Multiple-methods approaches to discourse topic structure

It is well known that people convey meaning in conversation not only with their choice of words, but also with the use of prosody: variations in pitch, loudness, and other characteristics of the sounds which can create different shades of meaning or even convey the opposite of what the words alone might suggest.

Current research on how prosody contributes to and/or conveys meaning in conversation suffers from fragmentation between approaches. Research in experimental phonology has identified common structural patterns for the ways pitch (i.e. intonation) and other prosodic characteristics are consistently used within languages, with an attempt to describe how these patterns map to consistent shades of meaning despite the varying lexical context in which they appear. Psycholinguistic studies have investigated the ways in which variations in prosody affect how listeners understand what they hear on a moment-to-moment basis. Studies in the perspective of Conversation Analysis (CA)/Phonetics of Talk-in-Interaction have described in great qualitative detail the ways in which prosodic variation is consistent across similar conversational functions, again with different content. However, these different research perspectives often operate independently of one another. Notably, phonetics-in-interaction research tends to neglect the quantitative aspect apparent in phonetic and psycholinguistic investigations and the attribution of an underlying abstract structure provided by intonational phonology, while phonological and psycholinguistic investigations often ignore the effects of a rich conversational context on the variation they observe.

Experiments carried out as part of my doctoral studies have investigated the role of prosodic variation in signalling topic change in spoken language, bringing together methods from phonetics, intonational phonology, and psycholinguistics. I have already begun communicating this research through participation in a variety of national and international conferences, and I have recently submitted a journal publication reporting some of these results.

With this fellowship, I intend to continue this interdisciplinary approach, bringing together my previous experimental findings with the detailed qualitative approach used in CA/Phonetics of Talk-in-Interaction. During the fellowship year, I would like to gain further training in CA techniques and apply them to my doctoral research. First, I intend to compare the findings of my controlled phonetic production experiment on topic change to the CA findings on similar interactional functions, in order to identify useful points of contact between intonation-phonological theory and CA. Second, I plan to pilot a perceptual experiment combining psycholinguistic methods with conversational data, following up an experiment conducted on more controlled data during my PhD research.

My research is of particular relevance to researchers working in the fields of intonational phonetics and phonology, psycholinguistics, and CA, by providing an interdisciplinary approach to a subject of interest to all of these groups. The use of multiple methods to investigate prosodic variation provides a unique opportunity to advance our understanding of the use of these phenomena in a variety of linguistic contexts and to bring together approaches which have historically been somewhat isolated from one another. The wider question of the means by which speakers and listeners share meaning in conversation may also interest anyone curious about cognition, human behaviour and interaction, and communication in general.

My priorities for the period of the fellowship are as follows: (1) to continue my interdisciplinary approach to the role of prosody in communication by expanding my knowledge in the field of CA, (2) to disseminate the results of my doctoral research through publications and conference attendance, and (3) to write a large-scale grant to support further research.