The paper addresses Russian pseudo-synonymous possessive constructions U X-a est’ Y ‘X has Y’ with overt verb ‘to be’ and U X-a Y ‘X has Y’ with zero copula. It outlines their possible interpretations, formulates the rules of their usage, and analyzes the relative difficulty of their acquisition by foreign language learners.
The results suggest that the usage of both constructions is regulated by pragmatic rules which determine their information structure and lexical co-occurrence. The construction with overt verb denotes possession of an object and stresses the fact of possession as opposed to non-possession. Hence, the verb easily becomes the focus of the utterance, and the construction itself is used only when speaking about objects which constitute non-trivial possession. Thus, certain semantic classes of objects that everybody possesses, such as body parts (eyes, nose) or other inseparable attributes of people (voice, name, surname, childhood), are excluded. The construction with zero copula has three possible interpretations and can therefore appear with all semantic classes of objects, but with certain classes it requires an attribute: 1) with most trivial entities, such as body parts, it denotes a characteristic of the object of possession and is used with an attribute; 2) with less trivial entities, such as personal possessions (clothes, dishes, etc.), it denotes actual possession at the moment of reference and is used with an attribute; 3) with non-trivial entities (such as palaces, yachts, etc.), it denotes permanent possession and its ability to be used without an attribute correlates with the degree of nontriviality of the object of possession.
The error analysis of the RULEC corpus demonstrates that the zero copula construction presents a greater challenge for non-native speakers: the number of errors in the use of the zero copula construction exceeds that of the overt verb construction by 2.5 times. It suggests that the ease of acquisition may correlate with the presence of negative rules: if a certain class of lexical items (such as body parts) is entirely forbidden, it reduces the number of wrong usages. Thus, the least number of errors occurred within a forbidden semantic class — inseparable attributes of a person (‘similar voices’, ‘small family’, ‘name and surname’).