reversal of merger

[manaster at umich.edu]

----------------------------Original message----------------------------
The claim that mergers can be reversed apparently started with
Halle and became one of the main arguments by generative
phonologists against phonemicists in the 1960's.  I discussed
this topic in my 1981 dissertation, which was intended as
broadbrush argument that generative phonology was wrong on
almost all points at issue a decade or two earlier.
I think it is easy to show that all the classic examples
were hopelessly wrong, and in fact most have been withdrawn.
Kiparsky for example debunked Postal's example from Mohawk,
Chomsky and Halle in SPE quietly take back what halle had said
earlier about early modern English.  I recently published in
Linguistique africaine a careful analysis and rebuttal of a
classic early example proposed by Paul Newman from Tera.
Etc., etc.
 
Of course, just as in syntax, today very few people seem
to know or care what teh original motivation for the
paradigms we take for granted was, and so it is an
uphill battle.  But one day truth will out--maybe.
 
Alexis MR