SUMMARY
The paper focuses on phonological similarities between Uralic languages. The study is based on a dataset which includes 33 word-prosodic and segmental features of 28 Uralic languages or main dialects, including all traditional subgroups of the language family. In statistical analysis clustering and dimension reduction techniques such as multidimensional scaling are applied. This methodology enables to explore distinctive subgroups of languages as well as calculate distances between languages and language groups. As a result we present a quantitative phonological typology. The main division appears between western and central-eastern phonological types of the Uralic languages. The detected phonological subgroups coincide with the traditional ones, i.e. Finnic, Saami, Mordvin, Mari, Permic, Hungarian, Ob-Ugric and Samoyedic. The Hungarian subgroup (Standard Hungarian, Csángó Hungarian) and the Ob-Ugric subgroup (Northern Mansi, Eastern Mansi, Northern Khanty, Eastern Khanty) are internally stable. However, their interrelation and relationship with other groups is ambiguous; according to our results, Hungarian is typologically closer to the Western Uralic language groups (i.e. Finnic and Saami), whereas Ob-Ugric languages form a distinct branch of Central-Eastern Uralic. In general, the results reveal a significant influence of multiple areal connections on the phonological formation of Uralic languages.