No CrossRef data available.
Article contents
Proto-Dravidian *c-: Toda t-
Published online by Cambridge University Press: 24 December 2009
Extract
1. That the fifteen or so Dravidian languages are to be considered a family is perhaps as obvious on first inspection as that the Romance languages are one. Phonetic correspondences, however, have with few exceptions not been worked out with any great clarity and exactness during the years since 1856, when Caldwell first published his comparative grammar of the Dravidian languages. The prime reason for this is, of course, the rarity of scholars who have worked or are working in this field; there can be named fewer than a score who have made first-rate contributions. Even so, the failure of scholars to state phonetic correspondences in definitive forms is surprising, since prima facie the material is easy to deal with and etymologically related groups of words leap to the eye. I need only instance the verb meaning ‘come’ which occurs in all the languages and which I have treated in Lg. 21. 184–213; or the noun meaning ‘worm,’ the forms of which are: Ta. puu ‘worm, maggot,’ Ma. pu
u ‘worm, caterpillar, maggot, grub, moth, mite,’ Kot. pu ‘worm,’ Tod. puf ‘worm, intestinal worm,’ Ka. pu
u, pu
a ‘worm, insect in general, snake,’ Koḍ. puḷu ‘worm,’ Tu. puri ‘worm, mite, moth, skin parasite,’ pura, puru ‘snail,’ Te. purugu, puruvu, pruvvu ‘worm, any insect or reptile,’ Kol. purre ‘worm,’ Go. pŭṛī ‘insect, worm,’ Kui pṛiu, pṛīu, piṛu ‘wingless insect, worm, maggot,’ Kur. pocgō ‘worm, caterpillar, fleshy larva, esp. of beetle,’ Malt, pocru ‘worm,’ Br. pū ‘worm, maggot, caterpillar,’ pul-mak(k)ī ‘tape-worm’ (makī ‘intestinal worm’ < Persian).
- Type
- Articles
- Information
- Bulletin of the School of Oriental and African Studies , Volume 15 , Issue 1 , February 1953 , pp. 98 - 112
- Copyright
- Copyright © School of Oriental and African Studies 1953
References
page 98 note 1 The sigilla for the languages are as follows: Dr. = Dravidian, PDr. = Proto-Dravidian, Ta. = Tamil, Ma. = Malayāḷam, Kot. = Kota, Tod. = Toda, Ka. = Old Kannāḍa, Koḍ. = Koḍagu (Coorg), Tu. = Tuḷu, Te. = Telugu (in the literary form of Sankaranarayana, P., A Telugu-English dictionary, Madras, 1927Google Scholar; B. = Brown, C. P., A Telugu-English dictionary, 2nd ed., Madras, 1903)Google Scholar, Kol. = Kolami, Go. = Gondi, Kur. = Kuru, Malt. = Malto, Br. = Brahui. Kui and Kuwi are not abbreviated. The bibliography, except for Telugu, is as in Language 21. 184, n. 1. Go. is quoted from a number of sources: Trench, without sigillum; W. = Williamson, H. D., Gondi grammar and vocabulary, London, n. d. (1890)Google Scholar; M. = A grammar of Maria Gondi as spoken by the Bison Horn or Dandami Marias of Bastar State, Jagdalpur, 1942Google Scholar; L. = A manual of Madia, Kedgaon, 1913Google Scholar (not available to me, but quoted sometimes by Burrow). Besides my fieldnotes on Kol. there is now available SR = P. Setumadhava Rao, A grammar of the Kolami language, Hyderabad (1950). Reference is made to (Sir) Ralph Lilley Turner, A comparative and etymological dictionary of the Nepali language, London, 1931, for some Indo-Aryan words. Indo-Aryan forms are quoted with the sigilla: IA = Indo-Aryan, Skt. = Sanskrit, Pkt. = Prakrit, Mar. = Marāṭhī, H. = Hindustānī.
The following notes on phonetics and phonemics will suffice for this paper. Tamil c is, according to J. R. Firth's description in A. H. Arden, A progressive grammar of common Tamil, 4th ed., Madras, 1934, which is based on modern educated speech, a ch-like affricate initially before short and long i and medially after r, and an s-like sibilant initially before other vowels than short and long i and intervocalically. Ka. c is a ch-affricate. Te. c is a ch-affricate before front vowels, a ts-affricate elsewhere. In the transcriptions of the non-literary languages Tod., Kot., Koḍ., and Tu., č is the ch-affricate; Tod. c is the ts-affricate. The accounts of Kur. and Malt, use c for a ch-affricate. Bray's account of Br. uses ch for the ch-affricate; I have simplified this writing to c. No guess is made about the phonetic nature of PDr. *c. Skt. c was an unaspirated palatal stop; most pronunciations of Skt. based on IA vernaculars use an unaspirated ch-affricate. Skt. ś is a palatal sibilant.
Not all the Toda phonemes need be described here, t, ṯ, ṭ are different phonemes, dental, apico-alveolar, and retroflex respectively; similarly for d, ḏ, ḍ. r and ṟ are two different phonemes, a voiced post-dental one-flap tremulant and a voiceless alveolar trill; these are essentially the values also of the Ta., Ma., Old Ka., and Old Te. phonemes transcribed with the same symbols. ṟ in Ta., Ma., and Old Ka. is a retracted fricative. Toda ṛ is a voiceless retroflex trill (corresponding usually to Ta. Ma. intervocalic ṭ). Toda has four phonemically different voiceless sibilants— s, s, š, ṣ, respectively post-dental, apico-alveolar, palato-alveolar of esh-type, and retroflex of esh-type; corresponds usually to PDr. l (see my paper A Dravidian etymology of the Sanskrit proper name Nala, University of California Publications in Classical Philology 12. 255–62, esp. 259, n. 27); š and ṣ are in part derived from r (cf. my paper The Dravidian verbs ‘come’ and ‘give,’ Lg. 21. 209), in part from ṟ, in part from ṟ. Toda
and ḷ are voiceless and voiced retroflex laterals respectively.
In etymologies of verbs two stems each are given for several of the languages; the second one is the past stem in Ta., Koḍ., Kol., and Kui, the stem for past and present in Kot. and Tod. Occasionally the past stem is given within parentheses for Ka. and Kur. Otherwise, as for nouns also, several forms given for one language are alternative forms given by the dictionaries.
page 99 note 1 Structure grammaticale des langues dravidiennes xi (Paris, 1946)Google Scholar: ‘II manque toujours le vocabulaire comparatif que Caldwell demandait dans la préface de sa 2e édition (1875).’
page 105 note 1 Evidence for three more Central Indian languages has been given me by Professor Burrow: Parji kic, Ollari kis, Naiki kicc (c = ts). The last language is very close to Kolami. The other two are more distant, and the combination of Kolami-Naiki, Parji, and Ollari, all with k-, speaks against borrowing from Gondi. This problem still remains and is even exacerbated.
page 105 note 2 L. V. Ramaswami Aiyar's treatment of these two groups of words for ‘fire,’ in Journal of the Department of Letters, University of Calcutta, 19. 8. 1–7, recognizes that they are two independent groups. He is primarily interested in the words with ti-. For the others he posits (p. 3, n. 1) a connexion with kāy ‘be hot’; this is hardly tenable, considering the initial c- in several of our forms.
page 106 note 1 For the IA words—Assamese, Bengali, Hindustānī, Sindhī, Gujarātī, Marāṭhī, Nepali, see Turner s.v jhāri.
page 107 note 1 Examples culled from the Tamil Lexicon: tattuvam, cattuvam < Skt. sattva-; tantati < Skt. saṃtati-; tāṉam < Skt. snāna-; tāṉai < Skt. sēnā-; ātāram < Skt. āsāra-; ātaṉan, ācaṉam < Skt. āsana-; tātaṉ < Skt. dāsa-; tāpataṉ < Skt. tāpasa-; tuccātaṉaṉ < Skt. Duḥśāsana-; taicatam < Skt. taijasa-. Examples of Ta. t- < Pkt. s- are not easily identified; they are, however, argued for by Ma. tattika ‘auspicious, holy’ < Pkt. satthiga- < Skt. svastika- (K. Godavarma, Indo-Aryan loanwords in Malayāḷam 88). The Lexicon suggests s.v.tumāṉam ‘jewel casket’ that we should compare Skt. śumbhāna-, which certainly does not have the required meaning in the Sanskrit dictionaries. L. V. Bamaswami Aiyar, Journal of Oriental Research, Madras, 3. 250 (1929), says: ‘in the colloquial dialects of the South … in adaptations of Sanskrit words with initial sibilants, t is substituted for the sibilant, e.g. Tam.: tamayam for samayam; taṅgati for saṅgati, etc., etc.’ He gives no example of Skt. ś. Godavarma, K., op. cit. 87 f., 110Google Scholar, gives numerous examples of old literary and modern colloquial Ma. t from Skt. s, but none from Skt. ś.
page 107 note 2 ‘Female buffalo’: Ta. erumai, Ma. erima, eruma, Kot. im, Tod. ïr, Ka. emme, Koḍ. emme, Tu erme Te. enumu (enu, enupa ‘of the buffalo’), Go. rmi, Maṛia ermī.
page 111 note 1 Probably related to Ta. eytu (past eyt-i-) ‘approach, reach, obtain, be suitable’; Ma. eytuka ‘get, obtain’; Ka. aydu id.,‘go to, join’; Tod. ïc,-, ïč- ‘reach’; Koḍ. ett-, ett-i- ‘arrive’; Kui e-p-, e-t- ‘arrive, reach, overtake, be sufficient’; Kuwi ejali id.
page 111 note 2 ‘Enter’ is so'ηg-, so'ηkt-, probably related to Go. sōṛītānā ‘enter, (lumbago) attacks’; Kui sōl-b-, sōt- ‘enter, penetrate.’
page 111 note 3 If it were not for the meaning, we would be tempted, because of the obvious phonological similarity, to connect this Kui verb with the following group: Ta. cē (cēpp-, cētt-) ‘dwell, lie, remain, sleep’; cēppu (past cēpp-i-) ‘abide, remain’; Ma. cēkuka, cēkkuka ‘roost’; Ka. kē (kēdu) ‘lie down, repose, copulate with’; Kol. ke-p-, ke-pt- ‘make (child) to sleep.’
page 112 note 1 Cf. Ta. Ma, Ka. Tu. kai (kay) ‘hand’: Te. cēyi id.; Ta. Ma. Ka. kai (kay) ‘be bitter’: Te. cdu ‘bitterness’; Ta, kayiṟu ‘rope,’ Ma. kayaṟu id.: Te. cēru ‘string, cord’ (but why not ṟ ?).