Newsgroups: sci.lang
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!news.alpha.net!uwm.edu!math.ohio-state.edu!howland.reston.ans.net!pipex!uknet!festival!edcogsci!iad
From: iad@cogsci.ed.ac.uk (Ivan A Derzhanski)
Subject: Re: languages with phonetic alphabets?
Message-ID: <D6GD4u.MJ1@cogsci.ed.ac.uk>
Organization: Centre for Cognitive Science, Edinburgh, UK
References: <1995Mar24.210241.81667@rs6000.cmp.ilstu.edu> <D61KLp.Kn0@cogsci.ed.ac.uk> <3lkcqk$aas@info.epfl.ch>
Date: Mon, 3 Apr 1995 09:31:40 GMT
Lines: 56

In article <3lkcqk$aas@info.epfl.ch> Dejan Djukic <Dejan.Djukic@circ.de.epfl.ch> writes:
>iad@cogsci.ed.ac.uk (Ivan A Derzhanski) wrote:
>> >tmoran@bix.com antau~nelonge skribis en sci.lang:
>> >: My son asked if there are any languages where each letter of the
>> >: alphabet has a single pronunciation.  Are there?
>> 
>> [...] yes, there are many such languages,
>> eg Finnish, Georgian, Hawai`ian, Serbian and Yiddish.
[...]
>If someone wants an alphabet to present ALL the phonemes
>(voices) used in a language, it would probably contain
>over 100 letters (signs), and 200 is more realistic figure.

That's not true.  The number of phonemes never exceeds 100, and seldom
gets close to that.  What you are thinking of is the number of
phonetically, not phonemically distinct sounds.

>The compromise is in this case the way to map close sounds to one letter.

Yes.  Where `close sounds' ideally means `allophones of one phoneme'.
If the original question was about languages in which allophones are
also consistently differentiated, then the answer is indeed `no', but
I don't think languages whose scripts map each letter to a unique
phoneme should be ruled out.  Representing regional and/or individual
variation would certainly be too impractical.

Your points about consonant assimilation and vowel nasalisation are
well taken.  So is the one about vowel length (not sure about pitch
-- is that not a feature of the syllable that just happens to be
realised on the peak?).

>In Italian, I think, the difference between the open and closed E is
>essential.  If one pronounces the "wrong" E, it will be perceived so.
>And both E-s are written with the same letter.

You'll note that I didn't list Italian as an example.  (Though there's
no shortage of other reasons for which it won't do: there's _c_, _g_,
_s_ and _z_ and a number of digraphs and trigraphs.)

>Another problem is the sound R.  In some words the accentuated
>syllable is made not with vowel, but with a vocalized R instead.
>That means that R is pronounced as shwa + R, [...]

Does it?  Is not /r/ itself the peak of the syllable?

>So, neither in Serbo-Croatian there is a perfect 
>map between the voices and letters, but the existing 
>writing system is probably the most that can be practically done.   

It comes close, anyway.

-- 
`"Haud oan there a meenit," says the king tae Joseph, "I've been thinkin."'
Ivan A Derzhanski (iad@cogsci.ed.ac.uk)    (J Stuart, _Auld Testament Tales_)
* Centre for Cognitive Science,  2 Buccleuch Place,   Edinburgh EH8 9LW,  UK
* Cowan House E113, Pollock Halls, 18 Holyrood Pk Rd, Edinburgh EH16 5BD, UK
