That pesky H again (was: varia)

At 11:30 am -0700 4/2/00, Ed Heil wrote:
>BP Jonsson wrote:
[...]
>> No no.  The problem with [h] is that it meets one of the criteria of
>> approximants: **lack of oral friction**.
>
>This kind of negative and vague "criterion" makes no sense.  One
>might as well and as usefully say that [h] meets one of the criteria
>for clicks, which is that it is not produced by oral articulation of a
>pulmonic egressive airstream, or that [h] meets one of the criteria
>for labial stops, which is that it is neither a velar or an alveolar
>stop.
I agree - it kind of makes [h] fit all sorts of categories!

>Surely the relevant "criterion" for an approximant would be: the
>articulators are close, but not close enough to produce turbulence
>(that would make a fricative).
That has certainly been my understanding of approximant also.  But even if
one takes a narrower definition of approximant to mean, as Philip says,
'lack of oral friction', then this is surely complementary to the
definition of fricative given by Dirk: a sound where friction is produced
in the oral cavity.  This means, as far as I see it, [h] and other glottal,
epiglottal & pharyngeal sounds cannot be described meaningfully as either
fricative _or_ approximant!  So what is [h]?

I maybe, I guess, an old-timer, but I can see no reason to depart from what
I have understood both 'fricative' and 'approximant' to mean.

>That would also explain why it is
>unmarked for approximants to be voiced: because an unvoiced
>approximant is virtually inaudible, due to the lack of said
>turbulence.
Amen.

>Under this criterion, [h] is a glottal fricative, as our friends in
>the IPA agree,
And until I'm given a plausible alternative explanation, I see no reason to
disagree with them.

[....]
>
>Granted, but we need not complicate this further by means of muddled
>definitions which obfuscate what might have been simple and clear  :)
But methinks, Ed, some people just hate things being clear & simple?

>>  In Ancient Greek /h/ is/was a prosody "voicelessness of initial
>> vowel",
>
>I did not know this.  Are you saying that the "(" mark did not in
>fact indicate an initial [h]?
I don't know exactly what Philip means, but I say that "(" marked initial
[h], but not initial /h/

I think the ancient Greeks were showing an intuitive awareness of the
function of [h] in their language when they chose to mark it with a
diacritic rather than a separate letter.  It seems to me to more simply
explain (1) many features of the language if [h] is regarded as a prosody
rather than a phoneme in ancient Greek, cf the word for "hair":

       H
Nom: trik-s  --> thriks

       H
Gen. trik-os  --> trikhos

and the verb "to have"

       H
pres. eko:   --> ekho:  (I have)

       H
fut.  ekso:  --> hekso: (I shall have)

A lot of irregularities suddenly disappear  :)

Ray.

(1) I try to split infinitives where-ever possible  :)






=========================================
A mind which thinks at its own expense
will always interfere with language.
                   [J.G. Hamann 1760]
=========================================