Throughout my career in the language industry, I have always heard about quality. One agency promises to provide the “best quality”, another client complains about the “poor quality” of the translated text… But what exactly is quality in translation?
The first answer that comes to mind when asked such a question is that a good translation is one that respects the source text. It’s great! It’s hard to disagree… although one can argue about the sourcing approach… but we’re not making much progress. It is a bit like a snake biting its own tail, and the question of quality, or rather, of what a good translation is, remains.
Translation Theories
The various theories of translation have been trying to answer this question for a long time, but they go in different directions without establishing an absolute truth, while each one contributes its own stone to the building.
The Tel Aviv School
The Tel Aviv school is a sociological approach to translation and it tells us that it is the social framework that defines what is translatable and what is not. While I cannot deny that the translator’s socio-cultural background is of considerable importance (it greatly influences his or her specialties, vocabulary, style…), the very idea of defining elements as untranslatable is repugnant to me. This also poses a big problem of quality when dealing with this type of material: non-translation, omission..
The interpretative trend
This is the famous theory of meaning. It implies that the meaning conveyed is paramount. Language only has a role in conveying the idea and the letter can therefore even be a hindrance to communication. While in a perfectly theoretical way I agree with this statement, the fact remains that the word cannot be forgotten. The letter is sometimes more than important: after all, clients do not give us glossaries for our own pleasure. Except for a few neologisms and acronyms, you can usually look up what you need in a dictionary. But to try to translate a slogan or a poem with just the idea and without the words is to betray the source text. And no: translating is not betraying!
The hermeneutical approach
Based on the work of Georges Steiner, this approach considers translation as an exact art and not as a science. This approach focuses on the author’s “will to say”, which the translator must restore. So she keeps the idea in mind, but does not dismiss the word. On the other hand, by focusing on the “meaning” it can include a bias since the translator is then interpreting the author’s thoughts with the risks of over-translation that this can generate.
The linguistic approach
This is the theory that tells us that any text must be considered in terms of the fundamental units of word, phrase and sentence. There is no denying this truth, our CAT tools push us to consider the word (through glossaries and as a measure of volume) and the sentence (the segments) as the basis of our work. But if you take this approach fully, it is also very problematic: what happens to the context and the undertones? Which synonyms should be used in which cases? If word transposition were sufficient, machine translation would no longer be a problem and statistical approaches would do very well. Modern engines, based on neural networks and Deep Learning, are now implementing sentiment analysis to improve translation, but this cannot be seen through a minimal unit like the word.
The literary approach
This approach makes the translator an author. Translation is no longer a linguistic operation, but a literary artistic operation. The words here are charged with energy, and it is this energy that the translator must restore. While this is true enough for editorial and literary texts, it will be difficult to adapt this idea in a technical or legal context. In any case, this approach provides us with little guidance for ensuring ideal quality and seems to rely more on a kind of instinct and inspiration which are certainly very present in reality, but which cannot be included in quality processes.
The semiotic approach
In this approach, three elements must meet: a sign, an object and an interpreter. It is therefore more or less a summary of the other approaches: the sign being the word, the object: the meaning conveyed and the interpreter being the translator with his prism of cultural adaptation. As you can see, I tend to favour this theory, although the others are not to be dismissed and this one is also incomplete and still does not solve our quality problem.
The commercial approach
Don’t look for a bibliographic reference, this approach is not academic, it comes from my imagination, or rather from my experience. Why? Simply because we are talking about professional translation and in this context the only thing that matters in the end (apart from finances) is customer satisfaction. So I say it loud and clear: a quality translation is a translation that suits its client!
While all of the above approaches have their positive and negative elements, the recipient is usually forgotten and, as with any product or service, it is the recipient who counts. It is therefore a matter of adapting the translation to the client’s requirements so that it is considered to be of high quality. This is obvious, but we often tend to forget it. Not all customers want the same thing, so there is not one quality but many qualities, the only valid and good one being the one expected by the customer at a given moment.
The person who simply wants to understand the document, the one who wants to adapt it, the one who wants to render it simply or accurately using the vocabulary of the brand for which he or she works… they may all start from the same text, but they have very different expectations. In each case, a different translation may be produced and judged to be good and of high quality, as it meets the client’s expectations.
This is why I insist on having as much information as possible about the context and use of the translation in order to draw up a proposal. My aim is simply to provide the most suitable quality.
What do you consider to be a good translation?