Non-veridicality in Habitual Context

Analysing the role of complex predicates in NPI Licensing

  • Vyom Sharma Indian Institute of Technology, Delhi


This paper presents a semantic-pragmatic analysis of the habitual aspect as a licensing context for Negative Polarity Items (NPIs). Further, we analyse why complex predicates (V1 + V2) in the habitual aspect form a better licensing context for NPIs than simple predicates. Habitual aspect can license certain NPIs in languages, in spite of being a non-Downward Entailing environment. Giannakidou (2002) argued that Veridicality, instead of Downward-Entailment (DE), should be the primary condition to characterize licensing contexts for NPIs. This paper attempts to further Giannakidou's (2002, 2011)  argument by proposing a Stalnakerian approach (Stalnaker, 1978) to define habitual aspect as an Iterative Pluractional and prove how it is non-veridical. Further we explore how the semantic compositionality of complex predicates makes the habitual context a better licensor for NPIs.