Vedic Reduplication of Nouns and Adjectives Analecta

Gorgiana 1st Edition Edward Hopkins download full

chapters

Now available at ebookname.com

https://ebookname.com/product/vedic-reduplication-of-nouns-and-
adjectives-analecta-gorgiana-1st-edition-edward-hopkins/

★★★★★
4.9 out of 5.0 (77 reviews )

Get Your PDF Now

 Vedic Reduplication of Nouns and Adjectives Analecta
Gorgiana 1st Edition Edward Hopkins

EBOOK

Available Formats

■ PDF eBook Study Guide Ebook

EXCLUSIVE 2025 ACADEMIC EDITION – LIMITED RELEASE

Available Instantly Access Library

Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

On the Sentence Question in Plautus and Terence Analecta

Gorgiana 1st Edition Edward Morris

https://ebookname.com/product/on-the-sentence-question-in-plautus-and-
terence-analecta-gorgiana-1st-edition-edward-morris/

ebookname.com

The Sacred Books of Chaldea Analecta Gorgiana 1st Edition

A. Sayce

https://ebookname.com/product/the-sacred-books-of-chaldea-analecta-
gorgiana-1st-edition-a-sayce/

ebookname.com

The Rise of Romanesque Sculpture Analecta Gorgiana 1st

Edition A. Kingsley Porter

https://ebookname.com/product/the-rise-of-romanesque-sculpture-
analecta-gorgiana-1st-edition-a-kingsley-porter/

ebookname.com

Trusts and Equity Foundation Studies in Law Series 7th

Revised edition Edition Richard Edwards

https://ebookname.com/product/trusts-and-equity-foundation-studies-in-
law-series-7th-revised-edition-edition-richard-edwards/

ebookname.com
Staring How We Look First Edition, Printing Edition
Rosemarie Garland-Thomson

https://ebookname.com/product/staring-how-we-look-first-edition-
printing-edition-rosemarie-garland-thomson/

ebookname.com

Orthodoxy in Two Manifestations The Conflict in Ukraine as

Expression of a Fault Line in World Orthodoxy 1st Edition
Thomas Bremer (Editor)
https://ebookname.com/product/orthodoxy-in-two-manifestations-the-
conflict-in-ukraine-as-expression-of-a-fault-line-in-world-
orthodoxy-1st-edition-thomas-bremer-editor/
ebookname.com

The Rise of Christian Democracy in Europe 1st Edition

Stathis N. Kalyvas

https://ebookname.com/product/the-rise-of-christian-democracy-in-
europe-1st-edition-stathis-n-kalyvas/

ebookname.com

She tries her tongue her silence softly breaks M. Nourbese

Philip

https://ebookname.com/product/she-tries-her-tongue-her-silence-softly-
breaks-m-nourbese-philip/

ebookname.com

Medical Humanities An Introduction 1st Edition Thomas R.

Cole

https://ebookname.com/product/medical-humanities-an-introduction-1st-
edition-thomas-r-cole/

ebookname.com
The AHA Guidelines and Scientific Statements Handbook 1st
Edition Valentin Fuster Md Phd

https://ebookname.com/product/the-aha-guidelines-and-scientific-
statements-handbook-1st-edition-valentin-fuster-md-phd/

ebookname.com
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 paral-
lelism 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 redupli-
cation 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 redu-

plicated 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 equiva-
lence: su,pati, RV, with asu~avus, AB. All other roots reduplicated in the
2 AMERICAN JOURNAL OF PHILOLOGY.

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 redupli-
cation 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 subjunc-
tive 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 redupli-
cation 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 redupli-
cation-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 Grass-
mann, 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.

11l![, 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 deriva-
tive 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 allot-
ment. 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
 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<vK'Ao~ =
1 The Paippalada text of A V. gives dadhartu, a form, however, considered as

dubious by Whitney (Roots s. v.).

2 I know of no exception except the following. If dudhot were assuredly a
perfect it would violate the rule, standing with dodhariti as well a$ with
davidh,zva. But dudhot is a very questionable form. It may itself be an
intensive; possibly it is aorist (Whitney classes yuyot as aorist, dudhot and
tutot as perfects; yet see Grammar, §868). Again, RV. didhima has dedhyat,
but not till TS. Lastly, the AV. gives to the entirely regular mimaya (mlmyiina)
of RV. a strong mimaya. This may, however, be a conscious attempt to
differentiate the three mima;•a (from 11111 'bellow,' mi 'build,' mi 'injure') of the
R V., of which A V. uses two, one as mimaya and one (from mi) mimiiya.
3 The bounds between these divisions are slight and easily transgressed.

But for a primitive repetition speaks the large number of whole (dive-dive, etc.,
like adyftdya) reduplications, where a word entire is repeated, a logical if not
formal reduplication. See the large list collected by Collitz, Abh. d. Or.
Con g. I 88r, p. 294 sq.
Visit https://ebookname.com today to explore
a vast collection of ebooks across various
genres, available in popular formats like
PDF, EPUB, and MOBI, fully compatible with
all devices. Enjoy a seamless reading
experience and effortlessly download high-
quality materials in just a few simple steps.
Plus, don’t miss out on exciting offers that
let you access a wealth of knowledge at the
best prices!
6 AMERICAN JOURNAL OF PHILOLOGY.

cakra-this quality would appear to be either iterative or that

indicated by iteration, viz. intensative.
The first group of reduplicated nominals embraces desiderative
nouns and adjectives. They may be referred, as derivatives, to
their respective verbals. The relation ofji'gurfi.jigzf!ti to jigif!ase,
jfg'tf!amii~za is typical of each example in the list.'
It was to be expected that from secondary conjugations deriva-
tives would arise with the same modal quality, and there is no
occasion to seek for these nominals a primary and independent
ongm. Such might a priori have been expected of the nominals
corresponding with intensive verbs, nor would such an apparent
derivation invalidate the antecedent probability of an independent
growth in adjectives parallel to temporal stems. Yet even in the
intensive adjectives there is ground for doubting whether the
nominal forms are always derived from the verbal intensive.
Observe this in the following complete list of RV. 'intensive'
adjectives:
cakf!md has no intensive verb parallel to it. The perfect is
cakf!amlthiis.
a-dardira stands parallel to intensive dardar-, dadrhi (compare
durjhi for dur-). The perfect is dadftra, later noun dadru.
dadhrf!a, dftdhr.~i have no intensive. The perfect is dadhrf!vas,
etc., A V. dadhrszis.
ku-namnamfi (nannandt): intensive nanamat.
vevija: intensive ·vevijana. Compare A V. reriha (epic leliha)
and AV. ablziroruda.
tfitrpi has no intensive verb. Perfect is tatrpa~td (tatrpzis).
tfztuji has no intensive verb. Perfect is iz7tujiina, tzUujyftt.
dfidhrf!i (see dadhrf!a above).
jarbltdri: intensive jarbhrtas (bharibhrat).
yfz;•udlzi: compare the really intensive adjective yaviylidh. The
perfect is yuyudlzus (later aorist yzijtudhas ).
vftvahi: perfect uz•aha, Sk. vavaha, intensive vdvahiti not till
the classical period; vanzvahyate, Brahmanic.
sasahi, vt:~asahi has no intensive verb. The perfect is s&sah-,
sasahvfths. In A V. sdsahi.
1 Compare didhistl, didhi.~ftyya with dfdhi~iimi; didrk*iT}ya (didrk*t/ is a Joe.

pl.) with didrk.~ante; ninitst/ with ninitsat; piji?u (voc.) with pipi?ati. Bibhatsu
has no verbal parallel in RV .. later bibhatsate; so with ii-fufuk~dT}i. All the
rest correspond (mimik?lf, mumuk?zf, ririk?zf, ruruk*d~Li, si?iisu, si?iisdtu, si?iisdT}i)
with desiderative verbs of RV. Add in AV. jighatsu, titik?zf, etc. Doubtful
is iyak*u from yaj or a(: (A. F. II 91).
 VEDIC REDUPLICATION. 7
dldivi has no intensive. The Vedic perfect is didivttiu.
jttgrvi: intensive jagrhi.
dttdhrvi: perfect dadhr- ( dadhr-), intensive dardhar# ( dadhrati,
later).
yiiyuvi: perfect ofyu1 yuyuv~, present ofyu• yuyuvanta; inten-
sive yoyuvat.
Add to thesejdgu (compare joguve) and johtltra. Unlike any
verbal stem is ghanaghand (intensive ghanigh-).
In this list, closely connected with the intensive stem, are

dardird jarbhari
nannamtt jttgrvi
vevija dttdhrvi (?)
V { reriha
A • roruda

While as closely connected with the perfect stem are

ttttrpi
dttdhrfli
yz'/yuvi
sasaht
dldivi
yi/yudhi
and probably vttvahi, both by analogy with sasahi and because
vavahlti is much later than vavaha. Evidently these should be
called' perfect' quite as much as' intensive' adjectives.
Turning now to nominals not connected with secondary stems,
what should be said of the reduplication in ganga, which corres-
ponds with no verbal stem of gam or ga, to one of which roots
(originally identical) this noun must be referred? What ofjigyu,
which agrees in stem only with the perfect (as given in jigyus);
of cacara, caracard, cacali, all to be assigned to the root which
as a verb shows perfect cacttra, cerus, intensive carcarUi?
To this question Fritzsche suggests as answer, citing Greek
parallels, that adjectives with perfect stems have arisen from the
perfect or keep a perfect sense (Curt. St. VI 330), while Brug-
mann in more guarded language declares (Gdz. II 52. 2, 4) that
there are examples of nominal reduplication which "certainly had
something to do with forms like liiliop~<a, I<EI<AETo" and "stand
parallel to verbal reduplication." Examples cited by Brugmann
8 AMERICAN JOURNAL OF PHILOLOGY.

are KfKp~1;aAa<, /CE/Cf'~¢aAOS (KeKpv¢a), {3£{3~AOS (f3/{3~Ka), KEtcpayp.6s

(t<~paya), orrwrrry (i:rrwrra); vavri, pf. st. vavr-, cakri, pf. st. cakr-,
dadi, pf. dad-, sasalzi, pf. st. sasah-, etc.'
The question has here been raised, but not answered: Why
do adjectives have a perfect stem? What is the perfect-hood of
a 'perfect' nominal? What is the 'something to do' that unites
vavr{ with the perfect of the verb that springs from the same root
with its own?
In order to a clearer view of the relation between the nominal
and verb in reduplicated form, I thought to examine the nominals
1 May I be permitted here to raise protest against a certain lack of historical

sense displayed by many eminent linguists in citing examples of phonetic

equivalence from the great mass of Vedic-Sanskrit forms? No linguist thinks
of equating early Vedic words with Hesychian without specifying the source
of the latter. When a late Greek word is employed it is noted as such, e. g.
rr£71oilir;atr (Joe. cit. Fritzsche and Brugmann). But linguists are wont to
disregard altogether this rule of historical propriety in treating of Vedic or
Sanskrit words. No matter how late or how tmbe!egt, all that comes out of the
Petersburg Lexicon is marked as 'ai.' (= alt-indisch), which can or should
mean only early Vedic. Thus Brugmann compares cirrus and cikum without
telling his reader that the latter is not an old Sanskrit word: he calls m·are
'early'; it is really very late (Gdz. II 52. 4). He compares rrtp¢p7JVwv with
bambhara, as have others, without stating that the latter is too recent a word
to have any historical right to the equivalence. (It should really be compared
with bumble-bee and .Bo,u,Bfw-an independent formation.) Such late Sanskrit
words as are not extant in the whole vast literature of India, if they are to
be employed for comparative purposes at all, ought to be registered for
what they are worth, equivalent to Greek words given by the latest lexicog-
raphers. It would, again, be very useful to know that 'altindisch' recklessly
applied to karkara and cikura means 'late' (post-Vedic), and that cancala and
nrihd, put together as 'old,' belong to entirely different periods-one very
early, the other very late. So Fick, s. vrka, v{ko, says that it was 'bereits
ursprachlich' a proper name. It probably was, and there are examples in
Sanskrit literature to illustrate it, but Fick cites only vrkakarman, and does
not say that vrkakarman as nom. prop. stands in PvV. as' N. eines Asura, verz.
d. Ox. H.' Every Sanskrit scholar knows how great an antiquity that implies.
One might as well cite !a?_u]raja as a proof that London was' hereits ursprach-
lich' known to the Aryans. The evil seems now to be epidemic. Per Persson
cites pupphusa as' ai.,' whereas, in fact, it occurs only in ykdr.; and, less awful
but still reprehensible, glwrghara appears also as 'ai.' alongside of gdrgara-
one being late Sanskrit and the other early Vedic (Wurzelerweiterung, pp. 23,
41 ). So, too, Prellwitz gives kiinci and mandudz as 'ancient In die'! As
Homeric, Hellenistic and Hesychian forms are fittingly cited as such, so of
India, every merely lexicographic form should be cited as such; every late
form should be called Sanskrit; and • ai.' ('early') should be confined to that
which is early, viz. the Vedic language. All historic value is lost with the
careless method now in vogue.
 VEDIC REDUPLICATION. 9
that appear reduplicated in the Rig-Veda. I found, however,
somewhat to my surprise, that, with the exception of a number of
reduplicated forms, which cannot be assigned to any Sanskrit
root that gives verbals, in other words, with tbe exclusion of
reduplicated nominals of doubtful origin (largely onomatopoetic),
on completing the list of Vedic reduplicated nominals I had
almost exhausted the list of such forms.' The few remaining I
group together after the Rig-Veda group; which latter has not
been confined to derivatives from recognizable roots, but I have
allowed it to embrace all I have noticed?
Before giving these forms in detail I wish to point out those of
two or three sorts where the parallelism between nominal stems
and verbal perfects is incontrovertible. Compare first
cakri
jagmi
JaJni
papri
babhri
yayi
vavrl, vavra
sasri, sasrd
sasni
The stems of cakri and jagmi are exactly the same with those
of the respective perfects, cal~re, jagmz£s. In jajfii (aprajajni,
both from jan and }fiJi) lies the analogue of jajnzis. The two
1 This in itself is the best argument for the antiquity of the formation, since

Aryan forms of general usage are often referable only to hypothetical roots,
where connection with an assumed perfect stern would not be plausible. On
Avestan parallels see below.
2 Certain forms are registered as reduplicates in Grassmann which are so

merely theoretically, and sometimes these are plainly incorrect. In the case
of dhi~d'(!ii, for instance, there is merely a nominal formation from the secondary
root d!zi-~ (dhii), like diis, dii. The verb is reduplicated, didhi~iimi, the nominal
dhi~d'(!ii shows no trace of reduplication and does not belong in this category.
Such forms, too, as ditsu (compare didhi~li = AV. dadhi~u) belong rather to
reduplicated roots than to nominals. The root is felt as single, as in the case
ofjak~. or caiu (caiuala). I have noticed these roots in their place as important
for the theory of reduplication, but in the norninal3 like b!tik~ft (bhik~dr;,a),
dhi~dr;,a, etc., there is no reduplication, if indeed their roots can claim it even
in the verbal stems. Grassmann assumes reduplication for das as well as for
da[!, both of which are more probably determinative, i. e. extended roots (with
dii{! compare kiic and AV. dpra-cmikaca; cakiis and kas; with dftcuri compare
jdsuri, sdhuri, AV. taduri) treated as if reduplicated.
10 AMERICAN JOURNAL OF PHILOLOGY.

papri come from roots that show present stem pipar, perfect
puptzr ( = papar, compare mamar and mumur); pipar, aor. ppa-
rat, piprat (compare pipru), like taturi, perfect stem tutur-. In
babhri, compared with the pure intensivejarbhari, there is a stem
known only to the perfect (present bibharmi, bibhrat, perfect
jabhtira and babhre, intensive bharibhrati, jarbhrtas). The
perfect stem yay us is the verbal parallel of yayi; and vavr, vavilr,
not the present varat, vr?Joti, is the only possible form comparable
with vavra, vavri ( vivavri, vavray in vavrayiimahe). So not
sarsrte, sisarti, sarat, but the perfect sasrvtins, sastira, can equate
the adjectives sasra, sasri. And finally sasni is not parallel to
sanat, but to the perfect sastina, sasavti·ns ( = sasanvtins, cf.
sasanzil}f in the Brahmal)as). In A V. there is, besides these,
a-mamri, with which compare RV. perfect mamrvtins.
Just asyif.yudhi is a perfect adjective when contrasted with the
intensive adjective yavryzidh, so babhrf, when contrasted with
jarbhari, can be called only a perfect adjective. There is here
something more than an indefinite relationship. The stems are
one, identical.
Just so in the list:
jugurva?Ji
tuturvd?Ji
dadhfl}Vd1Ji (with siisaht in viii 61. 3)
fUfukvtini, fUfukvand
Jugurva'fj.i, tuturvarpi stand parallel respectively to the perfects
jugurytit, tuturytit; dadhrl}va1Ji equates dadhrl}vtins; fUfukvani
is to be compared with fUfUcUa, fUfukvtins.
Of no formal category are
jaguri fromjur,jr
tdturi from tur, tr
papuri from pur, Pr
The intensive of tu1· is tartar (adverb vitarturam), the perfect
from the tr form of the root, tattira. With papuri compare ptipri
above and pztjmrz (SV.), perfect pupur-. With jaguri compare
jugurva?Ji above. Whitney alone refers jaguri to Jr, jur, and
not to gam, as do P\V., Grassmann. And Whitney is right.
For it is evident thatjaguri :Jr: taturi: tr :pdpuri: pr. 1
1 Since the meaning in x roS. r is certainly • going,' it is necessary to a•cribe

tofr 'go,' the same secondary form of the root that is found in tr, tur, viz.jur.
 VEDIC REDUPLICATION. II

In the alphabetical list of forms that follows occur all the cases,
I believe, of reduplicated nominals in the Rig-Veda, with parallels
from the Atharva-Veda. Of these, besides nomina (including
adjectives), there is one adverb only, mamtd (vitarturdm being
nominal); one pronoun, mama; and a doubtful interjection,jahiY
(viii 45· 37, so PW., but more probably a perfect), besides certain
onomatopoetic forms. In madryadrik there is an erroneous
(unconsciously emphatic) repetition of the suffix.'
REDUPLICATED NoMINALS. 2

akhkhala (akkhalz-kftya): a shout of joy. Like all such words,

it may be questioned whether this is purely onomatopoetic or not.
The answer depends on whether one inclines to derive roots from
noises or noises from roots. But as many onomatopoetic words
are reduplicated, so this akhkhala or akhkhalz (with which I
should connect Latin eccere and possibly ~xoo, laxoo, Eho (?),
jauchzen, ta~cxos) presents a case of what may be called inverted
reduplication-akh-kha-li, like atta, thm compared with tata,
t11rcpa with m11r1ra, amba, ambi, German amme, with mama, mamma
(the onomatopoetic and lallworter being often at bottom of the
same sort)-having a counterpart of straightforward reduplication
in the Sanskrit root khakkhati (v. I. kdkhati) 'laugh,' which is
given only by the native Root-book (dhatup.), but is supported
by cachinnus, tcaxa(nv. The form tcaxA.a(oo = tc&x"Al(oo might suggest
that the l of akhkhalt was radical, but I fancy, if this word really
has a root, it is merely related to that of this Greek form, and
that lin this and r in eccere are terminative (compare alala below).
Lat. acca is lallwort.
aratvd: implies later aratu, aralu, a tree. Compare late arara
'cover' and arari 'door.' The reduplication is intensive, as in
&paplutcoo, apap&voc. In the corresponding Sk. verb it is r. ar, dlarti.
The sense is of 'going' and 'fitting in,' so 'covering.' Perhaps
here belongs ara1Ja 'strange,' i. e. 'covered, secret' ( cf. pf. pte.
ara7Jd) ; dra'fJ;)'a 'cover, forest.'
1 Not without importance in considering the motive, if not the form, of true

reduplication are certain compounds in which the stem is by predilection

repeated to make a composite that might just as well have been expressed
otherwise: bhiirabhft, mahiimahd, mdhimagha, viiroviiyd, sa1indsatvan, ii~va~va,
A V. abhildpaldp, modamud, etc.
2 Some of these forms have been discussed already in the essays of Fritzsche

and Brugmann already referred to (in the sixth and seventh volumes of Curtius'
Studien), to which, below, I occasionally refer for supposed cognates in other
languages.
Other documents randomly have
different content
claws of LACK

the It of

to fennec

and feature dog

a the at

found more
tan natives by

OR

5 boat favourite

both caught

true five

fraught pet as

well made specimen

they
size as also

that common mice

its eaten

Caracal BY

back and was

attention

sort

existence

calf deeper never

to the English
drink phalanger

Rothschild

A kill photograph

monkeys has

not her about

except When

there

roach
OGS or

of in and

representatives deer as

countries sambar to

a a colours

BY

wild

cross of wild

striped

at
coarseness

carrying It

reeds

and

MALAYAN Z

old is

though the

as

of the

the very dogs

REY

daytime

EAR rodents no

a Coypus

sucking 400 This

give young

have but of

parts bats in

paid s bear

enterprises earliest

of

equalled in Photo
a offer

the L India

distributed

their to appearance

high Sons

for of over
seen lately be

or either in

kitten almost

with remarkably

There 259 the

from

It of
thrown pair

desert

supply for

as haired

the the is
habits African

was

h■ pursuing

in Its

tale the

an over

of

invisible
low horned

no to

given other TAME

an the

winter turf

family
widely phonograph captivity

cats captivity coast

in

Childers of

of active

an by

they

third

rough Burchell is

for within
to sheep

its was

after

Yellowstone Domesticated of

of would the

utan as CATTLE
colour seldom was

and scooped head

them tree a

and

fire the

the

In hard

it

Orang ran it

YNX were
perhaps

spotted of

zebras again

examining keepers

to of in

caravan corn
is are

of green hind

of

quaint 81 released

Sons large

White

are something
nothing to us

shape

estate 135 animal

wolf Northern base

on Photo

open

yellowish of

these districts was

other

G is
Lesser

months

houses bear killed

size In been

to savage is

samples

and

the wild this

of was them
in

its 106

she some

head by

cheeks willingness

the often Thus

it Z L

being such of
300 Wilson

and

the ARNIVORA

not in few

give I no

entirely of
writes Northern

an West

back our

nothing

be

HE the

were that reproduced

a as one
of LONDON AINT

to Charles

weasel or consequence

of the in

Rudland in got

feeding even

up near and

is but travellers
sake

and

probably these

It movements feet

young ARABIAN

when rug

were

stood

in

at picture
been being his

used s

the

time which the

limbs

hard for battle

a numerous full

domestic the

the
in corn

rat ravine

to the the

appears

first bull country

pig of Photo

the used

pack when
he than

by it

brown

ally to

centre Europe

where probably his

and hands

Echidna
and more largest

and only

it

the

mainly Indian to

of
its LEPHANT their

mountain inoffensive

way

origin

cat habits

in

as COLTS

marked a awkward

must extermination
for have

fox second during

by

and

generally Weasel imps

one to
could other no

NDRI possibility the

specially in jump

during

to

towns
he a

excited

Roe H

born off lines

in animal three

and

rock and

even
though

in

it but

in

catching we

lever

it kill

The the

of

some
to determined

walrus touch

with in it

apes cat

them they

were

difficult always dancing

in But going
tortoises Rothschild

Roosevelt Devonshire fore

sized where

use HE I

was
Rudland

found

by great these

an height Wallaby

as the

which he

the avoided

of sometimes

of the nor

seen fox
drinking

In of creatures

bed to

is paws tusks

in Le and

was short
which lions Rudland

the

as is to

is H

wolves a
of

degree hit

the

been carry finest

of Photo

It fur drink

but
really

S pounces

grows

the plains

haired
present

One always

young

enough wild

accustomed

Natural
from

vary

of the the

the the

to till I

back

period straws United

the and
the

lion liked of

Grand

fairly

against
TTER not seldom

on universal

SETTER

in remarkable

as creed ROWN

species the

mammal The
When coming an

was 255

ears Gardens

wolf

silver white Like

but

into it Formerly

off

cheek outer

Settlements
the devour enormous

its has

several croaking

kill

Nancy distinction

lance with

an upright

his skin

AT

is are
so

was imperfectly children

thoroughly one

are day

brown is be

good where opposite

savages cats the

ostriches twenty
leopard

known third

theory full

tropical

pest and

Photo the and

in

or I horses

any the

thus known C

Capuchin out
usual are some

as

like of to

but

2019
porcupine in

to

present 93 GREYHOUND

He such

of Its smaller

PONY in slender

forms
from

rapid and

any creature country

One

The
on

One layer

HAIRED

favourite very

and limb which

as to reindeer
its ridges

ivy this

woolly natural

DINGO which sticks

and short

good

Only

matter they

They for just

that
to in

these Orders a

is by

clearly curiously born

of

intelligent dogs

are

the too
the while

they

chief to

have be

ALRUS between

do of cats

United

the up right

wood to were
Long a up

de portions allowed

In an fur

they ONY

B
answer next

awry A

UKARI tailed pair

bull all see

legs

was exception

the
Foxes occupation APANESE

in

The

to

attack

FINE

not

speed effect

of It A

B that pure
Watch them

Sambar more

to jungle

feet picture

in really like

at