Retroflex sounds

[Sudalaimuthu Palaniappan]

In a message dated 98-06-21 11:37:48 EDT, witzel at FAS.HARVARD.EDU writes:

<< I forgot to mention that

 (A)  Southworth (1979: 201) has parallel development of *rt > .t  in
 Drav. and IA;

 see: Lexical evidence for early contacts between Indo-Aryan and Dravidian.
 In: M. Deshpande and P.E.  Hook. (eds.). Aryan and Non-Aryan in India. Ann
 Arbor : Center for South and Southeast Asian Studies, University of
 Michigan 1979, 191-233>>

The basis for this seems to be David McAlpin's Middle Elamite data, we can
defer  to Dr. Krishnamurti's view on this.

But, in the course of an unrelated research, I discovered some data which
seems to suggest *-r-, *-R- (or *-_t-), and *-l- become retroflexed in north
Dravidian. As far as I know, this has not been studied by Dravidian linguists.
(I am willing to be corrected on this. I have been trying to get some
linguists to review the data, to no avail.) The closest analysis is by Stephen
Hillyer Levitt in "The Alternation of r and l in Dravidian" in Indian
Linguistics, vol. 50, 1989, p. 130-147.

Levitt discusses how *-l- splits into *-l- and *-L- in South and Central
Dravidian and then uses the euphonic combination rules of Tolkappiyam to
explain the basis of Dravidian forms.

Examples for the retroflexion of *-r-/*-R-/*-l- in north Dravidian are as
follows. Here ".r" denotes retroflex r. # denotes item not in DEDR but found
in An Oraon-English Dictionary by A. Grignard or Malto-Hindi-English
Dictionary by B. P. Mahapatra)

DEDR 5400 Ta. vitir to scatter. Kur. bid.ra'AnA to scatter about in disorder,
Malt. biDrare to be dispersed

DEDR 5407 Ta. viracu to crowd together, join, unite. Malt. bir.ge to mix, add
(#Malt. bI.re crowd)

DEDR 1059 Ta. Or to consider, think, regard, etc.  #Kur. o.r'gnA to believe,
think, etc.

DEDR 3962  Ta. pari to separate, cut asunder, etc. Kui p.rApka to split open.
Kur. pa.rakhnA to separate or force asunder the two parts of some object.
Malt. par.ge to split.

DEDR 651 Ta. ural mortar. Ko. oLka.l mortar; oLka.l  kal a stone mortar.
#Malt. lo.ra a stone to grind spices.

(Note the Ko. form is closer to Vedic ulUkhala than ulakkai mentioned often,
which is in fact the pestle rather than mortar. My opinion is ulUkhala
probably stands for *ural(u)kal rather than *ulakkai.)

DEDR 3728 Ta. nURu to crush , reduce to powder. Malt. ni.rg/ni.rga - to
powder, grind.

DEDR 2639 Ta. cIRu to be infuriated, get angry with. #Kur. ci.rArnA  to lose
one's temper, to flare up. #Malt. ci.rAr to be angry.

DEDR 5538 Ta. vEli fence. #Malt. be.ri to build a fence, bE.ra - fence.

DEDR 833 Ta. eli rat, bandicoot. Te. eluka bandicoot. #Malt. e.rge a small
field rat.

The lack of realization of this has resulted in misidentification of the root
form for some Kurukh words such as a.rknA  "to knead with the hands, shampoo
by sqeezing from place to place, etc." with DEDR 63 Ta. aTaGku/aTakku "to
submit/press down" rather than correctly identifying it with DEDR 228 Ta.
arakku to rub with the palm of the hand or sole of the feet etc. A. Grignard
uses the word shampoo for massaging.

There are more items like these. But I hope these will be of some use to those
who want to model retroflexion in Dravidian.

Regards
S. Palaniappan