?
Size matters: towards a critical assessment of size-related effects in gender assignment in East Caucasian
In the vast literature on semantic principles of gender assignment, the effects of size play an important, even if secondary role. Typological surveys of nominal categorization as well as grammatical descriptions eventually indicate that, in individual languages, nouns designating small objects tend to belong to one gender and / or nouns designating big objects tend to belong to another. However, most often it remains unclear what counts as big, or small, which makes it difficult to both critically assess these claims and to apply them in a cross-linguistically uniform way. In this paper, we are trying to operationalize such claims and make them empirically verifiable. After discussing different guises under which size effects on gender assignment may make their way into grammar, we suggest methods to statistically test for their presence. We then apply these methods to the data from Archi, an East Caucasian language for which such claims have been made, as well as to the data from several other languages of the same family.