A Tale of Two Positivities and the N400: Distinct neural signatures are evoked by confirmed and violated predictions at different levels of representation

Abstract

Hierarchical prediction has been proposed as a fundamental computational principle underlying neurocognitive processing. Here we ask whether the brain engages distinct neurocognitive mechanisms in response to inputs that fulfill versus violate strong predictions at different levels of representation during language comprehension. Participants read three-sentence scenarios in which the third sentence constrained for a broad event structure (e.g. {Agent caution animate-Patient}). Low constraint contexts did not constrain for any specific event/lexical item. High constraint contexts additionally constrained for a specific event and lexical item (e.g. a twosentence context about a beach, lifeguards and sharks constrained for the event, {Lifeguards cautioned Swimmers} and the specific lexical item, "swimmers"). We measured ERPs on critical nouns that fulfilled and/or violated each of these constraints. We found clear, dissociable effects to fulfilled semantic predictions (a reduced N400), to event/lexical prediction violations (an increased late frontal positivity), and to event structure/animacy prediction violations (an increased late posterior positivity/P600). We argue that the late frontal positivity reflects a large change in activation associated with successfully updating the comprehender's mental model with new unpredicted input. We suggest that the posterior late positivity/P600 reflects an initial failure to incorporate new unpredicted input into the comprehender's mental model, and possibly a reanalysis of the previous inputs in attempts to revise this model. Together, these findings provide strong evidence that confirmed and violated predictions at different levels of representation manifest as distinct spatiotemporal neural signatures. and different tasks -all factors that are known to modulate the N400 and late positivity components. This made it difficult to compare the effects across different studies. However, based on a review of the early literature, Van Petten and Luka (2012) noted that, while both the late frontal and posterior positivity effects were associated with unexpected linguistic input, the main factor that distinguished between them was the plausibility of the resulting interpretation.This was confirmed in recent studies showing that, in high constraint contexts, the same individuals who produced a late frontal positivity effect on unexpected plausible critical words, produced a late posterior positivity effect on unexpected implausible critical words (in English

Neural evidence for the prediction of animacy features during language comprehension: Evidence from MEG and EEG Representational Similarity Analysis

“…And, indeed, as in many previous studies (e.g. Kuperberg, 2011, 2012;Szewczyk and Schriefers, 2011;Kuperberg, et al, 2019), the evoked N400 response on the subsequent noun was reduced when its animacy features matched (versus mismatched) the animacy constraints of the preceding verb. Moreover, a spatial similarity analysis on the subsequent noun confirmed that the spatial pattern of neural activity was more similar to plausible animate than inanimate nouns (Wang and Kuperberg, Unpublished).…”

“…We suggest that this was the first time point at which comprehenders were able to infer the full high-level event structure (e.g. <Agent cautioned animate noun>), and that they used this structure to generate top-down predictions of the semantic features linked to the animacy of upcoming arguments (Kuperberg and Jaeger, 2016;Kuperberg et al, 2019).…”

“…Each list contained 160 scenarios, with no verb being repeated in any of the lists (80 unique animate constraining and 80 unique inanimate constraining verbs). A detailed description of the precise counterbalancing scheme can be found in Kuperberg et al (2019).…”

Similar time courses for word form and meaning preactivation during sentence comprehension

“…It has been observed to vary with integration difficulty (Brouwer, Fitz, & Hoeks, ), conflict resolution (Vissers et al, ), language monitoring (Kolk, Chwilla, Van Herten, & Oor, ), and memory retrieval (see Van Petten & Luka, ). It has been proposed to relate to processes such as revision and repair when unpredicted input is encountered (Kuperberg & Wlotko, ). The posterior PNP/P600’s proposed relation to another ERP, the P3b, has also raised the possibility that it, too, may similarly be affected by the relevance of particular stimuli depending on experimental task (explicitly specified or implicitly perceived by participants; see Van Petten & Luka, , for a review).…”

Gesture's body orientation modulates the N400 for visual sentences primed by gestures

“…Of note, despite consistent reports (see Özyürek, 2014 for review), it remains an open question whether these N400 effects reflect the cost of semantically integrating gesture and language, or the differential level of semantic prediction from gesture to language and vice versa, or a combination of both processes. In fact, in language processing, although the N400 has been classically considered as a semantic integration index (Kutas & Hillyard, 1984), recent functional accounts of the component suggest either the prediction alternative (Bornkessel‐Schlesewsky & Schlesewsky, 2019; Kuperberg, Brothers, & Wlotko, 2020; Kutas & Federmeier, 2011; Lau, Phillips, & Poeppel, 2008), or a hybrid of both processes (Baggio & Hagoort, 2011; Nieuwland et al, 2020).…”