Spoken language is characterized by the occurrence of linguistic devices such as discourse markers (e.g. so, well, you know, I mean) and other so-called “disfluent” phenomena, which reflect the temporal nature of the cognitive mechanisms underlying speech production and comprehension. The purpose of this book is to distinguish between strategic vs. symptomatic uses of these markers on the basis of their combination, function and distribution across several registers in English and French. Through deep quantitative and qualitative analyses of manually annotated features in the new DisFrEn corpus, this usage-based study provides (i) an exhaustive portrait of discourse markers in English and French and (ii) a scale of (dis)fluency against which different configurations of discourse markers can be diagnosed as rather fluent or disfluent. By bringing together discourse markers and (dis)fluency under one coherent framework, this book is a unique contribution to corpus-based pragmatics, discourse analysis and crosslinguistic fluency research.

Liesbeth Degand, Université Catholique de Louvain:
Without any doubt, with DisFrEn, Ludivine Crible developed the largest, most diverse, and most inclusive spoken corpus annotated at the discourse level. This makes it an invaluable resource for further research, both for theoretical work and (computational) applications. The design of the annotation scheme, which is a topic of continuing research, is a crucial contribution to the field. The detailed way in which she motivates all her methodological decisions and procedures leaves the reader with a very transparent piece of work, making it possible to either replicate some of the studies, or to compare it with related work.