A question in the theory of translation has occurred to me: some words truly do not exist in certain languages, i.e. cannot be adequately expressed in a certain foreign language, but only approximated, or dropped, or described. For instance, the Hebrew words 'davka' and 'harey' (the particle, not the genitive plural of 'mountain') have no English counterparts, and are notoriously difficult to translate. This is obvious and well-known.
Naively, in the case of a text translated from English into Hebrew, we should not expect to find a single instance of either of these two words. But as a matter of fact, such translations do include these words, of course, even if (as I guess they do) in a lesser frequency than in original Hebrew texts. My question is roughly this: what are the conditions in which a translator would choose such words in translation? What sort of constructs or expressions in the source language typically produce such choices?
I figure it's been studied already. I'm just curious.
