Preposition-determiner interactions in a Northeastern Yiddish corpus

This article presents an empirical study of interactions between prepositions and definite determiners in a corpus of Northeastern Yiddish. In addition to the full and contracted forms of the definite determiner observed in other Germanic languages, Yiddish allows silent definite determiners after certain prepositions. While other Yiddish varieties display a full three-way alternation, in this Northeastern dialect the distribution of contracted and null determiners is phonologically conditioned: Null forms occur after nasal-final prepositions and contracted forms occur elsewhere. Therefore, semantically there is only a two-way contrast between full and reduced (i.e., contracted + null) forms. Two types of definiteness licensing that have been proposed to account for the distribution of full and contracted determiners in other Germanic languages—namely, anaphoricity and uniqueness—show a statistically significant but probabilistic influence on the choice between full and reduced forms in our corpus. Other factors, such as the gender and semantic class of the head noun, exhibit much stronger effects on the variation. The presence of modifiers also significantly affects P-D realization, although we find evidence countering the strict categorical claims made in previous analyses of the Yiddish alternation.