A Corpus-Based Description of Igbo Discourse Markers: Form, Function, and Frequency in Spoken and Written Registers

This dissertation presents a comprehensive corpus-based description of discourse markers in Igbo, characterising their formal properties, functional categories, and frequency distributions across spoken and written registers, and producing the first systematic inventory of Igbo discourse markers grounded in large-scale empirical data. Discourse markers constitute a functionally heterogeneous lexical class that plays a fundamental role in discourse organisation, interpersonal management, and inferential guidance, yet the systematic study of discourse markers in African languages, and Igbo specifically, remains in an early stage. The study constructs a balanced spoken and written Igbo corpus of approximately two million words, drawing on natural conversation recordings, broadcast media transcriptions, newspaper texts, academic writing, and literary prose. A discourse marker identification and annotation protocol is developed and applied by a team of trained Igbo linguists to the full corpus, yielding a dataset of over forty thousand discourse marker tokens. Functional classification is conducted using an expanded taxonomy derived from Fraser's discourse marker framework, supplemented by Igbo-specific functional categories identified inductively from the data. Findings document two hundred and twelve distinct Igbo discourse marker forms, substantially exceeding the inventory implied by existing descriptive grammars. Significant register-conditioned frequency differences are documented, with spoken registers showing approximately three times the discourse marker density of written registers. Several previously undescribed discourse marker forms unique to specific spoken sub-registers are identified and analysed. Keywords: Igbo discourse markers, corpus linguistics, register variation, pragmatics, spoken language.