Re: Initial /?/ (was: Number)

Lars Mathiesen wrote:

>> > I suspect there are plenty of languages with /?-/ : /0-/ contrast,
>> > provided they also contrast in other positions. Tonga, Samoa,
>> > Hawaiian to my knowledge. It's certainly audible in the flow of
>> > speech; in list pronunciation there might be a tendency, as in
>> > English, for automatic glottal onset.
>>
>> I kind of doubt this.  I'm just a native English speaker, so my opinion
>> isn't everything, but while I can easily distinguish [pa?e] from [pa.e],
>> I can barely tell the difference between [?a] and [a].  I wouldn't be
>> surprised if languages that almost have that contrast actually do
>> something extra to the vowels beginning without a glottal stop, like
>> beginning with a voiced [h] (IPA heng, hook-top h).
>
>Hmmm. I don't think I can produce an initial heng without a glottal
>stop creeping in before it. On the other hand, to suppress the glottal
>stop I can start with a voiceless, frictionless vowel (i.e., jaw and
>tongue positioned for the vowel, open glottis and pulmonic airflow)
>and then start the voicing.
You are probably both correct ;-)
>
>Anyway, to my ears the 'automatic' glottal stop before vowels in
>absolute initial position is lenis, so perhaps the contrast could be
>maintained by realizing an underlying consonant as fortis initially
>(which can probably only mean aspirated) and leniting it in most other
>positions.
That could well be; and is probably what older, impressionistic grammars are
referring to when they say that initial /?/ "sounds like a little (or
"suppressed") cough".