LINGUIST List 11.681

Mon Mar 27 2000

Disc: Underlying Schwa?

Editor for this issue: Karen Milligan <karenlinguistlist.org>

Directory

1. A.F. GUPTA, Re: 11.666, Disc: Underlying Schwa?
2. Wojcik, Richard H, RE: 11.658, Disc: Underlying Schwa?

Message 1: Re: 11.666, Disc: Underlying Schwa?

 "Geoffrey S. Nathan" <geoffnsiu.edu> said:

> And there are plenty of rules that
> aren't 'true on the surface' (which is a problem OT has been
> wrestling with for quite a while). Consider, for example, the
> contrast between 'police' and 'please', which, on the surface
> contrast in voicing of the /l/ (ignoring the irrelevant final
> consonant difference). In old-fashioned ordering terms,
> schwa-deletion counterfeeds liquid devoicing, and no surface-true
> theory could deal with this situation, ....

I think we have to always remember the dialectal variation in 
English. This should read " which, on the surface IN SOME DIALECTS 
contrast in voicing of the /l/ (ignoring the irrelevant final
consonant difference). "

It took me quite a while to work this one out, because for me the two 
words both normally have one syllable and the /l/ is devoiced as 
usual following the /p in both cases/. I don't think I can be the 
only one....

Anthea
 * * * * * * * * * * * * * *
Anthea Fraser GUPTA : http://www.leeds.ac.uk/english/$staff/afg
School of English
University of Leeds
LEEDS LS2 9JT
UK
 * * * * * * * * * * * *

Message 2: RE: 11.658, Disc: Underlying Schwa?

> 	I find this whole discussion about underlying schwa a little strange, but 
that may be because I can see no consensus in the linguistic community on how to 
define phonology. I think that we are talking about two very different concepts of 
"underlying" schwas, and we are not being clear about the difference. If we take 
"schwa" to be a phoneme, then the question seems to be about whether schwa can 
ever occur in the root representation of morphemes, where some morphological 
derivative contains a different vowel in its place. In that case, the schwa 
phoneme in the root might be said to "underlie" the schwa phoneme in the derivative. 
All this talk about whether "nation" and "native" are perceived as morphologically 
related is beside the point. Clear morphological relationships do exist, and 
there are clear phonemic correspondences between them. We just have to be clear 
in what sense phonemes can be said to underlie other phonemes across 
these relationships.
> 	
> 	Let's take the a/an alternation as an example. There should be no 
question that "a" and "an" are allomorphs of a morpheme. How do we handle 
that fact in a linguistic analysis? Is it a phonological relationship? I 
don't like to think so, but it all comes down to how we choose to define the 
term "phonology". Let's assume that "a" // is a schwa, and that it corresponds 
to the vowel /ae/ in "an". Does // underlie /ae/ here? We could argue either 
side of that one, but the argument wouldn't be very interesting. This is just 
one morpheme, not a morphological pattern. The fact is that // could 
plausibly underlie an /ae/ in English, and it is just serendipity that, if 
it doesn't, it doesn't. An accident of lexical history. We just need an 
example of a class of morphemes where we can distinguish a regular phonemic 
correspondence between // and /ae/, and that may be hard to come by. We 
should just be very careful about interpreting lacunae in morphological 
patterns as evidence of some kind of significant grammatical principle.
> 	
> 	Now let's talk about pure phonology, namely what happens to the vowel 
in "an" when it gets reduced to a schwa. In that case, we are not talking 
about a phonemic correspondence but about the pronunciation of the phoneme
 /ae/ in an unstressed context. The question here is a general one of whether 
vowel coloring can be superimposed on reduced vowels. If this question 
makes sense, then it pertains to a different sense of "underlying". The 
question has nothing at all to do with phonemic correspondences between 
related morphemes. It has to do with the pronunciation of vowels in 
unstressed syllables. This thread (not to mention phonological theory 
in general) has suffered greatly from an inability to distinguish 
phonemic correspondences between related morphemes from constraints on 
the pronunciation and perception of phonemes in a single morpheme. 
There is a big difference between choosing what phonemes to pronounce 
and actually pronouncing those phonemes.