Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.

Author ORCID Identifier


Open Access Dissertation

Document Type


Degree Name

Doctor of Philosophy (PhD)

Degree Program


Year Degree Awarded


Month Degree Awarded


First Advisor

Dr. Adrian Staub

Second Advisor

Dr. Charles Clifton

Third Advisor

Dr. Joonkoo Park

Fourth Advisor

Dr. Brian Dillon

Fifth Advisor

Dr. Shota Momma

Subject Categories

Cognitive Psychology | Cognitive Science | Psycholinguistics and Neurolinguistics


Language users predictively preactivate lexical units that appear to the comprehen- der to be likely to surface. Despite ample language experience and grammatical competence, it appears that language users tend to preactivate verbs in some contexts, called role-reversal contexts, that would create plausibility violations if they were to actually appear; these verbs assign thematic roles to their arguments in such a way that it leads to implausibility. These anomalous predictions provide a window into the mechanisms underlying lexical preactivation and are the case study that this dissertation focuses in on. This dissertation is an exploration of what linguistic information is effectively leveraged to modulate lexical expectations in sentence and discourse contexts (both appropriate and anomalous) and a comparison of the methodologies that are common to test for lexical preactivation.

Using a series of cloze experiments, I will argue that these anomalous predictions are the result of two mechanisms that co-contribute to spread preactivation to role- anomalous verbs. The first is a small but non-negligible “Bag of Words” associative priming mechanism, and the second is a more sophisticated mechanism that prioritizes pre-verbal arguments that are unsaturated, that is they are in need of a predicate, that I am calling a “Bag of not yet Saturated Arguments” mechanism. Though lexical predictions derived from sentence compositional meanings are indeed the most common, these less sophisticated mechanisms play an important role in lexical preactivation and contribute to the preactivation of both role-appropriate and role-anomalous verbs in real- time sentence processing.

By comparing cloze responses and reading times, I also argue in this dissertation that participants make both position-specific predictions, predictions about what words may occur in position n+1, and predictions that are expected to surface merely sooner-or-later in the context. I additionally argue that the N400 and eye tracking measures such as the first fixation duration are particularly reflective of this latter type of prediction, while cloze responses are less so. This explains the discrepancies observed between these common measures of lexical preactivation in role-reversed sentences.


Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.