Dravidian Studies: Selected Papers.

• Author: Zvelebil, Kamil V.

In every branch of scholarship there are several works which represent, so to say, an indispensable and permanent foundation of that particular discipline. In Dravidian studies, these works include, among others, R. Caldwell's Comparative Grammar (the interest in which is still far from being merely historical), Jules Bloch's attempt of 1946 at synthesis, some papers by L. V. Ramaswami Aiyar, etc. The works of M. B. Emeneau belong without any doubt to this category. Dravidian studies without Emeneau are unthinkable. Most of his writings - apart from several magnificent monographs - appeared in the form of articles. The most recent collection of his papers is the volume under review, dedicated to the memory of Edward Sapir, Emeneau's "teacher in modern linguistics."

The book, consisting of a brief preface by Emeneau himself, Krishnamurti's detailed critical introduction, and a bibliography of Emeneau's publications between 1979-1993, is arranged in two parts: twenty-three papers or chapters on "Dravidian linguistics" and four papers on "Toda studies." The twenty-three articles are arranged chronologically (1961-89); the studies on Toda enthnology and oral literature belong to the sixties.

Let me first say a few words on the valuable critical commentary by Bh. Krishnamurti. He offers his comments, taking the papers one-by-one - not in the chronological order but according to their subject-matter - starting with "comparative phonology."

Ad p. xv: The reconstructed PDr word for ear (DEDR 1977) is *kew-i. In Irula, this occurs as cevi, as in Tamil, which is the regular development. How to explain (unless these are loans) such Ir(ula) forms as kempu n. 'red color'; adj. 'red' (DEDR 1931 *cem-, etc.) or keruma n. 'cough' [greater than] kemma id. (1964)? Are these (and similar) forms in Ir. retentions of *k-, or instances of Nilgiri areal features, or loans from (most probably) Kannada?

Chapter 4, on "Brahui vowels," treats exhaustively the comparative phonology of Br. vowels, while chapter 16, on the development of PDr *z (DEDR r) in Brahui, "demonstrates the convergence of several PDr. phonemes into two overlapping sets of Brahui resultants...." The Umlaut treated in chapter 9 on a Kota vowel-shift evokes a similar and yet different phenomenon in Ir., cf. PSDr *oC-ay [greater than] *oC-e [greater than] *eC-e [greater than] Kota eC (e.g., *kotay [greater than] Ta. kutai : Ko. ker 'umbrella'). Ir. has a replacement of radical *a followed by Cay, by e, cf. *malay mountain : Ir. mele (cf. The Irula Language [1973], 12-13). Chapter 9 has received a detailed critical comment by Krishnamurti (pp. xvi-xvii). Emeneau's "Kodagu vowels" (chapter 10) is a very important paper showing that Kodagu developed two new vowel phonemes under the impact of preceding and following consonants and following vowels, i.e., back unrounded i and e (short and long). In Ir., the entire vowel system manifests a dichotomy, each non-low vowel having a corresponding centralized counterpart. The development may have been as follows: stage 1: emergence of centralized allophones of vowels conditioned by following retroflex, domal or alveolar consonants; stage 2: loss of conditioning factors; stage 3: emergence of new centralized vowel phonemes resulting in phonemic contrast, cf. kottu v. 'to chop into pieces' (DEDR 2091) : kottu n. 'neck', 'throat' (DEDR 7366 Ta. karuttu); ve:ru n. 'root'...