Re[8]: [tied] Re: PIE initial *a

At 4:58:01 PM on Thursday, May 15, 2008, Andrew Jarrette
wrote:

> "Brian M. Scott" <BMScott@...> wrote:

[...]

>> In RP the three are distinct: /A:/ in <father>, /A./ in
>> <pot>, and /æ/ in <cat>. (In case you're not familiar
>> with the ASCII IPA symbols, /A/ is script-a, and /A./ is
>> turned-script-a.) Almost all U.S. speakers have the same
>> vowel in <father> and <pot> (the father-bother merger),
>> notated /a/ and typically low central, contrasting with
>> /æ/ in <cat>; in SSE the vowels of <father> and <cat>
>> have merged as /a/ (generally realized, I believe, as
>> [a]), while <pot> has merged with <caught> as /O/.

> I would say that most, but not almost all, U.S. speakers
> have /a/, low central, in <father> and <pot>.

That's what I said. If you will read carefully, you'll see
that I did not say that almost all U.S. speakers have a low
central vowel in <father> and <pot>. I said that almost all
U.S. speakers have the *same* vowel in these two words. I
then went on to say that it is typically low central,
meaning for most speakers.

> And as I've said before, many U.S. speakers realize /æ/
> actually as /Ea/ or close to that, varying to /E:/, in
> _all_ positions

I'd write [E&] or [I&]; this is part of -- indeed, the
trigger for -- the Northern cities vowel shift, a chain
shift in the Inland North dialect.

There are also speakers in the Northeast for whom <can>
'able' and <(tin) can> are a minimal pair, the second being
realized something like [E&] by at least some speakers.

> if /a/ is taken to be low front (as officially in IPA),
> not low central.

If there is just one low, unrounded vowel phoneme, it is
perfectly correct to represent it as /a/, using the least
marked symbol in the right general area. Your 'officially
in IPA' really refers to the phones [a] and [A], which are
a somewhat different matter.

Brian