ABSTRACT
The Dravidian language family, the world’s fourth largest, consists of twenty-five languages spread over the South Asian subcontinent. It has four branches: South Dravidian with Tamil, Malaya-l.am, Irul.a, Kod. agu, Kota, Toda, Badaga, Kannad. a and Tulu; South-Central Dravidian with Telugu, Savara, Gon.d. i, Kon.d. a, Pengo, Mand. a, Ku-i and Ku-vi; Central Dravidian with Kolami, Naiki, Parji, Ollari and Gadaba; and North Dravidian with Ku˜r.ux, Malto and Brahui. Reports of other languages have appeared, but without adequate grammars we cannot determine whether these are new independent languages or simply dialects of ones already known. Indu and A
- we-have been
reported in South-Central Dravidian; Kuruba, Yerava, Yerukula, Kaikud. i, Korava, Koraga, Bellari and Burgundi in South Dravidian. Certain dialects of Gon.d. i and Ku-r.ux may prove under closer inspection to be independent languages. The Dravidian languages are spoken by approximately 220,000,000 people. Though concentrated in South India (see Map 37.1), the Dravidian languages are also
found in Maharashtra, Madhya Pradesh, Orissa, West Bengal and Bihar; and, outside India, in Sri Lanka, Pakistan, Nepal and the Maldives. The Dravidian languages share the South Asian subcontinent with three other language families: the Indo-Aryan branch of Indo-European, the Munda branch of Austro-Asiatic and Sino-Tibetan. Commerce and colonisation have carried some Dravidian languages, particularly Tamil, beyond South Asia to Burma, Indonesia, Malaysia, Fiji, Madagascar, Mauritius, Guyana, Martinique and Trinidad. The Eighth Schedule of the Indian Constitution (1951) mandates the creation of
states along linguistic lines, and accords official status to four Dravidian languages: Tamil in Tamil Nadu, Malaya-l.am in Kerala, Kannad. a in Karnataka and Telugu in Andhra Pradesh. These four have long histories, recorded in epigraphy and native literatures: Tamil dates from the second century BC; Kannad. a from the fourth century AD; Telugu from the seventh century AD; and Malaya-l.am from the tenth century AD. Starting with Caldwell’s (1875) Comparative Grammar of the Dravidian Languages,
linguists have reconstructed a fragment of Proto-Dravidian. This fragment incorporates
those features the Dravidian languages have in common and may be said to typify what is ‘Dravidian’ in a language. Proto-Dravidian has ten vowels, five short and five long: a, a-, i, ı-, u, u-, e, e-, o, o-. It has sixteen consonants, including an unusual system of stops contrasting in six points of articulation: labial, dental, alveolar, retroflex, palatal and velar, viz. p, t, R, t., c, k. Four nasals, m, n, n. , ñ; four resonants, l, l., r, z.; and two glides, v, y, complete the inventory of consonants. Alveolars, retroflexes and resonants do not occur word-initially. Caldwell’s Law describes the allophony of stops: they are voiceless when they occur initially or geminated, but voiced when they occur intervocalically
the e.g. (C1 2
alternates with (C1)V˘C2C3 as in the two stems of the verb ‘see’, *ka-n. - vs *kan. t.-. Though bisyllabic roots are occasionally indicated, reconstructed lexical roots are by and large monosyllabic. While any of the five vowel qualities may appear in a root, only a, i, u may appear in a derivative suffix. Dravidian morphology is transparent, agglutinating and exclusively suffixal. The
order of elements in a word is: lexical root, derivational suffix, inflectional suffix. ProtoDravidian has two parts of speech: noun and verb, both of which appear in simple and compound forms. Nouns inflect for case, person, number and gender. Proto-Dravidian has eight cases: nominative, accusative, sociative, dative, genitive, instrumental, locative and ablative. These eight are supplemented by postpositions, derived from independent nouns or non-finite verbs. Predicate nominals can be inflected to agree with their subjects, e.g. in Ancient Tamil -o-m marks the first person plural in na-m na-t.t.-o-m ‘we1 (are) countrymen2’. Proto-Dravidian has two numbers: singular and plural. Proto-Dravidian gender distinguishes animate and inanimate nouns on the basis of the natural gender of the referent, not ‘grammatical’ or conventional gender. Animate nouns may further be classified as honorific, masculine or feminine. A noun’s animacy helps determine other of its grammatical features: animates take the locative case marker *-it.am, inanimates *-il; most animates have the plural marker *-ir, inanimates *-kal.; the accusative case marker *-ay is obligatory for animates, but optional for inanimates. The very extensive system of compound nouns can be illustrated by the set of deictic pronouns, which contrast in four degrees: *ivan ‘this man’, *uvan ‘that man nearby’, *avan ‘that man yonder’, *evan ‘which, any man’. These are compound nouns, e.g. *avan ‘that man yonder’ consists of the nouns *a-‘that (one) yonder’ and *-(v)an ‘man’. Complex compound nouns are often translated into English as a sequence of numeral, adjective and noun; but the internal structure of these Dravidian expressions is that of a compound noun. Proto-Dravidian verbs are those forms that inflect for verbal categories such as tense
and mood. There are two tenses, past and non-past, and two moods, modal and indicative. From a formal viewpoint verbs are finite or non-finite. Finite verbs inflect for tense and subject-verb agreement. These inflections are overt, or, in the imperative and optative, covert. Proto-Dravidian has a constraint that limits the number of finite verbs in a sentence to a maximum of one: that lone verb stands at the extreme end of the sentence and commands all other verbs within. In effect, it brings the sentence to a close. All remaining verbs in the sentence must be non-finite. The first major set of non-finite verbs is defined as those which combine with a following verb, with or without other grammatical material coming between the two. In this set we find the infinitive, conjunctive participle and conditional. The second major set comprises all those non-finite verbs that combine with a following noun to form relative clauses and similar structures. Dravidian languages rely on a rich system of compound verbs to extend the somewhat limited set of simple verb forms. Lexical compound verbs supplement the lexicon by providing a complex morphosyntactic vehicle for combinations of lexical meanings which are not encoded in any single lexeme of the language. For example, the Tamil lexical compound kon. t.u vara ‘bring’ consists of the conjunctive participle of kol.l.a ‘hold’ and an inflected form of vara ‘come’. Auxiliary compound verbs, on the other hand, provide morphosyntactic vehicles for those verbal categories which are not encoded in any simple verb form of the language, e.g. perfect tense, benefactive voice. In this colloquial Kannad. a example the auxiliary verb iru ‘be’ conveys the perfect tense: na-n band(u) iddı-ni ‘I1 have3 come2’.
