Henry Fominyam, Radek Simik

• We provide an analysis of focus and exhaustive focus in the Grassfields Bantu language Awing. We show that Awing provides an exceptionally clear window into the syntactic properties of exhaustive focus. Our analysis reveals that the Awing particle lə́ (le) realizes a left-peripheral head which, in terms of its syntactic position in the functional sequence, closely corresponds to the Foc(us) head in standard cartographic analyses (e.g., Rizzi 1997). Crucially, however, we show that le is only used if the focus it associates with receives a presuppositional exhaustive (cleft-like) interpretation. Other types of focus are not formally encoded in Awing. In order to reflect this semantic specification of le, we call its syntactic category Exh rather than Foc. Another point of difference from what one would consider a “standard” cartographic Foc head is that the focus associated with le is not realized in its specifier but rather within its complement. More particularly, we argue that le associates with the closest maximalWe provide an analysis of focus and exhaustive focus in the Grassfields Bantu language Awing. We show that Awing provides an exceptionally clear window into the syntactic properties of exhaustive focus. Our analysis reveals that the Awing particle lə́ (le) realizes a left-peripheral head which, in terms of its syntactic position in the functional sequence, closely corresponds to the Foc(us) head in standard cartographic analyses (e.g., Rizzi 1997). Crucially, however, we show that le is only used if the focus it associates with receives a presuppositional exhaustive (cleft-like) interpretation. Other types of focus are not formally encoded in Awing. In order to reflect this semantic specification of le, we call its syntactic category Exh rather than Foc. Another point of difference from what one would consider a “standard” cartographic Foc head is that the focus associated with le is not realized in its specifier but rather within its complement. More particularly, we argue that le associates with the closest maximal projection it asymmetrically c-commands. The broader theoretical relevance of the present work is at least two-fold. First, our paper offers novel evidence in support of Horvath’s (2010) Strong Modularity Hypothesis for Discourse Features, according to which information structural notions such as focus cannot be represented in narrow syntax as formal features. We argue that the information structure-related movement operations that Awing exhibits can be accounted for by interface considerations, in the spirit of Reinhart (2006). Second, our data support the generality of the so-called closeness requirement on association with focus (Jacobs 1983), which dictates that a focus-sensitive particle be as close to its focus as possible (in terms of c-command). What is of special significance is the fact that Awing exhibits two different avenues to satisfying closeness. The standard one—previously described for German or Vietnamese and witnessed here for the Awing particle tśɔ’ə ‘only’—relies primarily on the flexible attachment of the focus-sensitive particle. The Awing particle le, in contrast, is syntactically rigid. For that reason, the satisfaction of closeness relies solely on the flexibility of other syntactic constituents.…