EQTIS specializes in quantitative corpus-based and process-based research of language use in different translation modes (written translation, interpreting, audiovisual translation). An important aspect of our research is the use of advanced statistics to measure the effect of different linguistic and contextual factors on translational choices simultaneously.
Research activities
Our research activities are sub-divided in three research strands:
- Corpus-Based Translation Studies We use parallel and comparable corpora to study language use and language variation in different translated genres, which are translated from different source languages. In recent years, we have explored the use of complex statistical models (logistic regression, corrrespondence analysis, hierarchical clustering) to find out which linguistic and contextual factors have a dominant effect on translation behavior. For more information on this strand, contact gert.desutter@ugent.be.
- Corpus-Based Interpreting Studies We are one of the few research centres in Europe who take a corpus-based approach to interpreting studies in an attempt to verify a number of received ideas about simultaneous interpreting, such as factors of cognitive load, effects of interpreting strategies and pragmatic marking, and to explore new dimensions of interpreting research, such as gender, and the comparison between interpreting and translation. For more information on this strand, contact bart.defrancq@ugent.be.
- Process-Based Translation Studies We use quantitative and qualitative findings from empirical research into translated and post-edited language use and investigate relations to both cognitive and social reading and translation process data such as translation onset time, pause behaviour, eye movements, translation competence level, interaction with translation technological tools, translation direction, and language and translation norms. For more information on this strand, contact sonia.vandepitte@ugent.be.
Research tools
Next to our publications and presentations, our team develops corpora, corpus tools and statistical techniques:
- Advanced interface for the Dutch Parallel Corpus: as part of the Comure project, an interface was built to query the Dutch Parallel Corpus. Contact Gert De Sutter for more information and access.
- Profile-Based Correspondence analysis: as part of the Comure project, dr. Koen Plevoets built a profile-based extension to the standard technique of correspondence analysis: https://cran.r-project.org/web/packages/corregp/.