21:17 Sep 5, 2009 

Couldn't agree more about the proviso! Too many answers on Kudoz are also plagued by clearly uncritical references to existing questionable (sometimes deeply dodgy) translations.

However at the level of Legifrance, and institutions with the kind of gravitas of the IGN, the EN versions of core texts can generally be relied on to have been drawn up with due expertise and care; and the specific para in question here cannot be faulted. Question of knowing which sources to trust, as you say.

20:35 Sep 5, 2009 

Yes, I agree with your point about established translations, with the proviso that they are not always right and, if the target language usage changes, they may need to be updated.

As an example, there is a very expensive English language dictionary of civil engineering with French translations provided of the main terms. This dictionary looks extremely authoritative until you read a bit and see how obvious it is that the author was a Frenchman with a good but very far from perfect grasp of English. I believe that this had led to a number of mistranslations being accepted into the "canon". On the other hand, the number of copies available 2nd hand on Amazon indicates that purchasers of this work swiftly realised its limitations.

18:39 Sep 5, 2009 

Good find. I agree that your source is better when considering the asker's term ("réseaux de nivellement"="levelling networks") taken in isolation, since it is an original native-language text from a clearly authoritative source, not a translation.

My point in referencing the IGN text, however, was equally to highlight that an established, authoritative translation of the passage in question is already available; indeed I take it as a general principle for official texts, especially legislation, that it is preferable (and more professional) to identify and use such established translations where they exist, rather than new translations.

17:36 Sep 5, 2009 

I think this is a better source than the IGN's translation and it supports Martin's case. I couldn't post it as a reference because the button disappeared.

"It should be pointed out that the final OSGM02 geoid patches represent fits to local levelling networks, ... Forsberg, R., M. G. Sideris: Geoid computations by the multi-band ... Symp. on Gravity, Geoid and Marine Geodesy, Tokyo, pp. ..."
www.ordnancesurvey.co.uk/gps/docs/osgm_report.doc