Kääntämisen universaaleja uudelleenkäännöksissä

The retranslation hypothesis (RH), postulating a growing closeness to the original as the work is repeatedly translated, rests upon an idea of change, of difference between the first and the subsequent translations. Translation universals, in contrast, are assumed to characterize translations in general, which posits a similarity among them. The claims laid by these hypotheses are thus opposite, and one could expect an analysis of a corpus of retranslations to support only one of the two – either difference or similarity, given that the evidence for the hypotheses is sought on the same level. In the present article, I investigate the validity of both hypotheses with regard to the same corpus of retranslations, consisting of the Finnish (re)translations of Dostoevsky’s Crime and Punishment (see Kuusi 2011). Following Deane (2011), I define the slippery concept of closeness with respect to the specific nature of the original. Based on Bakhtin's notion of polyphony, I argue that the benchmark relevant to closeness is Dostoevsky’s use of free indirect discourse (FID). The results do not confirm RH. Rather, they manifest a similarity among (re)translations, comprising of recurrent weakening or loss of FID that can be linked to the universals of translation.