I have a few experiments examining the effects of repeated mention and predictability on the production of prosodic prominence. Previous research has suggested that both repeated mention and predictability can affect prosody; however, these two factors are often confounded. Typically, words that are repeated are also the words that are most predictable. My research attempts to examine these to factors separately by setting up an experiment whereby repeated mention is what is unexpected. Early results suggest that both predictability and repeated mention have separate independent effects on prosody. The next question is where these effects are realized in the production stream. They could be referential, lexical, or segmental in nature.
I am interested in how people make use of information about the perspective of their interlocutors. While listening to or constructing an utterance, a person can either take an egocentric approach or consider his/her interlocutors' perspective. There are advantages and disadvantages to both approaches. In speech production, taking an egocentric approach may sacrifice clarity for ease of processing. Similarly, taking the listener's perspective may improve clarity by increasing processing load. My current studies attempt to test the hypothesis that difficultly in taking perspective may arise from suppression of perceptual information.