Somali and Prosodic Typology

Aims and significance of the project

Somali is said to have a pitch-accent system (Hyman 1981; Le Gac 2001), making it of inherent interest for theories of prosodic typology. Pitch-accent systems are notoriously difficult to characterise, as they fall in between stress and tone (Hyman 2006). Cushitic pitch-accent systems have the additional complication that the position and realisation of accent is grammatically determined. Mous (2009) argues that this is a unique typological property. Somali prosody is also relevant to the issue of what role intonation plays in languages with lexical pitch contrasts. Work like Hyman & Monaka (2011) argues that languages with pitch contrasts should have limited scope for intonation, but this claim is controversial: more work on different sentence types is needed. It is also controversial whether focus has the same sort of effect on intonation in these languages as in stress languages (Downing & Pompino-Marschall 2013). This project will aim to provide a complete analysis of Somali prosody, at the lexical and sentential level. It will pay particular attention to the effect of syntactic position and focus on pitch accent and the domain of pitch accent realization in sentences of different types and complexity. Because Somali pitch accent is understudied using modern analytic methods and has typologically unusual properties, this investigation will contribute to theories of prosodic typology.

The project also cooperates closely with the Somali ‘mother tongue’ program at GU, which provides one source of language consultants and contact with the local Somali community. One practical output of the project will be a booklet in Swedish contrasting Somali and Swedish prosodic systems, with advice on implementation in education.

Activities and implementation

Our investigation will necessarily need to go beyond the kind of data one can find in a thorough critical survey of the existing literature. To address our research questions, we will need to make extensive recordings of Somali speech and analyse it using modern phonetic and phonological techniques. The budget includes ca. 100 hours per year to support work with language consultants. Maarten Mous, professor of African Linguistics at Leiden University and an internationally recognized Cushitic specialist, is a partner in the project and will be involved in data collection and analysis. We will be advised for the phonetic analysis by phoneticians at GU and Leiden.

We will proceed step by step, collecting data in increasingly complex contexts, presenting our results regularly (once or twice a year) at international conferences and workshops. The plan of work is divided into, roughly, 6-month segments. In the first six months, we will collect and analyse prosody at the word level. In the next 6 months, we expand our corpus to more morphological contexts and begin the study of intonation. In the second year, we extend the corpus to include, first, simple declarative sentences, then different question and answer types, analysing both pitch-accent and intonation in these constructions. In the final year, we turn to an investigation of the prosody of complex sentences and the interaction of pitch-accent with other prosodies. In the final year and a half of the project, we will concentrate on writing up results, both for international journal articles and for a Swedish educational booklet. To promote international research into prosodic typology, we aim to organize two international workshops during the course of the project, one on the typology of stress-tone-accent systems, and one on the interaction of tone and intonation.