Laughter in French Spontaneous Conversational Dialogs

Laughter in French spontaneous conversational dialogs Brigitte Bigi, Roxane Bertrand Laboratoire Parole et Langage, CNRS, Aix-Marseille Universit´e, 5, avenue Pasteur, 13100 Aix-en-Provence, France {brigitte.bigi,roxane.bertrand}@lpl-aix.fr Abstract This paper presents a quantitative description of laughter in height 1-hour French spontaneous conversations. The paper includes the raw figures for laughter as well as more details concerning inter-individual variability. It firstly describes to what extent the amount of laughter and their durations varies from speaker to speaker in all dialogs. In a second suite of analyses, this paper compares our corpus with previous analyzed corpora. In a final set of experiments, it presents some facts about overlapping laughs. This paper have quantified these all effects in free-style conversations, for the first time. Keywords: laughter, conversation, analysis 1. Introduction Laughter can be conceived of as a non-linguistic or par- alinguistic event. However with the amount of interest for the conversational speech data, laughter appears as an essential component of human spoken interaction (Glenn, 2003). Frequency of occurrences of laugh items suggests that laughter, like spoken language, is a major form of human vocal communication (Provine and Fischer, 1989). The study (Provine, 1993) is a description of who laughs during naturally occurring conversations (speaker and/or audience). Most laugh episodes were recorded from sub- jects sitting at adjacent tables in dining areas, standing near the observer in a public meeting area or in a queue, or walking on a sidewalk or hallway. Among others, the au- thor observed that speakers laughed significantly more than their audience. He also found that when speakers laughed, they were more often female than male and when audi- ences laughed, they were also more likely to be female. Another study (Smoski and Bachorowski, 2003) has shown that both the sex and familiarity of a social partner influence the amount of laughter produced. The study in (Vettin and Todt, 2004) showed that participants frequently laughed af- ter their own verbal utterances. There are numerous studies showing that conversation in- cludes a substantial amount of overlapping speech, also called “cross-talk”, as for example (Shriberg et al., 2001; Heldner and Edlund, 2010). Furthermore, some studies (Smoski and Bachorowski, 2003; Laskowski and Burger, 2007) show that laughter tends to overlap with laughter. Im- portantly, laughing time and speaking time do not appear to be correlated across speakers (Laskowski and Burger, 2007), and laughter does in fact occur significantly with more overlap than speech; in relative terms, the ratio is 8.1% of meeting speech time versus 39.7% of meeting laughter time (estimated on ICSI meeting recorder corpus (Janin et al., 2003)). More recently, laughter in conversations has been analyzed as partner-specific adaptation and joint vocal action. (Trou- vain and Truong, 2012) focused on two task-oriented cor- pora: HCRC Map Task Corpus (Anderson et al., 1991), and the Diapix Lucid Corpus (Baker and Hazan, 2011) where two people are recorded while conversing to solve a ‘spot the difference’ task. They observed that laugh- ter as a cue for entrainment/convergence is mirrored by a number of laughs of conversational …

Content truncated to 500 words