Аннотация
The article discusses the theoretical foundations of the Matrix Language Frame (MLF) model and its application to the structure of Kazakh-Russian code-switching – the phenomenon mixing two or more languages within one discourse without necessarily changing the interlocutor or topic. The analysis is based on approximately eight hours of naturally occurring conversations collected from Kazakh-Russian bilinguals between 2015 and 2020. Using the core principles of the MLF model, the study identifies different types of bilingual constituents, and demonstrates key criteria for determining the Matrix Language in bilingual utterances. The analysis of the selected data identifies instances corresponding to the classic type of code-switching, with Kazakh serving as the Matrix Language. In this role, Kazakh provides systemic morphemes and governs the morpheme order within bilingual phrases. The data analysis highlights the types of constituents used in Kazakh-Russian code-switching: (1) embedded language bare forms – constituents composed of uninflected morphemes from the guest language; (2) embedded language islands – constituents consisting solely of guest language units; and (3) mixed constituents – constituents formed by combining elements from both the dominant Matrix Language and the guest language. The analyzed examples are consistent with the classic MLF pattern, in which the Matrix Language projects morpheme order and system morphemes in bilingual speech. The possible relationship between speakers’ proficiency in the Matrix Language and the predominant type of code-switching was not quantitatively tested in the present study and therefore remains a question for further empirical investigation. Keywords: code-switching, code-mixing, Matrix Language theory, Kazakh language, Russian language, bilingualism, Matrix Language Frame model