Code-Mixing Patterns in a Corpus-Based Megrelian-English Dictionary

This paper examines code-mixing patterns in a corpus-based Megrelian-English dictionary. Megrelian, an endangered Kartvelian language, is characterized by intensive contact with Georgian and, to a lesser extent, Russian. Using data from a morphologically annotated corpus of spoken Megrelian, the study analyzes the structural and functional properties of code-mixing, including insertion, alternation and lexicalized mixing.

The findings show that code-mixing is a systematic and productive phenomenon rather than random interference. Mixed forms are integrated into Megrelian morphosyntax and reflect both linguistic constraints and sociolinguistic factors such as age, region and communicative context.

The paper also demonstrates how such forms can be represented in a corpus-based bilingual dictionary using L2 tagging. This approach contributes to lexicographic practice and provides a more accurate representation of contemporary language use in endangered language documentation.