Abstract
In the Slavic languages, two models of the agreement of the predicate of a relative sentence with the vertex expressed by a personal pronoun are presented, during the relativization of the subject. The predicate can either stand in a personal-numeric form corresponding to a vertex pronoun, or receive a third person corresponding to a relative pronoun person.
The considered corpus data allow us to assert that the Slavic languages are not equally inclined to demonstrate these strategies. Face alignment with the apex is the least typical for the East Slavic languages and Czech.
It was also found that the frequency of face matching depends on the number and case of the vertex pronoun: the plural and nominative are more often accompanied by the use of the vertex-matched predicate form. The paper suggests that the distribution of concordance across languages is partly explained by the presence or absence of their inclination to use zero subjects: the choice of the form associated with the vertex is more typical for languages with frequency zero subjects (in the sample under consideration, for South Slavic and West Slavic).