LING 331 Lecture 35: LING331 December 5th
Document Summary
One response has been to become more skeptical of opacity as correctly described, genuine phonological phenomenon. This has been pursued in: cases where non-neutralization or near-neutralization implies that the output is not opaque, cases where the phonology involved is not productive. /b, d, g/ [p, t, k]/ # /b, d, g/ [b , d , g ]/ # Then [b , d , g ] can be included in the triggers for raising, and the process is not. As far as i know, most claimed phonological neutralizations have not been checked to see if they are really near-neutralizations. The polish example has been empirically attacked by nathan sanders3 as suffering from exceptions and lack of productivity. The same alternation exists in the feminine and neuter genitive, in which the plural is opaque. There are many approaches that have been tried in ot: Constraint conjunction (paul smolensky, robert kirchner, ania lubowicz, and others) Output-output correspondence (katherine crosswhite, laura benua, and others)
