Abbreviations

For more information about the lecture halls and seminar rooms, see our lecture room page. The names listed under "Technical Assistance" are student volunteers who will act as a contact person for technical questions of the lecturers and workshop speakers during the course or workshop.

Negative polarity items: corpus linguistics, semantics, and psycholinguistics

NPIs (ever, lift a finger) are expressions whose analysis minimally involves syntax, semantics, and pragmatics. NPIs are assumed to be licensed by negation, different degrees of negation being required by sub-classes of NPIs. There is also a fuzzy and incompletely understood borderline area of licensing environments that are not logically negative in an obvious sense (questions, subjunctives,...). Research on NPIs long proceeded within the confines of traditional theoretical subdisciplines, but recently methods from corpus linguistics and psycholinguistics offer new avenues of investigation. In this course we begin with a data-driven approach to review our current understanding of NPIs. After discussing automatic retrieval and classification of NPIs, we investigate characteristic classes of NPIs to see which properties and licensing environments from the literature fit which classes. This will lead to the formulation of new empirical questions about natural classes of NPIs in different languages, connecting the insights from theoretical linguistics and corpus linguistics. Corpora can provide interesting data to investigate existing theories, but they need to be checked carefully. In the second part of this course we will show how psycholinguistic methods help in (a) validating naturally occurring corpus data and (b) establishing theoretically important distinctions that are not reliably accessible to intuitions. We will discuss a number of experiments (questionnaire, reading time and ERP studies) designed to find out whether candidate expressions are NPIs at all, to determine the required strength of licensing environments and to test intervention effects. Other topics include the interaction of NPI licensing with NegRaising predicates, complement set anaphora readings induced by some quantifiers, and licensing across supposed syntactic boundaries. After reviewing experimental evidence we will return to the challenges these data present to existing theories.