ABSTRACT
I will take as my starting point the conclusion arrived at by Peter Koch in an article published in the volume on ‘Historical Semantics and Cognition’ (1999) which contains structural and cognitive studies. Koch admits that the ‘paths of semantic change’ outlined by him may have been used but do not have to have been used at all. ‘We have to realize that [the mode of presentation] is totally artificial in relation to what really happens in language change’ (Koch 1999: 296). Such a summary is more or less valid for most approaches to historical semantics. Paths of semantic change are designed on the basis of hypotheses which empirically remain without genuine verification. These hypotheses refer to a notion of language which can be characterized as a myth far apart from what really happens in language use (cf. Harris 1981).
