Prediction during sentence comprehension is more than a sum of lexical associations: the role of event knowledge

Click to increase image sizeClick to decrease image size

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. The view here should be understood as influenced by a number of proposals, from a variety of perspectives, in which structured representations of events are a key output of language comprehension (Van Dijk & Kintsch, 1983; Fillmore, 1982; Zwaan & Radvansky, 1998).

2. It should be noted that our use of the term “structure” is somewhat different from the way Chow et al. use the term. We focus on event-structural roles (e.g. the eater, the thing eaten, or at a more generalized level, Agent and Theme), while Chow et al. focus on syntactically defined grammatical roles (e.g. subject and object). Although these various types of structure are theoretically distinct, they are also highly correlated.

3. There is a potentially important difference between Chow et al.’s stimuli and most other sentences that trigger semantic P600s: while the latter are semantically anomalous (meals cannot devour), Chow et al.’s sentences are merely unusual (customers rarely serve waitresses). Given this, the conflict between event knowledge and syntactic form might not be as large in the Chow et al. situation.

4. Calculated by the near neighbours function available at http://lsa.colorado.edu using the topic space derived from the “general reading up to 1st year college” sample.

5. These values of cloze probability are in a range that would typically be considered “low constraint” in studies that manipulate the degree of semantic constraint from low to high (Kutas & Hillyard, 1984).

Additional information

Funding

This work was partially supported by the National Science Foundation [Grant Number 1431223] and by the National Institutes of Child Health and Human Development [Grant Number 1R03HD071094] to AEK.