Nordic Journal of African Studies 26(4): 403–430 (2017) Setswana Syllable Structure and Distribution Thapelo J. OTLOGETSWE University of Botswana, Botswana ABSTRACT This paper investigates the frequency and distribution of Setswana syllables by analysing the frequency and distribution of (orthographic representations of) Setswana syllables in a wordlist of about 59,000 words extracted from a corpus of over 13.6 million words. The study uses Wordsmith tools to calculate syllables found at word initial, word medial and word final positions of Setswana words. Each orthographic consonant is then paired to each of the five Setswana orthographic vowels to generate a list of potential Setswana syllables. Each potential syllable is then tested for frequency and distribution in the extracted wordlist. Non- occurring syllables are discarded, leaving an inventory of all Setswana syllables that make up Setswana words. The final part investigates the distribution of the allophones [l] and [d], as well as the phoneme /g/ and confirms, with some modification, previous analyses. Keywords: Setswana, syllable frequency, syllable distribution, phoneme, orthography. 1. INTRODUCTION Not much advanced analysis of Setswana syllable exists. Only basic analysis of the Setswana syllable structure has been undertaken (DALL, 1999; Cole, 1955:52). Much of the Setswana syllable research, however, is based on a theoretical assumption of what constitutes a Setswana syllable and largely ignores the large amount of borrowed terms which raise new challenges to long- held views. The paper studies a 13.6 million Setswana corpus of orthographic words using Oxford Wordsmith Tools to determine what constitutes a syllable in Setswana and how such syllables are distributed at either word initial, medial or final position. While syllabicity is usually perceived as a characteristic of speech, this study uses written text mainly because it is cheap to compile and easy to compute. The syllable position measured count is therefore of the Setswana orthographic word. The study uses quantitative and computational approaches to study language data to reveal patterns in language (Kessler & Treiman, 1997; Munthuli et al., 2015). The study measures syllable frequency and distribution and contributes towards a better understanding of Setswana phonotactics and syllable preference in Setswana word formation. Measuring syllable frequencies in a language is critical for many areas of linguistics, phonetics and speech technology.