Retroflex sounds

Bh.Krishnamurti bhk at HD1.VSNL.NET.IN
Mon Jun 22 15:38:41 UTC 1998

At 22:42 21/06/98 -0400, you wrote:
>Thanks for the clarification!
 >I believe in a similar scenario but I  am still at the stage of
>collecting data/proofs. Hence my questions, to go beyond Zevebil 1970

>To provide two somewhat "outrageous" scenarios:
>(1) If PDrav. _t, .t etc. are securely PDrav. (as also in Zvelebil/DED)
>is there any INTERNAL evidence in PDrav. that they might have developed
>(in *pre*-Proto-Drav.) from various Sandhi positions / particular phonetic

>I gather, not?

No; Hans Hock also asked me if we could somehow derive  all retroflexes and
alveolars [which do not occur word-initially= the basis of the hunch] from
sandhi of some liquid+dental. Only _t and .t are derivable in a few cases of
inflection and derivation by sandhi. What about l,.l, r,.z which also do not
occur word-initially? There is no way to derive them through sandhi. So
non-occurrence word-initially is only a phonetic constraint (the tongue-tip
can be raised for coronal articulation only after a syllabic segment) and
not reminiscent of a historical event.

My Hawaii paper, Patterns of sound change in Dravidian, is published in The
Yearbook of South Asian Languagea and Linguistics 1998 (ed. Rajendra Singh),
Sage, 1998. I have discussed some PD morphophonemics in it.

>(2) if  (1) does not work, and waht I also had in mind:  has anyone an
>opinion on:
> Nostratic **t  > PDrav. *.t,  etc.?
>A brief check of various sounds/and groups in : Illich-Svitych, Opyt...,
>Moskva 1971-6 follows.
>If I understand him correctly, he reconstructs Drav. retroflexes as
>conditioned by *intervocalic position* in Nostratic. I am somewhat
>hampered now as I have left my Russian Dict. in the office.
>I am quoting from Illich's dictionary (nith his no.s) , merely adding DED2
>numbers (hopefully correctly), as he only has DED1 page numbers (which I
>don't have here)
>        (Afro-Asiatic, AfrAs,  for his, older, Hamito_Semitic)
>        (NB. Nostr. .t, .k,   etc. are of course NOT retroflexes)
>Illich-Sv. no. 225 Drav. ke.t^ //  " padat'/to fall"
>**Nostr. .ket^, IE ? keid/k'ad ( O.Isl. hitta,  Skt. zad)
>Drav.   DED2 1124  Tam. ki.ta, Tel. ke.dayu
>no. 306 Drav. muu.t  // "konchat's/konec/ to end"
>Nostr. **muda, AfrAs. md, Alt. muda
>Drav. DED2 4922  Tam. mu.ti, Tel. muu.du (Krish. 466)
>no . 251 Drav. ne.t/naa.t  // "razyskivat', nakhodit'/ to seek"
>**Nostr. ? lewda", Ural. Lewda"
>Drav. (1. ) *neet   DED2 3766    Tam.  ne.tu,  Kodagu nee".d
>(2) *naa.t  DED2  3637  Tam. naa.tu, Tulu naa.du
>no. 205  Drav.  ku.d.d //  "malenkii/ small"
>**Nostr. .ku.t^, AfrAs. q(w).t, k(w).t, kt, Kartv. .ku.t, .ko.t
>Drav. ku.d.d  DED2 1670  Tam. ku.t.tam, Kui  guu.ta, Kurukh  gu.druu,
>Brahui  _ghu.d.du
>(As for Drav.  .d.d in mid-word position (inlaut), he presupposes some old
>suffix *-H : *-.tH > Drav.  .d.d )
>no. 194 Drav. ka.t/ka.t.ta //"ukhodit'/to pass through"  :
>Nostr. **.kaLa, IE?, Kartv. .kel, Ural. kad'a, Alt. k'ala Drav. : DED2 no.
>1109 'pass through' etc. Tam. ka.ta etc.
>no. 204 Drav. ku.t.t //  "tainyi/secret"
>Nostr. ** .kuLa,  AfrAs. q(w)l,  Kartv. .kwel, Alt. k'ula
>Drav. DED2 1675  Tam. ku.t.tu, Tel gu.t.tu etc.  'secret'
>no. 79 Drav. ka.n.t  // "samec/man, etc.,"
>Nostr. **ga"ndu   (a" = a + umlaut) , Drav. ka.n.t, Alt. g"andu"
>Drav.  DED2  1173 Tamil ka.n.ta_n, Tel. etc.
>no. 301 Drav. mi.n(.t)  // "zhenshchina/samka/ <<promiscuous>> woman"
>Nostr, **min'a", AfrAs. m(j)n, Ural. min'a"
>Drav.   DED2  4858  Kan. mi.n.di <<lusty female>>, Tulu mi.n.di, -- cf.
>Kota mi.n.d.n, Kan. mi.n.d.a; cf.  Brahui min.d; Skt. menaa  from N.
>no. 86  Drav. *ku.n.ta? //  rather doubtful. "serdce/heart"
>Nostr. **golH^, Kartv. gul-. Alt. gol(^) -
>Georgian gul, Old Uighur qo:l, Mong. gool, AfrAs some Chadian forms
>Drav. (?) DED2 1693  Tam. ku.n.ti, Kan. gu.n.dige, Tel. etc.
>(He has many more  cases of _r, .r,  -l, .l)
>Any opinion?   Or is that *too early* to ask?
>This is important:
>If one accepts the Nostraticists' position, then the Dravidians, too,
>started to retroflex their tongues only *after* entering the
>subcontinent.... as nearly everybody else.
>Michael Witzel                       witzel at
> I am not a Nostraticist. I looked at some basic vocabulary, like numerals,
personal pronouns and kinship terms and did not find much support for a long
range genetic relationship. I believe that the time tested comparative
method will fail in establishing long range genetic relaionships, because
aspects of  diffusion from other families in contact cannot be accounted
for, as different from gentic phenomena. I have just started reading Bob
Dixn's new book The rise and fall of languages. He has some interesting
insights on this question. Bh.K.
H.No. 12-13-1233, "Bhaarati"
Street 9, Tarnaka
Hyderabad 500017

More information about the INDOLOGY mailing list