Siswati Clefts: The Meeting Ground of Context and Contrast

Cleft structures have proved to be a tantalising structure, not easily expressible in current formalisms. Traditionally, they have been taken as a canonical contrastive-focus device, with clearcut presuppositional properties indicating dependence on context of the non-focussed part (REFS), yet, as has emerged from corpus collections and other recent work (Prince 1988, Hedberg 2000, Buler 2005), there are many divergent uses of clefts, indicating a much more complex relationship between structure and interpretational effect. Despite the subtle variation in the extent and precise nature of the context-dependence implied, structurally, the syntactic properties of clefts robustly display a great deal of cross-linguistic uniformity despite minor language-particular idiosyncrasies. And in this uniformity hides another puzzle. Quite generally, they display extensive parallelism with relative clauses, even to the extent of using the same morphological indicators as a relative clause sequence; and yet at one and the same time they display properties of left-dislocation. Siswati clefts display this straddling across relative-clause patterning and dislocation patterning particularly clearly. There are two main types of Siswati clefts; and they all display this split nature with attributes in common with left-peripheral structures on the one hand, and with relative clause structures on the other. The simple type involve first marking the “clefted” element with a preceding ngemarker (for class 1 nouns), and secondly an associated relative clause sequence relative clauses in Siswati are marked by a la-prefix as the first prefix in the verbal complex: