<?xml version="1.0" standalone="yes"?>
<Paper uid="W01-1625">
  <Title>Melodic cues to turn-taking in English: evidence from perception</Title>
  <Section position="1" start_page="0" end_page="0" type="abstr">
    <SectionTitle>
Abstract
</SectionTitle>
    <Paragraph position="0"> This paper presents a study of the effects of syntax and melodic configuration on turn-taking in Southern British English. Using dialogue materials, two perception experiments were carried out. In the first, subjects heard dialogue fragments in which syntactic completeness and melodic contour were systematically varied, and were asked whether they expected a subsequent turn exchange or not. In the second, subjects were presented with short speaker exchanges taken from the same material, and asked whether they thought the first speaker had intended to cede the turn or not. The results suggest that syntactic completion or non-completion is the main factor in predicting turn-taking behaviour. Only one melodic contour, the high level tone H* %, appears to operate as a turn holding device, regardless of whether the utterance is grammatically complete or not. The results of this study were found to be similar to those of a study of Dutch turn-taking.</Paragraph>
    <Paragraph position="1"> Introduction Most studies of the intonational cues to turn-taking have been carried out qualitatively within the theoretical framework of Conversation Analysis (e.g. Wells &amp; Macfarlane 1998, Selting 1996). An exception to this is the study by Ford and Thompson (1996), who found that turnchanges in American English mostly appear when melodic, syntactic and pragmatic completion coincide. Two recent studies of the melodic cues to turn-taking in Dutch (Caspers 2000, 2001) motivate the present study, which uses comparable English data to replicate as far as possible the perception experiments carried out in Caspers (2001). On the basis of the findings for Dutch we expected syntactic completeness to be the overriding predictor of a possible turn change. Where melody has an effect, we hypothesised that, as in Dutch, the high level tone was likely to signal more to come and that no subsequent turn change would be expected (Caspers 1998). We also hoped to gain some insight into possible similarities and differences between the two languages.</Paragraph>
  </Section>
class="xml-element"></Paper>