| Abstrakt: |
We have seen in this paper that there are a number of cases where a Case-marked phonologically empty element blocks rules with phonological effects. We have also seen instances where the influence of Case-marked trace fails. One can avoid an ad hoc marking of the distinction by observing that the relevant difference between these two types is one that can be incorporated into recent proposals by Selkirk, who assumes that the grammar contains both phonosyntactic phonological rules and phonological rules proper, with a mapping component linking the two blocks of rules. It turns out that the observed partitioning of rules into those that can and those that cannot see Case-marked empty elements coincides with the distinction between phonosyntactic and phonological rules in this framework. No ad hoc markings for rule-type are required, and Selkirk's mapping component defines the breaking point for the relevance of syntactic information to phonological processes. This appears to be a result worth striving for. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink