Indirect truth marking via backgrounding: evidence from Bantu

Abstract Verum focus has long been analysed as part of information structure, being accounted for within a theory of focus as generation of alternatives on a polarity value. Recently, however, this view has been challenged, with some authors arguing for a separation between verum and focus. The empirical debate focuses on two types of strategies: either verum is marked identically to focus (e.g., pitch accent in German, English) or through a dedicated strategy (e.g., dedicated markers in Gitksan, Bura, South Marghi). On the basis of the second type, it has been argued that the link between focus and verum is only superficial, with verum instead arising from a lexical operator VERUM. In this paper, we provide evidence in support of the traditional link between truth marking and focus, using original field data on the morphosyntactic expression of information structure in 11 Bantu (Niger-Congo) languages. By working from function to form, we show that these languages use a variety of grammatical strategies for truth expression, strategies that are used elsewhere in the information structural system for backgrounding. In this way, verum is not directly marked by focal accent but marked indirectly via backgrounding, which we formulate as the Backgrounding and Underspecification Thesis (BUT). On the basis of this Bantu evidence, we add a third type of strategy to the typology, where verum is marked indirectly via backgrounding. We use this updated typology to argue in favour of maintaining a conceptual link between verum, polarity focus, and predicated-centred focus more broadly.