The aim of the talk is to show that these two major linguistic models, viz. comparative and historical linguistics, share important properties in a number of ways. The main focus here will not be on the differences between them, but rather on their common characteristics. The comparison thus will be willingly simplified as to pay heed exclusively to what they have in common. Evidence for the hypothesis will be provided in a threefold way.

A first part will be dedicated to the what question, i.e. empirical aspects. These will mainly concern data from Romance languages, but some Germanic examples will be discussed as well.

 A second part will focus upon  the how question, i.e. methodological aspects.  Four questions will be addressed here (1) the relation between internal and external factors in language change (2) the "quality" problem of comparative and historical data (3) the question as to how language variation spreads, both in space and in time and (4) the notion of "correlation".

The last and third part will handle the why question, i.e. theoretical aspects. In this section the problem dealt with will be put into "a larger picture" in which a relation is established between linguistic and biological processes.