schwa-raising - formants

[Cecil Ward]

----------------------------Original message----------------------------
To clarify; when I raised this matter in my original post, I was very much
concerned only with deviation which is sufficiently marked that it "crosses
the gap between phonemes", if you understand me, from the viewpoint of an
albeit sensitized hearear who is a native speaker of some English dialect.
So my litmus test is the very hearer's "phonemic reading" of it, if you
like.

That's how Bob Rankin _heard_ his particular word, so that instance passes
my test, regardless of the precise phonetic realisation of Lise Doucett's
pronunciation. And a "phonemic-level distinction is possible in that case,
just as **"pritect" /prI/- and "protect" /pr@/- are distinct possible words.

I first noticed this phenomenon because when only _half listening_ to the
news, hearing a "wrong vowel" (phonemically distinct) made me start, as if
the speaker had selected a non-existent though possible word of English.

Certainly there are some instances of /I/ or /i/ where English orthography
would suggest some other vowel, possibly with an unstressed "schwa-class"
realisation. In my dialect "diprive" and "diny", possibly with /d'i/ rather
than /dI/) seems to be predominant. Similarly -age-words, with /I/ such as
villige, tonnage, footage, storage. However, these don't count for my
purposes, because there is no "deviance" from the norm, that _is_ the norm.

Taking Bob on trust on the cimunicate example, I don't see it as being
clearly be predictable by rule though, and so can't see it as an allophone.

Cecil Ward.