_pri_nce [phrIns], more than likely [pirins]
_Bru_nswick [brVnzwIk], probably [paransvik]~[baransvik] (not long -aa-,
I'd think)...

Tamil writing itself would not be constrianed by Unicode encoding...
Unicode encloding only constrains those of us who are trying to
digitally present Tamil... so far as we know, all
modern/contemporary/current needs for actualized Tamil should be
accomodated by the Unicode standard... (note: grantha/pali/sanskrit
extended letters, if they ever existed in Tamil, are not included in the
Unicode block of Tamil... largely due to the sponsoring bodies not
wanting or considering them)...

My assumption would be that Tamil speakers who have not been exposed to
many loanwords or "foreign sounds" will still follow the stereotypical
description of Tamil by: adding epenthetic vowels to break up clusters;
by avoiding voiced obstruents in initial position (given that voiced
obstruents are restricted in current colloquial Tamil); etc etc etc.
Given that the above words are non-native, they will probably be
"marked" in the Tamil speaker's mind as that and that loanwords (and
onomatopoeics) allow voiced obstruent onsets, it's possible.. though
more than likely they will still be _written_ with <p>...

The prior arguments that "because Tamil has [voiced obstruents in
initial position]" it will then readily accomodate other voiced initial
obstruents doesn't hold with me. While there are many who will be able
to incorporate these novel segments into their phonemic inventory, the
previous trends of loanword adaptation/nativization have shown that
Tamil *will* devoice these segments... there are of course, counter
examples of new segments being introduced into languages due to contact,
but.. who knows what Tamil'll do...

An example of "well Language X has Segment 1 in it's inventory, so it
should be able to accomodate it from Language Y" is pretty lame. A
concrete example is Spanish > English... a *very* common word for many
parts of the nation: "taco"... Current North American English (all
Englishes?) has [thakhow] (/takow/); source language (Spanish), however,
has [tako] (both unaspirated)... English, however has unaspirated
versions of /t/ and /k/, c.f. <tie> vs <sty>, <kill> vs. <skill>...
(note also that for most American English speakers, voiced obstruent
initials /b, d, g/ are actually lax and unvoiced...) Phonemically, they
are not constrastive... and they are environmentally conditioned...
normative English phonotactics disallows unvoiced unaspirated obstruent
initials... hence, we don't say "taco" like the source language...

ok.. enough caffeinated babbling...
-patrick

suzmccarth wrote:

> So back to my original question - does Spoken Tamil have initial
> consonant clusters such as "pri" (Prince) and "bru" (Brunswick)or
> not? Are these represented as "piri" and "pira" in Tamil because of
> their actual pronunciation in ST, because of the constraints of
> Unicode, because of the perception that Tamil writers have of Tamil
> pronunciation of these sounds or because of LT tradition? Why not
> put a a pulli over the initial consonant? (I do realize that this
> wouldn't differentiate the initial voiced/voiceless contrast.)
>
> Suzanne