A Hybrid, Theory-Driven Transcription System for the Study of Code-Switching

Abstract

Transcription plays a critical role in linguistic research, especially in studies of bilingual speech and code-switching. This paper introduces a custom-built transcription protocol designed to meet the theoretical and practical demands of analysing code-switching between typologically distinct languages—Afaan Oromoo and English—under the Matrix Language Frame (MLF) and 4-M models. The proposed system incorporates morpheme-level labelling, language-specific glossing, and clause segmentation in a layered and adaptable framework, drawing on principles from existing systems such as CHAT, ELAN, and the Leipzig Glossing Rules. The integrated system facilitates the identification of matrix and embedded languages while simultaneously capturing essential morphosyntactic features. It enables the empirical monitoring of code-switching phenomena through spreadsheet-based structuring, thereby supporting both qualitative and quantitative analyses. This transcription method improves current practices by connecting with theoretical ideas, addressing complex language issues, and offering a model that can be repeated for less-studied language pairs. The approach holds promise for future expansion, automation, and adaptation across diverse multilingual contexts.