Re: Another Daniels screwup [was: Re: Claims [was: Re: Drifting phonemes [was: Re: The AmE 'o' sound]]]

From: Bob Cunningham (exw6sxq_at_earthlink.net)
Date: 11/16/04 

Date: Tue, 16 Nov 2004 13:44:09 GMT

On 15 Nov 2004 18:43:02 -0800, k_over_hbarc@yahoo.com
(Andrew Usher) said:

> Bob Cunningham <exw6sxq@earthlink.net> wrote in message news:<eplgp0d0s2a8l5v4kqmh7pbvcihlrk3svb@4ax.com>...
 
> > > If [a] is supposed to be a front vowel, I don't think there is a
> > > phonemic difference.
 
> > My understanding is that a traditional Bostonian accent has
> > "cart/cat" pronounced [kat]/[k&t]. To me that pair
> > demonstrates contrastive distribution and shows that [a] and
> > [&] are in different phonemes in that dialect.
 
> OK.
 
Thank you.

> > > This difficulty is, I think, caused by the fact
> > > that the IPA notation for the low vowels is just wrong.
 
> > I don't see how you can say that any symbol assignment is
> > wrong. There is a low front region in the vowel
> > quadrilateral and IPA should be free to call that region
> > anything they want to call it, so long as they define it and
> > apply it consistently. In a similar way there is a region
> > of the vowel quadrilateral that's between open-mid and open
> > and IPA has chosen to call that region [&]. If there is no
> > reason to doubt that a user of IPA means open front when he
> > or she writes [a] and semi-open front when he or she writes
> > [a], how can you say it's wrong?
 
> My 'cot' vowel is central and my 'caught' is back. There ought to be a
> way to phonetically distinguish them. [a] vs. [A] seems to be the best
> way.

You are of course free to use any symbols you want to use,
but if you use [a] for a central vowel, it wouldn't hurt for
you to make it clear that you're not using the phonetic
alphabet of the International Phonetic Association, in which
[a] is a front vowel.

There is a way in Kirshenbaum ASCII IPA to show vowels for
which there are no explicit symbols defined. A low central
unrounded vowel can be written [vwl,low,ctr,urd], or
something close to that. Evan also has a couple of symbols
for ad hoc use. I forget what they are right now but let's
say one of them is "#". An orderly way to handle a low
central vowel is to predefine "#" as [vwl,low,ctr,urd], then
in the rest of that text you're free to use, and unambiguous
in using, [#] for that purpose.

All of that is no problem for me, because so far as I know I
have no [vwl,low,ctr,urd] in my idiolect. I also don't have
[a] ([vwl,low,frt,urd]) in my idiolect, except possibly in
diphthongs.
 
> Of course, the 'caught' vowel is /O/ phonemically, since we are agreed
> that that /O/ is that used in 'caught'.

"We" are not agreed on that. My "caught" is [kA:t], clearly
and unequivocally using a low back unrounded vowel.

> Therefore it is not necessary
> to use /A/ or /A./ in transcribing standard American English.
 
I wouldn't use phonemic notation in any case, but [A] is
very necessary in transcribing American English. My
experience persuades me that [A.] simply doesn't exist in
American English, although there are AUE contributors who
think it does. It certainly doesn't exist in my speech. It
also doesn't appear in Peter Ladefoged's write-up on
American English in the _Handbook of the International
Phonetic Association_. He shows [A] (IPA script "a") near
low back and his low front vowel is [&] (IPA digraph "ae").
He doesn't mention [a], the IPA low front vowel.

Strange to see he has no [O] in his vowel chart, and in his
transcription of "The North Wind and the Sun" he transcribes
"warm" and "cloak" with [o]. He says his discussion is
based on the speech of a 21-year-old female who has lived
all her life in Southern California. I find it hard to
believe that a Southern Californian would pronounce "warm"
and "cloak" with the same vowel. I also can't accept that
the vowels in "warm" and "cloak" belong to the same phoneme.

> > > If [&] is the front vowel (close enough), then [a] should be the
> > > central vowel in my (CINC) 'hot', and then [A] the back vowel in your
> > > 'hot'. This has the advantage that indeed [A.] (= British 'hot') would
> > > simply be a rounded version of [A], as it is supposed to be.
  
> > "Supposed to be" is open to question. According to the IPA
> > specification [A.] is simply a rounded [A], but Ladefoged,
> > in his _A Course in Phonetics Third Edition_ says that
> > turned script "a" (ASCII [A.]) is a lax partner of the tense
> > script "a" (ASCII IPA [A]). Specifically, Ladefoged says
 
> > British English [has] the tense vowel [A] as in
> > "calm, car card" in both open and closed syllables,
> > and they also have a lax vowel [A.] as in "cod,
> > common, con" [...] which occurs only in closed
> > syllables.
 
> I don't think he is using 'tense' and 'lax' correctly

There you go, see. I wonder if any two linguists fully
agree about what "tense" and "lax" mean.

My impression is that Peter Ladefoged's standing in the
linguistics community is such that his choice of words
shouldn't be dismissed lightly.

> (they should not be used to mean 'long' and 'short'.

But he doesn't say simply that. He presents pairs each
comprising a tense vowel and an related lax vowel ("beat,
bit"; "bait, bet"; and "boot, foot"), then he says

   In each of these pairs, the lax vowel is shorter,
   lower, and slightly more centralized than the
   corresponding tense vowel. There are no vowels
   that are very similar in quality to the remaining
   two lax vowels in most forms of American English,
   [&] as in "hat, cam" and [V] as in "hut, come."
   But both of these low lax vowels are shorter than
   the low tense vowel [A] as in "spa."

(As I usually do, I've transliterated IPA symbols to
Kirshenbaum ASCII IPA symbols.)

(Note that, abhorrent as it is to me to put things inside
quotes that aren't part of what's being quoted, I've done so
above because I'm quoting Professor Ladefoged, that's how he
did it, and I want to quote him as faithfully as I can.)

> It's true that the concepts coincide in American speech,
> though.). Lax vowels should be nearer the center of the
> vowel space, right?

There appears to be a consensus in that direction.
 
> > > In American English [A.:] is represented /O/, because we have a
> > > different distribution of back vowels. Failure to appreciate this also
> > > causes people to represent the 'horse/hoarse' vowel /Or/, where it
> > > clearly should be /or/ for American rhotic dialects, at least.

> > I don't think that's right. I've shown with formant
> > analysis that one AUE poster has [or] while I have [Or], and
> > we're both Americans and rhotic.
 
> I don't doubt that this is true phonetically. But phomemically, /Or/
> would be the vowel used in 'caught', which is certainly wrong.

"Certainly"? Dictionaries typically have the [O] symbol in
the pronunciation of "caught". In _Merriam-Webster's
Collegiate Dictionary 11th Edition_ their symbol in "caught"
is an "o" with a dot over it. They state in their "Guide to
Pronunciation" that that symbol is equivalent to IPA open-o,
which is equivalent to ASCII IPA [O].

(They also show that symbol in their only pronunciation of
"saw", which is repugnant to me because in my entire life
I've heard only a very few people pronounce it that way.
I've been used to hearing almost exclusively [sA:]. They
should at least give [sA:] an honorable mention.)

> I note that monophthongal /o/ is also observed before L, so I repeat
> 'hall whole whore', and I notice that 'whore' is slightly lower than
> 'whole', but much higher than 'hall'. This lowering before /r/ is also
> seen with /e/ /i/ and /u/ (as in 'error', 'mirror', and old-fashioned
> 'sure').
 
> Thus /or/ is the correct phonemic transcription.

In my humble opinion, despite my limited interest in
phonemes, the correct phonemic transcription is /Or/.

Relevant Pages

  • Re: Pronunciation symbols
    ... Kirshenbaum's ASCII IPA. ... pronunciation, I would write \!fe>re-\. ... use parens to bracket compound symbols, ... the vowel in 'bite' because the _Collegiate_ has 'i macron'. ...
    (alt.usage.english)
  • Re: The "u" and "v" in older written English is confvsing
    ... Quite a lot of people probably think the IPA does not understand IPA vowel symbols. ... Kirshenbaum has 6 levels -- nothing between lmd and low. ... Spelling in English of borrowings which had /wa/ in local languages can ...
    (sci.lang)
  • Re: Writing badly (a of Madchen)
    ... Much more common are those who can't pronounce English and instead ... misusing and misunderstanding the "Kirshenbaum IPA" ... pronouncing a sound, not to the "stupid" American ... An OPEN vowel is pronounced with the mouth more OPEN and the lips ...
    (alt.usage.english)
  • Re: Another Daniels screwup [was: Re: Claims [was: Re: Drifting phonemes [was: Re: The AmE o sound]]
    ... > phonemic difference. ... > that the IPA notation for the low vowels is just wrong. ... There is a low front region in the vowel ... I suspect what you really have in mind is that American ...
    (sci.lang)
  • Re: Am. Eng. "Nyeh" and "Nyello"
    ... American English when fellow bus passengers talk on their cell phones. ... regional distribution. ...  The vowel tends toward ... rising intonation. ...
    (sci.lang)