A Comparative Grammar of the Dravidian Languages (review)

This latest variation on Mikhail Andronov’s Drav-idian languages (Moscow: Nauka, 1970), A comparative grammar of the Dravidian languages, focuses on the comparative historical study of the Dravidian languages. In a case of unfortunate timing and bad luck, the book and its contents are largely overshadowed by Bh. Krishnamurti’s The Dravidian languages (Cambridge: Cambridge University Press, 2003), published the same year.

‘Phonetics’ (29–105) and ‘Morphology’ (107–302) constitute the bulk of the book, with a brief coda on Proto-Dravidian (PDr; 305–10). ‘Phonetics’ mixes phonetic and phonemic levels so that, for example, Tamil /-k-/ is represented intervocalically as [-x-] even in Old Tamil and PDr reconstructions, obscuring their allophonic relation. The discussion of subgrouping (25) posits seven subgroups rather than the four in current thinking. Syntax is avoided altogether because A apparently holds that comparative syntax is indistinguishable from typology (5). Thus we have no discussions of clause or sentence architecture, which hampers our understanding of what words may serve as predicates or how finite and nonfinite predicates are distributed in syntactic structures. A also seems wary of syntactic rules: he sets up echo words as (multiple) distinct parts of speech (108) rather than, say, the output of a postsyntactic rule, where the echo compound takes the same part of speech as the word echoed. In arguing against a particular etymology (163), A states that the coordinative compounding of pronominal bases is not attested in languages of the world, even though the South Dravidian language Kannada has such examples as avana-nanna snēhita ‘his (and) my friendship’.

Following an item-and-arrangement orientation, A concentrates on reconstructions of linguistic forms rather than linguistic rules or systems. In fact, the listing of data so occupies the bulk of the book that explicit grammatical arguments, either to support his own proposals or to engage those of others, are relatively scarce. For example, A dismisses Krishnamurti’s laryngeal hypothesis for lack of evidence, but then proposes (308) with a dearth of argumentation that Proto-Dravidian had but one finite verb (ā-‘become’). Most infelicities encountered in the text (e.g. expiration for aspiration, yotation for palatalization) do not hinder reading; sometimes, however, the train of thought turns fuzzy, as in the author’s treatment of the history of the present perfect in Pengo (224). The book’s presentation of the current state of comparative Dravidian linguistics is somewhat thin and undercut by the lack of engagement with many of the major proposals advanced during the past twenty years. When used in conjunction with other handbooks of the Dravidian languages, however, it can provide interested readers with much data on the individual languages and a lively dialogue with alternative comparative analyses of this family, as well as an able picture of Dravidian typology.