At 10:19:36 PM on Tuesday, May 18, 2004,
enlil@... wrote:
> Brian:
>> Indeed, a major point of the paper is that it
>> 'undermine[s] any direct phonetic licensing approach to
>> phonology, such as LICENSING BY CUE' (i.e., LBC).
> Does it really try to 'undermine' it
That's a direct quotation from the abstract. It is repeated
in nearly the same words in the last sentence on p. 74,
immediately following the observation (which I believe you
actually quoted in your previous post) that Lezgian violates
a major prediction of the LBC approach. The author clearly
believes that his evidence from Lezgian *does* undermine
LBC, whether he was trying to do so or not.
> or does it simply try to modify it to account for
> Lezghian? I understood the latter.
No. In the final section he looks at several arguments
purporting to show that his Lezgian evidence really is
compatible with LBC theory and rejects all of them. He also
mentions other recent research that 'has demonstrated other
problems with the LBC approach', mentioning in particular
evidence from Eastern Andalusian Spanish and Turkish. He
makes no attempt to modify the approach to account for his
data. What he does do is use comparative data to come up
with a historical explanation of final obstruent voicing in
Lezgian; so far as I can see, his point is that perfectly
natural historical changes can (and in this case did) lead
to synchronic violations of LBC theory, at least as
formulated by Steriade.
>> But if, as the author of the paper holds, LBC isn't
>> tenable, why get excited about matching an LBC option?
> I don't think that LBC is necessarily untenable. Do you?
> What's your opinion?
My knowledge of modern phonological theory barely qualifies
as minimal, but his argument looks pretty good to me. At
the very least LBC seems to need significant modification.
Brian