The latest edition of Lebende Sprachen, 2004/2, has just appeared. I am curious about an article by Dirk Siepmann, Siegen, entitled ‘High-profile Translation from the Mother Tongue into the Foreign Language: Effective Translation Strategies and Implications for Translation Theory and Translator Training’.
Lebende Sprachen does not give details of its authors, but here is a link with more information.
One main object of the article is to set out in detail strategies for creating one’s own corpus from the Internet (using what Siepmann calls the deep or invisible web, although I’m not sure that everything he mentions is strictly the invisible web), for instance an architecture corpus that one can search to determine if tympan or tympanum is the right word. Various other strategies for checking one’s English are discussed at length. These strategies are, as Siepmann himself says, useful for native speakers as well as those working into a foreign language. He also discusses intelligent guessing and suspicion of dictionaries.
The article creates a probably unintended impression that the translator would look up every word. I hope the archictecture text would be translated by someone who already knew the word tympanum. These are just examples to show how terms or phrases, or their absence, can be researched.
Another aim of the article is to comment critically on the value of translation theory. Siepmann comments on research based on observing students translating, although students have little experience of translation. And studies of the behaviour of professional translators all ‘abstract away from actual translation problems in their search for higher-order generalities’.
Of course, the article touches on the question of whether one should translate out of one’s native language. My usual answer to that question is that I would expect better results from an experienced non-native specialist translator, for example a legal or technical translator, in his or her field, than from most non-specialist native translators. However, the article gives examples of architecture, an estate agent’s brochure, a power of attorney (DE>FR), and a piece from the Frankfurter Rundschau with a literary touch, among others. If one translator is to translate all these different kinds of texts into the foreign language with a heavy use of corpora, that translator should have a day job (for instance at the University of Siegen?), because the hourly rate is not going to be good!
Incidentally, there is some rather peremptory advice on the course of study a translator should undertake:
bq. The non-native who intends to translate into English should work thoroughly and methodically through such textbooks as Smith/Klein-Braley (1985), Friederich (1969) and Gallagher (1982) (preferably in this order, which reflects an ascending scale of difficulty).
(I have my doubts about Friederich, but I have never worked through it).
These ideas are based on the idea that contrastive linguistics is a key to translation:
bq. In the rare event, however, that the client commissions a translation whose function differs from that of the source text, contrastive linguistics and translation science must part company. The relevance of this latter type of translation situation has been somewhat overstated by translation theorists …
In legal translation, many translations have a different function from the original. They are intended to inform a client or be read in a library rather than be presented to a court or read out to a defendant. Some books, such as Sarkevic’s, concentrate on, say, bilingual legislation in Canada or Switzerland – for me, that is work done by lawyers, not translators, in that it involves double and simultaneous drafting. Translating EU materials into new languages is a slightly different case, but somewhat different from my daily work.
Siepmann’s English is excellent. I imagine his French is too. I looked at the German Vollmacht (power of attorney) translated into French:
bq. Sometimes, however, it is not necessary to have recourse to an entire corpus, as in the case of highly stereotyped text types such as powers of attorney, where the download of two or three sample texts will usually do the trick.
Ha! This is highly suspicious. We are translating one legal system into another. Parts of a power of attorney may work as boilerplate, but others may have to be expressed in a more roundabout way. This is the corpus / Internet search problem of legal translators: just finding a phrase that sounds right is not necessarily good.
I will quote the end of the translation:
bq. Weitergehende Rechte sind mit dieser Vollmacht nicht verbunden.
Aucun autre droit n’est accordé au titre de la présente.
Der deutsche Text ist nur als Grundlage der französischen Übersetzung zu sehen.
Bei Meinungsverschiedenheiten über deren Auslegung ist ausschließlich der französische Text maßgebend.
En cas de différend [typo?] sur l’interprétation de la présente, seul le texte français (et non pas le texte-source allemand) fera foi.
This is quite surprising. Personally, I have never translated anything that is binding only under the target-language legal system. My translations have always been explanations of the German law. This approaches practising law without a licence. Perhaps one could risk it with a simple power of attorney, but I would be wary.
There are some isolated words at the bottom: ‘Pouvoir (F) /Procuration (CH)’, possibly meant to be removed or to be placed elsewhere. Of course, if one were working from DE to FR in Switzerland, the ground might be safer.
P.S. Many translators get very angry at the idea of people translating into a non-native language. This is common knowledge. Unreasoned flaming in the comments will be deleted.