Proceedings of the Annual Meeting of the Cognitive Science Society

One striking divergence between logical modeling and language processing emerges when evaluating the empty set in the interpretation of quantifiers. It has been shown that empty restrictor scenarios lead to presupposition failure, while empty scope scenarios lead to logically incorrect judgments and a processing delay of downward-entailing quantifiers. In this study, we tested whether empty restrictor scenarios would be evaluated as true or false rather than rejected as odd in a question-answering task. We contrasted two weak downward-entailing quantifiers (fewer than n, at most n-1) with two weak upward-entailing quantifiers (more than m, at least m+1) in two complex question embeddings that made the evaluation possible to various degrees. We found that in these embeddings, empty restrictors are less often rejected as odd as well as that the empty restrictors of downward-entailing quantifiers are often evaluated as false.