GSCL Pre-Conference Workshop
Contrastive Linguistics – Translation Studies – Machine Translation – what can we learn from each other?
September 27th 2011, Hamburg University
Contrastive Linguistics (CL), Translation Studies (TS) and Machine Translation (MT) have common grounds: They all work on the crossroads at which two or more languages meet. Recently, all three of them have shown a strong affinity towards using multilingual (parallel and comparable) corpora. In MT, for instance, parallel data collections serve as training material for translation models, but also for related issues from computational linguistics like multilingual grammar induction, automatic lexicography etc. Translation scholars use corpora striving for empirical models of the translation process (including translation strategies or specific properties of translated text). For professional translators, multilingual corpora serve as reference works, which enable quick and interactive access and information processing. Contrastive linguists use corpora both to empirically base their findings and to uncover differences in linguistic features which haven’t been studied before. Furthermore, multilingual corpora have found their way into lexicography and grammar writing.
Despite their inherent relatedness of interests, methodological exchange between the three disciplines is rare. For instance, when parallel corpora are used in CL or MT, factors like translation direction or translation properties and strategies are largely ignored. Also, especially MT is agnostic of dimensions like text type or register. On the other hand, the use of multilingual annotation and query techniques is often restricted to the most basic techniques in CL and TS – if applied at all.
The workshop aims to bring researchers together who work not only on the crossroads of languages but also of the three disciplines mentioned. It is intended to be a platform for interdisciplinary research to bridge the gaps between the disciplines and to further encourage hybrid machine translation and innovative ways of computer-aided language study and translation.
Hans Uszkoreit (DFKI & Saarland University, Saarbrücken & Berlin)
Oliver Čulo & Silvia Hansen-Schirra (Mainz University)
Michael Carl, Copenhagen Business School
Oliver Čulo, Mainz University,
Germersheim Andreas Eisele, Directorate-General for Translation, Luxemburg
Silvia Hansen-Schirra, Mainz University, Germersheim
Adam Kilgariff, Lexical Computing Ltd, and University of Leeds
Philipp Koehn, University of Edinburgh
Stella Neumann, RWTH Aachen
Sebastian Padó, University of Heidelberg
Reinhardt Rapp, Mainz University, Germersheim, and University of Leeds
Erich Steiner, Saarland University Hans Uszkoreit, DFKI & Saarland University
Špela Vintar, University of Ljubljana
Martin Volk, University of Zurich
Ruprecht von Waldenfels, University of Bern
Andreas Witt, IDS Mannheim
We invite extended abstract (2-4 pages, Times New Roman, 11pt) addressing one of the above listed research issues.
Submission deadline June 5th 2011
Notification of acceptance July 20th 2011
Submission of revised abstract July 20th 2011
Final version November 30th 2011
September 1st 2011
Extended abstracts should be submitted as PDF via the GSCL conference system (http://www.exmaralda.org/gscl/ ). There is a section “Workshop Crossroads” reserved for this workshop. Please submit your abstracts under this section, and please follow the submission guidelines of the GSCL main conference ( http://www.corpora.uni-hamburg.de/gscl2011/en/?Call_for_Papers ).
Once accepted, authors will receive a style sheet and submission details for the full papers. Full papers (8,000 – 12,000 words) will be due November 30th 2011. They will be published in a special issue of the open access journal “Translation: Computation, Corpora, Cognition”.
The journal „Translation: Computation, Corpora, Cognition“ is a new OpenAccess journal hosted by the University of Mainz, to be launched in November 2011. It is aimed to be a high-quality, competitive journal with double-blind reviews and a fast track to publication (with early access to accepted papers).