Newsgroups: sci.lang
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!uunet!mr.net!umn.edu!s6.math.umn.edu!kichenas
From: kichenas@s6.math.umn.edu (Satyanad Kichenassamy)
Subject: Re: languages with phonetic alphabets
Message-ID: <D6ttsD.Enu@news.cis.umn.edu>
Sender: news@news.cis.umn.edu (Usenet News Administration)
Nntp-Posting-Host: s6.math.umn.edu
Organization: University of Minnesota, Mathematics
References: <199504071537.KAA07224@dlep1.itg.ti.com> <3m3qc1$6ap@kralle.zdv.Uni-Mainz.DE> <D6q1JD.6xK@eskimo.com>
Date: Mon, 10 Apr 1995 16:05:15 GMT
Lines: 58

In article <D6q1JD.6xK@eskimo.com> rickw@eskimo.com (Richard Wojcik) writes:
>In article <3m3qc1$6ap@kralle.zdv.Uni-Mainz.DE>,
>Joerg Knappen <knappen@vkpmzd.kph.Uni-Mainz.DE> wrote:
>>In article <199504071537.KAA07224@dlep1.itg.ti.com>, sinh@dlep1.itg.ti.com
>(Arun Sinha) writes: 
>>:Hindi (the national languge of Bharat (India)), and most  Bharatiya
>>:languages are arranged phonetically.
>>
>>The devanagari alphabet (and other indic alphabets) has the remarkable
>>property of featuring some letters, which distinguish allophones of the
>>same phoneme, but don't represent true phonemes. This is the case with
>>the letters [ng] and [nj].
>>
>>There are other deviations from the phonemic principle.
>
>This is not necessarily a deviation from the phonemic principle, although
>the allophonic status of the sounds corresponding to those letters is
>correct.  The question is not whether the letters appear to correspond to
>distinct allophones, but whether writers associate the letters with the
>palatal or velar articulation.  It could be that they conceive of the
>letters as just conventional spellings for another sound before certain
>consonant letters.  To make a claim one way or the other, you need to look
>at the way devanagari writers make spelling errors or represent
>corresponding sounds in foreign words.  For example, would a devanagari
>writer transliterate Spanish "man~ana" with the palatal "nj" or with a
>combination of letters?  What kind of spelling errors do you find people
>making with these symbols?  The existence of the symbols alone is not
>sufficient to make the case.
>
>I should say that the devanagari alphabet may well be exceptional in these
>cases.  After all, India is the source of linguistic theory, and the
>ancients had a strong tradition of articulatory phonetics.  They understood
>the difference between phonemes and allophones.
>-- 
>Rick Wojcik  rickw@eskimo.com     Seattle (for locals: Bellevue), WA
>             http://www.eskimo.com/~rickw/

Quite right. One may add that the palatal nasal does not only occur
before other consonants:

in Sanskrit: jn~Ana
in Tamil:    n~AyiRu

I would certainly use n~a if writing man~ana in Tamil or Nagari.

Actually, ng too occurs vocalized in Tamil:
 Ex.:   ingngan_am

Finally, in Tamil `nAn_ku,' n_ is not ng (it is articulated midway
between the teeth and the palate).


                                Satyanad Kichenassamy
                                School of Mathematics
                                University of Minnesota
                                kichenas@math.umn.edu


