| Abstract: The article discusses the issue of elements of linguistic convergence in the process of arisal of new languages. In the author’s opinion, the importance of this issue grows in direct proportion to the amount of data we have on “mixed” languages (pidgins and creoles), despite the fact that it is frequently ignored or underestimated by traditionally-oriented scholars of comparative-historical linguistics. It is shown that the existing models of genetic classification, based on morphological and lexical (lexicostatistical) criteria, frequently do not work due to the impossibility of fully taking this issue into account. In addition, this problem is actual not only for originally non-related languages undergoing pidginization, but also for languages that belong to the same family whose divergence was at the same time accompanied by intense mutual contacts (this problem, after A. B. Dolgopolsky’s term, is named “the issue of the genealogical stump”). |