On 03/06/06, daniel prohaska <[log in to unmask]> wrote:
> [&] > [&:] => ME staff [st&f] > [st&:f] (further development in the 18th
> century > [sta:f]), similarly: glass, ask, path, laugh;
> [A] > [A:] => ME soft [sAft] > [sA:ft] (further development in the 18th
> century > [sO:ft]), likewise: off, cross, frost, cough, wrath,


From: Tristan Alexander McLeay
"[A]? Not [Q] (or even [O] in the first place)?"

Yes, [A] and not [Q] or [O].
[A] is the reflex of ME /O/ which became unrounded (lowered?) as part of the
second wave of vowel shifts in the late 16th century. So, in effect ME [O]
shifted > [A] (late 16th) and then back > [Q] (late 18th) in England Other
examples: lot, long, gosling, odd, lodge.

The unrounded quality of the 17th century vowel remained in most North
American varieties.


> While length in 'staff'-words is standardised in southern BE and RP, the
> short vowel has been taught in the 'soft'-words, though dialects and
> sociolects retain length, as in <officer> (dial.spel. <orf'cer> ['O:fs@])


"I think this is the norm in cot/caught-distinguishing American English,
isn't it? Except further generalised from what it was in British English?"

Possible, and I have no doubt that this is one reason for the merger in
quality. The loss of distinction was a wider development that set in when
NAE (North American English) lost phonemic vowel quantity.


> [A:] in <father> ['fA:D@] and <rather> ['rA:D@] can only be explained by
> mixed formes:
> ME father [a:] with long vowel > vowel shift (15th c.) > [E:]
> alternating with
> ME father [a] with short vowel > 2.nd vowel shift (16th/17th c.) > [&]
> Result:
> ['fE:D@r] (quantity) + ['f&D@r] (quality) = [f&:D@r] (3. vowel shift 2nd
> half of 18th c. > ['fa:D@(r)])

"Really? Are their dialects which distinguish /&/ and /A:/, "and *don't*
have /A:/ in "father"? "

I'd have to say - I don't know, but it does work the other way round in my
own dialect.
<Father> is a tricky word anyway, because it is frequently used in a
religious context, so a standard pronunciation is easier to displace
traditional dialect forms, what with the familiar word being <dad, daddy,
papa, pap, pa> etc.

My grandfather's Lancashire dialect has ['fe:D@r\], though my generation
says ['fa:D@]. Otherwise we have <bath>[b&T], <path> [p&T], <fast> [f&st],
and <laugh> [l&f] where RP has [A:].

"(I'd always assumed that the long vowel in "father" was just an exception
to the Great Vowel Shift, much like the vowel in "broad"---which is another
assumption of mine---influenced by respect for/the speech of one's father."

An exception, certainly, but one that can be explained by double forms. This
is the reason given in Pinsker "Historische Englische Grammatik" (Munich
1959) and it seems plausible.

It would be interesting to know whether there are traditional dialects that
undergo /&/ + /f s T/ > [A:] and retain [&] in <father>.


"I'd assumed that "rather" was part of the normal &>A: change which just
happened in an unexpected context---not so suprising given the number of
words which didn't change when it would've been expected.)

Since <rather> has the same vowel as <father> both in RP as well as in
traditional Lancashire dialect, I assume it went through the same
development as <father>. But of course it could also be a random change of
phoneme. Given the dialectal diversity of ME and early ModE I don't think
double forms and irregular levelling to be an out-of-the-world explanation.