UC Berkeley

The languages of the world encode possession in a variety of ways. In Slavic languages, possession on the level of the clause, or predicative possession, is represented by two main encoding strategies. Most Slavic languages, including those in the West and South Slavic sub-groupings, use a `have' verb comparable to English have and German haben. But Russian, an East Slavic language, encodes predicative possession only infrequently with its `have' verb imet'; instead, Russian uses a construction for predicative possession originating in a locative phrase, e.g. u menja est' kniga, which literally means `at me is a book' for `I have a book'.

This locative construction for predicative possession in Russian is often singled out as an aberrant construction in Slavic and attributed to contact-induced influence from Finnic languages. The opposite point of view is also put forth: that the locative construction for predicative possession in Russian is the original construction inherited from Late Proto-Slavic and the `have' verb used in other Slavic languages is merely a calque from Greek. Neither explanation is entirely satisfactory. As a matter of fact, early Slavic textual traditions, based on a comparison of textual examples from Old Church Slavic, Old Serbian and Croatian, Old Czech, and Early East Slavic, reveal that both a `have' verb and a locative construction for predicative possession were used in Late Proto-Slavic, alongside a third construction with the possessor encoded in the dative case.

The present-day distribution of encoding strategies in the Slavic languages is explained by tracing the different textual and population histories for multiple areas of Slavdom. Contacts with neighboring languages, especially neighboring non-Slavic languages, over the course of history influenced predicative possessive constructions (PPCs) in all areas of Slavdom. Because the Slavic languages spread rather rapidly over a vast geographic expanse, covering most of Eastern Europe in a matter of a few centuries in the latter half of the first millennium CE, the languages that different Slavic populations came into contact with were often quite different. In particular, languages in the western end of Slavdom were in contact with German-speaking populations to varying degrees of intensity; in the northeastern end of Slavdom, Early East Slavic assimilated and lived alongside large numbers of originally Finnic-speaking populations, who spoke languages closely related to Modern Finnish and Estonian.

In short, each Slavic language expanded usage of one of the three original encoding strategies for predicative possession already attested in Late Proto-Slavic and the encoding strategy that expanded brought it closer to usage in neighboring and historically substrate non-Slavic languages. Not only the form of the PPCs themselves came to parallel usage in neighboring non-Slavic languages, but the morphosyntactic and semantic properties of the Slavic constructions also converged with the PPCs used in areal languages. Additional support for the scenarios put forth in this dissertation comes from examination of factors outside the domain of predicative possession, including linguistic features other than predicative possession, textual histories and considerations of language standardization in different areas, socio-historical factors, and demographic factors. While this dissertation traces the development of one grammatical category - predicative possession - in the history of Slavic, the scenarios outlined are meant to contribute more generally to an understanding of linguistic change in the history of Slavic and how those changes reflect the influence of population processes on shaping the path of historical linguistic change.