Doctoral Dissertation Research: Active assignment of quantifier scope guides language processing

Some of the most interesting results in sentence processing are called illusions, cases where observed language behaviors do not align with the predictions of well-established grammatical theories (Vasishth, Brüssow, Lewis & Drenhaus 2008; Wagers, Lau & Phillips 2009). Analysis of these illusions shows not only that the real time language processing system builds and tolerates ungrammatical structures, but how it goes about the process of building such structures. Many theories have been developed which build elaborate processing systems around these illusion phenomena, framing them as ungrammatical structures that the parser must occasionally build in order to produce other structures compatible with the underlying grammar (Eberhard, Cutting & Bock 2005; Franck, Vigliocco & Nicol 2002;; Van Dyke & Lewis 2003; Lewis & Vasishth 2005). As the catalogue of sentence processing illusions has grown and our understanding of individual illusions has increased, the ability of various theories to predict all the illusion findings has been strained. Recent data from the negative polarity item (NPI) illusion shows that the illusion is not only sensitive to the properties of the NPI, but also only occurs when the NPI licensor is also a quantifier (Orth, Yoshida & Sloggett 2020). We consider the possibility that this illusion is created by the mechanism for processing quantifiers rather than it being the result of a polarity item specific process that requires quantification to satisfy some unrelated constraint.