Re: why can't the BBC

On May 24, 12:27 am, "John Atkinson" <johna...@xxxxxxxxxxx> wrote:
"Peter T. Daniels" <gramma...@xxxxxxxxxxx> wrote...
"John Atkinson" <johna...@xxxxxxxxxxx> wrote:
 "Peter T. Daniels" <gramma...@xxxxxxxxxxx> wrote...
"John Atkinson" <johna...@xxxxxxxxxxx> wrote:
"Peter T. Daniels" <gramma...@xxxxxxxxxxx> wrote...
"John Atkinson" <johna...@xxxxxxxxxxx> wrote:
"Richard Herring" <junk@[127.0.0.1]> wrote...
So what would be a minimal pair for /A/ versus /A:/ ?

Not sure what you're after.

A reason why someone would write /A:/ rather than just /A/.

Because it's a long vowel?
But if it doesn't contrast with a short /A/, then the : is
redundant
and must not appear in a phonemic transcription.
Just like Sanskrit, where e and o pattern like long vowels but
are
never written with macron.
Vowels in English fall into two classes.
There are the six "short" or "checked" vowels which are subject
to
the
phonotactic constraint that they can't occur in stressed
monosyllables
with no final consonants, and which, in South Brit (but not
American
or Scottish English), are characteristically of shorter
duration.
And
all the rest, the 13 "long" vowels and diphthongs, which can
appear
in stressed monosyllables with no final consonant, and which
are
characteristically of longer duration in South Brit.
Allophonic realizations have nothing to contribute to a phonemci
transcription.

Sorry, where did I say anything at all about allophones????
Because there is no contrast between A: and A, there is no phonemic
distinction betwen them, and the : merely indicates that it happens
to
be realized long. The length is not part of its phonemicity.
In other languages in which there is a distinction between long
and
short vowels, normal practice is for the long vowel phonemes are
written
long (e.g. using ":", or a macron, or by writing the vowel
double),
whether or not there is a corresponding short vowel,
In such cases they are not "long vowel phonemes" but simply
"vowel
phonemes," and while indicating the length is necessary in a
phonetic
transcription and wise in an orthography, it is wrong in a
phonemic
transcription.

Grabbing the first couple of texts I have handy and leafing
through
them
to find languages which have long vowels that don't correspond to
a
similar short vowel, I found the following:

Estonian (by Tiit-Rein Viitso)
Khanty (Daniel Abondolo)
Hungarian (Abondolo)
Selkup (Eugene Helimski)
Turoyo (Otto Jastrow)
Turkmen (Claus Schoenig)

All these mark the long vowel phonemes that don't have a
corresponding
short vowel in the same way as they do those that do. I guess all
these
authors are "wrong".
Are they explicitly operating withing a phonemic theory of
phonology,
or are they publishing in a post-SPE framework that, following
Halle,
rejects the level and entity of the phoneme entirely (and, IMHO,
foolishly)?

As is not uncommon in such descriptive works (the Routledge series on
Uralic and Turkic, as it happens), I don't think they ever explicitly
state that they're operating within any specific "theory". However,
they discuss, enumerate, and give lists of what they call "phonemes",
so
it's pretty clear they're not following Halle.
Then they can't be offered as examples of what phonological theorists
do, can they!

I'm not interested in what "phonological theorists" do!  (Well, that's
not true in general, but it's not what's relevant here.)  What's
relevant (to me) is the notations that descriptive linguists who find
the concept of "phoneme" useful (whether or not they subscribe to any
"theory" at all)

That's simply impossible. "Phoneme" is a theoretical notion.

I may be naive about distinctive feature theory (I know I am, so don't
bother telling me!), but I'll try:

There are five open or near-open vowels in RP (and other South Brit),
namely PART/PAT/PUTT/POT/PORT (/A:, &, V, A.,O:/ in Gimson-speak).
Possibly relevant distinctive features for them are long/short,
unrounded/rounded, and front/central/back.  Obviously you can't
distinguish them by one feature alone.  The only set of two features
which can distinguish all five are long/short and front/central/back:

There is no such set of two features, because features are binary.

For five vowels, you need three features.

/&/ is short front
/A:/ is long central
/V/ is short central
/O:/ is long back
/A./ is short back

("long front" is unoccupied, at least for varieties that lack the
BAD/LAD split.)

Yes, I know that for those who insist on binary features
front/central/back involves two features, front/nonfront and
back/nonback.  That's fine by me. (Then we have a minimal set of three
features.)

Who doesn't so insist?

(In Am E, which doesn't have long/short, where the COT vowel is
unrounded, and where (arguably) the STRUT vowel is less open, the four
"open" vowels can be distinguished using front/central/back and
rounded/unrounded -- I express no opinion at all as to whether this is a
good way to do Am E though.)

Anyway, the point of all this (such as it is) is that if one really
wanted to stick to SIGNIFICANT minimal contrasts, one would write
something like /&, a:, a, O:, O/ for RP.  But no one does this sort of
thing (except Smith-Trager?).  

We managed to win WWII on the basis of strict phonemic analyses. See
any of the "Structural Sketches" that emerged from the wartime
program, such as Hockett's Chinese, Bloch's Japanese, Hodge's Hausa
and Serbocroatian, Hall's French, Hungarian, and Melanesian Pidgin
English, Haas's Thai, Bloomfield's Russian, etc. etc.

"All redundancy is expunged" is a nice
principle which is largely honoured in the breech -- by most authors in
their descriptions of most languages -- including RP -- as my
mini-survey above showed.

read more »-

oh, dear
.

Relevant Pages

  • Re: Subtitutes for English /T/ and /D/
    ... Any case in which the vowel did not delete is a case in which the rule ... If the speakers are not articulating a vowel, ... when some phenomenon of interest is indistinguishable from the noise, ... which hard-encodes the sound change into the phoneme inventory is not ...
    (sci.lang)
  • Re: some more Irish vowels
    ... Mary is a label for a certain quality of vowel, ... Pronunciation, and the right column shows a rhotic pronunciation ... phoneme in) a set of words, the last sentence would be nonsensical. ... it indicates phonemes but perhaps intends a phonetic transcription. ...
    (sci.lang)
  • Re: Literary phonetic alphabet
    ... Furry has the NURSE vowel, which we've so far avoided. ... Are you claiming that they have the Mary phoneme (/eh / in T-S), and not the marry phoneme, either in your dialect or more generally in non-mergerers? ...
    (sci.lang)
  • Re: Phonemes
    ... I suggested that vowel length was a phoneme and he objected ... as a syllable feature rather than a segment feature (but it's easy ... languages as a supersegmental (I think it extends beyond syllables) ...
    (sci.lang)
  • Re: Phonemes
    ... >>> produce it as a back vowel, and which I perceive as the same phoneme. ... instead of they are indeed saying another word. ... > reflexes, especially with respect to vowels, while still perceiving ...
    (alt.usage.english)
Loading