Newsgroups: sci.lang
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!solaris.cc.vt.edu!insosf1.infonet.net!internet.spss.com!markrose
From: markrose@spss.com (Mark Rosenfelder)
Subject: Re: languages with phonetic alphabets?
Message-ID: <D64HvA.4t0@spss.com>
Sender: news@spss.com
Organization: SPSS Inc
References: <3kv7jk$61e@news1.delphi.com> <planders.578.2F732A45@mail.utexas.edu> <D62nrq.G1I@eskimo.com>
Date: Mon, 27 Mar 1995 23:42:45 GMT
Lines: 31

In article <D62nrq.G1I@eskimo.com>, Richard Wojcik <rickw@eskimo.com> wrote:
>Russian differs from English in that the spelling is can be linked to
>phonemes in a more predictable and consistent fashion, but there is very
>little one-to-one correspondence between phonemes and letters of the
>alphabet.  Just for starters, there is a complete mismatch between vowels
>and consonants.  Russian has only 6 vowel phonemes--i,y,u,e,o,a (5 if you
>count [i] and [y] as a single phoneme), but it has 10 vowel letters [...]
>Cyrillic consonants, reflecting an
>earlier phonemic pattern, simply fail to correspond to the
>palatalized/nonpalatalized dichotomy in the phonemic system.  Vowel letters
>are by and large used to signal whether or not a preceding consonant is
>palatalized.  At the end of a word, the "soft sign" (miagkii znak) is used
>to signal a final palatalized consonant.  So the vowel and consonant
>letters do not really correspond to Russian phonemes much of the time.

Technically you are correct-- there's not a 1-to-1 correspondence between
phonemes and letters-- but this is taking a very narrow view of a phonemic
script.  I'd say the Russian script *is* pretty phonemic.  It simply chooses
to mark the palatalization feature of the consonant using the following symbol,
instead of on the consonant itself.  This destroys the 1-to-1 correspondence,
but not the phonemicity of the script; the phonemic information is all there,
and the system thereby offers a considerable savings in graphemes.  A 
transliteration using a Roman or Roman-based script, including IPA, is more 
cumbersome without giving more information.

>It gets worse.  Not only is there a very active rule of vowel reduction,
>but the voicing of consonant clusters is usually based on the final
>consonant (contrary to English, which tends to base it on the initial
>consonant).  

If these are regular rules, why do they affect the phonemicity of the script?
