When can nonveridical preferential attitude predicates take questions?

A growing body of evidence suggests that whether or not attitude predicates may combine with question complements is determined at least in part by some of their semantic properties. Zooming in, nonveridical preferential predicates (NVPs) such as hope and fear have been claimed not to combine with questions, but this empirical generalization as well as its proposed explanation have been challenged by a number of counterexamples, whose properties-beyond the fact that they have the predicate combining with embedded questions-remain ill understood. By taking a closer look at the combinatorial and semantic properties of different NVPs from English, Japanese, Mandarin Chinese, Spanish, and Turkish, this work narrows down the semantic properties of attitude predicates that determine their combinatorial properties in a crosslinguistically informed and predictive way. We propose a new classification of NVPs based on two empirical properties: C(lausal)-distributivity and valence, and show that non-C-distributive NVPs (e.g., worry) as well as evaluatively negative and C-distributive NVPs (e.g., fear) can in fact combine with questions.