Academic knowledge presentation in MA theses: from corpus compilation to case studies of disciplinary conventions

This article discusses discipline-specific conventions in presenting academic knowledge in academic writing, an important first research text at many universities. Its empirical basis is a new corpus of South African MA theses (the ZAMA corpus) from Stellenbosch University. A pilot corpus of 100 MA theses and over 4 million words was compiled, paying special attention to disciplinary breadth from Humanities and Social Sciences to Natural Sciences and Engineering, including some interdisciplinary subjects (like Geography) and Law. Altogether between four and eight texts were chosen from 18 disciplines, which were later categorised deductively into six discipline types. Other factors taken into consideration were the socio-biographical diversity of authors, esp. ethnic/language by author's name (distinguishing English, Afrikaans, Black South African names, and a few others). The decisions of corpus compilation are carefully documented in order to obtain a stratified and balanced corpus for research in the South African university context. Descriptive statistical analyses were carried out to test whether discipline-specific frequency patterns could be found in three case studies using variables indicating metadiscourse conventions, such as modal auxiliaries, personal pronouns, and cohesive linkers.