Third European Seminar

Translation Equivalence -
Theory and Practice

Montecatini, Tuscany, Italy
October 16-18, 1997

- under the patronage of the European Commission -

"There is no point in denying the concept of translation. The fact is translation is a necessity on economic and general human grounds. Moreover, the fact of translation is a main challenge both to linguistics and to philosophy."

J.R.Firth: "Linguistic Analysis and Translation"

The Seminar was organised by TELRI in conjunction with the Tuscan Word Centre (TWC) in Pescia-Vellano.

• Aims of the Third European Seminar
• Seminar Programme
• Invited Speakers
• TELRI Partners

Aims of the Third European Seminar

Europe is multilingual and multicultural. The preservation of its national and ethnic identities is recognised as an important value. But European integration is equally important. A prospering Europe needs an information society that gives all its citizens the opportunity to communicate freely across language barriers and that gives them access to all modern electronic communication media.

More than other global regions Europe needs multilingual language technology that helps people to write texts in foreign languages, that makes translating more efficient and that supports information retrieval independently of the language used.

The Seminar will assess the achievements of existing translation software and try to propose new approaches that could radically improve their performance. Our experience so far suggests that we need a huge increase in linguistic knowledge before we can construct software that can deal successfully with natural language semantics, i.e. software that replicates the human faculty of understanding texts.

Translating a text means paraphrasing it in a different language. Translation equivalence is a relationship between translation units in a text and their equivalents in a translation, set up by an alignment process. Sometimes these aligned units are single words, but more often they are multi-word units, collocations, set phrases or even larger semantic units. The tradition of bilingual dictionaries and lexicons is to emphasise equivalence at the single word level and to neglect larger units. In order to identify, describe and align the larger translation units we need the information provided by corpora; comparable corpora in different languages give the data for the identification and description of units, while parallel or translated corpora store the insights of translators.

The Seminar will introduce a variety of approaches that will help us to provide the translation knowledge necessary for a new generation of translation tools. It will present the findings of the TELRI project on the multilingual Plato corpus and offer new ideas for the exploitation of parallel texts. It will demonstrate how the BRIDGE dictionaries can be used as a lexicon of translation equivalents.

(BRIDGE dictionaries are published by HarperCollins COBUILD section. They are a new "breed" of dictionary for learners of English. Based on the monlingual COBUILD learners' dictionaries, they are part-translated into the language relevant for the learners. For instance, the Brazilian Portuguese dictionary published in 1995 (ISBN: 0 00 370943-4) has the English headword and grammatical information, translated definitions and the original English example sentences. In this way, a "bridge" is built between the languages which avoids the oversimplicity of straight "translations" of headwords. Several other languages have been completed (eg Danish) and more are being prepared.)

TELRI, the Trans-European Language Resources Infrastructure, is a network of leading academic language technology institutions all over Europe (including the former Soviet Union). Its goal is to provide service (theoretical input, linguistc data, and computational solutions) for the development of multilingual language technology applications. It is a Concerted Action in the framework of the COPERNICUS Programme of the European Commission. TELRI is coordinated by Wolfgang Teubert [http://www.ids-mannheim.de/lexik/teubert.html]at the Institut für deutsche Sprache in Mannheim, Germany. 

NEWSeminar Programme:

For the Programme of the TELRI Plenary Meeting and Review Session before the Seminar, please click here [http://solaris3.ids-mannheim.de/~norbert/meeting.html].

Thursday, Oct 16

11.00 - 12.00 Young Researchers' Workshop:

11.00 - 11.20 Romana Kralikova, Prague: Tagging of Parallel Corpora
11.20 - 11.40 Inguna Greitane, Riga: Machine Translation and Multilingual Resources for Latvian
11.40 - 12.00 Magdalena Zawislawska, Warsaw: The meaning of the verb ?see? and its Polish equivalent ?widziec? in perspective of frame semantics.

12.00 - 12.30 Hamdam Arzikulov, Samarkand: Turkish Language Automaton - Existing Software and Turkish Language Resources
14.00 - 14.20 Renata Blatna, Prague: Performative Verbs in the Czech and English Version of Plato?s Republic. Text vs. Dictionary Equivalents
14.20 - 14.40 Maria Stambolieva, Sofia: A Linguist's Workbench
14.40 - 15.00 Stoyan Mihov, Sofia: Automatic Extraction of Translation Equivalents from Aligned Corpora of Legal Texts
15.00 - 15.20 Kemal Oflazer, Ankara: NLP/MT activities in Turkey.

15.20 - 15.40 Coffee Break

15.40 - 16.00 Cvetana Krstev, Belgrade: Title to be announced
16.00 - 16.20 Oliver Mason, Birmingham/Dan Tufis, Bucharest: Probabilistic Tagging in a Multi-Lingual Environment: Making an English Tagger understand Romanian
16.20 - 16.40 Ruta Marcinkeviciene, Kaunas: Words that only occur once in a Text: Hapax Legomena

- Here starts the official Seminar Programme -

17.30 Opening Addresses:

Wolfgang Teubert, Coordinator of the TELRI Concerted Action;

Ingegner Ginolo Ginori Conti, President of the Industrialist Association of Florence;

Poul Andersen, European Commission

18.45 Welcome Reception

Friday, Oct 17

 9.00 -  9.30 Ramesh Krishnamurthy, Birmingham: TRACTOR - TELRI Research Archive of Computational Tools and Resources
 9.30 - 10.00 Arne Zettersten, Copenhagen: Aspects of bilingual bridge dictionaries - the case of English-Danish
10.00 - 10.30 Alexandre Zoubov, Minsk: The Principles of Translation Equivalents Choice for Six-Language Pocket Computer Translation

10.30 - 10.50 Coffee Break

10.50 - 11.10 Industrial Software Demonstration 1: "The LOGOS Living Dictionary", LOGOS Group, Italy (Rodrigo Vergara)
11.10 - 11.30 Industrial Software Demonstration 2: TRADOS AG, Switzerland (Iris Jahnke)

11.30 - 12.00 N.N., Pisa: Title to be announced
12.00 - 12.30 Serge Yablonsky, St. Petersburg: RUSSICON Slavonic Language Resources and Software
12.30 - 13.00 Tatyana Yudina, Moscow: Thesaurus on Modern Life in Russia
13.00 - 13.30 Gianluca Mattioli, Bologna: Technical-Scientific Translation: An Industrial Necessity

13.30 - 14.30 Lunch Break

14.30 - 20.00 Visit to the "Opera del Dizionario Italiano" with a presentation of Prof. Beltrami (Director); and to the Accademia della Crusca, Villa di Castello, Florence. After the visit reception at the Tuscan Word Centre, followed by dinner at the nearby country restaurant "Goraiolo".

Saturday, Oct 18

 9.00 -  9.30 Laurent Romary, Nancy: Electronic (Bilingual) Dictionaries and Normalisation
 9.30 - 10.00 Michael Barlow, Houston, Tx.: Parallel Corpora (Working Title)
10.00 - 10.30 Anna Mauranen, Savonlinna: Sense and Form Relations in Parallel Corpora (Working Title)

10.30 - 10.45 Coffee Break

10.45 - 11.10 Daniel Ridings, Göteborg: Using Parallel Corpus in a Translator Training Program
11.10 - 11.30 Tamas Varadi, Budapest: Fishing for Translation Equivalents Using Grammatical Anchors
11.30 - 12.00 Elena Paskaleva, Sofia: Automatic Extraction of Translation Equivalents on Data-driven Platform - Wishes and Reality
12.00 - 12.30 Nancy Ide, Poughkeepsie, NY: Word Sense Disambiguation with Multi-Lingual Corpora
12.30 - 13.00 Mike Scott, Liverpool: Investigating Key Words in Two Languages

13.00 - 14.30 Lunch Break

14.30 - 15.00 Eva Hajicova, Prague: Translation Equivalents: Where neither a Dictionary nor a Corpus helps
15.00 - 15.30 Anatole Shaikevich, Moscow: Distributional Statistical Analysis of Parallel Corpora and Automatic Search for Translation Equivalents
15.30 - 16.00 Maria Stambolieva, Sofia: Context in Translation
16.00 - 16.30 Frantisek Cermak, Prague: Agreement and Disagreement. A Comparative Study of Czech and English in Translation Equivalents in Plato's Republic
16.30 - 16.50 Rema Rossini Favretti, Bologna: Using Comparable Corpora for the Analysis of Legal Language - The Bononia Legal Corpus
16.50 - 17.15 Sergio Viaggio, Vienna: Identity of Content over Equivalence of Form: A View of Translation from the Outer Ends of Communication

17.15 - 17.30 Coffee Break

17.30 Panel Discussion

20.00 Farewell Reception

Sunday, Oct 19

10.00 Florence Sightseeing Tour (optional)

Please note that last-minute changes to the Seminar programme may occur. These will be posted on the Information Board at the Seminar Office in Montecatini.
 
 

Invited Speakers:

Michael Barlow, Rice University, Houston, Tx., U.S.A.

Mattioli Gianluca, COFIMP, Bologna, Italy

Nancy Ide, Vassar College, Poughkeepsie, NY, U.S.A.

Ramesh Krishnamurthy, University of Birmingham, United Kingdom

Anna Mauranen, University of Joensuu, Savonlinna, Finland

Antoine Ogonowski, ERLI S.A., Charenton, France

Daniel Ridings, Göteborg University, Göteborg, Sweden

Rema Rossini Favretti, University of Bologna, Italy

Mike Scott, University of Liverpool, United Kingdom

Sergio Viaggio, Interpretation Section, United Nations Office, Vienna, Austria

Serge Yablonsky, Russicon Co., St. Petersburg, Russia

Tatyana N. Yudina, IS RUSSIA Project, Center for Information Research, Moscow, Russia

Arne Zettersten, University of Copenhagen, Denmark 

Back to Top of Page | Back to TELRI Homepage

Last alterations: October 9, 1997

This page was constructed by: TRS Stephan Fingerle & Joachim Hohwieler GdbR, Ludwigshafen/Rh., Germany