The Methodological Background of the Ural-Altaic Language Classification

In this paper, it is argued that the linguistic properties and affinities between the so-called Uralic/Finno-Ugric languages cannot be represented in terms of a family tree model. This means that the hypothesized genetic relationship between these languages is not sufficiently argued for. The early scholars of Finno-Ugric language comparison worked with word lists to proof genetic relationships. This was unmotivated, however. But also the sound laws that have been elaborated in Indo-Germanic linguistics cannot be copied one-to-one into Uralic/Finno-Ugric linguistics. The sound laws are not without exceptions in Uralic/Finno-Ugric and cannot be documented due to the lack of old documents. The lexical basic stock of Finno-Ugric cannot be isolated. There are substantial parallels with the Altaic languages, especially with (Old) Turkic. The number of lexical items of the basic vocabulary is not constant in the literature. This means that there is no methodologically well-established system to track down the items of the basic vocabulary. The morphological and grammatical domains display a number of serious divergences showing that a genetic relationship between the Uralic/Finno-Ugric languages is not very likely. From this it follows that the general, common features that certainly exist among the Uralic/Finno-Ugric languages (see the Appendix) can be explained the safest as a case of language contact. Such a theory of Sprachbund will not be elaborated in this paper, however.