January 27, 2014 3:41 pm
Terminology is not primarily about terms. It is fundamentally about concepts (and, of course, the terms used to denote those concepts). As translators, we have to understand the meaning behind the source text and render that meaning in another language. Most of the texts we translate are a mixture of language for general purposes (LGP) and language for special purposes (LSP), which present different challenges.
To translate LGP requires a thorough understanding of the source language, the subject matter, and the purpose of the original document and of the translation, and the ability to write well in the target language. Information that was implicit in the source text may need to be made explicit in the translation, and vice versa. Translating LSP requires complementary skills. When reading the source text, the translator identifies terms (ie, words or phrases with a precise meaning in the relevant technical field); identifies the concepts those terms denote; and renders
those concepts in the target language by using the equivalent terms. There is generally less variation in the vocabulary available to the translator, as the correct technical term should be used. However, considerations of register still apply – different terms may be used in texts for specialists or for a broader readership (eg, ‘myocardial infarction’ vs. ‘heart attack’) – and a noun may be replaced by a pronoun or even a verb. Therefore, a technical term appearing in a source sentence may not appear in the translation and vice versa.
Translation memory (TM) software helps translators work quickly and efficiently, by comparing every new sentence to be translated with all previously translated sentences. This saves time, as you do not have to re-translate what has already been translated, and enhances consistency. Sentence (or segment) matching is particularly useful for LGP, and concordance searching for LSP. A concordance search will show you every instance of the search string in the context of the original sentence, alongside the translation into your target language, indicating the source document.
When working in a large organisation, such as the European Commission, you can benefit not only from your own previous translations, but also from those of your fellow translators (both in-house and freelance). You can see that a particular translator rendered a particular term in a particular way in a particular document, but the TM won’t tell you why. If you find alternative translations of a technical term, which one should you choose? Did the translator research the term thoroughly and come up with the right solution, or might the translated term be wrong because the translator lacked the time, knowledge or search skills to find the right term?
This is where a terminology database, such as the EU’s InterActive Terminology for Europe (IATE; iate.europa.eu), comes in. Ideally, it should contain one – and only one – entry for every relevant concept. Unfortunately, IATE, which is available freely to the general public, contains many duplicate entries, but we are working on consolidating them.
Unless a term is self-explanatory, the concept it denotes should be defined in at least one language, which should be the anchor language for the entire entry. In IATE, the anchor language is usually English (and sometimes French) for general concepts; the language of the country concerned for country-specific concepts; and Latin for species of animals and plants, which are identified by their Latin names. Unfortunately, the anchor language is not identified as such in the public version of IATE. If definitions are added in other languages, they must define the same concept, and all terms for that entry, in all languages, must denote that concept.
If a terminologist or translator cannot find or write a suitable definition in the time available, a snippet of text – usually a sentence but sometimes several paragraphs – taken from a reliable published source, can be sufficient to identify the concept. For instance, a French text about nuclear power might explain that the coeur is the core of the reactor, whereas a medical text might explain that the coeur is the heart.
In this example, the homonyms come from different fields, so the domain labels (‘nuclear power station’ and ‘medical science’) would differentiate them. In other cases, a clear definition is needed to determine whether two terms are synonyms (and should therefore appear on the same entry) or not (in which case they should appear on different entries, and the difference between the concepts they denote should be made clear).
For instance, is fraud the same as swindling, or embezzlement the same as misappropriation? If not, what is the difference between them? Does évasion fiscale mean ‘tax evasion’ (which is illegal) or ‘tax avoidance’ (which is legal, although perhaps of doubtful morality)?
IATE not only provides you with terms and definitions, it also tells you where they were found. Every term and definition should have a reference to the source that
led the translator or terminologist to include it in the database. In a similar way to Google, IATE is an invaluable search tool, providing information that you would otherwise have great difficulty finding, but it is not intrinsically reliable. You should trust information not because you found it via IATE or Google, but because you trust the original source.
The public version of IATE has 24 languages – the 23 official EU languages, plus Latin – and it will soon have 25, with the addition of Croatian. It was created in 2004, subsuming and replacing the databases previously used by the various EU institutions. The oldest of these – the Commission’s Eurodicautom – was created in 1969, and was made available to external users in 1981, well before the advent of the worldwide web in 1990.
Eurodicautom provided the Commission’s translators with a single access point to its numerous multilingual glossaries, which had previously existed only on paper, thus greatly reducing search times. However, many of those glossaries merely juxtaposed equivalent terms without defining them. Such ‘legacy data’ may still be relevant today, so it is included in IATE, even though it doesn’t meet our current standards for recording terminological information. Some entries have no definitions, and some references (to paper sources predating the web) are so cryptic (because disk space was at a premium) that no one can decipher them. However, the terms may have the same reliability code (3 stars = ‘reliable’) as more recently researched data.
Tips for using the database
Most IATE users specify their target language, but there are advantages to selecting ‘Any’. First, the number of languages available and the amount of data provided are indicators of the quality of an entry. (The hit list provides a preview of the term reference, context, term note and definition fields.) An entry containing all 23 official EU languages and extensive data is likely to be the result of a recent terminology project, whereas an entry with just two languages and little data is likely to have been saved by a translator with little further checking. Entries with 4, 6, 7, 9 or 11 languages may date back to a time when that was the number of official languages.
Second, your target language may be missing but another language may refer to a source document published in the Official Journal of the European Union, which is also available in your language. If you follow the hyperlink to Eur-Lex, you can then switch languages and display the equivalent document in your target language, or view a bilingual display of any two available language versions. This is particularly useful for EU jargon, for which EU documents are the authoritative source. For non-EU-specific terminology, an original document from an authority in the relevant technical field might be more reliable than an EU document, which may be a translation.
Although the hit list gives a preview of certain fields, you need to click on ‘Full entry’ to see all the information available. If you searched for all target languages, you can then click on the code for your target language to move it to the top of the entry (above the source language). If you searched for a single target language, you can then ask to see all available languages. Full entry display provides access to the ‘Feedback’ link, so that you can inform IATE terminologists of any errors or omissions you find, which would be greatly appreciated. IATE will never be perfect, but is improving all the time. Used discerningly, it is an essential aid to translation between EU languages.
Article written by Tim Cooper, senior English Terminologist at the Commission in Brussels.
This article was first published in The Linguist, issue 52,4.
2,743 total views, 1 views today