Prosodic Incorporation of English Utterance Final Vocatives

Beckman and Pierrehumbert (1986) and Pierrehumbert and Hirschberg (1990) claim that final vocatives are prosodically separate from preceding matrix structures. We present data from two experiments on the prosodic incorporation of utterance-final vocatives in American English. One of these, based on dramatic reading of two scenes from a make-believe play, shows that in contexts approximating natural speech, final vocatives are prosodically incorporated into the matrix-structure intonation. By contrast, the second experiment, using decontextualized list readings, shows patterns similar to those of Beckman and Pierrehumbert (1986) and Pierrehumbert and Hirschberg (1990). We argue that final vocatives are not preceded by a prosodic break in natural speech. Further, we propose that dramatic readings are an important new tool in approximating natural speech in experimental situations.