Re: [tied] Sanskrit /r/

From: Piotr Gasiorowski
Message: 8818
Date: 2001-08-29

----- Original Message -----
From: liberty@...
Sent: Wednesday, August 29, 2001 6:28 AM
Subject: Re: [tied] Sanskrit /r/

[David:] > Rather than a retroflex 'r' triggering retroflex place assimilation might not an alveolar 'r' have triggered alveolar place assimilation?  Indians for whom English is a second language often substitute retroflex sounds for the ordinary alveolar 't', 'd' and 'n' of English and dental 't'and 'd' for the fricatives in 'thin' and 'then'.   If the Indo-Aryan dialects entering India had alveolar allophones for the dental stops caused by an alveolar 'r' and post-alveolar allophones after post-alveolar 'š' (*št > ST) might not these backed allophones of the dentals have been replaced with retroflex sounds by the Dravidian substrate speakers just as is done today?  The alveolar 'r' itself and alveolar 's' would not have been effected since Dravidian has its own alveolar 'r' and 's', if I'm not mistaken.  Therefore after a Dravidian and/or Munda substrate took up
speaking Indo-Aryan the result would be a language with dental reflexes of the dental allophones of the dental series, retroflex reflexes of all of the backed allophones of the dental series, and an alveolar 'r' and 's' spared the change of alveolars to retroflexes.  Isn't that how Vedic or Sanskrit is usually described and 'r' pronounced in the modern Indo-Aryan languages?
The pra:tis'a:kHya descriptions of the correct place of articulation for fricatives say only that they are articulated at the same points as the corresponding stops (that is, <s, s., s'> pattern with <t, t., c>. This suggests actual retroflex status for <s.>. The RUKI *š was no doubt originally some kind of (post)alveolar fricative sufficiently distinct from *s to become phonologised in the relevant branches. Place assimilation in *št, *z^d (*nizdo- > *niz^da- > ni:d.a-) and *šn is an Indo-Aryan speciality, which may have to do with a more retracted articulation of the fricative than in Iranian, which in turn might be explained by substrate influence (when already in India). Of course there may have been early Indo-Aryan dialects with their own developments (note the general merger of sibilants in Middle Indo-Aryan). I'm only speculating on the pronunciation of the "normative" variety of Old Indo-Aryan. I agree that the pattern of substrate influence was in general like what you say -- retraction reinterpreted as retroflexion.
I also agree that Old Indo-Aryan /r/ is not likely to have been fully retroflex in the technical sense; I only propose that it had a more retracted and more vowel-like articulation in non-prevocalic positions. By the way, Dravidian languages often have more than one "rhotic" phonemes, and in many modern Indo-Aryan languages there are two (an alveolar tap/trill contrasting with a retroflex flap) -- so the peninsula offers more than the "ordinary" rhotic :)

>> I have already commented on the pronunciation of Old Indic "v". I think most specialists would support the opinion that it was bilabial and that /w/ would be a more accurate transcription.
[David:] > This seems strange to me.  It was my understanding that the reflexes of I.E. *w in the modern Indo-Aryan languages, in Farsi and in Ossetic are all a labio-dental approximant with simultaneous velar approximation.  I searched the archives but wasn't able to find your comments on this.  May I impose upon you to direct me to them?
Rather than searching the archives again, let me just say that I reacted in the first place against the interpretation of Sanskrit <v> as phonetic [v] (a labiodental fricative, as in English). [w] for etymological *w seems not to be that rare in Iranian and Nuristani languages (though fine details of articulation are not always easy to recover from book descriptions). Other reflexes are found as well, including [v] and [b], the latter being also common in modern Indo-Aryan. A velarised labiodental approximant (let's symbolise it [V] here) is closely akin to a labial-velar approximant [w]. Historically, [w] is a rather unstable sound. Throughout Europe, [v] is the predominant modern reflex, though in most branches there is evidence of its development out of older [w].