Some alternatives? Event-related potential investigation of literal and pragmatic interpretations of some presented in isolation

Abstract

In sentence verification tasks involving under-informative statements such as Some elephants are mammals, some adults appear more tolerant to pragmatic violations than others. The underlying causes of such inter-individual variability remain however essentially unknown. Here, we investigated inter-individual variation in adults deriving the scalar inference "not all" triggered by the quantifier some. We first assessed the individual intolerance to pragmatic violations in adult participants presented with under-informative some-statements (e.g., Some infants are young). We then recorded event-related brain potentials in the same participants using an oddball paradigm where an ambiguous deviant word some presented in isolation had to be taken either as a match (in its literal interpretation "at least some") or as a mismatch (in its pragmatic interpretation "some but not all") and where an unambiguous deviant target word all was featured as control. Mean amplitude modulation of the classic P3b provided a measure of the ease with which participants considered some and all as deviants within each experimental block. We found that intolerance to pragmatic violations was associated with a reduction in the magnitude of the P3b effect elicited by the target some when it was to be considered a literal match. Furthermore, we failed to replicate a straightforward literal interpretation facilitation effect in our experiment which offers a control for task demands. We propose that the derivation of scalar inferences also relies on general, but flexible, mismatch resolution processes.