Searching for a universal constraint on the possible denotations of clause-embedding predicates

• One of the central questions in semantic research is whether there are any universal constraints on the possible denotations of lexical items of a certain grammatical category. • This question has been investigated most prominently in the domain of determiners like every, most, and some. • For instance, it has been proposed that all determiners are ‘conservative’ and that all monomorphemic determiners are ‘monotonic’ (Barwise and Cooper, 1981). • Recent work has explored possible semantic universals in the domain of clauseembedding predicates like know, agree, andwonder (Spector and Egré, 2015; Theiler et al., 2018; Uegaki, 2019; Steinert-Threlkeld, 2019). • Within this line of work, two basic questions can be distinguished: – The first is empirical: Which constraints, if any, do we find in the semantics of clause-embedding predicates? – The second is theoretical: What may explain the existence of such universal semantic constraints? • The present paper is primarily concerned with the first question. It also engages with the second question, but does not have the ambition to make a major contribution on this front. *This is a draft of a handout to be presented at Semantics and Linguistic Theory (SALT), which will be held as an online event mid August 2020. We welcome comments on any aspect of the proposal.