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

Mark J. Reed skrev:
 > On Dec 13, 2007 5:42 PM, T. A. McLeay
 > <conlang@...> wrote:
 >
 >> Mark J. Reed wrote:
 >>
 >>> Right. Not all Americans have the LOT/THOUGHT merger,
 >>> not even those
 >> who
 >>> have the PALM/LOT merger. This is why there are so
 >>> dadgum many lexical
 >> sets
 >>> in Wells's list; i'm almost certain that no single
 >>> dialect distinguishes
 >> all
 >>> of them.
 >> There's Americans without the PALM/LOT merger?
 >
 >
 > Well, probably some. But I was mixed up.
 >
 > In the general vicinity of "short o" and "flat a", Wells
 > defines six lexical sets: BATH, CLOTH, LOT, PALM, START,
 > and THOUGHT.
 >
 >  In General American English (henceforth GAE), the sounds
 >  of BATH and START are not found in this same vicinity, as
 >  BATH is merged with TRAP, and START is rhotic. So those
 >  sets drop from the list when considering only GAE.
 >
 >  That leaves CLOTH, LOT, PALM, and THOUGHT. In my
 >  particular 'lect, those four are *all* merged into a
 >  single vowel sound, so I get confused about where the
 >  divisions are for everyone else. IIRC the primary
 >  distinction is between LOT and THOUGHT. In GAE, CLOTH
 >  goes with THOUGHT and PALM goes with LOT, so you wind up
 >  with only two lexical - CLOTH/THOUGHT and LOT/PALM
 > - which are further merged into a single set for some
 >   speakers, such as yours truly. (Over in RP, CLOTH goes
 >   with LOT and PALM goes with START, which is why those
 >   are separate lexical sets - they always merge with
 >   something else, just not always the same something else
 >   - like BATH.)
 >
 > Of course these are general statements and therefore bound
 > to have exceptions.

I read 'bout all these things in Wells's original book
th'other day and can confirm you got it all right.

  - LOT and PALM (essentially renamed FATHer in the LPD) are
    one set for all non-Southern Americans.
  - THOUGHT and CLOTH is one set for most Eastern Americans
  - LOT-FATHer and THOUGHT-CLOTH are one set for most Western
    Americans (though Texas apparently counts as Southern.
    Y'all'd know that better than me...)

FYI the spelling-to-pronunciation rules for RP are
essentially as follows:

   - _o_ followed by a consonant letter other than _l, r, u,
     w_ is usually /Q/.
   - _o_ followed by _r_ + vowel letter is usually /Q/.
   - _a_ followed by _l_ may be any of /O, Q, A, &/ -- in the
     case of /O/ usually with silent _l_ (not a very helpful
     rule, I know...)
   - _a_ + _u, w_ is usually /O/, but sometimes /Q/ --
     especially when followed by _r_ + vowel letter.
   - _a_ followed by /f, s, T, r/ is normally /A/.
   - in older RP _a_ followed by /f, s, T, r/ was normally
     /O/ but now it is /Q/ for all young speakers.

My xenolect of English has the fourfold merger, essentially
consciously
acquired since my L1 lect has only three back vowels /u/,
/o/ and a phoneme varying between [Q] and [A], most
conveniently mapped as

   - /u/ = FOOT, GOOSE
   - /o/ = GOAT
   - /Q--A/ = LOT-FATHer-THOUGHT-CLOTH

My other L1 vowels mapped to English lexical sets are

  - /i/ = FLEECE
  - /e/ = KIT
  - /E/ = DRESS
  - /a--&/ = TRAP
  - /8/ = STRUT
  - /3\/ + /r/ = NURSE

Which leaves /y/ and /2/ 'unused' in my English, and the
sequences /Ej/, /Aj/, /oj/ serve for FACE, PRICE, CHOICE,
and I usually pronounce GOAT as [oB_o] but slip often. My
English is rhotic, so NEAR, SQUARE, START, NORTH are /ir/,
/Er/, /Ar/, /or/, of course. Unlike some other Swedish
speakers I can't reliably manipulate vowel length as a
separate parameter -- it is bound up with stress and
syllable structure.


/BP 8^)>
--
Benct Philip Jonsson -- melroch atte melroch dotte se
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   "C'est en vain que nos Josués littéraires crient
   à la langue de s'arrêter; les langues ni le soleil
   ne s'arrêtent plus. Le jour où elles se *fixent*,
   c'est qu'elles meurent."           (Victor Hugo)