Which makes sense, I guess, there's never such a thing as a perfect unassailable translation - at least not of a work of any complexity. Still, he sounds a bit like an armchair critic.

I liked the point about being unsure what is best, having a good writer for a translator or a less good one. Though it seems like the conundrum only exists because of an insufficiently defined concept of what a "good writer" is, and if one asks the more obvious question of whether you prefer a good translator or a less good one, the answer is in fact easy. And then of course some good writers are bad translators, and the other way around.

I rather doubt if this Herbert's translation was genuinely so much better than the later ones. It's possible, of course, but I don't think one should necessarily interpret Flaubert's reaction that way. It's not as if he could compare it to anything, and no doubt part of the reason why he insisted on Herbert's translation is that she could easily consult him, or he could tell her to change things. Those are advantages alright, but no doubt the others had other advantages. I guess he does say as much at the end, though, that Herbert's translation would merely show us a new way of falling short.