box set

[Donald M Lance]

on 2/4/02 5:46 PM, James A. Landau at JJJRLandau at AOL.COM wrote:

> In a message dated 01/31/2002 12:08:03 PM Eastern Standard Time,
> lancedm at MISSOURI.EDU writes:
>
>>> From: Benjamin Fortson <fortson at FAS.HARVARD.EDU>
>> .....
>>> Date: Thu, 31 Jan 2002 09:46:16 -0500
>> ......
>>> Well, even in "iced donut" you're not *really* "voicing halfway through a
>>> stop", since the main difference between voiced and voiceless stops is
>>> voice onset time, i.e., how long after release of the stop closure you
>>> start vibrating your vocal folds: it's quicker in the case of voiced
>>> stops. Thus what one writes or thinks of as a geminate d is no different
>>> from a geminate t until a few milliseconds after the closure.
>>
>> My ear and my speech mechanisms tell me that the /t/ at the end of the
> first
>> syllable is glottalized and the first syllable is temporally short, as
>> English demands, before the voiceless stop. And then, as Ben points out,
> the
>> voice onset follows immediately after release of the stop in the second
>> syllable, thereby indicating a preceding "voiced" stop -- phonotactic rules
>> of English. So the gemination here is emic rather than etic.
>>
>
> Just one question: what about aspirated versus unaspirated stops?
>
> - Jim Landau
>
Aspiration is irrelevant in the example being discussed.  A /t/ has to be
released to be aspirated. In a word like 'tea' the voice onset is delayed
during the period of aspiration.  If one "paused" between 'iced' and 'donut'
then there could be a little puff of aspiration after /aist/ -- same with
final -t in 'donut' -- but linguists tend to note aspiration only when a
following segment is affected.  In 'iced tea' of course aspiration would be
present in the second syllable, but not in 'iced dee' if there were such a
drink.  It is also possible, though not common, for progressive assimilation
to turn 'tea' into 'dee'.  /d/ is both unaspirated and not followed by delay
of voice onset, and you may take your pick as to which feature is redundant.
And /t/ following /s/ is unaspirated, as in 'steed'.
DMLance