This volume is the result of two years of concentrated reading, reflection and writing, from September 1999 to October 2001, with many years of research and study prior to it. This work differs somewhat in focus and scope from the other volumes in the Cambridge Language Surveys series. My focus is on historical and comparative aspects of the Dravidian languages, although I have not altogether neglected descriptive and typological ones. There have been several descriptive studies of the Dravidian languages, though they are not comprehensive, for instance, Andronov's Dravidian Languages (1970) and, more recently, The Dravidian Languages (1998) edited by Sanford B. Steever. The second book is more substantial, although it covers only ten out of twenty-six or so known Dravidian languages; it has two chapters each devoted to Tamil and Telugu, representing old and modern varieties. Caldwell's pioneering work of 1856 (revised in 1875 and 1919, reprinted 1956, 1961) has remained the solitary model in the field for nearly a century. Jules Bloch's Structure grammaticale des langues dravidiennes (1946; English translation by R. G. Harshe 1954) has a descriptive title, but a comparative focus; it deals with specific problems of morphology based on data of newly studied languages like Gondi, Kui, Kuṛux, Malto, Brahui etc. Zvelebil's survey monograph of 1990 is a broad sketch (156 pages) of major issues in phonology and morphology up to that period.

Introduction

A summary of conclusions from different chapters, which throw light on the subgrouping of the Dravidian languages adopted here, will be presented with a short review of earlier work on this topic. I will briefly speculate on the probable date of Proto-Dravidian. I have left out several questions like the original home of the Dravidians and whether they were the same people who ruled the Indus valley around 2500 BCE. The Proto-Dravidian culture that I have reconstructed, based on comparative vocabulary in section 1.2.2, must help archaeologists and linguistic archaeologists to sort out this problem. I have hinted at the future direction for research in Dravidian studies, comparative and typological.

Earlier attempts at subgrouping the Dravidian languages

Serious attempts at the subgrouping of the Dravidian languages proceeded hand in hand with the study of a number of new languages, mainly in central India, namely Kolami, Parji, Naiki, Ollari, Gadaba, Koṇḍa, Pengo etc. during 1950–75. Earlier, L. V. Ramaswami Aiyar, E. H. Tuttle and T. Burrow (till 1950) placed Telugu in South Dravidian (our South Dravidian I). I have noted that Telugu is genetically closer to its northern neighbours, Gondi, Koṇḍa, Kui, Kuvi, to which Pengo and Manḍa were added later. Burrow and Bhattacharya (1953: xi) have pointed out the close relationship among Parji, Ollari, Gadaba, Kolami and Naiki. They also speak of ‘many signs of special connection between Gondi–Koṇḍa and Kui–Kuvi’.

Introduction

The Dravidian lexicon may be divided into native and borrowed. I have drawn on the native lexicon to reconstruct Proto-Dravidian culture in section 1.2.2, based on DEDR (1984). The vocabulary is classified into semantic and functional categories like material culture, social organization (including kinship terms), flora and fauna, weather and water resources etc. A wide spectrum of native lexicon is thereby covered. Since the contact of Dravidian with Indo-Aryan goes back to c. 1500 BCE, a number of Dravidian words found their way into Indo-Aryan at all stages, Old, Middle and Modern. Southworth (1995: 264) estimates that 88 lexical items were borrowed into Sanskrit from Dravidian from the Proto-Indo-Iranian period down to Classical. Some aspects of this process have been described in chapter 1 (section 1.7). We find that NIA exhibits more of structural borrowing and less of lexical borrowing from Dravidian. This has been explained as a function of the absorption of a Dravidian substratum into Indo-Aryan from the earliest stages of contact, which gradually affected its grammatical structure over three millennia.

Indo-Aryan loanwords in South Dravidian I and II

The four languages of South Dravidian, after their separation from Proto-South Dravidian, developed their own writing systems (see chapter 3) and became vehicles of literature at different periods, Tamil from the early Christian era, Malayāḷam (an off-shoot of Tamil) from the twelfth century, Kannaḍa from the eighth century and Telugu from the eleventh.

Introduction

Nominals in Dravidian are a morphosyntactic class of words, which are inflected for case. Nominals consist of four subclasses: nouns, pronouns, numerals and adverbs of time and place. All but the adverbs are distinguished for gender and number, besides being inflected for case. Adverbs of time and place are morphologically nominal, since they carry case suffixes, but syntactically they function as adverbs, i.e. modifiers of the verb.

Noun stems can be simple, complex or compound. Simple stems are identical with the roots, e.g. *kāl ‘leg’, *ūr ‘village’. Complex stems result from the addition of derivational morphemes to verbal, adjectival, or nominal roots, e.g. *wil-ay ‘price’ from *wil ‘to sell’ (see section 5.8.1a), *weḷḷ-ay, *weḷ-u-ppu, *weṇ-may ‘whiteness’ from *weḷ / *weṇ ‘white’ [5496]. Compound stems consist of a minimum of two roots of which the head is a noun and the satellite a noun, an adjective or a verb, e.g. *kaṇ-ṇīr ‘tear’ (N + N), lit. ‘eye-water’ [1159b], *kiṯṯ-eli ‘a mouse’ (adj + N) (lit. ‘small rat’) [1594].

Gender and number: identification and definition

Gender is a system of classifying ‘nouns’ (other than time and place adverbs) in terms of certain semantic and formal properties, supported by their grammatical behaviour. In Dravidian, gender is distinguished in terms of the semantic categories [± human], [± male human], and marginally [± animate].

Introduction

In this chapter I treat certain parts of speech which are identified mainly syntactically, namely adjectives, adverbs and clitics. There are a few words which are basic adjectives and adverbs, e.g. in adjectives the three deictic bases ā ‘that’, ī ‘this’ and ē ‘what?’ occur only in attributive position before noun heads. There are also certain suffixes, which derive adjectives from nominals (nouns, pronouns, numerals etc.) and verbs (the relative participles), but all adjectives are identified as a class only by their syntactic function as qualifiers of noun heads. The exclusive basic adverbs that I can think of are reduplicated expressions, which function as manner adverbials, e.g. Te. gaṇṭa gaṇagaṇa mōgindi ‘the bell rang gaṇa gaṇa’, wāḍu gaḍagaḍa māṭlāḍatāḍu ‘he speaks fast’. Clitics are indeclinable. They are syntactic affixes, which can be added to any autonomous unit, i.e. word, phrase, clause, with various shades of meaning, e.g. *ā, added to declarative sentences to convert them into ‘yes–no’ questions, *ē (also *tān ‘self’ in Tamil), an emphatic particle meaning ‘only’, and *ō to express doubt or ‘either–or’ relationship, etc. Each language has created a host of such clitics, beside the inherited ones. These three parts of speech will be treated comparatively below.

Adjectives

Adjectives, like the other major parts of speech, nouns and verbs, can be defined in terms of certain universal semantic types and certain language-specific morpho-syntactic properties.

Introduction

The Dravidian languages retain most of the contrasts of Proto-Dravidian vowels and consonants. Proto-Dravidian had five vowels/i e a o u/ with length contrast /ˉ/. Alternatively, one can set up ten vowels for Proto-Dravidian, five short and five long. Diphthongs [ai au] can be treated as sequences of a vowel + a semiconsonant, i.e. /ay aw/ patterning with VC. There were seventeen consonants in Proto-Dravidian: six stops /p t ṯ ṭ c k/, four nasals /m n ṇ ñ/, two laterals /l ḷ/, one flap /r/, one retroflex frictionless continuant /ẓ/, three semivowels /w y H/. The last one is a laryngeal, which patterns phonologically with semivowels. All departures from this system can be traced to two sources: (a) certain sound changes within the historical period of individual languages or subgroups; (b) borrowing from contact languages, either of the same family or of a different family.

The five-vowel system is fairly stable in all subgroups; new centralized vowels have been added, through splits, to the list of phonemes of several of the Nilgiri languages, namely Toda, Iruḷa, Kuṟumba of South Dravidian I. Long vowels are less subject to change than the short vowels. Four of the stops have remained stable in the whole family: /p t ṭ k/. The affricate /c/ gets variable treatment in different subgroups, represented as [ts c s]. The alveolar /ṯ/ also has undergone change and has been eliminated as a distinctive unit outside South Dravidian.

Introduction

The unmarked order of constituents in a sentence is Subject–Object–Verb (SOV) in the Dravidian languages. A number of other features seem to flow from this dominant pattern: adjectives (including possessive nominals) precede noun heads, adverbs precede the verbs that they modify, complements precede the matrix clauses, auxiliaries follow the main verb, and postpositions rather than prepositions follow nominals. Dravidian languages have the nominative–accusative pattern with subject–predicate agreement, and not the absolutive–ergative. A sentence in Dravidian may be simple, complex or compound.

Simple sentences

A simple sentence is represented by two grammatical constituents, Subject + Predicate. The subject is either a Noun Phrase (NP) with the head noun in the nominative case, or a Postpositional Phrase (PP) with the head noun in the dative case (see discussion in section 9.2.5). The latter type is called the dative subject. The Predicate is represented either by a Verb Phrase (VP) or a NP. Both these may have PPs as complements or adjuncts. Sentences that have NP predicates are generally equative. The four types are illustrated below, taking (to the extent possible) one language from each subgroup:

(1) NP + VP

1. a. Ta. avaṉ va-nt-āṉ

2. [he come-past-3m-sg]

3. ‘he (informal) came’

4. b. Te. wāḍu wacc-ǣ-ḍu

5. [he come-past-3m-sg]

6. ‘he (informal) came’

7. c. Pa. tolen-kul verrar

8. [brother-pl come-past-3m-pl]

9. ‘the brothers came’

10. d. Kuṛ. tam-bas ke-cc-as

11. [he-gen-father die-past-3m-sg]

12. ‘his father died’

(2) NP + NP

1. a. Ta. avar eṉ āciriyar

2. [he-hon I-poss teacher-hon]

3. ‘he (polite) is my teacher’

4. …