Academic translating means, more or less, translating in order to make it sound bad. I am not joking. For quite a while now some of the more important University departments for translation have taught their students to stay as close to a text’s original syntax as possible, in order to make the reader never forget that he or she is reading a translation.Now, I am not party to what goes on in university departments for translation - important or not - although I do have a modest insight into what happens in more practice-oriented translation workshops. But I do think Kehlmann missed the point with this comment. What I hope is going on is that translators are learning to find their own fine line between fidelity and betrayal, and also to question the notions in the first place. There are many times, in my opinion and indeed in my work, when I too think it's wise to retain a semblance of German syntax in English, but there are others when I feel it's asking too much of readers. There are times when I do in fact want them to remember they're reading a translation and times when I want to give rhythm priority over content, or whatever.
One of the places where I think and talk about translation in a practical way is the no man's land translation lab in Berlin. Literary translators, either professional or amateur, old hands, beginners or even those still thinking about trying it out, can come along and bring texts they're working on for us to go through in the group. A while ago one of us was translating Kafka fragments, so we were looking at how on earth to put Kafka into English, with the added difficulty that this was unedited, unfinished Kafka. What on earth did he mean by that, what might that word have meant in 1914 Prague, and just how important is that modal particle? In this case, most of us agreed that the translator Ina Pfitzner needed to stay pretty close to the original syntax most of the time. But there are other Kafka translations - for other purposes - that veer further away, for instance Michael Hofmann's excellent version of "A Country Doctor". What Hofmann does is embellish ever so slightly, teasing out the gruesome humour. Whereas Joyce Crick's translation of the same story is closer to the original and therefore more useful for scholarly readers. Neither of the two is objectively better for an imaginary universal reader.
Anyway, there is a fascinating interview with Michelle Woods - someone who does teach translation theory - at World Literature Today, in which she talks about her forthcoming book on re-translating Kafka.