[EDLING:251] Re: phonology question

[Leo VanLier]

edling at ccat.sas.upenn.edu writes:
>The thing with words like Persian and Fusion is that the voicing happens
>when
>the spelling is with SI, not actually the letters TI.  That's what seems
>to
>make it so unique... and befuddling!

A perfect excuse for procrastination (given that I am late with several
papers...)
Here is my guess - I am sure someone will be able to shoot holes in it.

If the verb ends in /d/ (ex.: include), it will always have a voiced
fricative in the derived noun.
If the verb ends in /t/ (ex.: equate), the fricative will be voiced if
(and only if) the preceding syllable has [+approximant/liquid] [+V] – as
in [wei], [rei], [lei], or vice versa as in [er]. Presumably, in syllabic
structure terms, this combination is "heavy" and thus triggers voicing in
the following consonant, if that consonant is the onset of a new syllable
(e-kwei-zjen) – I won't attempt to insert IPA symbols.
It works not only with [ei] but also with [I] as in deride - derision (of
course there is also a vocalic laxing rule in effect here: from [ai] to
[I] - is this the famous tri-syllabic laxing rule?), [u] as in
allude-allusion and [o] as in erode- erosion.

LvL