Thanks Hans and Dipak for the clarification. Now I understand why Witzel referred to m/bruža without any explanation: splitting of sonorant clusters by an insertion of a homorganic voiced stop is a well-known phenomenon in Indo-European.

I also looked at Witzel's references to von Hinüber 1980 (in German) and 1989, but I still do not see where the ancient self-designation of the Burusho- people as mruža is attested. I appreciate any help in tracing the attestation of mruža.

Thanks,
Suresh.

On Sat, Nov 17, 2012 at 10:11 PM, Dipak Bhattacharya <dbhattacharya200498@yahoo.com> wrote:
OIA √brav (AiG.I 182) is said to be from Prim.Indo-Iranian √mrav retained in the Avesta. Wellknown phenomenon. See Aig lc for m>b in Greek
Best
DB

From: Suresh Kolichala <suresh.kolichala@GMAIL.COM>
To: INDOLOGY@liverpool.ac.uk
Sent: Saturday, 17 November 2012 9:52 PM
Subject: [INDOLOGY] m-/b- alternation in Burushaski

The list may recall our discussion on Narmada and Narbada alternation, and the possible sound change where intervocalic labial nasal transforms to become an oral stop along the following lines: -m- > -β̃- > -v/b-. 

However, I was surprised to find a word-initial alternation of m and b in one of the famous papers Witzel wrote in 1999 (he was very prolific in that year). He  says the following in this paper (http://www.people.fas.harvard.edu/~witzel/MT-Substrates.pdf): 

North of this area, at the northern bend of the Indus (Baltistan/Hunza),  Burushaski is spoken. However, the language and the tribal name are indirectly attested in this general area ever since the RV: *m/bruža  (mod. burušo) > Ved. Mūja-vant, Avestan Muža.

Is such m-/b- alternation a known Burushaski phenomenon? Witzel seems to suggest that Mruža was the the ancient self designation of the Burusho- people but, unfortunately, without any helpful references.

I appreciate any information or pointers on such transformation.

Thanks,
Suresh.

It is interesting that Brahui, a Dravidian language now found in the southwestern corner of Pakistan, also shows a similar transformation:

*mē(l)-  'over, above' > Br. bē-
*miṭ- 'to leap' > Br. *biṭ- 'to throw, let drop'

However, this change is apparently conditioned by a following front vowel (Dravidian nasals in Brahui, Krishnamurti 2001:121-126), and perhaps unrelated to the above alternation.