schwa-raising - formants

[Mark J. Jones]

----------------------------Original message----------------------------
Dear all,

more on raised schwa, I'm afraid.

Bob Rankin and Cecil Ward have both said basically the same thing: they
hear these vowels as /I/, and so they are /I/. I take this point, but I
don't think it stands up to scrutiny on a wider basis.

There are really two major aspects to the phonological identification of
any sound on which we can carry out phonetic investigations: production,
and perception.

It's clear that for both Bob and Cecil, the raised schwa counts as /I/,
i.e. perceptually they identify the vowel with /I/. I'm not sure what
Cecil's linguistic background is, but Bob is not a native speaker of the
same variety as me, and therefore his analysis is like saying that because
a speaker of a particular variety of English cannot hear the difference
between alveolar /t/ and dental /t/ in Tamil, there is no distinction in
Tamil. Or that because s/he hears French /t/ as English /d/, that is what
is French /d/ is phonologically. Clearly these comments are untenable for
the Tamil or French and maybe also therefore for differences between
varieties of English.

Production-wise, we can claim that if two things have a consistently
measurable difference (acoustic or otherwise) which cannot be attributed to
passive effects of surrounding segments etc., then they do not have the
same input to the speech production mechanism, i.e. for that speaker two
separate entities exist at the level of motor programming. It would be
normal to regard these as separate phonemes.

Essentially, it is all very well for me and others to bandy our opinions
around, but these things can be tested empirically, and really should be,
before we come to any conclusions.

For many speakers, raised schwa may be identified as /I/, and this would
lead presumably to a merger of raised schwa and /I/. But for others (like
me) these two things are not only acoustically, but also perecptually,
distinct.

Mark

Mark Jones
Department of Linguistics
University of Cambridge