOfUC; 2) carcar, cankram, badbadh; 3) ganigam, sanif}van (before two consonants ; see Grammar, §wo2). It will be seen that in some cases it is difficult to distinguish perfects from intensives, notably in the participle, where the forms, except for accent, are sometimes identical, and in some subjunctive forms where even the accent points to the form being rather perfect than intensive. But the accent is not fixed. Thus we have raraharJa with badbadhana, niraJJas; and the doubtful forms amzmet, fzitot, dudlzot, the last of which are reckoned by Whitney perfects or aorists; by Grassmann, intensives.' So the participle {fi{ujana is undoubtedly, as Whitney classifies it, an intensive (Grassmann 'perfect'), and yet it is of the same reduplication with that in the perfect {zifU'Vtts, and in jz7juvus jz7juvana (though the participle to filfttVus is {fifuvana !). The sense of present have at any period only an hypothetical (i. e. unbe!egte) reduplication of the aorist. The present i-reduplication is optional in vha(-, vava(-, both in RV. The same reduplication of radical a appears sporadically in the perfect vivakvllibS. 1

Compare Whitney, Gramm:u, §868.

2

Ibid., §§868, roo8, ror 3-

VEDIC REDUPLICATION.

3

these intensives is also such that they are in some cases better classed with perfects (Grammar, 1008, 1024). In a word, both in form and sense it is often difficult to decide whether we are dealing with a perfect or an intensive formation. If all reduplication had originally a common signification it would be natural to expect that perfect forms with strong reduplication-like that of the simple intensive-would often retain the force of that signification sufficiently to prevent the growth of a parallel intensive form, and that, when an intensive was made from the same root, it would not be of the simple, but of the middle or strong intensive form. With this expectation I have collected the examples of the strong perfect (i. e. with heavy reduplication) with the following result, as drawn from a tolerably complete list of such forms compared with the forms of intensives developed from the same source: klp, ciik?pc, no intensive (i. e. none known in literature). kan, ciikan, ciikananta (later cakana), no intensive. grdh,ja.grdhus, no intensive. ju,jujuvus, no intensive. tu, ft7fava, intensive only in the strong form tav'tfvat. tuj, tfltujana (both accents), tutujyfit, no other form of intensive. trp. tatr/Ja'!Ja, RV., tatrpus, AV., no intensive. trJJ, tafrl!ii1Jd, RV., tat-, pte., RV., tatrsus, A V., no intensive. dz-, dzdiyus, no intensive. dhz, dldhima, R V., no intensive till TS. dedhyat. drh, dadrha7Ja, adadrhanta, no intensive. dhr, dadhfira, dadhrc, intensive dardhartJi, R V.; later dadharti, dadhrati, TS. dhrJJ, dadhrszts, A V., dadharJJa, R V., no intensive. paj,papaje, isolated form, Whitney as perfect; PW. and Grassmann, intensive. pz-, pFpyana, pipytJs, no intensive. nam, naniima (p. nd-), intensive only in the for·m ndnnamlfi. blrz-, bzbhiiya, AB., no intensive. mah, miimllhas,t miimahana, etc., no (other) intensive. mi, see note below. mrj. mamrjus, intensive only marmrjat. 1 mamahmtta, cakramanta and the like are to be defined as reduplicated preterites. Further classification is otiose, for they are not pluperfects in sense. They lie between perfects and intensives. The first is in reduplication intensive, as the perfect itself is intensive, but in form it is a simple preterite.

4

AMERICAN JOURNAL OF PHILOLOGY.

mdnz![uS, intensive only marmrrat. ralz, rtirahd1Ja, no intensive (pf. accent, but PVV. 'intensive'). rak~, rtirak~ti~ld, no intensive. radlz, rdradhus, intensive rtiranddlzi (?). ran, rtird1Jas, riiran, etc. rablz, rilrabhe, rarablzma, no intensive. (vak) vailc, viivakre, RV., isolated, intensive avtivaczt (?). van, viivantha, viivana, vavandhi, intensive only in the derivative adj. vdnzvan. z1a[, vtivarzis, no intensive. vas, vavasana, no intensive. vrj. ·viivrje, vavrjus, intensive pte. in the form vanvrjat. ZJ?'f, viivarta, intensive varvartti and vanvarti ( avanvur). vrd/z, Vtivrd/ze, Viivtdlzadlzytii, 'ZHIV?'dlzenya, nO intensive. rad, [ii[adtis, [fi[adana, no (other I intensive. {lf, [iz[lWiina, {lt[UVUS, etc., no (other) intensive. sah, sastilza, no intensive. huj,, jllzzrf,a, A V., jiktcja, R V., no intensive. htfi,JiihT~ii1Ja, intensive only jarhrfianta. Some of these forms, despite their forbidding accent, are regarded by PW. and Grassmann as intensive. I follow "Whitney's allotment. Apart from the doubtful doubt of the perfect-hood of e. g. riirahii1Ja and the puzzle involved in rttradiina beside [ii[adus, there remains a sufficiently large number of cases to show that either no intensive, or, if any, a middle or a strong, not a simple intensive, stood beside strong perfects. The grammarians, for example, made an intensive stem riirablz- beside the ptrfect rdrabhe, rarabhe, but the ancient Vedan needed only his perfect, and kept his simple intensive for cases where he did not use a strong perfect; e. g. R V. cakar{ti gets a perfect only in the epic cakare; JB. ttitrasyate has as perfect tatrcisa; RV. jtigarti has no perfect at all; papatiti has only papUa; bfibadhana, only babadlze, etc. The perfect mamrr- (see list above) is associated with the intensive mannrr-; whereas mdmar.~a corresponds with miinqfjat, withal not till the Sutra period. \Nhen dadrM is found as intensive of dr the corresponding perfect is not da- but dadfira, with which relation contrast that of dii,;thit1Ja above, a strong perfect with no corresponding intensive. The word dadhtira presents an interesting study. This is the R V. strong perfect (with dadlzre), and at this time the corresponding intensive is only ddrdlzar.~i. But, although diidlztira was uniquely preserved 11l![,

VEDIC REDUPLICATION.

5

(it never was wholly abandoned), dadhara, in accordance with the usage of other cases, came later into use (JB.), and in this period, when dadhtira obtains and dtidhara is a mere archaism, arose the intensive dtidhrati, dtidharti (TS.). 1 That the strong perfect preceded the simple can be shown by some examples: R V. dtidhara later becomes dadhara; R V. jagrdh- later becomes jagrdh-; R V. mamrr- later becomes mamrr-. As a rule, then, in the early period a simple intensive and a strong perfect never go together! The perfect is by no means confined to its own peculiar endings. It may take with the same stem the endings of the present system, where belong many forms reckoned unnecessarily to the subjunctive (compare Grammar, §Srs). Again, it may be used in a sense not difterentiated from that of a present. Conversely, the intensive, confined, for the most part, to the present system, is often indistinguishable in sense from a present. It is plain from this that, so far as the verb goes, there is not a marked difference in the character of perfect and intensive reduplication. Now, in nominal reduplication, with which, as will be seen, that of verbs hangs closely together, the character cannot be temporal. Is, then, the character of verbal reduplication of entirely different sort? This is scarcely possible. Reduplication of every sort is not temporal, but qualitative. In view of the formal categories where reduplication is shown at its strongest, i. e. in intensives and desideratives, together with the oldest form of nominal reduplication, that employed for imitative purposes or to denote repeated action 3 -onomatopoetica and such forms as l