Linguistics and English Language

Discourse in corpus and experimental data [DisCorX]: bridging the methodological gap

Discourse analysis, understood here as the study of discourse-level processes such as coherence relations and their markers, is a rich multidisciplinary field of linguistic research that has been explored through cognitive corpus linguistics, natural language processing and psycholinguistic experiments (among others). These methods have contributed greatly to furthering our knowledge of the types of linguistic elements that help signal coherence relations, their forms and functions, the factors that influence the use of such signals and their effect on processing and comprehension.

Multi-method approaches are highly fruitful: corpus studies help describe a complex reality and provide hypotheses to be tested in more controlled experiments, where well-defined variables can be manipulated and further cognitive factors can be accessed. The combination of these methods seems to attract more and more attention (e.g. Mak et al., 2013; Zufferey & Gygax, 2015), although it remains fairly limited in the field. This methodological gap is mostly due to the different (and partly incompatible) techniques and requirements that each paradigm involves.

This workshop intends to gather linguists, NLP researchers, psycholinguists and psychologists who investigate spoken and written discourse through a (direct or indirect) combination of corpus-based and experimental methods. The focus is on coherence relations and their signals. The aim of the workshop is to share state-of-the-art research on discourse production and processing. In doing so, methodological issues will be discussed, covering the affordances and limitations of corpus-based and experimental approaches to discourse. We hope that this workshop will be an opportunity to share best practices and to further encourage multidisciplinarity and triangulation of results.

We encourage submissions of abstracts for 20-minute presentations covering the following topics:

  • categories of coherence relations;
  • categories of discourse signals;
  • polyfunctionality, ambiguity, information density;
  • “implicit” vs “explicit” relations;
  • discourse production;
  • processing and comprehension;
  • perception and acceptability.

All papers must involve both corpus-based (including NLP) and experimental methods, at least partly or indirectly. We also invite papers specifically targeting methodological issues:

  • what are the benefits and limitations of corpus-based approaches to discourse?
  • what are the benefits and limitations of experimental approaches to discourse?
  • how (far) can they be combined?

Abstracts can be up to 500-word long (excluding references). Please use editable formats (.doc/.docx) and submit your abstract via Easychair by 30th March 2020.

EasyChair for DisCorX 2020

Confirmed invited speakers:

  • Ted J. M. Sanders (Universiteit Utrecht)
  • Sandrine Zufferey (Universität Bern)

Important dates:

  • 31 March 2020: Deadline for abstract submission
  • 11 May 2020: Notification of acceptance
  • 15-16 October 2020: Workshop


Ludivine Crible


Discourse in corpus and experimental data [DisCorX]: bridging the methodological gap

Discourse analysis workshop

Room G32, Psychology Building, 7 George Square, Edinburgh, EH8 9JZ