Knott's empirical study of cue phrases has resulted in a taxonomic classification of some 150 cue phrases. The taxonomy can be viewed as a partial ordering on the set of phrases, and hence we can view cue phrase selection in terms of scalar implicature, following Hirschberg. We draw out some of the immediate consequences of this view, including the problematic status of the conventional/conversational distinction, and the proper treatment of high-level cue phrases. We conclude by considering how to test empirically some of the new predictions.
|Number of pages||8|
|Publication status||Published - 1996|