Tamil la-La-Za-ra (Re: Vedic l

[Jean-Luc Chevillard]

Peter M. Scharf a écrit :
> [....] One I'm a bit uncertain about it the so-called Vedic l, 
> sometimes called Marathi l, which occurs intervocalically in Rgvedic 
> dialect for retroflex d. I'm unsure whether this is an approximant 
> (like l) or a flap like .d (that is the Devanagari d with a dot below 
> as in Modern Hindi la.dakaa). Gujarati also has intervocalic flap 
> developing from .d. These flaps are not l's bur r's. The question is, 
> "Is (or was) the Vedic l a retroflex flap or a retroflex lateral 
> approximant?
>
> [...]
>
> One bit of evidence that might settle the question concerns whether 
> there are cases in Indic of the retroflex lateral approximant 
> occurring particularly in the intervocalic environment. Is the Tamil 
> retroflex l exclusively or particularly intervocalic?
In the case of (Classical) Tamil,
there is a group of 4 items to be considered:

la (ல): transcribed "la" in Tamil Lexicon and popularly called "small 
la" (ciṉṉa la) {cin_n_a la}

La (ள): transcribed "ḷa" {l.a} in Tamil Lexicon) and popularly called 
"big la" [periya la]

Za (ழ): transcribed "ḻa" {l_a} in Tamil Lexicon (BUT a popular way of 
writing it in newspapers is "zh") [[This sound is considered by some as 
unique to Tamil]]

ra (ர): transcribed "ra" in Tamil Lexicon and popularly called "small 
ra" (in order to distinguish it from ṟa {r_a} (ற, popularly called "big ra")

The Tolkāppiyam, which is the earliest extant Tamil grammar
describes these four items in two sūtra-s: TE95 and TE96

V.S.Rajam, in her 1981 Ph.D. dissertation (under the supervision of 
G.Cardona, U.Pennsylvania)
translates these two sūtras in the following way:

TE95: "As the front tongue rises and scrapes/brushes the palate,
rakāram (r) and Zakāram (ḻ) -- these two are born (op.cit. p.101)

TE96: "As the edge of the tongue swells and joins the root of the upper 
teeth
and (as it) strikes and scrapes/brushes the palate in that region,
lakaaram (l) and Lakaaram (L) are born (respectively). (op.cit. p.101)

It is quite difficult to know what the original pronunciation of those was,
at the time of Tolkāppiyam
because the technical terms used
(varuṭal "to scrape, to brush"; oṟṟal "to strike")
are not necessarily well understood nowadays
and because relying on modern Tamil pronunciation
for interpreting them is risky:
some dialects have merged "la" and "La", keeping them separate from "Za"
whereas other dialects have merged "La" and "Za", keeping them separate 
from "la"
(this was already the case in some areas at the time of the 
viiracoozhiyam (vīracōḻiyam)
a grammar probably dating back to the 11th cent.)

Sūtra TE95 is puzzling because
the same explanation is given for both "ra" and "Za"

Modern linguists usually describe
"ra" as a tap,
"Za" as an approximant,
and "la" and "La" as laterals (see Bh. Krishnamurti, The dravidian 
languages, 2003, p.62).

All of this brings more questions than answers
but the topic is quite interesting
in what it reveals about the genesis of those Tamil texts
(and we might be able to discuss it more at length next week in Potsdam)

-- Jean-Luc Chevillard


P.S. do you know what Renou is referring to
when he talks
about the role of the "extrémités de la langue" (jihvaantau)
in the faulty pronunciation of l
(See: "Terminologie grammaticale du Sanskrit", p.422,
jihvaantau "les deux extrémités de la langue": leur rôle dans la 
prononciation (fautive) de l [R. XIV 27 (785)]. )