CC BY
17
РАЗДЕЛ 4. ПЕРЕВОДОВЕДЕНИЕ SECTION 4. TRANSLATION STUDIES
УДК 81'253
А.-Р. Вуорикоски Как важно быть точным
Статья основана на корпусных данных выступлений, составленных автором во время пленарных заседаний Европейского парламента весной 1997 г. Эти данные стали основой диссертационного исследования, опубликованного в 2004 г. Целью работы было выявление новых путей подхода к вечной проблеме оценки качества перевода, в частности, уточнения концепции точности перевода.
Ключевые слова: корпусные данные, качество перевода, концепция точности перевода
Anna-Riita Vuorikoski Importance of being accurate
This paper is based on the corpus of speeches that the author compiled in the European Parliament plenary part-sessions in the spring of 1997. The corpus was complemented by the transcriptions of the same speeches, as produced by the EP services after the meetings. The aim of the Ph. D. research, published in 2004, was to find new ways of assessing the ever present issue of interpreting quality, focusing on the concept of accuracy in particular. This paper wishes to contribute to the discussion relating to the accuracy of Translation in a context that involves the use of Language for Special Purposes.
Keywords: corpus data, translation quality, concept accuracy of translation
1. Introduction
For the purposes of the present analysis, I understand the specific genre of the EU / EP discourse to be an example of a dense use of LSP, not only in the written documents, but in the spoken discourse as well. This claim is based on the fact that legal language is an essential element of the EU genre. Furthermore, legal language necessarily contains not only strictly legal terms, concepts and set phrases, but also a wide range of technical terms and concepts from any subject area that an EU directive will cover. The way the European laws are debated has features and characteristics that justify us calling the discourse a genre in its own right [see 7, p. 73-82). There are numerous examples of the use of LSP in the EU genre, eg. the budgetary debates, involving clearly defined terminology and set concepts.
Before entering a discussion of the Translation accuracy of the EP spoken discourse, it is appropriate to remind ourselves of what the European Parliament stands for. According to its home pages, the EP has three main roles:
- debating and passing European laws, with the Council;
- scrutinizing other EU institutions, particularly the Commission, to make sure they are working democratically;
- debating and adopting the EU's budget, with the Council.
Today, there are 28 Member States with one or more national languages. The number of official languages within the EP is 24. National languages are a most sensitive political issue within the European Union. Consequently, the EP has to accept the following principle:
In the European Parliament, all official languages are equally important: all parliamentary documents are published in all the official languages of the European Union (EU) and all Members of the European Parliament (MEP) have the right to speak in the official language of their choice. It also ensures everyone is able to follow and access the Parliament's.
The above principle presupposes a staff of translators and interpreters. The EU home pages provides the following statement relating to multilingualism within the EU:
With a permanent staff of around 1 750 linguists and 600 support staff, the European Commission has one of the largest translation services in the world. The Commission's interpretation service employs 600 staff interpreters, a pool of 3,000 freelance interpreters and 250 support staff.
Today it is possible to follow EP plenary debates via the internet, even live, in all the official languages, and to listen either to the original speaker or the interpreter.
2. The corpus
The research corpus was recorded in the EP plenary part-sessions while the author was working there as one of the Finnish interpreters recruited by the EP interpreting service. The Strasbourg plenary session was selected as a speech context for studying Translation accuracy in a representative setting.
A number of authors point at importance of the representativeness of a corpus [2, 3]. In designing the corpus, my aim was to collect a corpus that would illustrate some specific characteristics of the EP discourse from the point of view of Translation. Consequently, the corpus covers a wide range of topics that, according to my 'field work', I considered to be representative of the EP plenary session.
The design for collecting the corpus had the aim of following the principles of some of the leading scholars, according to which one of the main uses of a corpus is to identify what is central and typical in the language [5, p. 17, 102]; here we can add the words 'of a certain institution, such as the EP. '
The corpus consists of 122 original speeches, recorded on video tapes; the speeches were delivered in 4 languages (English, Finnish, German, Swedish) by 76 speakers (table 1).
Table 1
The corpus: number of speeches and their distribution across speakers and languages
Language of the original speech Number of speeches Number of speakers Number of words
English 56 35 22 113
German 45 27 16 400
Finnish 11 8 2692
Swedish 10 6 1917
Total 122 76 43122
As indicated above, the debates were selected with a view of having different types of agenda items, such as Oral Questions to the Council, a debate that is on the agenda of every EP part-session, as well as different topics, such as human rights, or antipersonnel landmines. For the distribution of the speeches in each language along the themes and topics, see table 2.
Table 2
The corpus: number of speeches and their distribution across the themes and speakers
Language of the original speech Number of speeches Themes and topics Number of speakers
English 56 12 35
German 45 10 27
Finnish 11 5 8
Swedish 10 6 6
Total 122 33 76
The EP plenary session agendas in 2014 are quite similar to the ones of 1997, the year when the data was compiled. During one day, Monday 14 April, the topics on the agenda cover a wide range of issues: a report on the European Investment Bank; equality between men and women in Europe; plant reproductive material law; the European System of Financial Supervision Review; report on public access
to documents; report on the activities of the Committee on Petitions; report on the future of Europe's horticulture sector; report on the eradication of torture in the world; report on Saudi Arabia, its relations with the EU and its role in the Middle East; and a report on assessing and setting priorities for EU relations with the Eastern Partnership countries.
To illustrate the nature of the EP genre, a sample speech is provided in Example 1. This speech was delivered by a UK Member of the Parliament in the plenary sitting.
Example 1
Foremost amongst our concerns are the report's proposals to merge the common foreign and security pillar into the Community pillar. Labour believes that the second and third pillar should remain intergovernmental but with greater involvement of the European Parliament. Equally, British Labour Members do not accept the desirability of giving the European Union a military competence, although we support the strengthening of the WEU. Mr. President, the Labour Party supports an extension of qualified majority voting but believes there is a need to retain unanimity in areas such as Treaty changes, budgetary policy, taxation, external border controls, and foreign and security issues. It should be noted that Labour supports co-decision for the European Parliament in all policy areas where there is qualified majority voting at Council.
The speech above illustrates the typical EP genre with its references 1) to the report that is being debated (in italics); 2) to the political party that the speaker represents (the British Labour); 3) key words of the report (Community pillar, second and third pillar, intergovernmental, military competence, qualified majority voting, co-decision).
3. Intertextuality between written and spoken texts
While the interpreters' task is to render spoken discourse in one language via the simultaneous oral translation mode into another language, it is important to bear in mind the high degree of intertextuality that is present between the written and spoken texts in the EU. Here I refer to the fact that the majority of the debates concern draft legislation, presented in the form of reports. The written background documents are present in the meetings, and the speakers refer to them either directly, by quoting the headings and codes in the document, or indirectly, by using the key words and concepts of the documents. By key concepts I mean the essential technical terms found in a document. According to one definition, 'the intellectual content of a <...> document is represented by the terminology used in the title, abstract and body of the document' [6].
The background documents, or parts of them, are produced in a number of the EU languages, and therefore they can be viewed as parallel texts that contain a high degree of equivalence between the key words, the terms and the various specific concepts and phrases in the various EU languages.
We may ask, how can an interpreter, being a non-specialist, or a non-expert, recognise the relevant elements of LSP? How do they know which are the key concepts and terms? Intertextuality in the form of the written parallel texts provides reliable support for the interpreters. Within the EP there is a clearly defined meeting routine, as well as plenty of material, including glossaries, both in printed and digital form, and in today's world, computers connected to appropriate links, all of which helps interpreters in their everyday work. Yet, even if printed agendas, printed questions to the Commission or to the Council, and printed reports are available to the interpreters, there is such a huge number of documents that an interpreter does not have the time to read and digest them properly; after all, one report may have up to 100 pages or more, and during one day, something like 10 reports may be discussed, like on Monday, 14th April, 2014. Still, the EP expects their interpreters to render the speeches accurately into their target language:
The interpreter's job is to ensure that speeches delivered in one of the official languages of the European Union are accurately rendered into the other official languages.
What exactly do they mean by 'accuracy'?
4. LSP and interpreting accuracy
Most authors on interpreting include 'accuracy' as an essential criterion of interpreting quality [4, 7, 1]. While the concept of accuracy has been discussed by a number of Translation scholars, there is no one single definition for interpreting accuracy that could be applied to all Translation situations. However, there is general agreement that the correct use of terms and concepts is one of the core elements of interpreting accuracy. This is confirmed by the customer satisfaction surveys that were carried out within the EU under the DG Interpreting in 2007. According to the response to these three surveys, terminology-related issues were clearly identified as something where improvement was needed.
4.1. One approach to interpreting accuracy
We may learn something about the 'real life accuracy' of interpreting by comparing the original speeches with the interpreters' versions in order to analyze the way interpreters render the LSP in the speech context of the European Parliament. After some preliminary pilot studies of the EP data I was convinced that rhetoric theories and argumentation theory offered an optimal tool for analysis. Starting with the classics, such as Aristotle, up to the present day European and American approaches, rhetoric analysis focuses on the sense of a message.
While classical theories of rhetoric have traditionally been used for drafting texts and speeches, today they are also used for analyzing extended texts, such as political speeches, party programmes and editorials. The analysis looks at the speech context, the audience in particular, as well as the way in which the speaker tries to persuade or convince his / her audience. His integrity, or ethos, is reflected, among other things, by his expertise, that is, his command of the topic he is discussing. This, again, is reflected in his use of LSP and the relevant concepts.
The speaker's integrity, and the way it is reflected in his choice of expressions, brings us back to the definition, quoted above, that helps us to locate the terms and key concepts of a text: 'the intellectual content of a <...> document is represented by the terminology used in the title, abstract and body of the document'. Obviously, the subject field of a speech determines the field of the terms and key concepts. Therefore it seemed necessary to pick the key words from each thematic field rather than treat the data as one large discourse entity.
4.2. Method of study
The four-language corpus has SI versions of each speech in one of the languages translated into the other three in order to avoid the problem of the singularity of interpreter performances. As each speech has been interpreted into three languages, on site, on five different dates as part of the normal EP routine, we can assume that this solution has reduced the problem of the inter-individual and intra-individual fluctuations in attentiveness, in cognitive factors, or any other specific individual competence. Furthermore, all the interpreters have passed the recruitment tests of the EP, a fact that should guarantee a standard level of competence. It was not possible to count the exact number of the interpreters that are at work in the recorded part-sessions; a rough estimate is 30 interpreters.
For the purpose of comparing the interpreters' versions with the originals, I started off by aligning the interpreters'versions below the original speech, as demonstrated in table 3.
Table 3
Example of the alignment
Original Now GMO legislation poses a particular problem
Finnish tama geeniteknologia se on erittäin suuri ongelma
[translation] [this gene technology it is a particularly big problem]
Swedish det har utgört ett särskilt problem
[translation] [that has created a particular problem]
German das ist problematisch
[translation] [that poses a problem]
It was not feasible to align each speech with the interpreters' versions in three languages - not single-handed, anyway. My solution was to transcribe each original speech, paying attention to those elements that I considered important from the point of view of first, the political rhetoric, and second, from the point of view of LSP. I listened to the interpreters and marked with a plus (+) the units that they conveyed accurately, and with a minus (-) the ones they did not convey accurately; the ones that were conveyed vaguely I marked with a dash (/).
In order to illustrate this method of analysing the accuracy of the interpreters' performances, the same text as in Example 1 has been aligned according to what was explained above (see Example 2).
Example 2
English original De Sv Fi
Foremost amongst our concerns are + + +
the report's proposals + + +
to merge the common foreign and security pillar + + +
into the Community pillar + + /
Labour believes - vi -
that the second and third pillar / - +
should remain intergovernmental / + -
but with greater involvement of the EP + + +
Equally - + -
British Labour Members - + -
The example shows that interpreters may omit key concepts of the EU genre, as well as elements of political rhetoric that is part and parcel of the genre. The 'Community pillar' is one such key concept that is only vaguely conveyed by the Finnish interpreter. References to the political groups, obviously key concepts as well, are an important element of EP rhetoric. The speech by this MEP is clearly a statement of position of one particular political group, presenting their stance against the positions presented in the report. We can say that the speech is not rendered accurately if such elements are omitted (cf. the EP definition of interpreting).
Key concepts, terms and acronyms abound in all the debates. We may analyse a few debates for their key concepts. The debates:
- on human rights
('human rights violations', 'slave labour', 'trafficking in human beings', 'Krajina', 'cocain seizures', 'independence', 'democracy');
- human rights: Iran, Bhutan, Algeria
('Bhutan', 'Nepalese', 'UNHCR', 'Nepalese-speaking [minority]', 'civil rights', 'demonstrators');
- antipersonnel landmines
('antipersonnel mines, 'APMs', 'CCW', 'UN Review Conference');
- on topical and urgent issues: Cuba
('Helms-Burton Law, 'Miami Connection');
- European programmes
('RETI, 'Committee of the Regions', 'Structural Fund [programmes]', 'Assembly of European Regions')
The proper names, acronyms, place names, political concepts and technical terms in the parentheses are all examples of concepts that were not rendered by the interpreters. In some cases the interpreter has not recognised the proper name, or place name, or acronym, and says something else instead; in place of the 'Helms-Burton Law' we hear 'American Law'; in place of 'Krajina' we hear 'Ukraine'; in place of 'CCW' we hear 'WEU' The number of interpreters' omissions of terms, concepts, proper names, place names and acronyms in this corpus is extremely hight. While omissions are the most frequent type of inaccuracy detected here, another typical solution on the part of the interpreters is to use a more generic expression instead of the specific term or concept used by the speaker.
5. Discussion
What does this tell about the accuracy of interpreting in the European Parliament? We like to try and find explanations to any phenomena that we find in our studies. Here, one of the obvious reasons for the omissions, or for not finding the correct equivalent in the target language, is the rate at which these speeches are delivered. Furthermore, many of the speeches have been written and are read out from script.
Political speech, and the EP speech culture, are genres in themselves, containing characteristics that interpreters may not always be aware of. An accurate rendering of this genre entails an understanding of the expectations of the audience, and the way specific key words, concepts, acronyms, abreviations, proper names, etc. are related to the intellectual content and knowledge structure of the EP genre. And in the case of the European Parliament discourse, an understanding of the function of LSP will enhance the accuracy of interpreting. This is where training comes in, both in the academic institutions as well as in-house.
Having said that, we have to accept the fact that not all speeches can be delivered 100% accurately. In the EU context, there are so many elements that cannot be controlled by the interpreters. However, an awareness of the role of LSP in specific contexts will help
interpreters to carry out an on-line rhetoric analysis of the key elements of a message, and focus their attention on rendering these elements.
Why do I think this is important? Simply because I think interpreters owe it to the speaker; I think their primary task is to convey the speaker's ethos to his audience.
References
1. Dam H. V., Engberg J. Assessing accuracy in consecutive interpreting: a comparison of semantic network analyses and intuitive assessments // Text and translation theory and methodology of translation / ed. by C. Heine, K. Schubert, H. Gerzymisch-Arbogast. Tübingen: Gunter Narr, 2006. P. 215-234.
2. Gile D. Methodological aspects of interpretation (and translation) research // Target. 1991. Vol. 3, № 2. P. 153-174.
3. Kalina S. Strategische Prozessebeim Dolmetschen: theoretische Grundlagen, empirische Fallstudien, didaktische Konsequenzen. Tübingen: Gunter Narr, 1998. 304 S.
4. Pöchhacker F. Introducing Interpreting Studies. London: Routledge, 2004. 252 p.
5. Sinclair J. Corpus, concordance, collocation. Oxford: Oxford Univ. Press, 1991. 179 p.
6. Handbook of terminology management / comp. S. E. Wright, G. Budin. Amsterdam; Philadelphia: J. Benjamins, 2001. Vol. 2: Application-oriented terminology management. 549 p.
7. Vuorikoski A.-R. A voice of its citizens or a modern tower of Babel? The quality of interpreting as a function of political rhetoric in the European Parliament. Tampere: Tampere Univ. Press, 2004. 281 s. (Acta Universitatis Tamperensis; 985).
