Recent work has shown that prosodic information has a number of important effects on human sentence processing, including effects on the semantic integration of clauses (Schafer 1997). In an effort to learn more about the nature of these prosodic effects, we ran an experiment that manipulated not only prosody, but also the type of initial subordinating conjunction, which we predicted might have a similar effect on processing. Our results confirmed previous work on the effects of conjunction type (Townsend & Bever 1978; Townsend 1983), and they replicated the basic pattern of the prosodic effect from Schafer (1997). However, a more in-depth analysis suggests (i) that prosody and conjunction type are playing different, interacting roles in this "makes sense" decision task, and (ii) that the prosodice effect may be more directly dependent on the gradient phonetic feature of phrase-final lengthening, rather than on an abstract phonological prosodic category.
|Number of pages||20|
|State||Published - 1 Aug 2005|