AV短视频

Spotlight on... Annarita Felici

Annarita Felici has been an associate professor of the FTI’s Department of Translation since 2014. She is head of the Italian Unit and teaches MA-level courses in practical translation, translation and revision, and legal and institutional translation from German and English into Italian. Her fields of research include translation studies, corpus linguistics and applied linguistics. Today, she will be talking to us about the CHEU-LEX corpus, a project that seeks to examine the linguistic influence of the European Union on Swiss legislation.

Lisez cet entretien en français

What is the CHEU-LEX corpus and how did you come up with this project?

The CHEU-LEX corpus is a parallel and comparable corpus of Swiss and EU legislation in the three official languages of the Confederation (French, German and Italian). Today, my research interests lie primarily in the area of legal and institutional translation and corpus methodology. My PhD dealt with the translation of performative utterances and deontic modality in EU legislative texts. Drawing on this background, I decided to investigate legal language in two multilingual environments that both rely on translation. Initially, I was inspired by the many studies that focus on eurolect, a language variety that emerges from the linguistic impact of the EU on several European national languages. At the same time, I came across some Swiss legal references that highlighted the progressive Europeanization of Swiss law as a result of bilateral agreements with the EU and the ‘autonomous adaptation’ (autonomer Nachvollzug). Thanks to funding from the , I decided to further investigate these topics and also take into consideration the role of translation and the languages that come into contact in the two institutions.

What are the main goals of the project?

The corpus was built for two main purposes. On the one hand, it was conceived to investigate the influence of EU drafting and translation practices on Swiss legislation, following in the wake – and expanding the scope – of similar initiatives in other EU countries and languages. On the other hand, it aims at providing a richly annotated multilingual resource to explore legislative language at several levels (macro-textual, lexical, morphosyntactic) and from different perspectives (monolingual, diachronic, cross-lingual and cross-textual). An additional goal was to implement a parallel resource that also included Italian, which in spite of being one of the three Swiss official languages, is generally less exploited than the other two idioms. As the corpus is aligned in the three languages, it can also be used in translation courses. CHEU-LEX was built on the open source version of Sketch Engine, a corpus management software, and is freely available on the .

Did you encounter any difficulties?

It took us almost a year to construct the corpus and it was quite demanding because we had to process the text in three languages. CHEU-LEX consists of bilateral agreements between Switzerland and the EU from 1972 to 2017, and Federal legislation representing the reception of these agreements, with a total of 792 texts and over 5M tokens in German, French and Italian. The corpus was developed with Professor Adriano Ferraresi, a colleague from the AV短视频 of Bologna-Forlì, and four students (three from Bologna and one from Geneva), who were in charge of text extraction, cleaning, alignment, POS and syntactic annotation. Although we used automatic software, it took us a great deal of time to segment and align the texts, and the corpus had to undergo several revisions and manual corrections to adjust for errors and discrepancies due to the structure of legal texts (e.g. the titles, sections and subsections of several articles), which is not always compatible with the output of the segmentation software. With the parallel alignment, we also had to make adjustments, particularly in German, as the sentence structure of the language resulted in more mismatches compared to the alignment of the French and Italian.

What are the initial results?

We finished constructing the corpus recently and I have taken an initial look at the Italian. The preliminary results do not seem to reveal any kind of relevant EU linguistic impact on the Swiss Italian, at least on the basis of frequencies. I plan to focus on the other languages and more qualitative and quantitative aspects in the months to come.

What other research projects are you working on?

I am also interested in accessible communication and I am working with other colleagues at the FTI on a side project on Multilingual Accessible Communication in Swiss Institutions. Governmental bodies and institutions are increasingly striving to communicate more clearly in the public sphere and most of them are investing in plain language and inclusive communication. We recently got in touch with the Federal Statistic Office to identify suitable ways for providing accessible information in all three official languages and we are working closely with those who produce the texts. Our goal is to exploit plain language in the drafting and translation processes, thus also looking at issues of quality.