Publications

Abstract

When perceiving the world around us, we are constantly integrating pieces of information. The integrated experience consists of more than just the sum of its parts. For example, visual scenes are defined by a collection of objects as well as the spatial relations amongst them and sentence meaning is computed based on individual word semantic but also syntactic configuration. Having quantitative models of such integrated representations can help evaluate cognitive models of both language and scene perception. Here, we focus on language, and use a behavioral measure of perceived similarity as an approximation of integrated meaning representations. We collected similarity judgments of 200 subjects rating nouns or transitive sentences through an online multiple arrangement task. We find that perceived similarity between sentences is most strongly modulated by the semantic action category of the main verb. In addition, we show how non-negative matrix factorization of similarity judgment data can reveal multiple underlying dimensions reflecting both semantic as well as relational role information. Finally, we provide an example of how similarity judgments on sentence stimuli can serve as a point of comparison for artificial neural networks models (ANNs) by comparing our behavioral data against sentence similarity extracted from three state-of-the-art ANNs. Overall, our method combining the multiple arrangement task on sentence stimuli with matrix factorization can capture relational information emerging from integration of multiple words in a sentence even in the presence of strong focus on the verb.

Abstract

Recent artificial neural networks that process natural language achieve unprecedented performance in tasks requiring sentence-level understanding. As such, they could be interesting models of the integration of linguistic information in the human brain. We review works that compare these artificial language models with human brain activity and we assess the extent to which this approach has improved our understanding of the neural processes involved in natural language comprehension. Two main results emerge. First, the neural representation of word meaning aligns with the context-dependent, dense word vectors used by the artificial neural networks. Second, the processing hierarchy that emerges within artificial neural networks broadly matches the brain, but is surprisingly inconsistent across studies. We discuss current challenges in establishing artificial neural networks as process models of natural language comprehension. We suggest exploiting the highly structured representational geometry of artificial neural networks when mapping representations to brain data.

Abstract

Background: Despite a pervasive cortico-centric view in cognitive neuroscience, subcortical structures including the thalamus have been shown to be increasingly involved in higher cognitive functions. Previous structural and functional imaging studies demonstrated cortico-thalamo-cortical loops which may support various cognitive functions including language. However, large-scale functional connectivity of the thalamus during language tasks has not been examined before. Methods: The present study employed meta-analytic connectivity modeling to identify language-related coactivation patterns of the left and right thalami. The left and right thalami were used as regions of interest to search the BrainMap functional database for neuroimaging experiments with healthy participants reporting language-related activations in each region of interest. Activation likelihood estimation analyses were then carried out on the foci extracted from the identified studies to estimate functional convergence for each thalamus. A functional decoding analysis based on the same database was conducted to characterize thalamic contributions to different language functions. Results: The results revealed bilateral frontotemporal and bilateral subcortical (basal ganglia) coactivation patterns for both the left and right thalami, and also right cerebellar coactivations for the left thalamus, during language processing. In light of previous empirical studies and theoretical frameworks, the present connectivity and functional decoding findings suggest that cortico-subcortical-cerebellar-cortical loops modulate and fine-tune information transfer within the bilateral frontotemporal cortices during language processing, especially during production and semantic operations, but also other language (e.g., syntax, phonology) and cognitive operations (e.g., attention, cognitive control). Conclusion: The current findings show that the language-relevant network extends beyond the classical left perisylvian cortices and spans bilateral cortical, bilateral subcortical (bilateral thalamus, bilateral basal ganglia) and right cerebellar regions.

Abstract

The language faculty is physically realized in the neurobiological infrastructure of the human brain. Despite significant efforts, an integrated understanding of this system remains a formidable challenge. What is missing from most theoretical accounts is a specification of the neural mechanisms that implement language function. Computational models that have been put forward generally lack an explicit neurobiological foundation. We propose a neurobiologically informed causal modeling approach which offers a framework for how to bridge this gap. A neurobiological causal model is a mechanistic description of language processing that is grounded in, and constrained by, the characteristics of the neurobiological substrate. It intends to model the generators of language behavior at the level of implementational causality. We describe key features and neurobiological component parts from which causal models can be built and provide guidelines on how to implement them in model simulations. Then we outline how this approach can shed new light on the core computational machinery for language, the long-term storage of words in the mental lexicon and combinatorial processing in sentence comprehension. In contrast to cognitive theories of behavior, causal models are formulated in the “machine language” of neurobiology which is universal to human cognition. We argue that neurobiological causal modeling should be pursued in addition to existing approaches. Eventually, this approach will allow us to develop an explicit computational neurobiology of language.

Abstract

The neural correlates of sentence production have been mostly studied with constraining task paradigms that introduce artificial task effects. In this study, we aimed to gain a better understanding of syntactic processing in spontaneous production vs. naturalistic comprehension. We extracted word-by-word metrics of phrase-structure building with top-down and bottom-up parsers that make different hypotheses about the timing of structure building. In comprehension, structure building proceeded in an integratory fashion and led to an increase in activity in posterior temporal and inferior frontal areas. In production, structure building was anticipatory and predicted an increase in activity in the inferior frontal gyrus. Newly developed production-specific parsers highlighted the anticipatory and incremental nature of structure building in production, which was confirmed by a converging analysis of the pausing patterns in speech. Overall, the results showed that the unfolding of syntactic processing diverges between speaking and listening.

Abstract

The neural infrastructure for sentence production and comprehension has been found to be mostly shared. The same regions are engaged during speaking and listening, with some differences in how strongly they activate depending on modality. In this study, we investigated how modality affects the connectivity between regions previously found to be involved in syntactic processing across modalities. We determined how constituent size and modality affected the connectivity of the pars triangularis of the left inferior frontal gyrus (LIFG) and of the left posterior temporal lobe (LPTL) with the pars opercularis of the LIFG, the anterior temporal lobe (LATL) and the rest of the brain. We found that constituent size reliably increased the connectivity across these frontal and temporal ROIs. Connectivity between the two LIFG regions and the LPTL was enhanced as a function of constituent size in both modalities, and it was upregulated in production possibly because of linearization and motor planning in the frontal cortex. The connectivity of both ROIs with the LATL was lower and only enhanced for larger constituent sizes, suggesting a contributing role of the LATL in sentence processing in both modalities. These results thus show that the connectivity among fronto-temporal regions is upregulated for syntactic structure building in both sentence production and comprehension, providing further evidence for accounts of shared neural resources for sentence-level processing across modalities.

Abstract

The neural representations for compositional processing have so far been mostly studied during sentence comprehension. In an fMRI study of sentence production, we investigated the brain representations for compositional processing during speaking. We used a rapid serial visual presentation sentence recall paradigm to elicit sentence production from the conceptual memory of an event. With voxel-wise encoding models, we probed the specificity of the compositional structure built during the production of each sentence, comparing an unstructured model of word meaning without relational information with a model that encodes abstract thematic relations and a model encoding event-specific relational structure. Whole-brain analyses revealed that sentence meaning at different levels of specificity was encoded in a large left frontal-parietal-temporal network. A comparison with semantic structures composed during the comprehension of the same sentences showed similarly distributed brain activity patterns. An ROI analysis over left fronto-temporal language parcels showed that event-specific relational structure above word-specific information was encoded in the left inferior frontal gyrus. Overall, we found evidence for the encoding of sentence meaning during sentence production in a distributed brain network and for the encoding of event-specific semantic structures in the left inferior frontal gyrus.

Abstract

Language is inherently multimodal. In spoken languages, combined spoken and visual signals (e.g., co-speech gestures) are an integral part of linguistic structure and language representation. This requires an extension of the parallel architecture, which needs to include the visual signals concomitant to speech. We present the evidence for the multimodality of language. In addition, we propose that distributional semantics might provide a format for integrating speech and co-speech gestures in a common semantic representation.

Abstract

How we combine minimal linguistic units into larger structures remains an unresolved topic in neuroscience. Language processing involves the abstract construction of ‘vertical’ and ‘horizontal’ information simultaneously (e.g., phrase structure, morphological agreement), but previous paradigms have been constrained in isolating only one type of composition and have utilized poor spatiotemporal resolution. Using intracranial recordings, we report multiple experiments designed to separate phrase structure from morphosyntactic agreement. Epilepsy patients (n = 10) were presented with auditory two-word phrases grouped into pseudoword-verb (‘trab run’) and pronoun-verb either with or without Person agreement (‘they run’ vs. ‘they runs’). Phrase composition and Person violations both resulted in significant increases in broadband high gamma activity approximately 300ms after verb onset in posterior middle temporal gyrus (pMTG) and posterior superior temporal sulcus (pSTS), followed by inferior frontal cortex (IFC) at 500ms. While sites sensitive to only morphosyntactic violations were distributed, those sensitive to both composition types were generally confined to pSTS/pMTG and IFC. These results indicate that posterior temporal cortex shows the earliest sensitivity for hierarchical linguistic structure across multiple dimensions, providing neural resources for distinct windows of composition. This region is comprised of sparsely interwoven heterogeneous constituents that afford cortical search spaces for dissociable syntactic relations.

Abstract

During communication in real-life settings, our brain often needs to integrate auditory and visual information, and at the same time actively focus on the relevant sources of information, while ignoring interference from irrelevant events. The interaction between integration and attention processes remains poorly understood. Here, we use rapid invisible frequency tagging (RIFT) and magnetoencephalography (MEG) to investigate how attention affects auditory and visual information processing and integration, during multimodal communication. We presented human participants (male and female) with videos of an actress uttering action verbs (auditory; tagged at 58 Hz) accompanied by two movie clips of hand gestures on both sides of fixation (attended stimulus tagged at 65 Hz; unattended stimulus tagged at 63 Hz). Integration difficulty was manipulated by a lower-order auditory factor (clear/degraded speech) and a higher-order visual semantic factor (matching/mismatching gesture). We observed an enhanced neural response to the attended visual information during degraded speech compared to clear speech. For the unattended information, the neural response to mismatching gestures was enhanced compared to matching gestures. Furthermore, signal power at the intermodulation frequencies of the frequency tags, indexing non-linear signal interactions, was enhanced in left frontotemporal and frontal regions. Focusing on LIFG (Left Inferior Frontal Gyrus), this enhancement was specific for the attended information, for those trials that benefitted from integration with a matching gesture. Together, our results suggest that attention modulates audiovisual processing and interaction, depending on the congruence and quality of the sensory input.

Abstract

Words are not processed in isolation; instead, they are commonly embedded in phrases and sentences. The sentential context influences the perception and processing of a word. However, how this is achieved by brain processes and whether predictive mechanisms underlie this process remain a debated topic. Here, we employed an experimental paradigm in which we orthogonalized sentence context constraints and predictive validity, which was defined as the ratio of congruent to incongruent sentence endings within the experiment. While recording electroencephalography, participants read sentences with three levels of sentential context constraints (high, medium, and low). Participants were also separated into two groups that differed in their ratio of valid congruent to incongruent target words that could be predicted from the sentential context. For both groups, we investigated modulations of alpha power before, and N400 amplitude modulations after target word onset. The results reveal that the N400 amplitude gradually decreased with higher context constraints and cloze probability. In contrast, alpha power was not significantly affected by context constraint. Neither the N400 nor alpha power were significantly affected by changes in predictive validity.

Abstract

There is evidence from both behavior and brain activity that the way information is structured, through the use of focus, can up-regulate processing of focused constituents, likely to give prominence to the relevant aspects of the input. This is hypothesized to be universal, regardless of the different ways in which languages encode focus. In order to test this universalist hypothesis, we need to go beyond the more familiar linguistic strategies for marking focus, such as by means of intonation or specific syntactic structures (e.g., it-clefts). Therefore, in this study, we examine Makhuwa-Enahara, a Bantu language spoken in northern Mozambique, which uniquely marks focus through verbal conjugation. The participants were presented with sentences that consisted of either a semantically anomalous constituent or a semantically nonanomalous constituent. Moreover, focus on this particular constituent could be either present or absent. We observed a consistent pattern: Focused information generated a more negative N400 response than the same information in nonfocus position. This demonstrates that regardless of how focus is marked, its consequence seems to result in an upregulation of processing of information that is in focus.

Abstract

In Swedish, words are associated with either of two pitch contours known as Accent 1 and Accent 2. Using a psychometric test, we investigated how listeners judge pitch accent violations while interpreting discourse. Forty native speakers of Central Swedish were presented with auditory dialogues, where test words were appropriately or inappropriately accented in a given context, and asked to judge the correctness of sentences containing the test words. Data indicated a statistically significant effect of wrong accent pattern on the correctness judgment. Both Accent 1 and Accent 2 violations interfered with the coherent interpretation of discourse and were judged as incorrect by the listeners. Moreover, there was a statistically significant difference in the perceived correctness between the accent patterns. Accent 2 violations led to a lower correctness score compared to Accent 1 violations, indicating that the listeners were more sensitive to pitch accent violations in Accent 2 words than in Accent 1 words. This result is in line with the notion that Accent 2 is marked and lexically represented in Central Swedish. Taken together, these findings indicate that listeners use both Accent 1 and Accent 2 to arrive at the correct interpretation of the linguistic input, while assigning varying degrees of relevance to them depending on their markedness.

Abstract

Previous research on the effect of perturbed auditory feedback in speech production has focused on two types of responses. In the short term, speakers generate compensatory motor commands in response to unexpected perturbations. In the longer term, speakers adapt feedforward motor programmes in response to feedback perturbations, to avoid future errors. The current study investigated the relation between these two types of responses to altered auditory feedback. Specifically, it was hypothesised that consistency in previous feedback perturbations would influence whether speakers adapt their feedforward motor programmes. In an altered auditory feedback paradigm, formant perturbations were applied either across all trials (the consistent condition) or only to some trials, whereas the others remained unperturbed (the inconsistent condition). The results showed that speakers’ responses were affected by feedback consistency, with stronger speech changes in the consistent condition compared with the inconsistent condition. Current models of speech-motor control can explain this consistency effect. However, the data also suggest that compensation and adaptation are distinct processes, which are not in line with all current models.

Abstract

In this contribution, the following four questions are discussed: (i) where is meaning?; (ii) what is meaning?; (iii) what is the meaning of mechanism?; (iv) what are the mechanisms of meaning? I will argue that meanings are in the head. Meanings have multiple facets, but minimally one needs to make a distinction between single word meanings (lexical meaning) and the meanings of multi-word utterances. The latter ones cannot be retrieved from memory, but need to be constructed on the fly. A mechanistic account of the meaning-making mind requires an analysis at both a functional and a neural level, the reason being that these levels are causally interdependent. I will show that an analysis exclusively focusing on patterns of brain activation lacks explanatory power. Finally, I shall present an initial sketch of how the dynamic interaction between temporo-parietal areas and inferior frontal cortex might instantiate the interpretation of linguistic utterances in the context of a multimodal setting and ongoing discourse information.

Abstract

In this Review, I propose a multiple-network view for the neurobiological basis of distinctly human language skills. A much more complex picture of interacting brain areas emerges than in the classical neurobiological model of language. This is because using language is more than single-word processing, and much goes on beyond the information given in the acoustic or orthographic tokens that enter primary sensory cortices. This requires the involvement of multiple networks with functionally nonoverlapping contributions

Abstract

Prediction in language has traditionally been studied using
simple designs in which neural responses to expected
and unexpected words are compared in a categorical
fashion. However, these designs have been contested
as being ‘prediction encouraging’, potentially exaggerating
the importance of prediction in language understanding.
A few recent studies have begun to address
these worries by using model-based approaches to probe
the effects of linguistic predictability in naturalistic stimuli
(e.g. continuous narrative). However, these studies
so far only looked at very local forms of prediction, using
models that take no more than the prior two words into
account when computing a word’s predictability. Here,
we extend this approach using a state-of-the-art neural
language model that can take roughly 500 times longer
linguistic contexts into account. Predictability estimates
fromthe neural network offer amuch better fit to EEG data
from subjects listening to naturalistic narrative than simpler
models, and reveal strong surprise responses akin to
the P200 and N400. These results show that predictability
effects in language are not a side-effect of simple designs,
and demonstrate the practical use of recent advances
in AI for the cognitive neuroscience of language.

Abstract

Human language processing involves combinatorial operations that make human communication stand out in the animal kingdom. These operations rely on a dynamic interplay between the inferior frontal and the posterior temporal cortices. Using source reconstructed magnetoencephalography, we tracked language processing in the brain, in order to investigate how individual words are interpreted when part of sentence context. The large sample size in this study (n = 68) allowed us to assess how event-related activity is associated across distinct cortical areas, by means of inter-areal co-modulation within an individual. We showed that, within 500 ms of seeing a word, the word's lexical information has been retrieved and unified with the sentence context. This does not happen in a strictly feed-forward manner, but by means of co-modulation between the left posterior temporal cortex (LPTC) and left inferior frontal cortex (LIFC), for each individual word. The co-modulation of LIFC and LPTC occurs around 400 ms after the onset of each word, across the progression of a sentence. Moreover, these core language areas are supported early on by the attentional network. The results provide a detailed description of the temporal orchestration related to single word processing in the context of ongoing language.

Abstract

How does the human brain combine a finite number of words to form an infinite variety of sentences? According to the Memory, Unification and Control (MUC) model, sentence processing requires long-range feedback from the left inferior frontal cortex (LIFC) to left posterior temporal cortex (LPTC). Single word processing however may only require feedforward propagation of semantic information from sensory regions to LPTC. Here we tested the claim that long-range feedback is required for sentence processing by reducing visual awareness of words using a masking technique. Masking disrupts feedback processing while leaving feedforward processing relatively intact. Previous studies have shown that masked single words still elicit an N400 ERP effect, a neural signature of semantic incongruency. However, whether multiple words can be combined to form a sentence under reduced levels of awareness is controversial. To investigate this issue, we performed two experiments in which we measured electroencephalography (EEG) while 40 subjects performed a masked priming task. Words were presented either successively or simultaneously, thereby forming a short sentence that could be congruent or incongruent with a target picture. This sentence condition was compared with a typical single word condition. In the masked condition we only found an N400 effect for single words, whereas in the unmasked condition we observed an N400 effect for both unmasked sentences and single words. Our findings suggest that long-range feedback processing is required for sentence processing, but not for single word processing.

Abstract

This dataset, colloquially known as the Mother Of Unification Studies (MOUS) dataset, contains multimodal neuroimaging data that has been acquired from 204 healthy human subjects. The neuroimaging protocol consisted of magnetic resonance imaging (MRI) to derive information at high spatial resolution about brain anatomy and structural connections, and functional data during task, and at rest. In addition, magnetoencephalography (MEG) was used to obtain high temporal resolution electrophysiological measurements during task, and at rest. All subjects performed a language task, during which they processed linguistic utterances that either consisted of normal or scrambled sentences. Half of the subjects were reading the stimuli, the other half listened to the stimuli. The resting state measurements consisted of 5 minutes eyes-open for the MEG and 7 minutes eyes-closed for fMRI. The neuroimaging data, as well as the information about the experimental events are shared according to the Brain Imaging Data Structure (BIDS) format. This unprecedented neuroimaging language data collection allows for the investigation of various aspects of the neurobiological correlates of language.

Abstract

Speakers are influenced by the linguistic context: hearing one syntactic alternative leads to an increased chance that the speaker will repeat this structure in the subsequent utterance (i.e., syntactic priming, or structural persistence). Top-down influences, such as whether a conversation partner (or, interlocutor) is present, may modulate the degree to which syntactic priming occurs. In the current study, we indeed show that the magnitude of syntactic alignment increases when speakers are interacting with an interlocutor as opposed to doing the experiment alone. The structural persistence effect for passive sentences is stronger in the presence of an interlocutor than when no interlocutor is present (i.e., when the participant is primed by a recording). We did not find evidence, however, that a speaker’s syntactic priming magnitude is influenced by the degree of their conversation partner’s priming magnitude. Together, these results support a mediated account of syntactic priming, in which syntactic choices are not only affected by preceding linguistic input, but also by top-down influences, such as the speakers’ communicative intent.

Abstract

Interactions between top-down and bottom-up information streams are integral to brain function but challenging to measure noninvasively. Laminar resolution, functional MRI (lfMRI) is sensitive to depth-dependent properties of the blood oxygen level-dependent (BOLD) response, which can be potentially related to top-down and bottom-up signal contributions. In this work, we used lfMRI to dissociate the top-down and bottom-up signal contributions to the left occipitotemporal sulcus (LOTS) during word reading. We further demonstrate that laminar resolution measurements could be used to identify condition-specific distributed networks on the basis of whole-brain connectivity patterns specific to the depth-dependent BOLD signal. The networks corresponded to top-down and bottom-up signal pathways targeting the LOTS during word reading. We show that reading increased the top-down BOLD signal observed in the deep layers of the LOTS and that this signal uniquely related to the BOLD response in other language-critical regions. These results demonstrate that lfMRI can reveal important patterns of activation that are obscured at standard resolution. In addition to differences in activation strength as a function of depth, we also show meaningful differences in the interaction between signals originating from different depths both within a region and with the rest of the brain. We thus show that lfMRI allows the noninvasive measurement of directed interaction between brain regions and is capable of resolving different connectivity patterns at submillimeter resolution, something previously considered to be exclusively in the domain of invasive recordings.

Abstract

A commonly held assumption in cognitive neuroscience is that, because measures of human brain function are closer to underlying biology than distal indices of behavior/cognition, they hold more promise for uncovering genetic pathways. Supporting this view is an influential fMRI-based study of sentence reading/listening by Pinel et al. (2012), who reported that common DNA variants in specific candidate genes were associated with altered neural activation in language-related regions of healthy individuals that carried them. In particular, different single-nucleotide polymorphisms (SNPs) of FOXP2 correlated with variation in task-based activation in left inferior frontal and precentral gyri, whereas a SNP at the KIAA0319/TTRAP/THEM2 locus was associated with variable functional asymmetry of the superior temporal sulcus. Here, we directly test each claim using a closely matched neuroimaging genetics approach in independent cohorts comprising 427 participants, four times larger than the original study of 94 participants. Despite demonstrating power to detect associations with substantially smaller effect sizes than those of the original report, we do not replicate any of the reported associations. Moreover, formal Bayesian analyses reveal substantial to strong evidence in support of the null hypothesis (no effect). We highlight key aspects of the original investigation, common to functional neuroimaging genetics studies, which could have yielded elevated false-positive rates. Genetic accounts of individual differences in cognitive functional neuroimaging are likely to be as complex as behavioral/cognitive tests, involving many common genetic variants, each of tiny effect. Reliable identification of true biological signals requires large sample sizes, power calculations, and validation in independent cohorts with equivalent paradigms.

SIGNIFICANCE STATEMENT A pervasive idea in neuroscience is that neuroimaging-based measures of brain function, being closer to underlying neurobiology, are more amenable for uncovering links to genetics. This is a core assumption of prominent studies that associate common DNA variants with altered activations in task-based fMRI, despite using samples (10–100 people) that lack power for detecting the tiny effect sizes typical of genetically complex traits. Here, we test central findings from one of the most influential prior studies. Using matching paradigms and substantially larger samples, coupled to power calculations and formal Bayesian statistics, our data strongly refute the original findings. We demonstrate that neuroimaging genetics with task-based fMRI should be subject to the same rigorous standards as studies of other complex traits.

Abstract

New linguistic information must be integrated into our existing language system. Using a novel experimental task that incorporates a syntactic priming paradigm into artificial language learning, we investigated how new grammatical regularities and words are learned. This innovation allowed us to control the language input the learner received, while the syntactic priming paradigm provided insight into the nature of the underlying syntactic processing machinery. The results of the present study pointed to facilitatory syntactic processing effects within the first days of learning: Syntactic and lexical priming effects revealed participants’ sensitivity to both novel words and word orders. This suggested that novel syntactic structures and their meaning (form–function mapping) can be acquired rapidly through incidental learning. More generally, our study indicated similar mechanisms for learning and processing in both artificial and natural languages, with implications for the relationship between first and second language learning.

Abstract

Semantic unification during sentence comprehension has been associated with amplitude change of the N400 in event-related potential (ERP) studies, and activation in the left inferior frontal gyrus (IFG) in functional magnetic resonance imaging (fMRI) studies. However, the specificity of this activation to semantic unification remains unknown. To more closely examine the brain processes involved in semantic unification, we employed simultaneous EEG-fMRI to time-lock the semantic unification related N400 change, and integrated trial-by-trial variation in both N400 and BOLD change beyond the condition-level BOLD change difference measured in traditional fMRI analyses. Participants read sentences in which semantic unification load was parametrically manipulated by varying cloze probability. Separately, ERP and fMRI results replicated previous findings, in that semantic unification load parametrically modulated the amplitude of N400 and cortical activation. Integrated EEG-fMRI analyses revealed a different pattern in which functional activity in the left IFG and bilateral supramarginal gyrus (SMG) was associated with N400 amplitude, with the left IFG activation and bilateral SMG activation being selective to the condition-level and trial-level of semantic unification load, respectively. By employing the EEG-fMRI integrated analyses, this study among the first sheds light on how to integrate trial-level variation in language comprehension.