Dravidian "c"-"y" alternation (was Re: solution to the "kuyava' etymology)

[Palaniappa]

In a message dated 98-01-20 07:53:15 EST, bhk at HD1.VSNL.NET.IN writes:

<< >As for the process of one-way change of "-c-" to "-y-", is it universal?
Then,
 >can somebody explain the following? In DED, we find the following.

 >*kAy,  to grow hot, burn, etc. - Ta. kAy; Ma. kAyuka to be hot, To. ko.y-
 (ko.c-) to be hot; Tu. kAyuni to be hot; Go. kAsAnA to become hot.
 >*koy, to pluck leaves, flowers, reap, harvest, etc. - Ta. koy; Ko. koy-
(koc-)
 >to cut; To. kwIy- (kwIs-) to pluck fruits, Go. kOiyAna harvest wheat or
crops,
 >etc., (M) koidAna to reap; Kur.  khoynA (khoss-) to cut down grass and the
 >like with the sickle, mow, reap
 >*kAy and *koy are PDr forms reconstructed by Dr. Krishnamurthy in his TVB.
 >Clearly, we have here "y" changing to "c"/"s"/"t".  In what way, is the
 >"kuyava"/"kucava" problem different, if we assume "kuy" is the base?

 Go. ka:s is a restuctured stem from the past ka:si:, just like Modern Telugu
 has ka:s-tu:; I discussed elsewhere how the roots have incorporated tense
 morphemes and got restructured. (Also discussed in TVB); in the case of koy
 what you find in parenthese are past stems with c/t in Sandhi. Kota koy-t
 --> koc; so  also Toda; Kur. khoss- is past stem; -ss- is past tense marker
 in Kur-Malto.
  >>

I think there is one unambiguous way to check if the change of "y" > "c" can
happen in Dravidian. If we compare loan words and how they are pronounced in
Dravidian, we can see this. Let us consider the following words.

1. Greek ionia? > Sanskrit/Tamil yavana(n2) > Tamil cOn2akan2
2. Sanskrit yakSI > Tamil iyakki > Tamil icakki
3. Sanskrit vyAdhi > Tamil viyAti > Tamil vicAti
4. Sanskrit gAGgeya > Tamil kAGkEyan2 > Tamil kAGkEcan2

Items 1, 2, and 3 as well as "kAGkEyan2" can be checked in the Madras
University Lexicon. The form "kAGkEcan" occurs as part of a name of a town in
Northern Sri Lanka called "kAGkEcan tuRai".  So the "c"-"y" variation could be
a two-way street.

No wonder then G. S. Rao, a Dravidian linguist and a former student of Dr.
Krishnamurthy, in page 14 of his book "A Comparative Study of Dravidian Noun
Derivatives" (published in 1991) says, "When we say that in ProtoDravidian *c
and *y alternate in the rootfinal position we imply that it is a synchronic
fact of Proto Dravidian whose traces we can find in the modern descendant
dialects of the parent speech." Please note the use of the words "*c and *y
alternate" as opposed to  "*c weakens to *y".

In fact, later in the book (p.98), he gives the following etymology:

1. Ta. acai 'to be weary', acar 'to become faint', ayar-ppu, ayar-vu
'forgetfulness', Ma. ayar-ppu 'forgetfulness, swoon' *ayar-p-(41).

(41) refers to DED entry. If the weakening hierarchy of "c" > "y" were to be
strictly followed would he not have to give "*acar-p-" as the root?

In his book Telugu Verbal Bases, Dr. Krishnamurthy in connection with the
words meaning "to weave, weaving" (Ta. ney tr. to weave; necavu, neyavu
weaving; Te. nEyu to weave; ne/Eta weaving DED 3103) gives the root as PDr.
*nec-/*ec->*ne/Ey-. But P.S. Subrahmanyam, in p. 198 of his book "Dravidian
Comparative Phonology" (published in 1983) reconstructs a Proto-Dravidian root
*ney- for the same. (However, he does not give "c" as a reflex of Proto
Dravidian *-y-.)

P. S. Subrahmanyam gives a reflex "-j-" in Kui for Proto-Dravidian "*-y-". We
also know that Sanskrit "ja" is often transformed into Tamil "ca". This is not
due to someone deliberately following Literary Tamil conventions. Illiterate
persons often will say "rOcA" for "rojA" and "mavarAcan2" for "mahArAjan". Can
one postulate a theoretical sequential operation of these two processes as
explaining this "y"> "c" change? (Sanskrit vijAti and vyAdhi become Tamil
"vicAti".)

I have heard some Jaffna Tamil speakers pronounce "y" as "j". For instance,
the name "yOgEsvaran" was pronounced as "jOgEsvaran". In fact, some even
transliterate Tamil "ya" as "ja" as an earlier posting in the Indology list
did. I wonder what caused "yAzppANam" to become "Jaffna" - Tamil's
pronunciation as "jA" or Portuguese/Dutch/English naturalization of Tamil "yA"
as their "jA".


Regards

S. Palaniappan