Stavros Skopeteas (Bielefeld): From OV to VO: Observables and linguistic structure

Language contact is an interesting empirical situation because it allows insights about the kind of linguistic entities that are transferred between grammars (Matras & Sakel 2007). For instance, a change from OV to VO is reported for several languages in contact situations: Quechua in contact with Spanish, Southern Uto-Aztecan languages in contact with Mayan languages, urban varieties of Nubian in contact with Egyptian Arabic, Karaim in contact with Russian and Lituanian, etc. The linearization of verbs and objects is certainly a reasonable observable but not necessarily the transferred entity at issue. The crucial question is whether changes at this layer are part of a larger change in the structure of the encompassing projections.

This presentation deals with Caucasian Urum, an Anatolian dialect of Turkish spoken on the Small Caucasus (Georgia); crucially, the exposition of the young speakers to Russian increases in the recent years. In the current stage of the language, a difference in the syntax of two speaker generations is observable in speech production (studied in a parallel corpus of narratives by 32 speakers): the most frequent word order in Old Urum is OV, while the most frequent order in Young Urum is VO. In an acceptability study with speakers of these generations we examined the speakers' intuitions with respect to the linearization of constructions of the type [αP α[βP β[γP γ]]]: (a) Time, Manner, Verb; (b) Matrix Verb, Embedded Verb, Object.

The acceptability data indicates that Urum does not change from head-final to head-initial syntax, but from a head-final origin to a grammar with head movement (in the sense of Haider & Rosengren 2003). Furthermore, a comparison with control groups of Turkish and Russian speakers reveals that the observed developments can only be explained if we assume that speakers actively participate to the structure-building processes - even in language contact situations. A relieving conclusion: the impact of language contact is not reducible to the replication of perceived observables.

References
Matras, Yaron, and Jeanette Sakel 2007, Investigating the mechanisms of pattern replication in language convergence. Studies in Language 31.4, 829-865.

Haider, Hubert & Inger Rosengren 2003. Scrambling: Nontriggered Chain Formation in OV Languages. Journal of Germanic Linguistics 15,3: 203 - 267.