Dravidian I and We and You

8/8/2019 Dravidian I and We and You

1/10

Dravidian 'I', 'We' and 'You'Author(s): Edwin H. TuttleSource: The American Journal of Philology, Vol. 49, No. 4 (1928), pp. 334-342Published by: The Johns Hopkins University PressStable URL: http://www.jstor.org/stable/290044

Accessed: 06/08/2010 06:32

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at

http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at

http://www.jstor.org/action/showPublisher?publisherCode=jhup.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and extend access to The

American Journal of Philology.

http://www.jstor.org
http://www.jstor.org/stable/290044?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/action/showPublisher?publisherCode=jhuphttp://www.jstor.org/action/showPublisher?publisherCode=jhuphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/290044?origin=JSTOR-pdf


2/10

DRAVIDIAN ' I , ' WE' AND 'YOU '.The second number of Dravidic Studies, issued by the Uni-

versity of Madras (1919), contains a discussion of Dravidian'I' and 'we' by K. V. Subbaiya, with remarks by M. Collins.Subbaiya's work shows an amazing disregard of scientificmethods. He gives word-lists meant to prove that ancient eis "preserved in all the languages" (p. 14), and that ancient ais "preserved in all the Dravidian languages" (p. 15). Fromthese lists it appears that "all the Dravidian languages" arethe southern tongues; the northern ones are cast into the outerdarkness of "minor dialects" (p. 8). The first word in thea-list is Tamil a (cow), which corresponds to Malto &i: ven-geance is mine, saith the North. If these words came fromAryan avi (sheep), Tamil has a < *au in accord with the noun-ending a, a variant of -avu. In the e-list we find Tamil per(name) with its cognates, Kanara . . .", Tulu ". . ."Telugu peru. In order to get a bit of padding for the e-list,Subbaiya ignores Tamil pejar (the older form of per!), Kanarahesaru and pesar, Tulu pudar, Kodagu peda. The Kanara,Tamil and Tulu forms are recorded under pesar in Kittel'sDictionary, one of Subbaiya's sources; Caldwell discusses themtwice in his Comparative Grammar, also one of Subbaiya'ssources. It may be a question whether Telugu peru represents*prer < *pedaror a Tamil-like developmentof *pitar to *plecar;but the basic short vowel is self-evident.'

From Brahui k7an-,nominative i, Gondi nd-, nom. and, Kuinia-, nom. dnu, Kurukh-Malto er-, nom. en, Kanara en-, nom.an(u), Tamil en-, nom. jdn, Telugu na-, nom. enu, Tulu jen-,nom. jdn (I), we may infer a basis *en having short e, withlengthening in the nominative because of emfasis. Subbaiyaassumes that the root-vowel was a more open sound, like thevowel of English hat. This theory is founded on the correspond-ence of Telugu e, in a small number of words, to Kanara d andTamil d or jd. Subbaiya admits, at the end of his work, that a

1I use j for consonant-i; a grave accent marks variable or uncertainlength; e is a very open e-sound; e = e in bakery. Tamil is transcribedin accordance with native Ispelling.

334


3/10

DRAVIDIAN 'I ' 'WE ', AND 'YOU '.

lengthened basic e produced a and ja in Kanara and Tamilinterrogatives, but fails to see that the admission spoils thetheory of a special vowel in 'I' and a few other words. Thetrue explanation of Telugu e, where it corresponds to Tamil aor ja, presumably lies in the lengthening of a short e. In manyEuropean tongues we find a development of ea or ia or ie frome, with or without lengthening. In early Dravidian, where shorte was apparently similar to the e of English let, the lengthenedvowel was sometimes confused with the ordinary close e, andsometimes underwent breaking or became a more open sound,and thus produced ja or a.In Kurukh-Malto the inflectional stem of en has l for nbefore a g-suffix. In Tamil the stem en- may become enn-before vowels, apparently by conformation to noun-stems likecann- (eye) beside the nominative kan. The same variation isfound in Kanara en(n)-, which has been changed to nan(n)-under the influence of the nominative ndnu, a form explainedbelow. In Brahui, which regularly has a for ancient short eand e for ancient weak-strest ai, the dative-accusative kane isapparently equivalent to a blending of the Tamil accusativeennai and the Tamil dative enakcku: the k-formant of the dativeis prefixt, just as the ordinarily prefixt cum appears as a suffixin Latin mecum. The Brahui genitive k7ana for *and < *end)has taken k from the objective kane. In Gondi, Kui and Telugu,where stress-displacement is common, the genitive *ena hasbecome nd (with a second suffix in the Gondi genitive navor),and na- is used as the inflectional stem. The genitive *ena,represented in five of the six main divisions of Dravidian, seemsto be extremely ancient. In Tamil the nn of ennai probablyshows that the accusative-ending was a much later addition.Brahui z < *en agrees with Brahui xasun (red) correspondingto Kurukh xeso, Tamil ce, cem, cej, ce (< *kenso?). Gondiand < *enen has the variants anna and nannd : anna is parallelwith Gondi talla, a variant of tala (head); initial n could havecome from the general stem or from a lost plural *nam. Theform *enen may be compared with Latin mem, tete, sese, andSwaheli mimi (I), wewe (you), jeje (he). Kui has iaru-Kanara ivaru < ivar < *ihar (these); prdu < *pradS< *pras= Malay bras (rice); tlau== Tamil talai < *talas (head); andit has likewise added a final vowel in anu < *en. It is note-

335


4/10


5/10

DRAVIDIAN 'I', 'WE', AND 'YOU '.

lacking in Kanara, Gondi, and Brahui; that it was probablyunknown in early Dravidian, and that it was perhaps borrowedfrom Kolarian. Subbaiya repeats Konow's theory with approval.From Trench's Grammar (1919) we learn that Gondi dis-tinguishes inclusive 'we' and exclusive ' we' in pronouns andin verb-endings. Kanara has pronouns that represent in form,tho not in sense, the distinction as made elsewhere. Brahuinan (we) corresponds in form to the Kurukh-Malto inclusiveplural nam. We may infer that the distinction was a basicfeature of Dravidian. Its meaning has been lost in Kanara andBrahui, just as the meaning of the neuter-ending -d is lost inBrahui, altho the form is kept.Kurukh and Malto have exclusive em-, nom. em, and inclusivenam-, nom. ndm. The initial n corresponds to the generalDravidian stem of 'you' (singular): un-, in-, n- or nin-.Apparently the compounding began with the general stems, notwith the nominatives: weak stress reduced *(n)inem or *unel,to nam-, from which the form ndm was constructed on the modelof em beside em-. The stem of ndm seems to lack parallels inGondi-Kui; but it is representedin the four other main divisionsof Dravidian, and may therefore be a very early development.2Kanara has em(m)-, nom. am (later avu), and nam(m)-,nom. navu. The difference of meaning is lost, but the formsare parallel with those of Kurukh-Malto, modern -vu beingregularly used for older -m, as in maravu < maram= Tamilmaram (tree). An old form *ndm, correspondingto later ndvu.does not seem to be recorded in Kanara: its absence might heheld to show that the meaning was unsettled in the literaryperiod, and literary use of the form was therefore avoided untila later time when the special meaning was forgotten.Tamil has em(m)-, nom. jam, inclusive nam(m)-, nom. nam,and exclusive erqkal-,nom. nda7kal. Obsolete jam is commonlydescribed as a mere variant of ndm, but Subbaiya tells us thatearly writers distinguish exclusive ja5mand inclusive nim. Thiswelcome bit of news shows that Konow's view of the historic

21 am not aware that the foregoing theory of nam has been printedbefore. Collins rightly rejects Subbaiya's theory that the n of Tamilnvm (as well as that of nan) may be an internal development, andassumes that narm is ancient, but does not try to explain it.

337


6/10

AMERICAN JOURNAL OF PHILOLOGY.development is sadly awry (L. S. I., Vol. 4, p. 293). Subbaiyafails to see the bearing of his discovery. The Kanara-Tamilformations are fundamentally the same as those of Kurukh-Malto, Tamil jam being derived from *em in accord with thesingular jdrn< *en. After jam was confused with nam, a newexclusive form was made with the ending -al : *ja,rkcal,whichchanged to ndrkal when jam went out of use. The nominativewas so unlike the stem eqkcal-hat the change of form did notaffect the stem.

Telugu has exclusive m,a-,nom. (n)emu (later memu), andinclusive mana-, nom. manamu. The final vowel of the nomina-tive is a mechanical addition parallel with the treatment of oldfinal consonants in Kui and modern Kanara. We may assumean early inclusive *ndm(u) with the genitive *nama. Stress-displacements changed *nama thru *mand to mana. The dis-cordant *nam(u) could not be kept beside the general stemmana-, so its ending was added to the stem to form a new nomi-native. The influence of *nrdm(u) changed em(u) to nem(u );that of ma- produced memu. The change of the stem *emd- toma- was parallel with na- < *ena-.Tulu has exclusive je?rkculu,nd inclusive nam-, nom. nama.From Tulu mara = Tamil maram (tree), it may be supposedthat *em developt to *e or *ja, and *ndm to *nd. These nomi-natives were so different from the stems, *em- and nam-, thatthey were given up. Exclusive *em- formed a nominative *em-kulu (= Tamil *emkal-), and then the suffix was extended tothe entire stem. A blending of *nd and nam- produced *namd,which later became nama.

Brahui has nan as the nominative and general stem. Proba-bly *em became a simple (nasalized?) long vowel and was lostbecause of conflict with the same word-form derived from *en.The inclusive plural *nd (< *ndm) replaced the derivative of*em; the stem *nam- was changed to nan- under the influenceof the singular kan- or older *an-, and then the stem-formreplaced *nd because of conflict with nma Tamil una (your).Gondi has exclusive mar-, om. ammat, inclusive ap1o(-Hindiaplog), and the equivalent verb-endings -am (exclusive), -at(inclusive), according to Trench. The Linguistic Survey givesother nominatives: am(m)ot, (m)ammdt, mamet. The verb-endings seem to imply as lost forms *ama < *Omemend *at=

338


7/10

DRAVIDIAN 'I', 'WE', AND ' YOU'.Kui adzu (adzu?), which combined to produce ammat and itsvariants after foren aplo replaced the native word. The formswith initial m show the influence of the stem md- (< *emd-).Kui has exclusive ma-, nom. dmu, in accord with the singular,and inclusive amma-, nom. ddzu. The stem amm?d-was appar-ently reduced from *namma- under the influence of both nomi-natives. Kui dz, which should perhaps rather be transcribed asreverted dz, seems to come mainly from dr, dr or dl. In Kanaraa checking nasal is commonly lost except after a main-strestshort vowel (A. J. Ph., Vol. 48, p. 268), and parallel develop-ments are found in Telugu, Tulu and Gondi. If we assume thtatadiu has lost a nasal, its source might be *dndru or *andriurepresenting *dniru, a compound of anu (I) and iru (you).Kui ddzTumay thus be connected with the ending of Tuluamma.dlu (fathers), appead.lu (mothers). As an independentdevelopment, we should expect Tulu *jddzi < *endri < *enir,but -ddlu could represent *endri conformed to the plural-ending-lu (a common variant of -kulu). Tulu has dJn~ as a variantof dzanv (what) from the e-basis with prefixt d, showing thatlengthened e might make medial d beside initial jt.

From the various forms of 'you' (singular), Brahui n-, nom.ni, Gondi ni-, nom. immad,Kui ni-, nom. mnu,Kurukh-Maltonirq-,nom. nin, Kanara nin-, nom. n(nu), Tamil (n)un-, nin-,nom. ni, Telugu ni-, nom. (n)ivu, Tulu ni(n)-, nom. i, togetherwith the Brahui verb-ending -s, we may infer basic *is and thenominative *is. As Brahui has bat Tamil vej (mouth), froma Dravidian basis *wds (derived from Aryan *os?), and bdmus(nose) beside Malto muso (nose), it appears that an ancientfinal s became i or zero in Brahui as elsewhere, and that anancient intervocalic s was kept in Brahui as in Kurukh-Malto.The verb-ending -s seems to represent *isi(s) or *zsz(s), whichreplaced the simple form when final s began to disappear.

The nominative *is became i in Tulu, and likewise in thenorthernmost and southernmost tongues, Brahui, Kurukh-Malto,Kanara-Tamil. Outside of Tulu, *i became ni, with n from thegeneral stem (explained below). In Kurukh-Malto and Kanarathe ending of en or its derivative was added to ni. In Gondi-Kui and Telugu apparently *is became *isu with the usualaddition of -u after a consonant. Telugu has v in (n)Zvu < *isu

339


8/10

AMERICAN JOURNAL OF PHILOLOGY.as in van- < *awan < *asan (him), viru < *iwar < *ihar(these); initial n came from the stem ni-. Corresponding toTamil kcanavu dream), Telugu has kala with -a < *-a < *-au,the I being due to dissimilation in the verb kalagan- (dream):in Telugu tala < *talas (head) we may assume either -a < *-< *-au < *-asu < *-as or a loss of -s after weak vowels before*is became *isu. In Kui, where *talas changed thru *taldsu totldu, inu replaced *u < *isu under the influence of anu (I).Gondi imma is a plural used for the singular, with -a fromana (I). The loss of final short vowels in Gondi makes it hardto find evidence for or against an early change of *is to *isu inGondi. It is true that Gondi has talad- ui tldu (head), andkaud-, nom. kcav (ear) correspondingto Tamil kct-, nom. katu.But au may have been kept if au became a in tala, just as aiand ai show differing treatments in Greek; or the stem of 'ear'may have had av when au changed to a. From certain pointsof likeness in Tulu and G6ndi-Kui, we might suppose that Tului came from *is thru *isu and *uL; but such a development,beside the ending - < *-ai < *-as in Tulu tarQ (head), wouldimply two different treatments of -s, depending on a differencein the stress of preceding vowels. There is ground for thinkingthat Kui has -a < *-as with weak stress beside -au < *-asu< *-as with strong stress.In Brahui and Kurukh-Malto, where ancient intervocalic sis kept, the normal change of *is to *z and the loss of s beforeconsonants caused the stem *is- to become *i- before vowels byanalogy. In the other Dravidian tongues there was a regularloss of s between vowels. We thus have to deal with a generalstem *i- instead of *is-. The genitive *ia was an unstable form:it tended to become *z or *e, but the need of keeping the endinghindered such a contraction. Under the influence of *ena (my),*a was changed to *ina. Open a is the most sonorous of thevowels; i is one of the least sonorous. In early western Romanica weak a was regularly kept, while other weak vowels were oftenlost. Weak stress caused Dravidian *ina to become na, the formkept in Brahui; and *n- became a general equivalent of *in-.A blending of the two stems produced nin-, which is representedor kept in Kurukh-Malto, Kanara-Tamil and Tulu. Tamil un-is an expansion of *n-, with u taken from the end of precedingwords; nun- is a blend of *n- and un-. In the central tongues,

340


9/10

DRAVIDIAN 'I', 'WE ', AND 'YOU'.Gondi-Kui and Telugu, assimilation changed *ina to *ini, whichbecame nmand produced the stem ni-. Tulu has the genitivenina, and the dative nilkke (= Tamil unackku) the a of *nakkbecame i under the influence of z or nin-.

Two basic forms of plural 'you' are widely represented inDravidian, *isir and *im. The equivalent verb-endings generallycontain r, not m: probably *im was a later formation, modeledon *em, the plural of *en. In Brahui, Kurukh, Malto, Kanara,Tamil, the n- of the singular has been added to the plural.Apparently *isir became *zr in Brahui and Kurukh-Malto underthe influence of normal *i < *is; elsewhere the medial s waslost as a regular sound-change.Brahui num, used as nominative and stem, stands for older*um Tamil um-; the change of *im to *um may have beenparallel with Tamil un- < *in-, explained above, or perhaps iwas changed directly to u by contact with a labial, as in Tulupudar < *pitar (name). The nominative *z < *zm conflictedwith the singular *i < *s, and was therefore replaced by thestem-form *im.

Gondihas the verb-ending-an beside the pronounana < *enen(I). We may assume end-stress in and, against -an< *-ana< *enen with stronger stress on the first vowel than on thesecond. The plural of -an is -am < *dma < *-emem. Gondi*im became *i; and the singular *is also made *i, either directlyor thru *isu and *iu as explained above, with a contraction of*u to *! in accord with Gondi er=-Kui iaru < *ihar (these).The leveling caused *im to be used for the nominative, andallowed it to become the singular, which was then conformed toand or anna : imma. The plural immadthas adopted the endingof ammat (we). The general stem is m-- < *imi- < *im-,parallel with that of the singular. Kui has a corresponding mi-,with the nominative iru < *isir.

Kurukh and Malto have nim-, nom. nim, with n- from nin.Kanara likewise has nim-, nom. nim (later nivu). Tamil has(n)um- < *im- as just explained, also urvkal-and a nominativeni7qkal for *nim < *im) modeled on erlkal-,nom. ndarkal we);and a nominative nir, which produced analogic i for i in thevariants nijir and nivir. The ending of these forms resemblesiru (two), as Caldwell remarks. Telugu has mi-, nom. iru in3

341


10/10

342 AMERICAN JOURNAL OF PHILOLOGY.accord with Kui; also mrruwith m from the stem, and (m)zraluwith a noun-ending -lu added. Tulu has ir as nominative andstem, and nikculubased on the stem of the singular. In TuluZ< *is and Zr< *isir the basic forms seem to show less changethan in any other Dravidian tongue. EDWIN H. TUTTLE.

NORTH HAVEN, CONN.

Dravidian I and We and You

Documents

Transcript of Dravidian I and We and You