Re: Vowel quality // was My new IPA resource

Adrian wrote:
>> Incidentally I wouldn't mind discussing rhoticity, palatisation and
>> labialisation, because such qualities are usually neglected when
>> discussing vowel quality and there's a lot I don't know. I understand
>
Well, I realize you're coming at this from an Australian POV, and I from a
more or less "standard" American POV, and AFAICT there's almost no
one-to-one mapping between the two :-), but anyway....

Rhoticity, palatalization and labialization are more likely, IMO, to be
features of consonants.  Rhoticity in particular seems a slippery term,
since so many sounds can be classified-- in a _phonemic_ analysis, where
experts may differ-- as rhotic.

>1. The /}:/ vowel as in "who", is that labialised, i.e. [}_w:]?
>   It's undeniable that /}:/ shares a quality with [o_w] that I can
>   only describe as 'melodic', and I'm guessing that this quality
>   is equatable with labialisation.
According to the descriptions at www.phon.ucl.ac.uk/home/sampa/home.htm  [}]
is "close central rounded" (barred u, the rounded counterpart of barred i).
So I assume that that [_w] represents a [u]-like offglide/diphthongization?
I wouldn't call it a "labialization feature", since a rounded vowel is by
definition labialized.  (The [h] portion, in a close phonetic transcription,
might well be marked with the labialized diacritic (subscript small script w
thingy IIRC). Personally I see no relationship with [o_w] which is lower and
more back.
>
>2. What is the distinguishing vowel quality of the vowel in "all"?
>   There's definitely something there that isn't an axis of the
>   quadrilateral, but I can't put a name to it.
In most std.US, it's [O] IPA turned c, Cardinal 6, followed by a velarized
[l] which may or may not produce a schwa off-glide on the vowel (only in a
rather slow, citiation form for me).
>
>3. Am I right that in the same way that [u] = [w=] so too
>   [}`] = [r\=] ?
I'd question the [w=] writing-- [w] is an approximant and consonantal, [u]
is a vowel; there seems to be a difference in tongue position. There might
well be a relationship but that would be language specific (as in Sanskrit
where you might have /dwV../ vs. /duC.../) and would be a matter of
phonological, not phonetic, analysis.  There _is_ a subscript diacritic
(inverted breve IIRC) for "non-syllabic" vowel, usually used to show the
off-glides of diphthongs e.g. [a_U].  Some people, as a sort of short-hand
convenience, might write [a] plus superscript _w_, but I think that's
technically incorrect.

As to [}`] etc.-- I don't have (X)SAMPA consonants in front of me so don't
know what [r\=] means.  Is it supposed to be the American stressed vowel as
in "bird"? that's usually [3^]-- which the SAMPA site calls "rhotic open mid
central unrounded" (I'd prefer "retroflexed" to "rhotic", since the tongue
tip is slightly curled back.)  Or the Czech syllabic trill (?) as in "Brno"?
That I don't know.

There is a difference between my "bod" (slang for "body") [bA:d] and "bard"
[bA^:d.] (that's d-dot for retroflexed d)-- the [A] is retroflexed too--
more correctly [A] plus a superscript (offglide) IPA symbol for American r
(upside down r?) for which I don't recall the SAMPA. In any case, the vowel
is noticeably different.   In terms of English, we might call this a
"rhotic" effect, but if those were e.g. French words ("bade", "barde") I
suspect the vowel would be transcribed identically in both [bad], [baRd]--
just a guess. French [R], though a "rhotic" doesn't seems to affect
preceding vowels in the way AmEngl. r does.
>
>4. Describing the quality of a real vowel with the quadrilateral
>   is sometimes like describing the location of the bottom of a
>   volcanic vent with a global positioning satellite!
I don't see why.  The Cardinal Vowels, in particular, have fixed locations
and "authorized" pronunciations.  They can be slightly modified by raising,
lowering, fronting, backing, within their regions.  So a lowered [u] and a
raised [U] are theoretically distinct, though very close. (It's also true
that IPA changes-- when I learned it there was no symbol "reversed e" (SAMPA
@\ )-- I have no idea how that sounds.......)

I'm interested
>   in talking about - and trying to visualise - the topography of the
>   true vowel diagram which exists in multiple dimensional space and
>   has axes for palatisation, rhoticity, labialisation and more.
I think these might be more a question of how one analyzes one's phonetic
data-- is a sequence [...kju...] an instance of palatal-k plus [u], or k
plus palatal-u... or is it just a cluster k-j-u?  Instrumental analysis
might clarify, but IMO  it's more a  language-specific problem. Consider
Russian-- supposedly it's the consonants that are palatalized, but it's
usually shown by the vowel symbols.

There is the problem, too, that some varieties of English are spoken with
greater general muscular tension in the entire vocal tract. (Merkin is
supposedly notoriously slack in that respect-- South African in my
experience is at the other extreme, such that the front vowels, in
particular, sound as if they've been raised one entire IPA cell--
impressionistically, my [æ] is their [E], my [E] is their [e], my [I] is
their [i] etc).

>In such a diagram, palatisation, I expect, would bulge in the region
>   of closed front vowels, rhoticity around closed central vowels, and
>   labialisation around closed back vowels. Confirm, deny or
>   supplement?
Not necessarily, IMO.  You can palatalize or labialize (onset or offset)
just about any vowel.  And if the tongue tip is free to move (as it tends
not to be with the high central vowels, for me at least) you can retroflex
(rhoticize?) most of the vowels. A Tamil-speaking professor I had, spoke
Engl. with pronounced [j] and [w] onsets to initial [i,e] and [u,o]
respectively.  And since, like many Indians, he used his retroflex t and d
in place of our alveolars, the preceding vowels tended to be retroflexed
too.  (Generally an agreeable accent; and impressively, he could distinguish
three tap/trill [r]s-- dental, alveolar and retroflexed.)
>
>The reason for these questions is that I'm in the middle of writing a
>web page on the subject of my pronunciation and I want to get these
>details right.
Might I suggest, for those (like me) who aren't entirely at home in
(X)SAMPA, you include some graphics (handwritten if need be) or pdf's using
_real_ IPA???

Hope this helps.