Phonology question

[Arnold M. Zwicky]

On Mar 14, 2008, at 1:35 PM, Scot LaFaive wrote:

>> Where doesn't it work?
>
> It seems like the principle doesn't work for some prefixes, such as
> "distaste," but perhaps it isn't supposed to work there. I honestly
> know
> some about it, though phonology wasn't a large part of my program
> and we
> were merely told about the principle and that it works. Are there more
> intricacies about it?

the short version is that the principle works within particular
phonological domains -- for example, it doesn't work in general
between phonological words (so the final [s] of "pass" in "pass
coffee" doesn't syllabify with the following [k], which therefore is
aspirated).  the phonological domains are established in part on
morphological and syntactic grounds.

in particular, certain prefixes in english have secondary stress and
are semantically transparent *and* mark off phonological domains.
this would be the "dis-" of "distaste" (vs. the "dis-" of "disturb"),
which has secondary stress, is semantically transparent, and
constitutes a separate phonological domain (so that the /t/ of "taste"
is aspirated).

final complexity: in casual/fast speech, the domains can be re-
calculated.  you can find renditions of "distaste" syllabified as
di.staste (with a resulting unaspirated allophone of /t/).

arnold

------------------------------------------------------------
The American Dialect Society - http://www.americandialect.org