Re: OT: YAEPT: English low vowels (was briefly: Re: Y/N variants (< OT: English a...

ROGER MILLS wrote:
> Mark Reed wrote:
>
>>> As to which vowel I have in "palm", the answer is "none of the above".
>>>  For me, "all" words form a lexical set unto themselves:  All, awl,
>>> ball, Bali, balm (doesn't rhyme with bomb), bawl, call, calm, caul,
>>> Collie, doll, dolly, fall, folly, gall, golly, hall, holly, jolly,
>>> Kali, lolly, mall, maul, Molly, Nall, pall, palm, Paul, Polly, qualm,
>>> Raleigh, Rawls, solly, tall, trawl, vol, wall, y'all.
>>>
> A little adjustment for my sets:
> 1.These have /O/ [O]: All, awl, ball,  bawl, call, caul, fall, gall, hall,
> mall, maul pall, Paul Rawls, tall, trawl, wall, y'all; add Saul. And
> Raleigh, I think (at least most of the time)-- also walk, talk (no L of
> course), balk , caulk, fault (yes L in those)
For the first group, I agree, except that I use [o:] i.e. NORTH/THOUGHT.
The only Raleighs I've known were /r&ili/ i.e. "Rayly".

"Balk" and "fault" have /O/ i.e. LOT. "Caulk" I've never heard before
but could only imagine joins LOT as well. In general,
  o: > O / _lC[-voice]
(i.e. o: > O before l followed by a voiceless consonant).

"Bald" and any similar words have /@u\/ [Ou] i.e. GOAT. In general,
  o: > OU / _lC[+voice]
This could also be seen as just the result of allophonic lengthening
based on voice. Many people (especialy my age and younger) avoid this
change to "bald" to avoid making it a homophone with "bold", or say
/b&ld/ instead.

Neither change happens if there is a morphological boundary in between,
so "balled" is different from bald/bold/bowled.

> 2. These have /a/ most likely [A] for me: bomb, Bali, collie, doll, dolly,
> folly, golly, holly, jolly, Kali, lolly, Molly, Polly, solly (dim. of
> Sol[omon]?), vol (as in U.Tenn?)
These divide into two groups, one with /O/ i.e. LOT and one with /a:/
i.e. START/CALM.

2a. These have /O/ for me: bomb, collie, dolly, folly, golly, holly,
jolly, lolly, Molly, Polly, solly.

2a(1). These can have [O] or [Ou] in them, but the vowels do not
contrast: doll, vol, poll, pole, toll, roll, role, stole, mole, fowl,
bowl. They differ in how I recite them (but never so that there's
minimal pairs), but when I speak in conversation there's no difference.

2b. These have /a:/ for me: Bali, Kali.

> 3. closer to /a/ than /O/ : balm, palm, qualm etc (see previous post re the
> l+C cluster, which maybe should be specifically /-lm#/)
These have /a:/ for me. The "l" is merely orthographic.

> 4. varies between /a/ and /O/: water (/a/ is native, /O/ acquired after
> teasing by posh Easterners at prep school :-((((( )
/o:/.

And some extra categories: These have /O/=LOT for me: multi-, alter,
molt, malt, pulse, false, salt, sultan, sultana

These have [Ou]=/@u\/=GOAT for me: old, bald, Mulder, Malden.

(There was a pun when I was younger about the X-Files: one of the
investigators was called "Mouldy", which is a minimal pair with
"Mulder". I can't actually remember how it was set up, or what happened
to Scully.)

I might have forgotten something interesting; /l/ scares almost all
vowels that come before it. This is just what happens to the
back/unrounded central vowels.

> Amazing the traffic that YAEPTs generate!!!! but it's fun.........
Indeed.

--
Tristan.