Corpus construction and annotation of labour law texts: methodology and comparative case study on Kenya and Tanzania.

This dissertation develops and tests a methodology for building and annotating a legal corpus of labour law texts. Its primary aim is to integrate a corpus-based discourse analytical approach into the broader Worlds of Labour (WoL) project, a research initiative that investigates legal segmentation in labour law systems, with particular attention to countries of the Global South, adopting the methodological approach of leximetrics. The annotation framework was designed to reflect the structural features of legislative texts and is intended to serve as a resource for future interdisciplinary research in law and language. To test the analytical potential of this methodology, a case study was conducted combining leximetrics and Corpus-Assisted Dis-course Studies (CADS) to explore the representation of female labour in legal texts of two East African countries, Kenya and Tanzania, between 1902 and 2011. Drawing on the WoL database and its coding infrastructure, two annotated sub-corpora were compiled for each country: one focused on general labour legislation (WoL-K1, WoL-T1), and one on legal texts specifically concerning women’s labour (WoL-K2, WoL-T2). These corpora were analysed using Sketch Engine to uncover discursive dimensions of legal segmentation in recurring expressions, struc-tural patterns, and narrative strategies concerning women and work. The findings show that, within the selected texts, women are consistently depicted as passive subjects, typically framed in terms of vulnerability, protection, or dependency, rather than agency and participation. The study highlights colonial patterns in the marginalisation of female labour and stresses the value of discourse analysis in legal research. It concludes by offering the proposed methodology as a replicable tool for future interdisciplinary work on legal discourse, to complement quantitative leximetric approaches with qualitative insights.