Raising of unstressed vowels in Eng from schwa to /I/

[Jim Rader]

----------------------------Original message----------------------------
>
> In the US it was a feature of some rural dialects, and still is to an
> extent, but the environment is different.  Word-final schwa (the one
> that picks up an -r in much of the world) is [I] (the same small-cap
> i) in these dialects.  You see it reflected orthographically as <y> in
> attempts to reproduce rustic, period pronunciation:  Arizona -->
> Arizony, sasparilla --> sasparilly, umbrella --> umbrelly, Alabama -->
> Alabamy, and such.  All the cases I can think of are at least
> trisyllabic.  Doing this with "Cuba, sofa, panda" seems very strange
> to me.  There is also an [ae] "ash" that raises to [I] in words like
> Arkansas --> ArkInsas; this may be the sort of thing your SIddam is
> showing.
>
> I tend to think of these as a 19th century phenomenon, but I'm not an
> Anglicist and could be corrected on that.  Nor can I picture how
> someone who says "umbrelly" might pronounce "communicate", so these
> may be entirely unrelated phenomenoa.
>
> Bob Rankin
> U. of Kansas

Final schwa is perhaps a different category entirely.  It has distinctive
permutations in varieties of English on both sides of the Atlantic, such
as hyperrhotacism and "Bristol l,"  that may have to do with constraints
on final schwa or may have some completely different origin.  There is
also a batch of American toponyms/ethnonyms, most acquired from
North American French, that have final [O] or less often [A] and
sometimes [ey] where one might have expected reduction to schwa.
The original vowel in French was presumably usually [a] or [A], which
after English stress placement on a non-final syllable was not reduced.
>From the top of my head, examples of ethnonyms:  Chickasaw,
Choctaw, Quapaw, Chippewa; ethnonyms and toponyms: Arkansas,
Omaha, Ioway (though usually Iowa with schwa), Utah (from Spanish,
presumably), Wichita.  There must be many others.

As for the final vowel of <umbrelly>, note that in some more traditional
"rural" varieties of American English and I believe still in Southern
American English the suffix <-y> was pronounced [I].  I've wondered if
[I] for schwa doesn't involve morphologizing in some cases.  Of course,
[i] in "General American" pronunciation (a convenient fiction, I know)
has traditionally been realized as [I]/schwa in "Missouri" or "Cincinnati"
in "folk speech."  The "proper" pronunciation of <Missouri>, i.e.,
whether final schwa, [I], or [i] was once the subject of furious debate.
There is a well-known article on the matter by the late Allan Walker
Read in the 1933 volume of _American Speech_.

Jim Rader