I think another, and related, part of the problem, is that project managers ofen have no knowledge of the target language we as translators turn in.

They might use some term mining tool like SDLs MultiTerm Extract to come up with a list of terms that aren´t really terms. I´ve had jobs where everything from prepositions to adverbs to actual terms were included in a termbase I absolutely HAD to follow (otherwise the PM, who had no knowledge of my target language, Norwegian, would come after me with QA-reports stating that I had not used the terminology correctly). Needless to say, complying with term lists such as this makes the language in my translation extremely cumbersome and not exactly fluent and idiomatic.

So I think that a misguided effort to bring as much terminology as possible to the translator might be worse than not giving the translator anything at all. Also, if a review is really just a QA check that is run against a term base, very funny results might be the outcome. Sadly this kind of “review” seems to be taking over – as you point out a review by a (human!) specialist with the target language as a first language, should ideally be the case.

Håvard

I just discovered your blog after checking the cross-references of my blog.

I just wanted to comment that, even though I posted this second “sacred cow” about the importance of translation edition, my latest interview/chat with an industry expert basically proposes the opposite, almost the elimination of reviewers/editors, so we all have all kinds of ideas about this.

Anyway, I am adding your blog’s link to my list of recommended blogs, as I have found some very interesting ideas here, too.

Cheers,

Ivette
http://lapsustranslinguae.wordpress.com/