A predictive coding framework for rapid neural dynamics during sentence-level language comprehension

Abstract

There is a growing literature investigating the relationship between oscillatory neural dynamics measured using electroencephalography (EEG) and/or magnetoencephalography (MEG), and sentence-level language comprehension. Recent proposals have suggested a strong link between predictive coding accounts of the hierarchical flow of information in the brain, and oscillatory neural dynamics in the beta and gamma frequency ranges. We propose that findings relating beta and gamma oscillations to sentence-level language comprehension might be unified under such a predictive coding account. Our suggestion is that oscillatory activity in the beta frequency range may reflect both the active maintenance of the current network configuration responsible for representing the sentence-level meaning under construction, and the top-down propagation of predictions to hierarchically lower processing levels based on that representation. In addition, we suggest that oscillatory activity in the low and middle gamma range reflect the matching of top-down predictions with bottom-up linguistic input, while evoked high gamma might reflect the propagation of bottom-up prediction errors to higher levels of the processing hierarchy. We also discuss some of the implications of this predictive coding framework, and we outline ideas for how these might be tested experimentally.

Grammatical predictions reveal influences of semantic attraction in online sentence comprehension: evidence from speeded forced-choice sentence continuations

“…Finally, given the amount of emphasis on predictive mechanisms in current sentence processing research (e.g. Altmann & Mirkovic, 2009;Brothers, Swaab, & Traxler, 2015;Farmer, Brown, & Tanenhaus, 2013;Federmeier, 2007;Fine, Jaeger, Farmer, & Qian, 2013;Garrod & Pickering, 2015;Kim & Lai, 2012;Levy, 2008;Lewis & Bastiaansen, 2015;Van Petten & Luka, 2012; see Kuperberg & Jaeger, 2016 for a review), we believe that this method may be valuable in addressing a variety of research questions regarding how language users leverage predictive processes to help resolve conflicts online (e.g. in cases of co-reference resolution or syntactic ambiguity resolution).…”

What do we mean by prediction in language comprehension?

“…Unlike evoked ERP or MEG responses, which index phase-locked activity that is time-locked to specific events (Luck, 2014), and which are therefore best suited to detecting facilitation when a new incoming stimulus appears, low frequency oscillatory activity may be better suited for capturing top-down predictive neural activity (for general discussion, see Arnal & Giraud, 2012; Engel & Fries, 2010; Weiss & Mueller, 2012, and for recent discussion in relation to language comprehension, see Lewis & Bastiaansen, 2015). These studies generally used simple contexts that constrained strongly (versus weakly) for the perceptual features of new inputs.…”

“…Jordan & Rumelhart, 1992). Finally, it has been proposed that this type of hierarchical actively generative architecture is instantiated at the neural level in the form of predictive coding (Friston, 2005, 2008, 12 see Lewis & Bastiaansen, 2015 and Kuperberg, under review, for discussion in relation to the neural basis of language comprehension), although it is important to recognize that the most direct evidence for predictive coding in the brain comes from Rao and Ballard’s (1999) initial descriptions within the visual system. Given these considerations, we believe that this type of multi-representational hierarchical actively generative architecture can potentially provide a powerful bridge across the fields of computational linguistics, psycholinguistics and the neurobiology of language, and we hope that, by sketching out its principles, this will stimulate cross-disciplinary collaboration across these areas.…”

Basic composition and enriched integration in idiom processing: An EEG study.

“…Despite the different frequency ranges of lower and middle gamma bands, and the different time development of the effects observed (sustained in time, Hald et al, 2006;Penolazzi et al, 2009;Wang et al, 2012;or short lasting, Monsalve et al, 2014), according to a recent interpretation of these changes the increase in gamma power would reflect reverberatory activity triggered by a successful match "between the pre-activation of the neural representation of the predicted word, and the neural representation of the actually incoming word" (Wang et al,p.11). This explanation (see also Lewis & Bastiaansen, 2015) processes involved in processing highly predictable strings. In idiomatic contexts, once the idiom is recognised, the processing of the remaining idiom constituents may be shallower if not absent and this led to the power differences on W2.…”

“…In fact, power increase in lower and middle gamma frequency bands (approximately between 35-75 Hz) has been associated with successful semantic unification mechanisms (e.g., Hald, Bastiaansen & Hagoort, 2006;Penolazzi, Angrilli & Job, 2009;Wang, Zhu & Bastiaansen, 2012). Lewis and Bastiaansen (2015) proposed that such activity "reflects a match between strong top-down predictions and bottom-up linguistic input" (p.159). As we mentioned, Rommers et al (2013) directly contrasted the EEG activity in idiomatic and literal sentence processing in the time-frequency domain.…”