An experimental investigation of the homogeneity of conjunctions

Abstract

Nominal conjunctions in English are associated with the all or nothing homogeneity inference. There are two main approaches to explain the interpretations: the first derives homogeneity as implicatures (Magri, 2014; Bar-Lev and Fox, 2020), the second provides a trivalent account (Schwarzschild, 1993; Križ, 2015; Križ and Spector, 2021). The two accounts differ in their predictions on the existence of gaps and the status of positive and negative sentences in non-uniform contexts. We report two experiments investigating adults’ interpretation of nominal conjunctions in English. The experiments reveal that participants provide gappy judgments for sentences with conjunctions, and the gaps for positive and negative sentences in non-uniform contexts are symmetric. We discuss how the findings of the two experiments are expected on a trivalent approach to homogeneity and the open challenges they pose for the implicature approach.