The Evaluability Hypothesis

Although the field of polarity is well researched, this monograph offers a new take on polarity sensitivity that both challenges and incorporates previous theories. Based primarily on Swedish data, it presents new solutions to long-standing problems, such as the non-complementary distribution of NPIs and PPIs in yes/no-questions and conditionals, long distance licensing by superordinate elements, and the occurrence of polarity items in wh-questions. It is argued that polarity sensitivity can be understood in terms of evaluability. Lacking any immediate predecessor in the literature, evaluability refers to the possibility of accepting or rejecting an utterance as true in a communicative exchange. Intriguingly, the evaluable status of a clause is shown to have syntactic correlates in Swedish, mirrored in the configuration of the C-domain. This book is of interest to scholars studying the interplay between syntax, semantics and pragmatics, particularly those working on negation and polarity.