1. UCL Home
  2. Instituts de recherche
  3. Institut Langage et Communication
  4. Recherche
  5. Projets en cours
  6. Fluency and disfluency markers
  7. General
  8. Research (Recherche)
  9. Discourse markers as (dis)fluency markers (Crible Ludivine)

Discourse markers as (dis)fluency markers (Crible Ludivine)

IL&C Louvain-La-Neuve, Mons

 

Context

          PhD Candidate : Crible Ludivine, ludivine.crible@uclouvain.be 

Promoters: L. Degand & G. Gilquin

Scientific comity: S. Zufferey & A-C. Simon

Period : 2013-2017

 Discourse markers as fluencemes in FRL1/ENL1

The aim of the present work package is to investigate the specific contribution of discourse markers (DMs) to native speakers’ fluency in English and French, thus adopting a contrastive approach.
We will compare several types of registers in the two languages, on a scale of formal-informal speech (e.g. casual conversations, news broadcast, radio interviews). The different texts in the corpus are distinguished on a number of situational parameters such as the degree of preparation. This factor is expected to have an influence on speakers’ fluency behavior, as unprepared speech usually means that the speaker has to pay more attention to the content of his/her speech than to its form (or, to use Kempen’s (1977) terminology, the focus is on “conceptualization” rather than on “formulation”)

Research questions

    1. DMs do not occur in isolation. Fluent or disfluent stretches of discourse result from a combination of fluencemes. DMs should thus be viewed as one type of fluenceme among others.
    2. DMs frequency and distribution differ throughout the scale of cognitive load (Roberts & Kirsner 2000) imposed on a situation. DMs are very sensitive to the different amount of pressure the context imposes on the speaker.
    3. Unprepared speech requires more intense content planning (Koch & Oesterreicher 2001), which results in more disfluent stretches of discourse (marked by DMs) than (semi-)prepared discourse. In general, frequency of DMs and other fluencemes will vary systematically according to situational parameters.

Hypotheses

  1. DMs do not occur in isolation. Fluent or disfluent stretches of discourse result from a combination of fluencemes. DMs should thus be viewed as
    one type of fluenceme among others.
  2. DMs expressing metatextual functions (e.g. marking boundaries, framing the conversation, foregrounding and backgrounding information) are more
    likely to lead to disfluent passages because discourse planning implies a heavy cognitive load (Roberts & Kirsner 2000).
  3. Unprepared speech requires more intense content planning (Koch & Oesterreicher 2001), which results in more disfluent stretches of discourse
    (marked by DMs) than (semi-)prepared discourse.

Main lines of investigation

  1. Comparison between French and English. We will start by establishing similar functions in the two languages and then see by which markers they are fulfilled. Given the multifunctionality of discourse markers, we expect similar functions in French and English to be fulfilled by DMs that are not usually viewed as equivalents of each other.
  2. Comparison between fluent and disfluent functions. The DMs retrieved from hesitant passages according to the methodology described above will be extracted from the whole corpora, and their use in fluent and disfluent contexts will be compared. It will be investigated whether fluent DMs display typical features that distinguish them from disfluent DMs, and whether these features tend to be similar or different in French and English.
  3. Comparison between prepared and unprepared speech. Unrehearsed speech tends to be characterised by a large number of disfluencies, because of the pressure of online speech production. Prepared speech, on the other hand, is less subject to such pressure, but it can still include markers of (dis)fluency, either to imitate spontaneous speech (e.g. in political speeches, as a rhetorical device) or to create short breaks that will help listeners digest the information. By comparing prepared and unprepared speech in French and English, we want to see how DMs facilitating the speaker’s online planning differ from those that are essentially directed at the hearer. The impact of other situational parameters will also be examined.

Publications

Crible, L. 2014. Reaching cross-linguistic comparability across eight speech situations : a challenge for corpus design. Paper presented at the 7th Biennial IVACS Conference, June 19-21 2014, Newcastle, UK.

Crible, L. 2014. Selection and functional description of DMs in French and English : towards crosslinguistic and operational categories for contrastive annotation. Paper presented at the International workshop « Pragmatic markers, discourse markers and modal particles : what do we know and where do we go from here ? », October 16-17 2014, Como, Italy.

Bolly, C., Crible, L., Degand, L. & Uygur-Distexhe, D. 2014. Towards a Model for Discourse Marker Annotation in spoken French : from potential to feature-based discourse markers. Paper presented at the International workshop « Pragmatic markers, discourse markers and modal particles : what do we know and where do we go from here ? », October 16-17 2014, Como, Italy.

Crible, L. 2014. Analyse contrastive des marqueurs pragmatiques en situation de faible interactivité. Étude pilote sur un corpus d’interviews anglais-français. Poster’s Day, Institute for Language & Communication (IL&C). April 30rd 2014, Louvain-la-Neuve, Belgium.