Re: Is this a realistic phonology?

> I don't think all Sanskritist would agree with your pronunciation. Indeed,
> I was told by a Sanskritist that it was _not_ the American 'er' of water.
> Without recourse to time-travel one cannot, of course, ever be absolutely
> certain on such matters.
You're right, of course.  The reconstructed pronunciation of ancient languages
is,
in general, an area rife with disagreements and disputes.

The vowel in question is described as the sound you makes if you open your
mouth, curl your tongue up so that the tip is pointing to the top of your head
and then say, "AH" /a/.  When I do this, it invariably comes out *sounding* like
the rhotic sound in "water," or "turn" and so forth.

Although I (and many Americans) articulate this sound a little differently
(I tend to curl or slightly round my lips for both /r/ and /R/, I've also
heard that some speakers "bunch" their tongue instead of placing it in a
retroflex position) there are phoneticians who teach non-native English
speakers to pronounce the American R much the same as the Sanskrit
tutorial did in the example above. Although these people may not be
articulating their R's the same way I do, their speech still *sounds*
the same as mine.  We learn language through mimicking sounds,
not through exact articulation.  If I learn a new language and it
*sounds* right to native speakers, then they aren't going to trouble
me about whether or not my unvoiced stop is dental oralveolar.)

Then again, from what I've seen online, the pronunciation(s) of
R's in English is also an area of much contention ; )

According to the same tutorial, the consonant R is a flap pronounced with
the tongue in that same position (a retroflex flap, I think.  IPA terminology
still makes my head twirl.)

Basically, in this system of phonetic classification,
/i/ is to /j/ as
/u/ is to /w/ as
/R/ (in rhotic "turn") is to /r/

> Indeed, IIRC in the later Indic languages the sound was realized as /ir/ or
> /ri/.  Certainly in ancient Greek the PIE syllabic r gave way to /ar/ or
> /ra/ in most dialects & to /or/ or /ro/ is some others; in Latin it gave
> way to /or/.  In all these languages the /r/ is held to have been the
> apical trill (it's certainly hard to see why most of their modern
> descendants have the apical trill if the ancient languages didn't).  I
> personally would find these all rather odd developments if the PIE &
> Sanskrit syllabic r was the American 'er' of water.
<aside

After actually making these sounds, I could easily imagine how a proto-
language that once contained the /R/ vowel and the English /r/
could evolve so that the latter became a flap or trill (instead of an
approximant) and then was moved forward.  Likewise,
I can also see how the former would mutate into /ri/, /or/ etc.
Let us remember that most speakers of IE languages may not
be descended from native IE speakers, but from people who
acquired IE languages after having been acquired themselves
(i.e. conquered).  When this occurs, difficult to pronounce
elements are generally the first to be dropped or transformed.

end of aside>


Whether correct or incorrect, the Sanskrit phonology in the
tutorial I found provides a very intriguing alternative to the IPA system
(which is, admittedly, more scientific).  Basically, the cardinal vowels
are all the vocalic versions of the apporximants /h/, /j/, /r/, /l/ and /w/.
All of the other vowels are treated as combinations or hybrids of these
five.

I'll post the link if I can find it.

Have a good one,

Mathew