At 3:26 am -0600 7/3/99, Tom Wier wrote:
>Nik Taylor wrote:
>> > >> [bb] is an implosive bilabial stop /`b/.
>> > >> [z] is a voiceless alveolar click /t!/ (occasionally nasalized in lazy
>> > speech).  It is an implosive alveolar stop /`d/ in some traditional
>> > dialects.
>> Hmm, you have phonemic implosives for bilabial and sometimes alveolar,
>> but it's allophonic for velar?  Unlikely.
>I'm not sure about that -- I mean, if you take the simple case
>of the history of English, we used to have only two phonemic
>nasals, /m/ and /n/, where [N] was merely an allophone of
>/n/ before velar stops.  Some sounds just tend not to be
>phonemic in lots of languages (like [N]).   The point being,
>of course, that just because there's a tendency for features to
>spread to all places of articulation, that doesn't mean they
>have to.

But nasality is somewhat different from implosion. Nasal sounds are very
common as syllable finals; as such they will often be subject to phonemic
variation as they tend to become homorganic with the consonant onset of the
succeeding syllable.  That, after all, is how /N/ began life, so to speak,
in English and it still hasn't achieved separate phonemic status in every
dialect of Brit English.

But the behavior of implosives are surely likely to more analogous to that
of plosives. As far as I'm aware, implosion is more common for bilabial and
dental/ alveolar sounds but palatal, velar and uvular implosives are
certainly known.  But, like Nik, I find the scenario phonemic bilabial
implosives,  alveolar implosives are phonemic in some dialects and that
uvular (I assume that is what post-velar is) implosive is an allophonic
variant of the uvular plosive before fromt vowels an unlikely scenario.

>I find the whole idea of implosives unlikely, but hey, that's
>just me. :)

I think so :)
They are not uncommon in many African languages and I believe they are
known elsewhere.

>> > There are 7 vowels:
>> > [i]                     [u]
>> > [e]=/E/    [eh]=/V/     [o]     [r]=/R/
>> >            [a]
>> > Main questions:
>> > ! Is it naive to call [r] a vowel?  There is no consonant form in the
>> > language.
>> It's not a *true* vowel.  The proper term is syllabic r (assuming you
>> mean the sound in some dialects in watER), but it acts like a vowel.
>But doesn't he say about that /r/ = [R]?  Well, I favor the theory that
>postvocalic /r/ in rhotic dialects of English, at least, is really not a
>at all, but really just the remnants of some former truly consonantal /r/
>that has
>left its mark on preceding vowels.  Much like the loss of truly consonantal
>/n/ in French produced a whole series of phonemically nasalized vowels.

Here I agree with Tom.  The r-colored vowels so typical of American speech
and which occur in quite a few Brit English dialects surely are vowels with
retroflexion which has resulted from the _loss_ of the following consonant
/r/ in an manner entirely analogous to the development of nasal vowels in
French and other languages where the vowel has nasality after the following
nasal consonant has fallen silent.  The ER in 'water' seems to me a
retroflex shwa rather than the true syllabic /r/ that one finds, e.g. in

And indeed Sahla did say that /r/ = [R] which I take to be the uvular r.
If it is the uvular trill, then it can certainly be syllabic in a post- or
inter-consonantal position.

While agreeing with Tom on this last point, I must say I by and large go
along with Nik's comments generally.