Some Origins of Kota -j(-).

• Author: EMENEAU, M. B.

Murray Emeneau here continues his inquiries into the phonological history of the lesser-known Dravidian languages.

1. A statement of conditioned developments of [[blank].sup.*]1 in the Dravidian language Kota has not yet been undertaken except in a preliminary way. That [[blank].sup.*]-1(1)i is represented by Ko.-j in several items has been noted by several scholars; the nearest approach to complete statement is that by Subrahmanyam 1983: 87, [sections]6.3.5; and 404, [sections]34.1.2. However, it is possible to make a more complete collection of material from DEDR (and to add from later-collected field work an entry that is found only in the Nilgiri languages) and to attain more precision in the statement of conditioned developments.

2. In three items (a complete collection, it is thought), after i and before -y (which is [less than] *-i) *1/11 [greater than] l.

   1572 gily 'extreme fear'; cf. Ka. Te. gili Ta. kili 'fear'.

   4181 pily 'Toda or Kurumba witchcraft'; cf. Ta. pilli 'sorcery, magic', To. pily 'Toda sorcery'.

   Kota vily 'Rhododendron nilagiricum (or R. arboreum, var. nilagirica)' has been found in a recent research of my records; it is joined by the recent recording of Badaga (Hock.) billi id. and Alu Kurumba bille-mara id ([less than]*villay, with a different formative suffix *-ay). 4553 To. pisx id. (so I recorded it) would attest formation with a different formative suffix (*vil[l]-Vnk-); a very recent recording [1] yielded both pisx and pisk, the latter presumably [less than] *vil(l)-Vkk-. This item, then, requires for Kota a reconstruction *villi. Since the botanical item is found within Dravidian-speaking subcontinental India only in the Nilgiri micro-area, Badaga billi, which is not attested in Kannada, must be a borrowing from one of the more aboriginal Nilgiri languages; Kota would seem, so far, to be the source.

3. Earlier statements of the type *-li[greater than] Ko.-j now have to be revised to take account of the conditioned development which has just been set forth as *-1(1)i [greater than] Ko. -ly, when the preceding vowel is *i. When the vowel that precedes is other than *i (there are no occurrences with *1), *-1(1)i [greater than] Ko. -j; the preceding vowel is *a, *e, *e, *o, *o, *u, *u, (no occurrence with *a). When the vowel is *e, there is some complication in the statement, in that *e]li[greater than] Ko.-e]yj, which Subrahmanyam 1983: 87 attempts to solve by a statement involving metathesis (viz., "after e, *li...