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

Predictive language processing is often studied by measuring eye movements as participants look at objects on a computer screen while they listen to spoken sentences. The use of this variant of the visual world paradigm has shown that information encountered by a listener at a spoken verb can give rise to anticipatory eye movements to a target object, which is taken to indicate that people predict upcoming words. The ecological validity of such findings remains questionable, however, because these computer experiments used two-dimensional (2D) stimuli that are mere abstractions of real world objects. Here we present a visual world paradigm study in a three-dimensional (3D) immersive virtual reality environment. Despite significant changes in the stimulus material and the different mode of stimulus presentation, language-mediated anticipatory eye movements were observed. These findings thus indicate prediction of upcoming words in language comprehension in a more naturalistic setting where natural depth cues are preserved. Moreover, the results confirm the feasibility of using eye-tracking in rich and multimodal 3D virtual environments.

Abstract

When talking, speakers continuously monitor and use the auditory feedback of their own voice to control and inform speech production processes. When speakers are provided with auditory feedback that is perturbed in real time, most of them compensate for this by opposing the feedback perturbation. But some speakers follow the perturbation. In the current study, we investigated whether the state of the speech production system at perturbation onset may determine what type of response (opposing or following) is given. The results suggest that whether a perturbation-related response is opposing or following depends on ongoing fluctuations of the production system: It initially responds by doing the opposite of what it was doing. This effect and the non-trivial proportion of following responses suggest that current production models are inadequate: They need to account for why responses to unexpected sensory feedback depend on the production-system’s state at the time of perturbation.

Abstract

Speaking is a complex motor skill which requires near instantaneous integration of sensory and motor-related information. Current theory hypothesizes a complex interplay between motor and auditory processes during speech production, involving the online comparison of the speech output with an internally generated forward model. To examine the neural correlates of this intricate interplay between sensory and motor processes, the current study uses altered auditory feedback (AAF) in combination with magnetoencephalography (MEG). Participants vocalized the vowel/e/and heard auditory feedback that was temporarily pitch-shifted by only 25 cents, while neural activity was recorded with MEG. As a control condition, participants also heard the recordings of the same auditory feedback that they heard in the first half of the experiment, now without vocalizing. The participants were not aware of any perturbation of the auditory feedback. We found auditory cortical areas responded more strongly to the pitch shifts during vocalization. In addition, auditory feedback perturbation resulted in spectral power increases in the θ and lower β bands, predominantly in sensorimotor areas. These results are in line with current models of speech production, suggesting auditory cortical areas are involved in an active comparison between a forward model's prediction and the actual sensory input. Subsequently, these areas interact with motor areas to generate a motor response. Furthermore, the results suggest that θ and β power increases support auditory-motor interaction, motor error detection and/or sensory prediction processing.

Abstract

Participants’ performance differs when conducting a task in the presence of a secondary individual, moreover the opinion the participant has of this individual also plays a role. Using EEG, we investigated how previous interactions with, and evaluations of, an avatar in virtual reality subsequently influenced attentional allocation to the face of that avatar. We focused on changes in the alpha activity as an index of attentional allocation. We found that the onset of an avatar’s face whom the participant had developed a rapport with induced greater alpha suppression. This suggests greater attentional resources are allocated to the interacted-with avatars. The evaluative ratings of the avatar induced a U-shaped change in alpha suppression, such that participants paid most attention when the avatar was rated as average. These results suggest that attentional allocation is an important element of how behaviour is altered in the presence of a secondary individual and is modulated by our opinion of that individual.

Abstract

Low-frequency neural entrainment to rhythmic input
has been hypothesized as a canonical mechanism
that shapes sensory perception in time. Neural
entrainment is deemed particularly relevant for
speech analysis, as it would contribute to the extraction
of discrete linguistic elements from continuous
acoustic signals. However, its causal influence in
speech perception has been difficult to establish.
Here, we provide evidence that oscillations build temporal
predictions about the duration of speech tokens
that affect perception. Using magnetoencephalography
(MEG), we studied neural dynamics during
listening to sentences that changed in speech rate.
Weobserved neural entrainment to preceding speech
rhythms persisting for several cycles after the change
in rate. The sustained entrainment was associated
with changes in the perceived duration of the last
word’s vowel, resulting in the perception of words
with different meanings. These findings support oscillatory
models of speech processing, suggesting that
neural oscillations actively shape speech perception.

Abstract

Neural oscillations may be instrumental for the tracking and segmentation of continuous speech. Earlier work has suggested that delta, theta and gamma oscillations entrain to the speech rhythm. We used magnetoencephalography and a large sample of 102 participants to investigate oscillatory entrainment to speech, and observed robust entrainment of delta and theta activity, and weak group-level gamma entrainment. We show that the peak frequency and the hemispheric lateralisation of the entrainment are subject to considerable individual variability. The first finding may support the involvement of intrinsic oscillations in entrainment, and the second finding suggests that there is no systematic default right-hemispheric bias for processing acoustic signals on a slow time scale. Although low frequency entrainment to speech is a robust phenomenon, the characteristics of entrainment vary across individuals, and this variation is important for understanding the underlying neural mechanisms of entrainment, as well as its functional significance.

Abstract

Syntactic binding refers to combining words into larger structures. Using EEG, we investigated the neural processes involved in syntactic binding. Participants were auditorily presented two-word sentences (i.e. pronoun and pseudoverb such as ‘I grush’, ‘she grushes’, for which syntactic binding can take place) and wordlists (i.e. two pseudoverbs such as ‘pob grush’, ‘pob grushes’, for which no binding occurs). Comparing these two conditions, we targeted syntactic binding while minimizing contributions of semantic binding and of other cognitive processes such as working memory. We found a converging pattern of results using two distinct analysis approaches: one approach using frequency bands as defined in previous literature, and one data-driven approach in which we looked at the entire range of frequencies between 3-30 Hz without the constraints of pre-defined frequency bands. In the syntactic binding (relative to the wordlist) condition, a power increase was observed in the alpha and beta frequency range shortly preceding the presentation of the target word that requires binding, which was maximal over frontal-central electrodes. Our interpretation is that these signatures reflect that language comprehenders expect the need for binding to occur. Following the presentation of the target word in a syntactic binding context (relative to the wordlist condition), an increase in alpha power maximal over a left lateralized cluster of frontal-temporal electrodes was observed. We suggest that this alpha increase relates to syntactic binding taking place. Taken together, our findings suggest that increases in alpha and beta power are reflections of distinct the neural processes underlying syntactic binding.

Abstract

When we comprehend language, we often do this in rich settings in which we can use many cues to understand what someone is saying. However, it has traditionally been difficult to design experiments with rich three-dimensional contexts that resemble our everyday environments, while maintaining control over the linguistic and non-linguistic information that is available. Here we test the validity of combining electroencephalography (EEG) and Virtual Reality (VR) to overcome this problem. We recorded electrophysiological brain activity during language processing in a well-controlled three-dimensional virtual audiovisual environment. Participants were immersed in a virtual restaurant, while wearing EEG equipment. In the restaurant participants encountered virtual restaurant guests. Each guest was seated at a separate table with an object on it (e.g. a plate with salmon). The restaurant guest would then produce a sentence (e.g. “I just ordered this salmon.”). The noun in the spoken sentence could either match (“salmon”) or mismatch (“pasta”) with the object on the table, creating a situation in which the auditory information was either appropriate or inappropriate in the visual context. We observed a reliable N400 effect as a consequence of the mismatch. This finding validates the combined use of VR and EEG as a tool to study the neurophysiological mechanisms of everyday language comprehension in rich, ecologically valid settings.

Abstract

Learning often occurs in communicative and collaborative settings, yet almost all research into the neural basis of memory relies on participants encoding and retrieving information on their own. We investigated whether learning linguistic labels in a collaborative context at least partly relies on cognitively and neurally distinct representations, as compared to learning in an individual context. Healthy human participants learned labels for sets of abstract shapes in three different tasks. They came up with labels with another person in a collaborative communication task (collaborative condition), by themselves (individual condition), or were given pre-determined unrelated labels to learn by themselves (arbitrary condition). Immediately after learning, participants retrieved and produced the labels aloud during a communicative task in the MRI scanner. The fMRI results show that the retrieval of collaboratively generated labels as compared to individually learned labels engages brain regions involved in understanding others (mentalizing or theory of mind) and autobiographical memory, including the medial prefrontal cortex, the right temporoparietal junction and the precuneus. This study is the first to show that collaboration during encoding affects the neural networks involved in retrieval.

Abstract

Readers and listeners actively predict upcoming words during language processing. These predictions might serve to support the unification of incoming words into sentence context and thus rely on interactions between areas in the language network. In the current magnetoencephalography study, participants read sentences that varied in contextual constraints so that the predictability of the sentence-final words was either high or low. Before the sentence-final words, we observed stronger alpha power suppression for the highly compared with low constraining sentences in the left inferior frontal cortex, left posterior temporal region, and visual word form area. Importantly, the temporal and visual word form area alpha power correlated negatively with left frontal gamma power for the highly constraining sentences. We suggest that the correlation between alpha power decrease in temporal language areas and left prefrontal gamma power reflects the initiation of an anticipatory unification process in the language network.

Abstract

Using magnetoencephalography, the current study examined gamma activity associated with language prediction. Participants read high- and low-constraining sentences in which the final word of the sentence was either expected or unexpected. Although no consistent gamma power difference induced by the sentence-final words was found between the expected and unexpected conditions, the correlation of gamma power during the prediction and activation intervals of the sentence-final words was larger when the presented words matched with the prediction compared with when the prediction was violated or when no prediction was available. This suggests that gamma magnitude relates to the match between predicted and perceived words. Moreover, the expected words induced activity with a slower gamma frequency compared with that induced by unexpected words. Overall, the current study establishes that prediction is related to gamma power correlations and a slowing of the gamma frequency.