Teasing apart coercion and surprisal: Evidence from eye-movements and ERPs

Abstract

Previous behavioral and electrophysiological studies have presented evidence suggesting that coercion expressions (e.g., began the book) are more difficult to process than control expressions like read the book. While this processing cost has been attributed to a specific coercion operation for recovering an event-sense of the complement (e.g., began reading the book), an alternative view based on the Surprisal Theory of language processing would attribute the cost to the relative unpredictability of the complement noun in the coercion compared to the control condition, with no need to postulate coercion-specific mechanisms. In two experiments, monitoring eye-tracking and event-related potentials (ERPs), respectively, we sought to determine whether there is any evidence for coercion-specific processing cost above-and-beyond the difficulty predicted by surprisal, by contrasting coercing and control expressions with a further control condition in which the predictability of the complement noun was similar to that in the coercion condition (e.g., bought the book). While the eye-tracking study showed significant effects of surprisal and a marginal effect of coercion on late reading measures, the ERP study clearly supported the surprisal account. Overall, our findings suggest that the coercion cost largely reflects the surprisal of the complement noun with coercion specific operations possibly influencing later processing stages.

Quasi-compositional mapping from form to meaning: a neural network-based approach to capturing neural responses during human language comprehension

“…This view seems largely consistent with a recent study controlling for surprisal between coercion and non-coercion conditions (e.g. 'John began/ bought the book') and observing no differences in N400s nor any other significant ERP differences [9]. These authors suggested, however, that the absence of ERP correlates should not be taken to indicate that coercion did not take place, royalsocietypublishing.org/journal/rstb Phil.…”

“…Other researchers do view compositionality not as a method but as a factual claim, open to empirical testing [4]. For those who see compositionality as a factual claim, a possible empirical research programme (related to the question raised in the next paragraph) is to investigate whether the operations postulated to enable compositional analyses of natural language sentences are reflected in online measures of human language comprehension [5][6][7][8][9].…”

Toward an optimal code for communication: The case of scientific English

“…complementizers or relativizers; Sikos et al 2017) or condensation (e.g. coercion; Delogu et al 2017) in lexico-grammatical contexts that are highly predictive. Such studies show that language users seem to be concerned about informativity and try to modulate the amount of information in transmission by their linguistic choices.…”

Expectation-based Comprehension: Modeling the Interaction of World Knowledge and Linguistic Experience

“…Another important implication of the presented perspective on surprisal involves its mapping to Event-Related Potentials (ERPs). In the electrophysiological domain, surprisal has often been linked to the N400 component (Delaney-Busch, Morgan, Lau, & Kuperberg, 2017;Delogu, Crocker, & Drenhaus, 2017;Frank et al, 2015), the amplitude of which is inversely related to the expectancy of a word; the less expected a word, the higher N400 amplitude (Kutas, Lindamood, & Hillyard, 1984). Although the N400 component was previously taken to be an index of integrative processing (i.e., the updating of an unfolding utterance representation), it has recently been linked to the process of lexical retrieval, which is facilitated by lexical and contextual priming; that is, N400 amplitude is reduced if word-associated conceptual knowledge is preactivated in memory (Brouwer, Fitz, & Hoeks, 2012;Kutas & Federmeier, 2000;Lau, Phillips, & Poeppel, 2008;van Berkum, 2009van Berkum, , 2012.…”