3. Negative and positive polarity items

In this chapter, we discuss the distribution and lexical properties of common varieties of negative polarity items (NPIs) and positive polarity items (PPIs). We establish first that NPIs can be licensed in negative, downward entailing, and nonveridical environments, and within the NPI class we distinguish two main distributional patterns: an NPI with a broad distribution in the environments just mentioned, and a more narrow NPI appearing only in negative contexts. We also note that within the broad class there are scalar and non-scalar NPIs. We examine to what extent the scalarity approach (originating in Kadmon & Landman 1993) can explain the basic patterns, and conclude that this approach, by reducing all polarity phenomena to one source, fails to capture that a significant number of NPIs and PPIs are not scalar, and that there is variation within the scalar class between broader (any) and narrow NPIs (either, minimizers). Another fundamental problem with pure scalarity is that it predicts weaker effects (contradictions, presupposition failures) with ill-formed NPIs than is actually the case. We then consider the variation approach pursued by Giannakidou (1998 and sequel) which posits two possible lexical sources of polarity: (a) scalarity (Giannakidou 2007a), which mostly yields weaker ill-formedness, and (b) referential deficiency of expressions that end up becoming NPIs. The variation approach is more consistent with the diversity of NPIs, and extends easily to PPIs, which, as a class, appear to be non-scalar (Szabolcsi 2004; Ernst 2009).