Reworking the language network

Unveiling altered connectivity between cognitive networks and cerebellum in schizophrenia

Cognitive functioning is a crucial aspect in schizophrenia (SZ), and when altered it has devastating effects on patients' quality of life and treatment outcomes. Several studies suggested that they could result from altered communication between the cortex and cerebellum. However, the neural correlates underlying these impairments have not been identified. In this study, we investigated resting state functional connectivity (rsFC) in SZ patients, by considering the interactions between cortical networks supporting cognition and cerebellum. In addition, we investigated the relationship between SZ patients' rsFC and their symptoms. We used fMRI data from 74 SZ patients and 74 matched healthy controls (HC) downloaded from the publicly available database SchizConnect. We implemented a seed-based connectivity approach to identify altered functional connections between specific cortical networks and cerebellum. We considered ten commonly studied resting state networks, whose functioning encompasses specific cognitive functions, and the cerebellum, whose involvement in supporting cognition has been recently identified. We then explored the relationship between altered rsFC values and Positive and Negative Syndrome Scale (PANSS) scores. The SZ group showed increased connectivity values compared with HC group for cortical networks involved in attentive processes, which were also linked to PANSS items describing attention and language-related processing. We also showed decreased connectivity between cerebellar regions, and increased connectivity between them and attentive networks, suggesting the contribution of cerebellum to attentive and affective deficits. In conclusion, our findings highlighted the link between negative symptoms in SZ and altered connectivity within the cerebellum and between the same and cortical networks supporting cognition.

Structural network topology associated with naming improvements following intensive aphasia therapy in post-stroke aphasia

A stroke can disrupt the finely tuned language network resulting in aphasia, a language impairment. Though many stroke survivors with aphasia recover within the first 6 months, a significant proportion have lasting deficits. The factors contributing to optimal treatment response remain unclear. Some evidence suggests that increased modularity or fragmentation of brain networks may underlie post-stroke aphasia severity and the extent of recovery. We examined associations between network organization and aphasia recovery in sixteen chronic stroke survivors with non-fluent aphasia following 35 h of Multi-Modality Aphasia Therapy over 10 days and 20 healthy controls who underwent imaging at a single timepoint. Using diffusion-weighted scans obtained before and after treatment, we constructed whole-brain structural connectomes representing the number of probabilistic streamlines between brain regions. Graph theory metrics were quantified for each connectome using the Brain Connectivity Toolbox. Correlations were examined between graph metrics and speech performance measured using the Boston Naming Test (BNT) at pre-, post- and 3-months post-intervention. Compared to controls, participants with stroke demonstrated higher whole-brain modularity at pre-treatment. Modularity did not differ between pre- and post-treatment. In individuals who responded to therapy, higher pre-treatment modularity was associated with worse performance on the BNT. Moreover, higher pre-treatment participation coefficients (i.e., how well a region is connected outside its own module) for the left IFG, planum temporale, and posterior temporal gyri were associated with greater improvements at post-treatment. These results suggest that pre-treatment network topology may impact therapeutic gains, highlighting the influence of network organization on post-stroke aphasia recovery.

Verbal short term memory contribution to sentence comprehension decreases with increasing syntactic complexity in people with aphasia

Sentence comprehension requires the integration of linguistic units presented in a temporal sequence based on a non-linear underlying syntactic structure. While it is uncontroversial that storage is mandatory for this process, there are opposing views regarding the relevance of general short-term-/working-memory capacities (STM/WM) versus language specific resources. Here we report results from 43 participants with an acquired brain lesion in the extended left hemispheric language network and resulting language deficits, who performed a sentence-to-picture matching task and an experimental task assessing phonological short-term memory. The sentence task systematically varied syntactic complexity (embedding depth and argument order) while lengths, number of propositions and plausibility were kept constant. Clinical data including digit-/ block-spans and lesion size and site were additionally used in the analyses. Correlational analyses confirm that performance on STM/WM-tasks (experimental task and digit-span) are the only two relevant predictors for correct sentence-picture-matching, while reaction times only depended on age and lesion size. Notably increasing syntactic complexity reduced the correlational strength speaking for the additional recruitment of language specific resources independent of more general verbal STM/WM capacities, when resolving complex syntactic structure. The complementary lesion-behaviour analysis yielded different lesion volumes correlating with either the sentence-task or the STM-task. Factoring out STM measures lesions in the anterior temporal lobe correlated a larger decrease in accuracy with increasing syntactic complexity. We conclude that overall sentence comprehension depends on STM/WM capacity, while increases in syntactic complexity tax another independent cognitive resource.

Functional integration and segregation during semantic cognition: Evidence across age groups

Semantic cognition is underpinned by ventral anterior temporal lobe (vATL) which encodes knowledge representations and inferior frontal gyrus (IFG), which controls activation of knowledge based on the needs of the current context. This core semantic network has been validated in substantial empirical findings in the past. However, it remains unclear how these core semantic areas dynamically communicate with each other, and with other neural networks, to achieve successful semantic processing. Here, we investigated this question by testing functional connectivity in the core semantic network during semantic tasks and whether these connections were affected by cognitive ageing. Compared to a non-semantic task, semantic tasks increased the connectivity between left and right IFGs, indicating a bilateral semantic control system. Strengthened connectivity was also found between left IFG and left vATL, and this effect was stronger in the young group. At a whole-brain scale, IFG and vATL increased their coupling with multiple-demand regions during semantic tasks, even though these areas were deactivated relative to non-semantic tasks. This suggests that the domain-general executive network contributes to semantic processing. In contrast, IFG and vATL decreased their interaction with default mode network (DMN) areas during semantic tasks, even though these areas were positively activated by the task. This suggests that DMN areas do not contribute to all semantic tasks: their activation may sometimes reflect automatic retrieval of task-irrelevant memories and associations. Taken together, our study characterizes a dynamic connectivity mechanism supporting semantic cognition within and beyond core semantic regions.

Lateral frontoparietal effective connectivity differentiates and predicts state of consciousness in a cohort of patients with traumatic disorders of consciousness

Neuroimaging studies have suggested an important role for the default mode network (DMN) in disorders of consciousness (DoC). However, the extent to which DMN connectivity can discriminate DoC states-unresponsive wakefulness syndrome (UWS) and minimally conscious state (MCS)-is less evident. Particularly, it is unclear whether effective DMN connectivity, as measured indirectly with dynamic causal modelling (DCM) of resting EEG can disentangle UWS from healthy controls and from patients considered conscious (MCS+). Crucially, this extends to UWS patients with potentially "covert" awareness (minimally conscious star, MCS*) indexed by voluntary brain activity in conjunction with partially preserved frontoparietal metabolism as measured with positron emission tomography (PET+ diagnosis; in contrast to PET- diagnosis with complete frontoparietal hypometabolism). Here, we address this gap by using DCM of EEG data acquired from patients with traumatic brain injury in 11 UWS (6 PET- and 5 PET+) and in 12 MCS+ (11 PET+ and 1 PET-), alongside with 11 healthy controls. We provide evidence for a key difference in left frontoparietal connectivity when contrasting UWS PET- with MCS+ patients and healthy controls. Next, in a leave-one-subject-out cross-validation, we tested the classification performance of the DCM models demonstrating that connectivity between medial prefrontal and left parietal sources reliably discriminates UWS PET- from MCS+ patients and controls. Finally, we illustrate that these models generalize to an unseen dataset: models trained to discriminate UWS PET- from MCS+ and controls, classify MCS* patients as conscious subjects with high posterior probability (pp > .92). These results identify specific alterations in the DMN after severe brain injury and highlight the clinical utility of EEG-based effective connectivity for identifying patients with potential covert awareness.

Neural interactions in the human frontal cortex dissociate reward and punishment learning

How human prefrontal and insular regions interact while maximizing rewards and minimizing punishments is unknown. Capitalizing on human intracranial recordings, we demonstrate that the functional specificity toward reward or punishment learning is better disentangled by interactions compared to local representations. Prefrontal and insular cortices display non-selective neural populations to rewards and punishments. Non-selective responses, however, give rise to context-specific interareal interactions. We identify a reward subsystem with redundant interactions between the orbitofrontal and ventromedial prefrontal cortices, with a driving role of the latter. In addition, we find a punishment subsystem with redundant interactions between the insular and dorsolateral cortices, with a driving role of the insula. Finally, switching between reward and punishment learning is mediated by synergistic interactions between the two subsystems. These results provide a unifying explanation of distributed cortical representations and interactions supporting reward and punishment learning.

Functional and Effective Connectivity Underlying Semantic Verbal Fluency

Semantic verbal fluency (SVF) impairment is present in several neurological disorders. Although activation in SVF-related areas has been reported, how these regions are connected and their functional roles in the network remain divergent. We assessed SVF static and dynamic functional connectivity (FC) and effective connectivity in healthy participants using functional magnetic resonance imaging. We observed activation in the inferior frontal (IFG), middle temporal (pMTG) and angular gyri (AG), anterior cingulate (AC), insular cortex, and regions of the superior, middle, and medial frontal gyri (SFG, MFG, MidFG). Our static FC analysis showed a highly interconnected task and resting state network. Increased connectivity of AC with the pMTG and AG was observed for the task. The dynamic FC analysis provided circuits with connections similarly modulated across time and regions related to category identification, language comprehension, word selection and recovery, word generation, inhibition of speaking, speech planning, and articulatory planning of orofacial movements. Finally, the effective connectivity analysis provided a network that best explained our data, starting at the AG and going to the pMTG, from which there was a division between the ventral and dorsal streams. The SFG and MFG regions were connected and modulated by the MidFG, while the inferior regions formed the ventral stream. Therefore, we successfully assessed the SVF network, exploring regions associated with the entire processing, from category identification to word generation. The methodological approach can be helpful for further investigation of the SVF network in neurological disorders.

Revealing the spatiotemporal brain dynamics of covert speech compared with overt speech: A simultaneous EEG-fMRI study

Covert speech (CS) refers to speaking internally to oneself without producing any sound or movement. CS is involved in multiple cognitive functions and disorders. Reconstructing CS content by brain-computer interface (BCI) is also an emerging technique. However, it is still controversial whether CS is a truncated neural process of overt speech (OS) or involves independent patterns. Here, we performed a word-speaking experiment with simultaneous EEG-fMRI. It involved 32 participants, who generated words both overtly and covertly. By integrating spatial constraints from fMRI into EEG source localization, we precisely estimated the spatiotemporal dynamics of neural activity. During CS, EEG source activity was localized in three regions: the left precentral gyrus, the left supplementary motor area, and the left putamen. Although OS involved more brain regions with stronger activations, CS was characterized by an earlier event-locked activation in the left putamen (peak at 262 ms versus 1170 ms). The left putamen was also identified as the only hub node within the functional connectivity (FC) networks of both OS and CS, while showing weaker FC strength towards speech-related regions in the dominant hemisphere during CS. Path analysis revealed significant multivariate associations, indicating an indirect association between the earlier activation in the left putamen and CS, which was mediated by reduced FC towards speech-related regions. These findings revealed the specific spatiotemporal dynamics of CS, offering insights into CS mechanisms that are potentially relevant for future treatment of self-regulation deficits, speech disorders, and development of BCI speech applications.

Distributed Sensitivity to Syntax and Semantics throughout the Language Network

Human language is expressive because it is compositional: The meaning of a sentence (semantics) can be inferred from its structure (syntax). It is commonly believed that language syntax and semantics are processed by distinct brain regions. Here, we revisit this claim using precision fMRI methods to capture separation or overlap of function in the brains of individual participants. Contrary to prior claims, we find distributed sensitivity to both syntax and semantics throughout a broad frontotemporal brain network. Our results join a growing body of evidence for an integrated network for language in the human brain within which internal specialization is primarily a matter of degree rather than kind, in contrast with influential proposals that advocate distinct specialization of different brain areas for different types of linguistic functions.

Effects of computerized cognitive training on functional brain networks in patients with vascular cognitive impairment and no dementia

AIMS

Previous neuroimaging studies of vascular cognitive impairment, no dementia (VCIND), have reported functional alterations, but far less is known about the effects of cognitive training on functional connectivity (FC) of intrinsic connectivity networks (ICNs) and how they relate to intervention-related cognitive improvement. This study provides comprehensive research on the changes in intra- and inter-brain functional networks in patients with VCIND who received computerized cognitive training, with a focus on the underlying mechanisms and potential therapeutic strategies.

METHODS

We prospectively collected 60 patients with VCIND who were randomly divided into the training group (N = 30) receiving computerized cognitive training and the control group (N = 30) receiving fixed cognitive training. Functional MRI scans and cognitive assessments were performed at baseline, at the 7-week training, and at the 6-month follow-up. Utilizing templates for ICNs, the study employed a linear mixed model to compare intra- and inter-network FC changes between the two groups. Pearson correlation was applied to calculate the relationship between FC and cognitive function.

RESULTS

We found significantly decreased intra-network FC within the default mode network (DMN) following computerized cognitive training at Month 6 (p = 0.034), suggesting a potential loss of functional specialization. Computerized training led to increased functional coupling between the DMN and sensorimotor network (SMN) (p = 0.01) and between the language network (LN) and executive control network (ECN) at Month 6 (p < 0.001), indicating compensatory network adaptations in patients with VCIND. Notably, the intra-LN exhibited enhanced functional specialization after computerized cognitive training (p = 0.049), with significant FC increases among LN regions, which correlated with improvements in neuropsychological measures (p < 0.05), emphasizing the targeted impact of computerized cognitive training on language abilities.

CONCLUSIONS

This study provides insights into neuroplasticity and adaptive changes resulting from cognitive training in patients with VCIND, with implications for potential therapeutic strategies.

From task structures to world models: what do LLMs know?

In what sense does a large language model (LLM) have knowledge? We answer by granting LLMs 'instrumental knowledge': knowledge gained by using next-word generation as an instrument. We then ask how instrumental knowledge is related to the ordinary, 'worldly knowledge' exhibited by humans, and explore this question in terms of the degree to which instrumental knowledge can be said to incorporate the structured world models of cognitive science. We discuss ways LLMs could recover degrees of worldly knowledge and suggest that such recovery will be governed by an implicit, resource-rational tradeoff between world models and tasks. Our answer to this question extends beyond the capabilities of a particular AI system and challenges assumptions about the nature of knowledge and intelligence.

A speech fluency brain network derived from gliomas

The brain network of speech fluency has not yet been investigated via a study with a large and homogenous sample. This study analysed multimodal imaging data from 115 patients with low-grade glioma to explore the brain network of speech fluency. We applied voxel-based lesion-symptom mapping to identify domain-specific regions and white matter pathways associated with speech fluency. Direct cortical stimulation validated the domain-specific regions intra-operatively. We then performed connectivity-behaviour analysis with the aim of identifying connections that significantly correlated with speech fluency. Voxel-based lesion-symptom mapping analysis showed that damage to domain-specific regions (the middle frontal gyrus, the precentral gyrus, the orbital part of inferior frontal gyrus and the insula) and white matter pathways (corticospinal fasciculus, internal capsule, arcuate fasciculus, uncinate fasciculus, frontal aslant tract) are associated with reduced speech fluency. Furthermore, we identified connections emanating from these domain-specific regions that exhibited significant correlations with speech fluency. These findings illuminate the interaction between domain-specific regions and 17 domain-general regions-encompassing the superior frontal gyrus, middle frontal gyrus, inferior frontal gyrus and rolandic operculum, superior temporal gyrus, temporal pole, inferior temporal pole, middle cingulate gyrus, supramarginal gyrus, fusiform gyrus, inferior parietal lobe, as well as subcortical structures such as thalamus-implicating their collective role in supporting fluent speech. Our detailed mapping of the speech fluency network offers a strategic foundation for clinicians to safeguard language function during the surgical intervention for brain tumours.

Remapping and Reconnecting the Language Network after Stroke

Here, we review the literature on neurotypical individuals and individuals with post-stroke aphasia showing that right-hemisphere regions homologous to language network and other regions, like the right cerebellum, are activated in language tasks and support language even in healthy people. We propose that language recovery in post-stroke aphasia occurs largely by potentiating the right hemisphere network homologous to the language network and other networks that previously supported language to a lesser degree and by modulating connection strength between nodes of the right-hemisphere language network and undamaged nodes of the left-hemisphere language network. Based on this premise (supported by evidence we review), we propose that interventions should be aimed at potentiating the right-hemisphere language network through Hebbian learning or by augmenting connections between network nodes through neuroplasticity, such as non-invasive brain stimulation and perhaps modulation of neurotransmitters involved in neuroplasticity. We review aphasia treatment studies that have taken this approach. We conclude that further aphasia rehabilitation with this aim is justified.

Hierarchical dynamic coding coordinates speech comprehension in the brain

Speech comprehension requires the human brain to transform an acoustic waveform into meaning. To do so, the brain generates a hierarchy of features that converts the sensory input into increasingly abstract language properties. However, little is known about how these hierarchical features are generated and continuously coordinated. Here, we propose that each linguistic feature is dynamically represented in the brain to simultaneously represent successive events. To test this 'Hierarchical Dynamic Coding' (HDC) hypothesis, we use time-resolved decoding of brain activity to track the construction, maintenance, and integration of a comprehensive hierarchy of language features spanning acoustic, phonetic, sub-lexical, lexical, syntactic and semantic representations. For this, we recorded 21 participants with magnetoencephalography (MEG), while they listened to two hours of short stories. Our analyses reveal three main findings. First, the brain incrementally represents and simultaneously maintains successive features. Second, the duration of these representations depend on their level in the language hierarchy. Third, each representation is maintained by a dynamic neural code, which evolves at a speed commensurate with its corresponding linguistic level. This HDC preserves the maintenance of information over time while limiting the interference between successive features. Overall, HDC reveals how the human brain continuously builds and maintains a language hierarchy during natural speech comprehension, thereby anchoring linguistic theories to their biological implementations.

The Importance of Linguistic Factors: He Likes Subject Referents

We report the results of one visual-world eye-tracking experiment and two referent selection tasks in which we investigated the effects of information structure in the form of prosody and word order manipulation on the processing of subject pronouns er and der in German. Factors such as subjecthood, focus, and topicality, as well as order of mention have been linked to an increased probability of certain referents being selected as the pronoun's antecedent and described as increasing this referent's prominence, salience, or accessibility. The goal of this study was to find out whether pronoun processing is primarily guided by linguistic factors (e.g., grammatical role) or nonlinguistic factors (e.g., first-mention), and whether pronoun interpretation can be described in terms of referents' "prominence" / "accessibility" / "salience." The results showed an overall subject preference for er, whereas der was affected by the object role and focus marking. While focus increases the attentional load and enhances memory representation for the focused referent making the focused referent more available, ultimately it did not affect the final interpretation of er, suggesting that "prominence" or the related concepts do not explain referent selection preferences. Overall, the results suggest a primacy of linguistic factors in determining pronoun resolution.

Lexical-Semantic Content, Not Syntactic Structure, Is the Main Contributor to ANN-Brain Similarity of fMRI Responses in the Language Network

Representations from artificial neural network (ANN) language models have been shown to predict human brain activity in the language network. To understand what aspects of linguistic stimuli contribute to ANN-to-brain similarity, we used an fMRI data set of responses to n = 627 naturalistic English sentences (Pereira et al., 2018) and systematically manipulated the stimuli for which ANN representations were extracted. In particular, we (i) perturbed sentences' word order, (ii) removed different subsets of words, or (iii) replaced sentences with other sentences of varying semantic similarity. We found that the lexical-semantic content of the sentence (largely carried by content words) rather than the sentence's syntactic form (conveyed via word order or function words) is primarily responsible for the ANN-to-brain similarity. In follow-up analyses, we found that perturbation manipulations that adversely affect brain predictivity also lead to more divergent representations in the ANN's embedding space and decrease the ANN's ability to predict upcoming tokens in those stimuli. Further, results are robust as to whether the mapping model is trained on intact or perturbed stimuli and whether the ANN sentence representations are conditioned on the same linguistic context that humans saw. The critical result-that lexical-semantic content is the main contributor to the similarity between ANN representations and neural ones-aligns with the idea that the goal of the human language system is to extract meaning from linguistic strings. Finally, this work highlights the strength of systematic experimental manipulations for evaluating how close we are to accurate and generalizable models of the human language network.

Strong Prediction: Language Model Surprisal Explains Multiple N400 Effects

Theoretical accounts of the N400 are divided as to whether the amplitude of the N400 response to a stimulus reflects the extent to which the stimulus was predicted, the extent to which the stimulus is semantically similar to its preceding context, or both. We use state-of-the-art machine learning tools to investigate which of these three accounts is best supported by the evidence. GPT-3, a neural language model trained to compute the conditional probability of any word based on the words that precede it, was used to operationalize contextual predictability. In particular, we used an information-theoretic construct known as surprisal (the negative logarithm of the conditional probability). Contextual semantic similarity was operationalized by using two high-quality co-occurrence-derived vector-based meaning representations for words: GloVe and fastText. The cosine between the vector representation of the sentence frame and final word was used to derive contextual cosine similarity estimates. A series of regression models were constructed, where these variables, along with cloze probability and plausibility ratings, were used to predict single trial N400 amplitudes recorded from healthy adults as they read sentences whose final word varied in its predictability, plausibility, and semantic relationship to the likeliest sentence completion. Statistical model comparison indicated GPT-3 surprisal provided the best account of N400 amplitude and suggested that apparently disparate N400 effects of expectancy, plausibility, and contextual semantic similarity can be reduced to variation in the predictability of words. The results are argued to support predictive coding in the human language network.

Limited but specific engagement of the mature language network during linguistic statistical learning

Statistical learning (SL) is the ability to detect and learn regularities from input and is foundational to language acquisition. Despite the dominant role of SL as a theoretical construct for language development, there is a lack of direct evidence supporting the shared neural substrates underlying language processing and SL. It is also not clear whether the similarities, if any, are related to linguistic processing, or statistical regularities in general. The current study tests whether the brain regions involved in natural language processing are similarly recruited during auditory, linguistic SL. Twenty-two adults performed an auditory linguistic SL task, an auditory nonlinguistic SL task, and a passive story listening task as their neural activation was monitored. Within the language network, the left posterior temporal gyrus showed sensitivity to embedded speech regularities during auditory, linguistic SL, but not auditory, nonlinguistic SL. Using a multivoxel pattern similarity analysis, we uncovered similarities between the neural representation of auditory, linguistic SL, and language processing within the left posterior temporal gyrus. No other brain regions showed similarities between linguistic SL and language comprehension, suggesting that a shared neurocomputational process for auditory SL and natural language processing within the left posterior temporal gyrus is specific to linguistic stimuli.

Disentangling neuroplasticity mechanisms in post-stroke language recovery

A major objective in post-stroke aphasia research is to gain a deeper understanding of neuroplastic mechanisms that drive language recovery, with the ultimate goal of enhancing treatment outcomes. Subsequent to recent advances in neuroimaging techniques, we now have the ability to examine more closely how neural activity patterns change after a stroke. However, the way these neural activity changes relate to language impairments and language recovery is still debated. The aim of this review is to provide a theoretical framework to better investigate and interpret neuroplasticity mechanisms underlying language recovery in post-stroke aphasia. We detail two sets of neuroplasticity mechanisms observed at the synaptic level that may explain functional neuroimaging findings in post-stroke aphasia recovery at the network level: feedback-based homeostatic plasticity and associative Hebbian plasticity. In conjunction with these plasticity mechanisms, higher-order cognitive control processes dynamically modulate neural activity in other regions to meet communication demands, despite reduced neural resources. This work provides a network-level neurobiological framework for understanding neural changes observed in post-stroke aphasia and can be used to define guidelines for personalized treatment development.

Driving and suppressing the human language network using large language models

Transformer models such as GPT generate human-like language and are predictive of human brain responses to language. Here, using functional-MRI-measured brain responses to 1,000 diverse sentences, we first show that a GPT-based encoding model can predict the magnitude of the brain response associated with each sentence. We then use the model to identify new sentences that are predicted to drive or suppress responses in the human language network. We show that these model-selected novel sentences indeed strongly drive and suppress the activity of human language areas in new individuals. A systematic analysis of the model-selected sentences reveals that surprisal and well-formedness of linguistic input are key determinants of response strength in the language network. These results establish the ability of neural network models to not only mimic human language but also non-invasively control neural activity in higher-level cortical areas, such as the language network.

Phase-encoded fMRI tracks down brainstorms of natural language processing with subsecond precision

Natural language processing unfolds information overtime as spatially separated, multimodal, and interconnected neural processes. Existing noninvasive subtraction-based neuroimaging techniques cannot simultaneously achieve the spatial and temporal resolutions required to visualize ongoing information flows across the whole brain. Here we have developed rapid phase-encoded designs to fully exploit the temporal information latent in functional magnetic resonance imaging data, as well as overcoming scanner noise and head-motion challenges during overt language tasks. We captured real-time information flows as coherent hemodynamic waves traveling over the cortical surface during listening, reading aloud, reciting, and oral cross-language interpreting tasks. We were able to observe the timing, location, direction, and surge of traveling waves in all language tasks, which were visualized as "brainstorms" on brain "weather" maps. The paths of hemodynamic traveling waves provide direct evidence for dual-stream models of the visual and auditory systems as well as logistics models for crossmodal and cross-language processing. Specifically, we have tracked down the step-by-step processing of written or spoken sentences first being received and processed by the visual or auditory streams, carried across language and domain-general cognitive regions, and finally delivered as overt speeches monitored through the auditory cortex, which gives a complete picture of information flows across the brain during natural language functioning. PRACTITIONER POINTS: Phase-encoded fMRI enables simultaneous imaging of high spatial and temporal resolution, capturing continuous spatiotemporal dynamics of the entire brain during real-time overt natural language tasks. Spatiotemporal traveling wave patterns provide direct evidence for constructing comprehensive and explicit models of human information processing. This study unlocks the potential of applying rapid phase-encoded fMRI to indirectly track the underlying neural information flows of sequential sensory, motor, and high-order cognitive processes.

The Geometry of Low- and High-Level Perceptual Spaces

Low-level features are typically continuous (e.g., the gamut between two colors), but semantic information is often categorical (there is no corresponding gradient between dog and turtle) and hierarchical (animals live in land, water, or air). To determine the impact of these differences on cognitive representations, we characterized the geometry of perceptual spaces of five domains: a domain dominated by semantic information (animal names presented as words), a domain dominated by low-level features (colored textures), and three intermediate domains (animal images, lightly texturized animal images that were easy to recognize, and heavily texturized animal images that were difficult to recognize). Each domain had 37 stimuli derived from the same animal names. From 13 participants (9F), we gathered similarity judgments in each domain via an efficient psychophysical ranking paradigm. We then built geometric models of each domain for each participant, in which distances between stimuli accounted for participants' similarity judgments and intrinsic uncertainty. Remarkably, the five domains had similar global properties: each required 5-7 dimensions, and a modest amount of spherical curvature provided the best fit. However, the arrangement of the stimuli within these embeddings depended on the level of semantic information: dendrograms derived from semantic domains (word, image, and lightly texturized images) were more "tree-like" than those from feature-dominated domains (heavily texturized images and textures). Thus, the perceptual spaces of domains along this feature-dominated to semantic-dominated gradient shift to a tree-like organization when semantic information dominates, while retaining a similar global geometry.

Domain-general and language-specific contributions to speech production in a second language: an fMRI study using functional localizers

For bilinguals, speaking in a second language (L2) compared to the native language (L1) is usually more difficult. In this study we asked whether the difficulty in L2 production reflects increased demands imposed on domain-general or core language mechanisms. We compared the brain response to speech production in L1 and L2 within two functionally-defined networks in the brain: the Multiple Demand (MD) network and the language network. We found that speech production in L2 was linked to a widespread increase of brain activity in the domain-general MD network. The language network did not show a similarly robust differences in processing speech in the two languages, however, we found increased response to L2 production in the language-specific portion of the left inferior frontal gyrus (IFG). To further explore our results, we have looked at domain-general and language-specific response within the brain structures postulated to form a Bilingual Language Control (BLC) network. Within this network, we found a robust increase in response to L2 in the domain-general, but also in some language-specific voxels including in the left IFG. Our findings show that L2 production strongly engages domain-general mechanisms, but only affects language sensitive portions of the left IFG. These results put constraints on the current model of bilingual language control by precisely disentangling the domain-general and language-specific contributions to the difficulty in speech production in L2.

Left-Hemisphere Cortical Language Regions Respond Equally to Observed Dialogue and Monologue

Much of the language we encounter in our everyday lives comes in the form of conversation, yet the majority of research on the neural basis of language comprehension has used input from only one speaker at a time. Twenty adults were scanned while passively observing audiovisual conversations using functional magnetic resonance imaging. In a block-design task, participants watched 20 s videos of puppets speaking either to another puppet (the dialogue condition) or directly to the viewer (the monologue condition), while the audio was either comprehensible (played forward) or incomprehensible (played backward). Individually functionally localized left-hemisphere language regions responded more to comprehensible than incomprehensible speech but did not respond differently to dialogue than monologue. In a second task, participants watched videos (1-3 min each) of two puppets conversing with each other, in which one puppet was comprehensible while the other's speech was reversed. All participants saw the same visual input but were randomly assigned which character's speech was comprehensible. In left-hemisphere cortical language regions, the time course of activity was correlated only among participants who heard the same character speaking comprehensibly, despite identical visual input across all participants. For comparison, some individually localized theory of mind regions and right-hemisphere homologues of language regions responded more to dialogue than monologue in the first task, and in the second task, activity in some regions was correlated across all participants regardless of which character was speaking comprehensibly. Together, these results suggest that canonical left-hemisphere cortical language regions are not sensitive to differences between observed dialogue and monologue.

Side-by-side regions in dorsolateral prefrontal cortex estimated within the individual respond differentially to domain-specific and domain-flexible processes

A recurring debate concerns whether regions of primate prefrontal cortex (PFC) support domain-flexible or domain-specific processes. Here we tested the hypothesis with functional MRI (fMRI) that side-by-side PFC regions, within distinct parallel association networks, differentially support domain-flexible and domain-specialized processing. Individuals (N = 9) were intensively sampled, and all effects were estimated within their own idiosyncratic anatomy. Within each individual, we identified PFC regions linked to distinct networks, including a dorsolateral PFC (DLPFC) region coupled to the medial temporal lobe (MTL) and an extended region associated with the canonical multiple-demand network. We further identified an inferior PFC region coupled to the language network. Exploration in separate task data, collected within the same individuals, revealed a robust functional triple dissociation. The DLPFC region linked to the MTL was recruited during remembering and imagining the future, distinct from juxtaposed regions that were modulated in a domain-flexible manner during working memory. The inferior PFC region linked to the language network was recruited during sentence processing. Detailed analysis of the trial-level responses further revealed that the DLPFC region linked to the MTL specifically tracked processes associated with scene construction. These results suggest that the DLPFC possesses a domain-specialized region that is small and easily confused with nearby (larger) regions associated with cognitive control. The newly described region is domain specialized for functions traditionally associated with the MTL. We discuss the implications of these findings in relation to convergent anatomical analysis in the monkey.NEW & NOTEWORTHY Competing hypotheses link regions of prefrontal cortex (PFC) to domain-flexible or domain-specific processes. Here, using a precision neuroimaging approach, we identify a domain-specialized region in dorsolateral PFC, coupled to the medial temporal lobe and recruited for scene construction. This region is juxtaposed to, but distinct from, broader PFC regions recruited flexibly for cognitive control. Region distinctions align with broader network differences, suggesting that PFC regions gain dissociable processing properties via segregated anatomical projections.

TMS disruption of the lateral prefrontal cortex increases neural activity in the default mode network when naming facial expressions

Recognizing facial expressions is dependent on multiple brain networks specialized for different cognitive functions. In the current study, participants (N = 20) were scanned using functional magnetic resonance imaging (fMRI), while they performed a covert facial expression naming task. Immediately prior to scanning thetaburst transcranial magnetic stimulation (TMS) was delivered over the right lateral prefrontal cortex (PFC), or the vertex control site. A group whole-brain analysis revealed that TMS induced opposite effects in the neural responses across different brain networks. Stimulation of the right PFC (compared to stimulation of the vertex) decreased neural activity in the left lateral PFC but increased neural activity in three nodes of the default mode network (DMN): the right superior frontal gyrus, right angular gyrus and the bilateral middle cingulate gyrus. A region of interest analysis showed that TMS delivered over the right PFC reduced neural activity across all functionally localised face areas (including in the PFC) compared to TMS delivered over the vertex. These results suggest that visually recognizing facial expressions is dependent on the dynamic interaction of the face-processing network and the DMN. Our study also demonstrates the utility of combined TMS/fMRI studies for revealing the dynamic interactions between different functional brain networks.

The role of multimodal cues in second language comprehension

In face-to-face communication, multimodal cues such as prosody, gestures, and mouth movements can play a crucial role in language processing. While several studies have addressed how these cues contribute to native (L1) language processing, their impact on non-native (L2) comprehension is largely unknown. Comprehension of naturalistic language by L2 comprehenders may be supported by the presence of (at least some) multimodal cues, as these provide correlated and convergent information that may aid linguistic processing. However, it is also the case that multimodal cues may be less used by L2 comprehenders because linguistic processing is more demanding than for L1 comprehenders, leaving more limited resources for the processing of multimodal cues. In this study, we investigated how L2 comprehenders use multimodal cues in naturalistic stimuli (while participants watched videos of a speaker), as measured by electrophysiological responses (N400) to words, and whether there are differences between L1 and L2 comprehenders. We found that prosody, gestures, and informative mouth movements each reduced the N400 in L2, indexing easier comprehension. Nevertheless, L2 participants showed weaker effects for each cue compared to L1 comprehenders, with the exception of meaningful gestures and informative mouth movements. These results show that L2 comprehenders focus on specific multimodal cues - meaningful gestures that support meaningful interpretation and mouth movements that enhance the acoustic signal - while using multimodal cues to a lesser extent than L1 comprehenders overall.

Bridging the Divide: Brain and Behavior in Developmental Language Disorder

Developmental language disorder (DLD) is a heterogenous neurodevelopmental disorder that affects a child's ability to comprehend and/or produce spoken and/or written language, yet it cannot be attributed to hearing loss or overt neurological damage. It is widely believed that some combination of genetic, biological, and environmental factors influences brain and language development in this population, but it has been difficult to bridge theoretical accounts of DLD with neuroimaging findings, due to heterogeneity in language impairment profiles across individuals and inconsistent neuroimaging findings. Therefore, the purpose of this overview is two-fold: (1) to summarize the neuroimaging literature (while drawing on findings from other language-impaired populations, where appropriate); and (2) to briefly review the theoretical accounts of language impairment patterns in DLD, with the goal of bridging the disparate findings. As will be demonstrated with this overview, the current state of the field suggests that children with DLD have atypical brain volume, laterality, and activation/connectivity patterns in key language regions that likely contribute to language difficulties. However, the precise nature of these differences and the underlying neural mechanisms contributing to them remain an open area of investigation.

Multilingual Language Diversity Protects Native Language Production under Different Control Demands

The use of multiple languages has been found to influence individuals' cognitive abilities. Although some studies have also investigated the effect of multilingualism on non-native language proficiency, fewer studies have focused on how multilingual experience affects native language production. This study investigated the effect of multilingualism on native language production, specifically examining control demands through a semantic Go/No-Go picture naming task. The multilingual experience was quantified using language entropy, which measures the uncertainty and diversity of language use. Control demands were achieved by manipulating the proportion of Go (i.e., naming) trials in different conditions. Results showed that as control demands increased, multilingual individuals exhibited poorer behavioral performance and greater brain activation throughout the brain. Moreover, more diverse language use was associated with higher accuracy in naming and more interconnected brain networks with greater involvement of domain-general neural resources and less domain-specific neural resources. Notably, the varied and balanced use of multiple languages enabled multilingual individuals to respond more efficiently to increased task demands during native language production.

Atrophy in behavioural variant frontotemporal dementia spans multiple large-scale prefrontal and temporal networks

The identification of a neurodegenerative disorder's distributed pattern of atrophy-or atrophy 'signature'-can lend insights into the cortical networks that degenerate in individuals with specific constellations of symptoms. In addition, this signature can be used as a biomarker to support early diagnoses and to potentially reveal pathological changes associated with said disorder. Here, we characterized the cortical atrophy signature of behavioural variant frontotemporal dementia (bvFTD). We used a data-driven approach to estimate cortical thickness using surface-based analyses in two independent, sporadic bvFTD samples (n = 30 and n = 71, total n = 101), using age- and gender-matched cognitively and behaviourally normal individuals. We found highly similar patterns of cortical atrophy across the two independent samples, supporting the reliability of our bvFTD signature. Next, we investigated whether our bvFTD signature targets specific large-scale cortical networks, as is the case for other neurodegenerative disorders. We specifically asked whether the bvFTD signature topographically overlaps with the salience network, as previous reports have suggested. We hypothesized that because phenotypic presentations of bvFTD are diverse, this would not be the case, and that the signature would cross canonical network boundaries. Consistent with our hypothesis, the bvFTD signature spanned rostral portions of multiple networks, including the default mode, limbic, frontoparietal control and salience networks. We then tested whether the signature comprised multiple anatomical subtypes, which themselves overlapped with specific networks. To explore this, we performed a hierarchical clustering analysis. This yielded three clusters, only one of which extensively overlapped with a canonical network (the limbic network). Taken together, these findings argue against the hypothesis that the salience network is preferentially affected in bvFTD, but rather suggest that-at least in patients who meet diagnostic criteria for the full-blown syndrome-neurodegeneration in bvFTD encompasses a distributed set of prefrontal, insular and anterior temporal nodes of multiple large-scale brain networks, in keeping with the phenotypic diversity of this disorder.

A personalized cortical atlas for functional regions of interest

Advances in functional MRI (fMRI) allow mapping an individual's brain function in vivo. Task fMRI can localize domain-specific regions of cognitive processing or functional regions of interest (fROIs) within an individual. Moreover, data from resting state (no task) fMRI can be used to define an individual's connectome, which can characterize that individual's functional organization via connectivity-based parcellations. However, can connectivity-based parcellations alone predict an individual's fROIs? Here, we describe an approach to compute individualized rs-fROIs (i.e., regions that correspond to given fROI constructed using only resting state data) for motor control, working memory, high-level vision, and language comprehension. The rs-fROIs were computed and validated using a large sample of young adults (n = 1,018) with resting state and task fMRI from the Human Connectome Project. First, resting state parcellations were defined across a sequence of resolutions from broadscale to fine-grained networks in a training group of 500 individuals. Second, 21 rs-fROIs were defined from the training group by identifying the rs network that most closely matched task-defined fROIs across all individuals. Third, the selectivity of rs-fROIs was investigated in a training set of the remaining 518 individuals. All computed rs-fROIs were indeed selective for their preferred category. Critically, the rs-fROIs had higher selectivity than probabilistic atlas parcels for nearly all fROIs. In conclusion, we present a potential approach to define selective fROIs on an individual-level circumventing the need for multiple task-based localizers.NEW & NOTEWORTHY We compute individualized resting state parcels that identify an individual's own functional regions of interest (fROIs) for high-level vision, language comprehension, motor control, and working memory, using only their functional connectome. This approach demonstrates a rapid and powerful alternative for finding a large set of fROIs in an individual, using only their unique connectivity pattern, which does not require the costly acquisition of multiple fMRI localizer tasks.

Neurobiological mechanisms for language, symbols and concepts: Clues from brain-constrained deep neural networks

Neural networks are successfully used to imitate and model cognitive processes. However, to provide clues about the neurobiological mechanisms enabling human cognition, these models need to mimic the structure and function of real brains. Brain-constrained networks differ from classic neural networks by implementing brain similarities at different scales, ranging from the micro- and mesoscopic levels of neuronal function, local neuronal links and circuit interaction to large-scale anatomical structure and between-area connectivity. This review shows how brain-constrained neural networks can be applied to study in silico the formation of mechanisms for symbol and concept processing and to work towards neurobiological explanations of specifically human cognitive abilities. These include verbal working memory and learning of large vocabularies of symbols, semantic binding carried by specific areas of cortex, attention focusing and modulation driven by symbol type, and the acquisition of concrete and abstract concepts partly influenced by symbols. Neuronal assembly activity in the networks is analyzed to deliver putative mechanistic correlates of higher cognitive processes and to develop candidate explanations founded in established neurobiological principles.

A social-semantic working-memory account for two canonical language areas

Language and social cognition are traditionally studied as separate cognitive domains, yet accumulative studies reveal overlapping neural correlates at the left ventral temporoparietal junction (vTPJ) and the left lateral anterior temporal lobe (lATL), which have been attributed to sentence processing and social concept activation. We propose a common cognitive component underlying both effects: social-semantic working memory. We confirmed two key predictions of our hypothesis using functional MRI. First, the left vTPJ and lATL showed sensitivity to sentences only when the sentences conveyed social meaning; second, these regions showed persistent social-semantic-selective activity after the linguistic stimuli disappeared. We additionally found that both regions were sensitive to the socialness of non-linguistic stimuli and were more tightly connected with the social-semantic-processing areas than with the sentence-processing areas. The converging evidence indicates the social-semantic working-memory function of the left vTPJ and lATL and challenges the general-semantic and/or syntactic accounts for the neural activity of these regions.

Driving and suppressing the human language network using large language models

Transformer models such as GPT generate human-like language and are highly predictive of human brain responses to language. Here, using fMRI-measured brain responses to 1,000 diverse sentences, we first show that a GPT-based encoding model can predict the magnitude of brain response associated with each sentence. Then, we use the model to identify new sentences that are predicted to drive or suppress responses in the human language network. We show that these model-selected novel sentences indeed strongly drive and suppress activity of human language areas in new individuals. A systematic analysis of the model-selected sentences reveals that surprisal and well-formedness of linguistic input are key determinants of response strength in the language network. These results establish the ability of neural network models to not only mimic human language but also noninvasively control neural activity in higher-level cortical areas, like the language network.

The State of Cognitive Control in Language Processing

Understanding language requires readers and listeners to cull meaning from fast-unfolding messages that often contain conflicting cues pointing to incompatible ways of interpreting the input (e.g., "The cat was chased by the mouse"). This article reviews mounting evidence from multiple methods demonstrating that cognitive control plays an essential role in resolving conflict during language comprehension. How does cognitive control accomplish this task? Psycholinguistic proposals have conspicuously failed to address this question. We introduce an account in which cognitive control aids language processing when cues conflict by sending top-down biasing signals that strengthen the interpretation supported by the most reliable evidence available. We also provide a computationally plausible model that solves the critical problem of how cognitive control "knows" which way to direct its biasing signal by allowing linguistic knowledge itself to issue crucial guidance. Such a mental architecture can explain a range of experimental findings, including how moment-to-moment shifts in cognitive-control state-its level of activity within a person-directly impact how quickly and successfully language comprehension is achieved.

Disruptions in modular structure and network integration of language-related network predict language performance in temporal lobe epilepsy: Evidence from graph-based analysis

OBJECTIVE

Temporal lobe epilepsy (TLE) is a network disorder that alters the total organization of the language-related network. Task-based functional magnetic resonance imaging (fMRI) aimed at functional connectivity is a direct method to investigate how the network is reorganized. However, such studies are scarce and represented mostly by the resting-state analysis of the individual connections between regions. To fill this gap, we used a graph-based analysis, which allows us to cover the total language-related network changes, such as disruptions in an integration/segregation balance, during a language task in TLE.

METHODS

We collected task-based fMRI data with sentence completion from 19 healthy controls and 28 people with left TLE. Using graph-based analysis, we estimated how the language-related network segregated into modules and tested whether they differed between groups. We evaluated the total network integration and the integration within modules. To assess intermodular integration, we considered the number and location of connector hubs-regions with high connectivity.

RESULTS

The language-related network was differently segregated during language processing in the groups. While healthy controls showed a module consisting of left perisylvian regions, people with TLE exhibited a bilateral module formed by the anterior language-related areas and a module in the left temporal lobe, reflecting hyperconnectivity within the epileptic focus. As a consequence of this reorganization, there was a statistical tendency that the dominance of the intramodular integration over the total network integration was greater in TLE, which predicted language performance. The increase in the number of connector hubs in the right hemisphere, in turn, was compensatory in TLE.

SIGNIFICANCE

Our study provides insights into the reorganization of the language-related network in TLE, revealing specific network changes in segregation and integration. It confirms reduced global connectivity and compensation across the healthy hemisphere, commonly observed in epilepsy. These findings advance the understanding of the network-based reorganizational processes underlying language processing in TLE.

Evolutionary neuroanatomical expansion of Broca's region serving a human-specific function

The question concerning the evolution of language is directly linked to the debate on whether language and action are dependent or not and to what extent Broca's region serves as a common neural basis. The debate resulted in two opposing views, one arguing for and one against the dependence of language and action mainly based on neuroscientific data. This article presents an evolutionary neuroanatomical framework which may offer a solution to this dispute. It is proposed that in humans, Broca's region houses language and action independently in spatially separated subregions. This became possible due to an evolutionary expansion of Broca's region in the human brain, which was not paralleled by a similar expansion in the chimpanzee's brain, providing additional space needed for the neural representation of language in humans.

Artificial intelligence based multimodal language decoding from brain activity: A review

Decoding brain activity is conducive to the breakthrough of brain-computer interface (BCI) technology. The development of artificial intelligence (AI) continually promotes the progress of brain language decoding technology. Existent research has mainly focused on a single modality and paid insufficient attention to AI methods. Therefore, our objective is to provide an overview of relevant decoding research from the perspective of different modalities and methodologies. The modalities involve text, speech, image, and video, whereas the core method is using AI-built decoders to translate brain signals induced by multimodal stimuli into text or vocal language. The semantic information of brain activity can be successfully decoded into a language at various levels, ranging from words through sentences to discourses. However, the decoding effect is affected by various factors, such as the decoding model, vector representation model, and brain regions. Challenges and future directions are also discussed. The advances in brain language decoding and BCI technology will potentially assist patients with clinical aphasia in regaining the ability to communicate.

Language disorders in patients with striatal lesions: Deciphering the role of the striatum in language performance

The classical neural model of language refers to a cortical network involving frontal, parietal and temporal regions. However, patients with subcortical lesions of the striatum have language difficulties. We investigated whether the striatum is directly involved in language or whether its role in decision-making has an indirect effect on language performance, by testing carriers of Huntington's disease (HD) mutations and controls. HD is a genetic neurodegenerative disease primarily affecting the striatum and causing language disorders. We asked carriers of the HD mutation in the premanifest (before clinical diagnosis) and early disease stages, and controls to perform two discrimination tasks, one involving linguistic and the other non-linguistic stimuli. We used the hierarchical drift diffusion model (HDDM) to analyze the participants' responses and to assess the decision and non-decision parameters separately. We hypothesized that any language deficits related to decision-making impairments would be reflected in the decision parameters of linguistic and non-linguistic tasks. We also assessed the relative contributions of both HDDM decision and non-decision parameters to the participants' behavioral data (response time and discriminability). Finally, we investigated whether the decision and non-decision parameters of the HDDM were correlated with brain atrophy. The HDDM analysis showed that patients with early HD have impaired decision parameters relative to controls, regardless of the task. In both tasks, decision parameters better explained the variance of response time and discriminability performance than non-decision parameters. In the linguistic task, decision parameters were positively correlated with gray matter volume in the ventral striatum and putamen, whereas non-decision parameters were not. Language impairment in patients with striatal atrophy is better explained by a deficit of decision-making than by a deficit of core linguistic processing. These results suggest that the striatum is involved in language through the modulation of decision-making, presumably by regulating the process of choice between linguistic alternatives.

Prodromal language impairment in genetic frontotemporal dementia within the GENFI cohort

OBJECTIVE

To identify whether language impairment exists presymptomatically in genetic frontotemporal dementia (FTD), and if so, the key differences between the main genetic mutation groups.

METHODS

682 participants from the international multicentre Genetic FTD Initiative (GENFI) study were recruited: 290 asymptomatic and 82 prodromal mutation carriers (with C9orf72, GRN, and MAPT mutations) as well as 310 mutation-negative controls. Language was assessed using items from the Progressive Aphasia Severity Scale, as well as the Boston Naming Test (BNT), modified Camel and Cactus Test (mCCT) and a category fluency task. Participants also underwent a 3 T volumetric T1-weighted MRI from which regional brain volumes within the language network were derived and compared between the groups.

RESULTS

3% of asymptomatic (4% C9orf72, 4% GRN, 2% MAPT) and 48% of prodromal (46% C9orf72, 42% GRN, 64% MAPT) mutation carriers had impairment in at least one language symptom compared with 13% of controls. In prodromal mutation carriers significantly impaired word retrieval was seen in all three genetic groups whilst significantly impaired grammar/syntax and decreased fluency was seen only in C9orf72 and GRN mutation carriers, and impaired articulation only in the C9orf72 group. Prodromal MAPT mutation carriers had significant impairment on the category fluency task and the BNT whilst prodromal C9orf72 mutation carriers were impaired on the category fluency task only. Atrophy in the dominant perisylvian language regions differed between groups, with earlier, more widespread volume loss in C9orf72, and later focal atrophy in the temporal lobe in MAPT mutation carriers.

CONCLUSIONS

Language deficits exist in the prodromal but not asymptomatic stages of genetic FTD across all three genetic groups. Improved understanding of the language phenotype prior to phenoconversion to fully symptomatic FTD will help develop outcome measures for future presymptomatic trials.

Polygenic risk for schizophrenia and the language network: Putative compensatory reorganization in unaffected siblings

Language-related symptoms, such as disorganized, impoverished speech and communicative behaviors, are one of the core features of schizophrenia. These features most strongly correlate with cognitive deficits and polygenic risk among various symptom dimensions of schizophrenia. Nevertheless, unaffected siblings with genetic high-risk fail to show consistent deficits in language network (LN), indicating that either (1) polygenic risk has no notable effect on LN and/or (2) siblings show compensatory changes in opposing direction to patients. To answer this question, we related polygenic risk scores (PRS) to the region-level, tract-level, and systems-level structure (cortical thickness and fiber connectivity) of LN in 182 patients, 48 unaffected siblings and 135 healthy controls. We also studied the relationships between symptoms, language-related cognition, social functioning and LN structure. We observed a significantly lower thickness in LN (especially the Broca's, Wernicke's area and their right homologues) in patients. Siblings had a distinctly higher thickness in parts of the LN and a more pronounced small-world-like structural integration within the LN. Patients with reduced LN thickness had higher PRS, more disorganization and impoverished speech with lower language-related cognition and social functioning. We conclude that the genetic susceptibility and putative compensatory changes for schizophrenia operate, in part, via key regions in the Language Network.

Microstructural Properties of the Cerebellar Peduncles in Children with Developmental Language Disorder

Children with developmental language disorder (DLD) struggle to learn their native language for no apparent reason. While research on the neurobiological underpinnings of the disorder has focused on the role of cortico-striatal systems, little is known about the role of the cerebellum in DLD. Cortico-cerebellar circuits might be involved in the disorder as they contribute to complex sensorimotor skill learning, including the acquisition of spoken language. Here, we used diffusion-weighted imaging data from 77 typically developing and 54 children with DLD and performed probabilistic tractography to identify the cerebellum's white matter tracts: the inferior, middle, and superior cerebellar peduncles. Children with DLD showed lower fractional anisotropy (FA) in the inferior cerebellar peduncles (ICP), fiber tracts that carry motor and sensory input via the inferior olive to the cerebellum. Lower FA in DLD was driven by lower axial diffusivity. Probing this further with more sophisticated modeling of diffusion data, we found higher orientation dispersion but no difference in neurite density in the ICP of DLD. Reduced FA is therefore unlikely to be reflecting microstructural differences in myelination in this tract, rather the organization of axons in these pathways is disrupted. ICP microstructure was not associated with language or motor coordination performance in our sample. We also found no differences in the middle and superior peduncles, the main pathways connecting the cerebellum with the cortex. To conclude, it is not cortico-cerebellar but atypical olivocerebellar white matter connections that characterize DLD and suggest the involvement of the olivocerebellar system in speech acquisition and development.

The human language system, including its inferior frontal component in "Broca's area," does not support music perception

Language and music are two human-unique capacities whose relationship remains debated. Some have argued for overlap in processing mechanisms, especially for structure processing. Such claims often concern the inferior frontal component of the language system located within "Broca's area." However, others have failed to find overlap. Using a robust individual-subject fMRI approach, we examined the responses of language brain regions to music stimuli, and probed the musical abilities of individuals with severe aphasia. Across 4 experiments, we obtained a clear answer: music perception does not engage the language system, and judgments about music structure are possible even in the presence of severe damage to the language network. In particular, the language regions' responses to music are generally low, often below the fixation baseline, and never exceed responses elicited by nonmusic auditory conditions, like animal sounds. Furthermore, the language regions are not sensitive to music structure: they show low responses to both intact and structure-scrambled music, and to melodies with vs. without structural violations. Finally, in line with past patient investigations, individuals with aphasia, who cannot judge sentence grammaticality, perform well on melody well-formedness judgments. Thus, the mechanisms that process structure in language do not appear to process music, including music syntax.

Phase-encoded fMRI tracks down brainstorms of natural language processing with sub-second precision

The human language system interacts with cognitive and sensorimotor regions during natural language processing. However, where, when, and how these processes occur remain unclear. Existing noninvasive subtraction-based neuroimaging techniques cannot simultaneously achieve the spatial and temporal resolutions required to visualize ongoing information flows across the whole brain. Here we have developed phase-encoded designs to fully exploit the temporal information latent in functional magnetic resonance imaging (fMRI) data, as well as overcoming scanner noise and head-motion challenges during overt language tasks. We captured neural information flows as coherent waves traveling over the cortical surface during listening, reciting, and oral cross-language interpreting. The timing, location, direction, and surge of traveling waves, visualized as 'brainstorms' on brain 'weather' maps, reveal the functional and effective connectivity of the brain in action. These maps uncover the functional neuroanatomy of language perception and production and motivate the construction of finer-grained models of human information processing.

No evidence of theory of mind reasoning in the human language network

Language comprehension and the ability to infer others' thoughts (theory of mind [ToM]) are interrelated during development and language use. However, neural evidence that bears on the relationship between language and ToM mechanisms is mixed. Although robust dissociations have been reported in brain disorders, brain activations for contrasts that target language and ToM bear similarities, and some have reported overlap. We take another look at the language-ToM relationship by evaluating the response of the language network, as measured with fMRI, to verbal and nonverbal ToM across 151 participants. Individual-participant analyses reveal that all core language regions respond more strongly when participants read vignettes about false beliefs compared to the control vignettes. However, we show that these differences are largely due to linguistic confounds, and no such effects appear in a nonverbal ToM task. These results argue against cognitive and neural overlap between language processing and ToM. In exploratory analyses, we find responses to social processing in the "periphery" of the language network-right-hemisphere homotopes of core language areas and areas in bilateral angular gyri-but these responses are not selectively ToM-related and may reflect general visual semantic processing.

Using in vivo functional and structural connectivity to predict chronic stroke aphasia deficits

Focal brain damage caused by stroke can result in aphasia and advances in cognitive neuroscience suggest that impairment may be associated with network-level disorder rather than just circumscribed cortical damage. Several studies have shown meaningful relationships between brain-behaviour using lesions; however, only a handful of studies have incorporated in vivo structural and functional connectivity. Patients with chronic post-stroke aphasia were assessed with structural (n = 68) and functional (n = 39) MRI to assess whether predicting performance can be improved with multiple modalities and if additional variance can be explained compared to lesion models alone. These neural measurements were used to construct models to predict four key language-cognitive factors: (i) phonology; (ii) semantics; (iii) executive function; and (iv) fluency. Our results showed that each factor (except executive ability) could be significantly related to each neural measurement alone; however, structural and functional connectivity models did not explain additional variance above the lesion models. We did find evidence that the structural and functional predictors may be linked to the core lesion sites. First, the predictive functional connectivity features were found to be located within functional resting-state networks identified in healthy controls, suggesting that the result might reflect functionally specific reorganization (damage to a node within a network can result in disruption to the entire network). Second, predictive structural connectivity features were located within core lesion sites, suggesting that multimodal information may be redundant in prediction modelling. In addition, we observed that the optimum sparsity within the regularized regression models differed for each behavioural component and across different imaging features, suggesting that future studies should consider optimizing hyperparameters related to sparsity per target. Together, the results indicate that the observed network-level disruption was predicted by the lesion alone and does not significantly improve model performance in predicting the profile of language impairment.

Semantic reconstruction of continuous language from non-invasive brain recordings

A brain-computer interface that decodes continuous language from non-invasive recordings would have many scientific and practical applications. Currently, however, non-invasive language decoders can only identify stimuli from among a small set of words or phrases. Here we introduce a non-invasive decoder that reconstructs continuous language from cortical semantic representations recorded using functional magnetic resonance imaging (fMRI). Given novel brain recordings, this decoder generates intelligible word sequences that recover the meaning of perceived speech, imagined speech and even silent videos, demonstrating that a single decoder can be applied to a range of tasks. We tested the decoder across cortex and found that continuous language can be separately decoded from multiple regions. As brain-computer interfaces should respect mental privacy, we tested whether successful decoding requires subject cooperation and found that subject cooperation is required both to train and to apply the decoder. Our findings demonstrate the viability of non-invasive language brain-computer interfaces.

Non-literal language processing is jointly supported by the language and theory of mind networks: Evidence from a novel meta-analytic fMRI approach

Going beyond the literal meaning of language is key to communicative success. However, the mechanisms that support non-literal inferences remain debated. Using a novel meta-analytic approach, we evaluate the contribution of linguistic, social-cognitive, and executive mechanisms to non-literal interpretation. We identified 74 fMRI experiments (n = 1,430 participants) from 2001 to 2021 that contrasted non-literal language comprehension with a literal control condition, spanning ten phenomena (e.g., metaphor, irony, indirect speech). Applying the activation likelihood estimation approach to the 825 activation peaks yielded six left-lateralized clusters. We then evaluated the locations of both the individual-study peaks and the clusters against probabilistic functional atlases (cf. anatomical locations, as is typically done) for three candidate brain networks-the language-selective network (Fedorenko, Behr, & Kanwisher, 2011), which supports language processing, the Theory of Mind (ToM) network (Saxe & Kanwisher, 2003), which supports social inferences, and the domain-general Multiple-Demand (MD) network (Duncan, 2010), which supports executive control. These atlases were created by overlaying individual activation maps of participants who performed robust and extensively validated 'localizer' tasks that selectively target each network in question (n = 806 for language; n = 198 for ToM; n = 691 for MD). We found that both the individual-study peaks and the ALE clusters fell primarily within the language network and the ToM network. These results suggest that non-literal processing is supported by both i) mechanisms that process literal linguistic meaning, and ii) mechanisms that support general social inference. They thus undermine a strong divide between literal and non-literal aspects of language and challenge the claim that non-literal processing requires additional executive resources.

The domain-separation language network dynamics in resting state support its flexible functional segregation and integration during language and speech processing

Modern linguistic theories and network science propose that language and speech processing are organized into hierarchical, segregated large-scale subnetworks, with a core of dorsal (phonological) stream and ventral (semantic) stream. The two streams are asymmetrically recruited in receptive and expressive language or speech tasks, which showed flexible functional segregation and integration. We hypothesized that the functional segregation of the two streams was supported by the underlying network segregation. A dynamic conditional correlation approach was employed to construct framewise time-varying language networks and k-means clustering was employed to investigate the temporal-reoccurring patterns. We found that the framewise language network dynamics in resting state were robustly clustered into four states, which dynamically reconfigured following a domain-separation manner. Spatially, the hub distributions of the first three states highly resembled the neurobiology of speech perception and lexical-phonological processing, speech production, and semantic processing, respectively. The fourth state was characterized by the weakest functional connectivity and was regarded as a baseline state. Temporally, the first three states appeared exclusively in limited time bins (∼15%), and most of the time (> 55%), state 4 was dominant. Machine learning-based dFC-linguistics prediction analyses showed that dFCs of the four states significantly predicted individual linguistic performance. These findings suggest a domain-separation manner of language network dynamics in resting state, which forms a dynamic "meta-network" framework to support flexible functional segregation and integration during language and speech processing.

Abnormal resting-state function within language network and its improvement among post-stroke aphasia

Several studies with resting-state magnetic resonance imaging (rs-fMRI) have examined functional impairments and plasticity within language network in patients with post-stroke aphasia (PSA). However, there is still ubiquitous inconsistency across these studies, partly due to restricted to very small sample size and the absence of validation with follow-up data. In the current study, we aimed at providing relatively strong evidence to support functional impairments and its reorganization in PSA. Here, the amplitude of low frequency fluctuations (ALFF) and functional connectivity were used to assess functional alterations of PSA with moderate sample size at baseline (thirty-five PSA patients and thirty-five healthy controls). Functional abnormalities at baseline were observed whether improved among sixteen follow-up patients. Compared with controls, PSA at baseline presented decreased ALFF in the left inferior frontal gyrus (IFG) and decreased functional connectivity of the left IFG with the bilateral supplementary motor area (SMA) and right superior temporal gyrus (STG). The decreased ALFF in IFG, decreased IFG-SMA and IFG-STG connectivity were enhanced among follow-up patients and was synchronized with language-performance improvement. Our results revealed reduced intrinsic neural activity and inter-connections within language network in PSA, which would be normalized synchronously as the improvement of language performance.