ABSTRACT
The interest-relativity of explanation may make translation easier at the beginning (if Karl says ‘gavagai’, and the relevant cue is always a rabbit, we translate ‘gavagai’ as rabbit without worrying about the possibility of such translations as ‘rabbithood again’ or ‘undetached rabbit-part’); but if what we are doing is forcing our pattern of thought on a linguistic scheme which doesn’t have structural similarities to ours, one which doesn’t ‘map’ onto ours under any mapping that is ‘natural’ (to us), we would expect to pay a price later, when we try to extend our first translations of isolated utterances to a scheme for interpreting the whole language.
