...Washington
Forthcoming in Proceedings of the 6th International Conference-Seminar of Tamil Studies, held in Kuala Lumpur, November, 1987. An earlier version of this paper was presented at the annual meetings of the American Oriental Society, New Haven, March, 1986.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
...cases
In fact all Dravidian literary languages are described by native grammarians as having eight cases:

`There are eight cases, viz., nominative, accusative, instrumental, dative, ablative, genitive, locative and vocative according to the native grammarians of Tamil (Tol. 546, 547 and Nannul 290), Malayalam (Lilatilakam S. 22), Kannada (SMD. 103) and Telugu (Bala vyakaranamu 5.1).' (Shanmugam 1971:250)

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
...presented.
I shall not attempt to go beyond the morphology and syntax of case in Tamil and try to formulate an overall semantic analysis for each case morpheme/postposition. There is a need here not only to determine what semantic distinctions are involved, but also what the surface-structure categories are, since there is not even agreement in this area. Since the Tamil case/postpositional system seems to involve many more contrasts than seem to be minimally necessary according to analysts of case systems in general (cf. Fillmore, 1968:24, who posits six cases minimally), I shall not attempt to fit this analysis into a `universalist' framework.

One must also confront here a problem that comes up in all analyses of case systems, namely, whether something is a `true' case marker, or `just' a postposition. Underlying many analyses of Dravidian systems is an uneasiness in dealing with the genitive, since it seems to stand midway between case and postposition, or to show characteristics of both. There seems to be a somewhat universal notion that case is to be understood as consisting of those bound morphemes that do not occur elsewhere in the language, whereas postpositions are independent, non-bound free forms that cannot be attached directly to stems of nouns or pronouns but must follow some case marker. They supposedly can (in most instances in the Dravidian languages at least) be easily shown to be derived from nouns or verbs; deverbal postpositions usually require the case-marker that the source verb requires. Case markers are supposedly bound and do not occur elsewhere in the language, although they can sometimes be traced historically (or derivationally) to some other morpheme in the language. Thus, Caldwell, for example, describes the Dravidian system as follows:

``All case-relations are expressed by means of postpositions, or postpositional suffixes. Most of the postpositions are, in reality, separate words; and in all the Dravidian dialects, retain traces of their original character as auxiliary nouns. Several case-signs, especially in the more cultivated dialects, have lost the faculty of separate existence, and can only be treated now as case-terminations; but there is no reason to doubt that they are all postpositional nouns originally." (Caldwell 1961:253).

Lyons, to quote one analyst of case, feels that the distinction is basically irrelevant, since it is only a surface category:

``Whether the term `case' should be extended beyond its traditional application, to include prepositions as well as inflexional variation, is also a question of little importance. The difference between inflexional variation and the use of prepositions is a difference in the `surface' structure of languages. What is of importance, from the point of view of general linguistic theory, is the fact that the `grammatical' and `local' functions traditionally held to be inherent in the category of case can be no more sharply distinguished in those languages which realize them by means of prepositions than they can in languages in which they are realized inflexionally." (Lyons 1968:303).
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
...uses
Tolkappiyanar seems to have favored analyzing instrumental and sociative as separate cases, but later commentators, e.g. Cenavaraiyar (14th century) was opposed to this on the grounds that the two suffixes were for the most part in free variation, and because they were not considered separate in Sanskrit (Shanmugam 1971:250). Caldwell (1856, repr. 1961) felt that sociative and instrumental were quite different and could not always be interchanged:
`[T]he Dravidian social ablative, as some have called it, or rather, as it should be termed, the conjunctive case, though it takes an important position in the Dravidian languages, has been omitted in each dialect from the list of cases, or added on to the instrumental case, simply because Sanskrit knows nothing of it as separate from the instrumental. The conjunctive, or social, stands in greater need of a place of its own in the list of cases in these languages than in Sanskrit, seeing that in these it has several case-signs of its own, whilst in Sanskrit it has none.' (Caldwell 1961:278).
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
...kaiyaal).
In an earlier version of this paper I gave an example of what I thought was a use of the sociative marker oode as an instrumental marker, as in
kayyoode saappidunga `eat with your hand'
I was ignoring the fact that sociative use of oode in this example expresses not instrumentality but `immediacy', i.e. it expresses the idea of eating `on the run'. This construction is an elipsis for a fuller expression ``kayyoode kayyumaa" (cf. Schiffman 1979:21 for a more complete description of this idiom).
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
...obvious
As it was in fact to earlier scholars: `Dravidian grammarians have arranged the case system of their nouns in the Sanskrit order, and in doing so have done violence to the genius of their own grammar.' (Caldwell 1961:277)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
...records
S. V. Shanmugam has noted, as we have seen, that in the time of Tolkappiyam there were some commentators who felt that the Sanskrit model should not be followed, but by the time of Cenavaraiyar (14th century) arguments in favor of the Sanskrit model were more persuasive. (Shanmugam 1971:250).
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
...locative,
Asher (1982:111) states this very succinctly: ``Though ablative has often been recognised as one of the case forms, it would equally well be analysed as locative (kitte or -le) + a postposition (-runtu `source'). Note in this connection that -runtu is cognate with iruntu, the past participle of the verb iru `be'."
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
...all,
The nominative is conspicuous in its lack of ability to take postpositions, whereas the accusative, which may at first glance seem to be unrelated to postpositional usage, actually has a number of de-verbal postpositions such as poola `like', tavira `besides', seettu `together' that cannot be ignored. The vocative I exclude totally from the case system, for afore-mentioned reasons.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Harold Schiffman
4/26/1999