Prediction during language comprehension: what is next?

Prediction is often regarded as an integral aspect of incremental language comprehension, but little is known about the cognitive architectures and mechanisms that support it. We review studies showing that listeners and readers use all manner of contextual information to generate multifaceted predictions about upcoming input. The nature of these predictions may vary between individuals owing to differences in language experience, among other factors. We then turn to unresolved questions which may guide the search for the underlying mechanisms. (i) Is prediction essential to language processing or an optional strategy? (ii) Are predictions generated from within the language system or by domain-general processes? (iii) What is the relationship between prediction and memory? (iv) Does prediction in comprehension require simulation via the production system? We discuss promising directions for making progress in answering these questions and for developing a mechanistic understanding of prediction in language.

Predicting Definite and Indefinite Referents During Discourse Comprehension: Evidence from Event-Related Potentials.

Linguistic predictions may be generated from and evaluated against a representation of events and referents described in the discourse. Compatible with this idea, recent work shows that predictions about novel noun phrases include their definiteness. In the current follow-up study, we ask whether people engage similar prediction-related processes for definite and indefinite referents. This question is relevant for linguistic theories that imply a processing difference between definite and indefinite noun phrases, typically because definiteness is thought to require a uniquely identifiable referent in the discourse. We addressed this question in an event-related potential (ERP) study (N = 48) with preregistration of data acquisition, preprocessing, and Bayesian analysis. Participants read Dutch mini-stories with a definite or indefinite novel noun phrase (e.g., "het/een huis," the/a house), wherein (in)definiteness of the article was either expected or unexpected and the noun was always strongly expected. Unexpected articles elicited enhanced N400s, but unexpectedly indefinite articles also elicited a positive ERP effect at frontal channels compared to expectedly indefinite articles. We tentatively link this effect to an antiuniqueness violation, which may force people to introduce a new referent over and above the already anticipated one. Interestingly, expectedly definite nouns elicited larger N400s than unexpectedly definite nouns (replicating a previous surprising finding) and indefinite nouns. Although the exact nature of these noun effects remains unknown, expectedly definite nouns may have triggered the strongest semantic activation because they alone refer to specific and concrete referents. In sum, results from both the articles and nouns clearly demonstrate that definiteness marking has a rapid effect on processing, counter to recent claims regarding definiteness processing.

Can You Hear What's Coming? Failure to Replicate ERP Evidence for Phonological Prediction.

Prediction-based theories of language comprehension assume that listeners predict both the meaning and phonological form of likely upcoming words. In alleged event-related potential (ERP) demonstrations of phonological prediction, prediction-mismatching words elicit a phonological mismatch negativity (PMN), a frontocentral negativity that precedes the centroparietal N400 component. However, classification and replicability of the PMN has proven controversial, with ongoing debate on whether the PMN is a distinct component or merely an early part of the N400. In this electroencephalography (EEG) study, we therefore attempted to replicate the PMN effect and its separability from the N400, using a participant sample size (N = 48) that was more than double that of previous studies. Participants listened to sentences containing either a predictable word or an unpredictable word with/without phonological overlap with the predictable word. Preregistered analyses revealed a widely distributed negative-going ERP in response to unpredictable words in both the early (150-250 ms) and the N400 (300-500 ms) time windows. Bayes factor analysis yielded moderate evidence against a different scalp distribution of the effects in the two time windows. Although our findings do not speak against phonological prediction during sentence comprehension, they do speak against the PMN effect specifically as a marker of phonological prediction mismatch. Instead of an PMN effect, our results demonstrate the early onset of the auditory N400 effect associated with unpredictable words. Our failure to replicate further highlights the risk associated with commonly employed data-contingent analyses (e.g., analyses involving time windows or electrodes that were selected based on visual inspection) and small sample sizes in the cognitive neuroscience of language.

How 'rational' is semantic prediction? A critique and re-analysis of.

In a recent article in Cognition, Delaney-Busch et al. (2019) claim evidence for 'rational', Bayesian adaptation of semantic predictions, using ERP data from Lau, Holcomb, and Kuperberg (2013). Participants read associatively related and unrelated prime-target word pairs in a first block with only 10% related trials and a second block with 50%. Related words elicited smaller N400s than unrelated words, and this difference was strongest in the second block, suggesting greater engagement in predictive processing. Using a rational adaptor model, Delaney-Busch et al. argue that the stronger N400 reduction for related words in the second block developed as a function of the number of related trials, and concluded therefore that participants predicted related words more strongly when their predictions were fulfilled more often. In this critique, I discuss two critical flaws in their analyses, namely the confounding of prediction effects with those of lexical frequency and the neglect of data from the first block. Re-analyses suggest a different picture: related words by themselves did not yield support for their conclusion, and the effect of relatedness gradually strengthened in othe two blocks in a similar way. Therefore, the N400 did not yield evidence that participants rationally adapted their semantic predictions. Within the framework proposed by Delaney-Busch et al., presumed semantic predictions may even be thought of as 'irrational'. While these results yielded no evidence for rational or probabilistic prediction, they do suggest that participants became increasingly better at predicting target words from prime words.

Concurrent use of animacy and event-knowledge during comprehension: Evidence from event-related potentials.

In two ERP experiments, we investigated whether readers prioritize animacy over real-world event-knowledge during sentence comprehension. We used the paradigm of Paczynski and Kuperberg (2012), who argued that animacy is prioritized based on the observations that the 'related anomaly effect' (reduced N400s for context-related anomalous words compared to unrelated words) does not occur for animacy violations, and that animacy violations but not relatedness violations elicit P600 effects. Participants read passive sentences with plausible agents (e.g., The prescription for the mental disorder was written by the psychiatrist) or implausible agents that varied in animacy and semantic relatedness (schizophrenic/guard/pill/fence). In Experiment 1 (with a plausibility judgment task), plausible sentences elicited smaller N400s relative to all types of implausible sentences. Crucially, animate words elicited smaller N400s than inanimate words, and related words elicited smaller N400s than unrelated words, but Bayesian analysis revealed substantial evidence against an interaction between animacy and relatedness. Moreover, at the P600 time-window, we observed more positive ERPs for animate than inanimate words and for related than unrelated words at anterior regions. In Experiment 2 (without judgment task), we observed an N400 effect with animacy violations, but no other effects. Taken together, the results of our experiments fail to support a prioritized role of animacy information over real-world event-knowledge, but they support an interactive, constraint-based view on incremental semantic processing.

Anticipating words during spoken discourse comprehension: A large-scale, pre-registered replication study using brain potentials.

Numerous studies report brain potential evidence for the anticipation of specific words during language comprehension. In the most convincing demonstrations, highly predictable nouns exert an influence on processing even before they appear to a reader or listener, as indicated by the brain's neural response to a prenominal adjective or article when it mismatches the expectations about the upcoming noun. However, recent studies suggest that some well-known demonstrations of prediction may be hard to replicate. This could signal the use of data-contingent analysis, but might also mean that readers and listeners do not always use prediction-relevant information in the way that psycholinguistic theories typically suggest. To shed light on this issue, we performed a close replication of one of the best-cited ERP studies on word anticipation (Van Berkum, Brown, Zwitserlood, Kooijman & Hagoort, 2005; Experiment 1), in which participants listened to Dutch spoken mini-stories. In the original study, the marking of grammatical gender on pre-nominal adjectives ('groot/grote') elicited an early positivity when mismatching the gender of an unheard, highly predictable noun, compared to matching gender. The current pre-registered study involved that same manipulation, but used a novel set of materials twice the size of the original set, an increased sample size (N = 187), and Bayesian mixed-effects model analyses that better accounted for known sources of variance than the original. In our study, mismatching gender elicited more negative voltage than matching gender at posterior electrodes. However, this N400-like effect was small in size and lacked support from Bayes Factors. In contrast, we successfully replicated the original's noun effects. While our results yielded some support for prediction, they do not support the Van Berkum et al. effect and highlight the risks associated with commonly employed data-contingent analyses and small sample sizes. Our results also raise the question whether Dutch listeners reliably or consistently use adjectival inflection information to inform their noun predictions.

Definitely saw it coming? The dual nature of the pre-nominal prediction effect.

In well-known demonstrations of lexical prediction during language comprehension, pre-nominal articles that mismatch a likely upcoming noun's gender elicit different neural activity than matching articles. However, theories differ on what this pre-nominal prediction effect means and on what is being predicted. Does it reflect mismatch with a predicted article, or 'merely' revision of the noun prediction? We contrasted the 'article prediction mismatch' hypothesis and the 'noun prediction revision' hypothesis in two ERP experiments on Dutch mini-story comprehension, with pre-registered data collection and analyses. We capitalized on the Dutch gender system, which marks gender on definite articles ('de/het') but not on indefinite articles ('een'). If articles themselves are predicted, mismatching gender should have little effect when readers expected an indefinite article without gender marking. Participants read contexts that strongly suggested either a definite or indefinite noun phrase as its best continuation, followed by a definite noun phrase with the expected noun or an unexpected, different gender noun phrase ('het boek/de roman', the book/the novel). Experiment 1 (N = 48) showed a pre-nominal prediction effect, but evidence for the article prediction mismatch hypothesis was inconclusive. Informed by exploratory analyses and power analyses, direct replication Experiment 2 (N = 80) yielded evidence for article prediction mismatch at a newly pre-registered occipital region-of-interest. However, at frontal and posterior channels, unexpectedly definite articles also elicited a gender-mismatch effect, and this support for the noun prediction revision hypothesis was further strengthened by exploratory analyses: ERPs elicited by gender-mismatching articles correlated with incurred constraint towards a new noun (next-word entropy), and N400s for initially unpredictable nouns decreased when articles made them more predictable. By demonstrating its dual nature, our results reconcile two prevalent explanations of the pre-nominal prediction effect.

Dissociating activation and integration of discourse referents: Evidence from ERPs and oscillations.

A key challenge in understanding stories and conversations is the comprehension of 'anaphora', words that refer back to previously mentioned words or concepts ('antecedents'). In psycholinguistic theories, anaphor comprehension involves the initial activation of the antecedent and its subsequent integration into the unfolding representation of the narrated event. A recent proposal suggests that these processes draw upon the brain's recognition memory and language networks, respectively, and may be dissociable in patterns of neural oscillatory synchronization (Nieuwland & Martin, 2017). We addressed this proposal in an electroencephalogram (EEG) study with pre-registered data acquisition and analyses, using event-related potentials and neural oscillations. Dutch participants read two-sentence mini stories containing proper names, which were repeated or new (ease of activation) and semantically coherent or incoherent with the preceding discourse (ease of integration). Repeated names elicited smaller N400 and Late Positive Component amplitude than new names, and also an increase in theta-band (4-7 Hz) synchronization, which was largest around 240-450 ms after name onset. Discourse-coherent names elicited an increase in gamma-band (60-80 Hz) synchronization compared to discourse-incoherent names. This effect was largest around 690-1000 ms after name onset and exploratory beamformer analysis suggested a left frontal source. We argue that the initial activation and subsequent discourse-level integration of referents can be dissociated with event-related EEG activity, and are associated with respectively theta- and gamma-band activity. These findings further establish the link between memory and language through neural oscillations.

Dissociable effects of prediction and integration during language comprehension: evidence from a large-scale study using brain potentials.

Composing sentence meaning is easier for predictable words than for unpredictable words. Are predictable words genuinely predicted, or simply more plausible and therefore easier to integrate with sentence context? We addressed this persistent and fundamental question using data from a recent, large-scale (n = 334) replication study, by investigating the effects of word predictability and sentence plausibility on the N400, the brain's electrophysiological index of semantic processing. A spatio-temporally fine-grained mixed-effect multiple regression analysis revealed overlapping effects of predictability and plausibility on the N400, albeit with distinct spatio-temporal profiles. Our results challenge the view that the predictability-dependent N400 reflects the effects of either prediction or integration, and suggest that semantic facilitation of predictable words arises from a cascade of processes that activate and integrate word meaning with context into a sentence-level meaning. This article is part of the theme issue 'Towards mechanistic models of meaning composition'.

Distinguishing Old From New Referents During Discourse Comprehension: Evidence From ERPs and Oscillations.

In this EEG study, we used pre-registered and exploratory ERP and time-frequency analyses to investigate the resolution of anaphoric and non-anaphoric noun phrases during discourse comprehension. Participants listened to story contexts that described two antecedents, and subsequently read a target sentence with a critical noun phrase that lexically matched one antecedent ('old'), matched two antecedents ('ambiguous'), partially matched one antecedent in terms of semantic features ('partial-match'), or introduced another referent (non-anaphoric, 'new'). After each target sentence, participants judged whether the noun referred back to an antecedent (i.e., an 'old/new' judgment), which was easiest for ambiguous nouns and hardest for partially matching nouns. The noun-elicited N400 ERP component demonstrated initial sensitivity to repetition and semantic overlap, corresponding to repetition and semantic priming effects, respectively. New and partially matching nouns both elicited a subsequent frontal positivity, which suggested that partially matching anaphors may have been processed as new nouns temporarily. ERPs in an even later time window and ERPs time-locked to sentence-final words suggested that new and partially matching nouns had different effects on comprehension, with partially matching nouns incurring additional processing costs up to the end of the sentence. In contrast to the ERP results, the time-frequency results primarily demonstrated sensitivity to noun repetition, and did not differentiate partially matching anaphors from new nouns. In sum, our results show the ERP and time-frequency effects of referent repetition during discourse comprehension, and demonstrate the potentially demanding nature of establishing the anaphoric meaning of a novel noun.

Do 'early' brain responses reveal word form prediction during language comprehension? A critical review.

Current theories of language comprehension posit that readers and listeners routinely try to predict the meaning but also the visual or sound form of upcoming words. Whereas most neuroimaging studies on word prediction focus on the N400 ERP or its magnetic equivalent, various studies claim that word form prediction manifests itself in 'early', pre-N400 brain responses (e.g., ELAN, M100, P130, N1, P2, N200/PMN, N250). Modulations of these components are often taken as evidence that word form prediction impacts early sensory processes (the sensory hypothesis) or, alternatively, the initial stages of word recognition before word meaning is integrated with sentence context (the recognition hypothesis). Here, I comprehensively review studies on sentence- or discourse-level language comprehension that report such effects of prediction on early brain responses. I conclude that the reported evidence for the sensory hypothesis or word recognition hypothesis is weak and inconsistent, and highlight the urgent need for replication of previous findings. I discuss the implications and challenges to current theories of linguistic prediction and suggest avenues for future research.

Large-scale replication study reveals a limit on probabilistic prediction in language comprehension.

Do people routinely pre-activate the meaning and even the phonological form of upcoming words? The most acclaimed evidence for phonological prediction comes from a 2005 Nature Neuroscience publication by DeLong, Urbach and Kutas, who observed a graded modulation of electrical brain potentials (N400) to nouns and preceding articles by the probability that people use a word to continue the sentence fragment ('cloze'). In our direct replication study spanning 9 laboratories (N=334), pre-registered replication-analyses and exploratory Bayes factor analyses successfully replicated the noun-results but, crucially, not the article-results. Pre-registered single-trial analyses also yielded a statistically significant effect for the nouns but not the articles. Exploratory Bayesian single-trial analyses showed that the article-effect may be non-zero but is likely far smaller than originally reported and too small to observe without very large sample sizes. Our results do not support the view that readers routinely pre-activate the phonological form of predictable words.

Neural Oscillations and a Nascent Corticohippocampal Theory of Reference.

The ability to use words to refer to the world is vital to the communicative power of human language. In particular, the anaphoric use of words to refer to previously mentioned concepts (antecedents) allows dialogue to be coherent and meaningful. Psycholinguistic theory posits that anaphor comprehension involves reactivating a memory representation of the antecedent. Whereas this implies the involvement of recognition memory or the mnemonic subroutines by which people distinguish old from new, the neural processes for reference resolution are largely unknown. Here, we report time-frequency analysis of four EEG experiments to reveal the increased coupling of functional neural systems associated with referentially coherent expressions compared with referentially problematic expressions. Despite varying in modality, language, and type of referential expression, all experiments showed larger gamma-band power for referentially coherent expressions compared with referentially problematic expressions. Beamformer analysis in high-density Experiment 4 localized the gamma-band increase to posterior parietal cortex around 400-600 msec after anaphor onset and to frontotemporal cortex around 500-1000 msec. We argue that the observed gamma-band power increases reflect successful referential binding and resolution, which links incoming information to antecedents through an interaction between the brain's recognition memory networks and frontotemporal language network. We integrate these findings with previous results from patient and neuroimaging studies, and we outline a nascent corticohippocampal theory of reference.

On predicting form and meaning in a second language.

We used event-related potentials (ERP) to investigate whether Spanish-English bilinguals preactivate form and meaning of predictable words. Participants read high-cloze sentence contexts (e.g., "The student is going to the library to borrow a . . ."), followed by the predictable word (book), a word that was form-related (hook) or semantically related (page) to the predictable word, or an unrelated word (sofa). Word stimulus onset synchrony (SOA) was 500 ms (Experiment 1) or 700 ms (Experiment 2). In both experiments, all nonpredictable words elicited classic N400 effects. Form-related and unrelated words elicited similar N400 effects. Semantically related words elicited smaller N400s than unrelated words, which however, did not depend on cloze value of the predictable word. Thus, we found no N400 evidence for preactivation of form or meaning at either SOA, unlike native-speaker results (Ito, Corley et al., 2016). However, non-native speakers did show the post-N400 posterior positivity (LPC effect) for form-related words like native speakers, but only at the slower SOA. This LPC effect increased gradually with cloze value of the predictable word. We do not interpret this effect as necessarily demonstrating prediction, but rather as evincing combined effects of top-down activation (contextual meaning) and bottom-up activation (form similarity) that result in activation of unseen words that fit the context well, thereby leading to an interpretation conflict reflected in the LPC. Although there was no evidence that non-native speakers preactivate form or meaning, non-native speakers nonetheless appear to use bottom-up and top-down information to constrain incremental interpretation much like native speakers do. (PsycINFO Database Record

Can structural priming answer the important questions about language?

Structural priming makes a valuable contribution to psycholinguistics, but it taps into implicit memory representations and processes that may differ from what is deployed during online language processing. As a result, the strength of inductive inference regarding linguistic representation is rather limited. We question whether implicit memory for language can and should be equated with linguistic representation or with language processing.

Understanding Counterfactuality: A Review of Experimental Evidence for the Dual Meaning of Counterfactuals.

Cognitive and linguistic theories of counterfactual language comprehension assume that counterfactuals convey a dual meaning. Subjunctive-counterfactual conditionals (e.g., 'If Tom had studied hard, he would have passed the test') express a supposition while implying the factual state of affairs (Tom has not studied hard and failed). The question of how counterfactual dual meaning plays out during language processing is currently gaining interest in psycholinguistics. Whereas numerous studies using offline measures of language processing consistently support counterfactual dual meaning, evidence coming from online studies is less conclusive. Here, we review the available studies that examine online counterfactual language comprehension through behavioural measurement (self-paced reading times, eye-tracking) and neuroimaging (electroencephalography, functional magnetic resonance imaging). While we argue that these studies do not offer direct evidence for the online computation of counterfactual dual meaning, they provide valuable information about the way counterfactual meaning unfolds in time and influences successive information processing. Further advances in research on counterfactual comprehension require more specific predictions about how counterfactual dual meaning impacts incremental sentence processing.

Neural correlates of verbal joint action: ERPs reveal common perception and action systems in a shared-Stroop task.

Recent social-cognitive research suggests that the anticipation of co-actors' actions influences people's mental representations. However, the precise nature of such representations is still unclear. In this study we investigated verbal joint representations in a delayed Stroop paradigm, where each participant responded to one color after a short delay. Participants either performed the task as a single actor (single-action, Experiment 1), or they performed it together (joint-action, Experiment 2). We investigated effects of co-actors' actions on the ERP components associated with perceptual conflict (Go N2) and response selection (P3b). Compared to single-action, joint-action reduced the N2 amplitude congruency effect when participants had to respond (Go trials), indicating that representing a co-actor's utterance helped to dissociate action codes and attenuated perceptual conflict for the responding participant. Yet, on NoGo trials the centro-parietal P3 (P3b) component amplitude increased for joint-action, suggesting that participants mapped the stimuli onto the co-actor's upcoming response as if it were their own response. We conclude that people represent others' utterances similarly to the way they represent their own utterances, and that shared perception-action codes for self and others can sometimes reduce, rather than enhance, perceptual conflict.

Incremental comprehension of spoken quantifier sentences: Evidence from brain potentials.

Do people incrementally incorporate the meaning of quantifier expressions to understand an unfolding sentence? Most previous studies concluded that quantifiers do not immediately influence how a sentence is understood based on the observation that online N400-effects differed from offline plausibility judgments. Those studies, however, used serial visual presentation (SVP), which involves unnatural reading. In the current ERP-experiment, we presented spoken positive and negative quantifier sentences ("Practically all/practically no postmen prefer delivering mail, when the weather is good/bad during the day"). Different from results obtained in a previously reported SVP-study (Nieuwland, 2016) sentence truth-value N400 effects occurred in positive and negative quantifier sentences alike, reflecting fully incremental quantifier comprehension. This suggests that the prosodic information available during spoken language comprehension supports the generation of online predictions for upcoming words and that, at least for quantifier sentences, comprehension of spoken language may proceed more incrementally than comprehension during SVP reading.