Auditory and speech-perception difficulties are common in post-stroke aphasia (PSA), but cortical responses during passive listening to Mandarin speech remain underexplored. Seventeen Mandarin-speaking individuals with PSA and 18 matched healthy controls listened to auditory oddball blocks containing /ba/ standards and /pa/ deviants during functional near-infrared spectroscopy (fNIRS). Although these syllables instantiate an aspiration contrast strongly cued by voice onset time (VOT), standard and deviant trials were combined within a single block-level condition. The resulting measures therefore indexed activation during speech-containing blocks relative to silent inter-block intervals rather than phonemic discrimination, categorization, or VOT-specific processing. Seven theory-driven channels covering bilateral auditory, sensorimotor, inferior frontal, and dorsolateral prefrontal regions were analyzed using three HbO indices: β coefficient, peak amplitude, and temporal centroid. The PSA group showed an overall reduction in task-related HbO activation, reflected by a significant Group effect for the β coefficient. Peak amplitude also showed significant Group and Group × Channel effects, although no individual channel remained significant after correction for multiple comparisons. Temporal centroid showed no significant group or interaction effects. These findings provide preliminary evidence for reduced block-evoked cortical recruitment during passive auditory exposure to Mandarin speech sounds in PSA. Because the analysis did not include a direct standard-deviant or active auditory control comparison, the findings do not establish a selective impairment of phonemic discrimination, categorization, or VOT processing. Because patients were assessed 6-7 months after stroke onset, the activation differences may reflect both persistent impairment and recovery-related reorganization during the early chronic stage.

Social engagement is a critical process for children's language and social development and is altered in neurodevelopmental conditions such as autism. Although previous EEG studies have investigated correlates of children's social development in closely controlled experimental studies, few studies have examined how the brain reflects meaningful behavioral states, such as social engagement, within naturalistic parent-child interaction. This study collected EEG from typically developing toddlers of varying language ability and their parent (n = 49, 24-44 months) during the Social EEG paradigm; dyads completed contexts of dyadic interaction (book reading and puzzle play) and baseline (movie watching) during EEG recording. Videos of the interaction session were coded for social engagement using moment-by-moment behavioral coding and applied to the EEG data. As predicted, child EEG alpha power decreased (p < .001) and theta power increased (p = .01) during social engagement relative to baseline movie engagement. Further, the same pattern was replicated in an independent sample (n = 24, 25-47 months) including children diagnosed with autism, who showed the same pattern of power. These results demonstrate for the first time that EEG power is associated with naturally occurring moments of parent-child social engagement across a range of social/linguistic abilities and ages. This opens avenues for study of more naturalistic and meaningful developmental constructs.

Language production in clinical populations varies by both neurological condition and discourse task. This study analyzes connected speech from 1394 individuals with left or right hemisphere damage, Traumatic Brain Injury, Mild Cognitive Impairment, dementia, or no neurological diagnosis. Participants completed narrative, procedural, and picture-based tasks. Using Open Brain AI to extract linguistic features across phonology, morphology, syntax, semantics, lexicon, and readability, we applied mixed-effects models to examine effects of diagnosis, task, and their interaction. Results showed that task type strongly shapes linguistic output and interacts with diagnostic group across nearly all domains. Narrative tasks elicited more complex syntax and aspectual morphology, while procedural tasks prompted simpler grammatical structures and picture-based descriptions produced greater lexical diversity. Uniform Manifold Approximation and Projection further revealed clear clustering by task type, reflecting a continuum of contextual constraint ranging from structured picture descriptions and procedural instructions to free speech and open-ended narratives. This gradient illustrates a shift in linguistic scaffolding: highly structured tasks tightly constrain the target lexicon and syntax, whereas open-ended tasks require speakers to independently generate and organize narrative content. These findings show that clinical language tasks are not interchangeable. Treating them as equivalent may reduce diagnostic accuracy and weaken comparisons with healthy norms. We propose an Open Brain AI task-specific framework for clinical language assessment to improve identification of linguistic impairments and guide targeted intervention.

Multitasking as a cognitive ability is significant, and yet few studies have investigated how to improve multitasking, especially by language experience. To investigate how language multitasking may enhance domain-general multitasking, this fMRI study compared interpreting students and matched bilingual controls in a non-linguistic dual-task (auditory Task 1, visual Task 2) and examined how their neural data may relate to behavioral performance. Behaviorally, interpreting students showed higher response-time costs in Task 1 but lower costs in Task 2, resulting in faster responses to dual-task trials. At the neural level, interpreting students exhibited greater dual-task-specific activation in the left postcentral gyrus, SMA, and cerebellum (areas subserving sensory processing, response selection, and motor-cognitive integration). Notably, greater engagement of these regions was correlated with greater Task 1 accuracy decrements in interpreting students but not in controls. These findings, together with slower Task 1 responses, indicate that interpreting students may have strategically activated task-set information from both tasks at trial onset. This strategy, while increasing load on motor-cognitive brain regions and coming at a cost to Task 1 performance, accelerated Task 2 responses by pre-activating subsequent task-set information. Hyperactivation was also observed in the left STG and bilateral MTG, with left STG activity negatively correlated with Task 2 time costs only in interpreting students, suggesting that interpreting may enhance recruitment of language-related networks for faster switching. The selectivity and consistency of these brain-behavior correlations demonstrate the significant effects of interpreting on domain-general multitasking, wherein interpreting students strategically trade initial response speed for overall efficiency.

How do our brains represent events in language? Recent work has shown that when we comprehend state-change events, multiple representations corresponding to different states of an object are simultaneously activated. As the sentence unfolds and the object is cued for retrieval, areas of the brain associated with representational conflict show greater activation in proportion to how dissimilar the object-states are to each other. Before being retrieved, however, these object-states have to be actively maintained in the brain. What are the neural mechanisms by which we do so? Across two EEG experiments, we present evidence that, during sentence processing, the maintenance of multiple object-states in state-change events is supported by a neural signal that has been frequently linked to visual working memory: left temporal theta-gamma phase-amplitude coupling.

The aim of this study was to investigate the activation of grammatical gender and its modulation by semantic information, such as biological sex and associated stereotypes. Using a gender-priming paradigm we explored the presence and modulation of the gender-congruency effect (GCE) through behavioral and electrophysiological measures (e.g., N400 and P300) in Spanish native and Italian-Spanish bilingual speakers. Primes included nouns with biological gender (e.g., cocineros MAS - male cooks), arbitrary gender referring to inanimate entities (e.g., ventanas FEM - windows); these nouns could be associated with a gender stereotype (e.g., bomberos MAS - male firefighters; faldas FEM - skirt) or neutral. Targets were gender-marked pronouns (éstos MAS / éstas FEM - these). Participants decided the gender of the pronoun. A GCE for both biological and arbitrary gender nouns was found, supporting the notion that grammatical gender is active at the lexical level and its activation is automatic, even in absence of relevant semantic information. Stereotypical primes enhanced the GCE, particularly among bilinguals, indicating that gender stereotypes can strengthen lexical-semantic activation in non-native speakers. Moreover, unlike Italian, both male and female Spanish speakers did not process the masculine form as inclusive, suggesting a shift toward more gender-sensitive representations in Spanish. Finally, while bilinguals showed evidence of cross-linguistic gender coactivation, native speakers displayed deeper semantic engagement, as reflected in stronger GCEs behaviorally. Overall, these findings demonstrate that grammatical gender processing in Spanish operates at a pure lexical level, interacts with semantic and sociolinguistic factors, and is sensitive to individual differences.

Using a picture-word interference paradigm combined with electroencephalography (EEG), this study investigated how healthy aging affects the temporal dynamics of lexical selection and phonological encoding during spoken word production. Younger and older adults named pictures while word frequency, syllable frequency, and phonological relatedness between targets and distractors were systematically manipulated. Behaviorally, older adults exhibited larger word frequency and syllable frequency effects but reduced phonological facilitation relative to younger adults, indicating age-related changes in both lexical and phonological processing. ERP analyses revealed stage-dependent age differences across successive stages of lexical selection, phonological encoding, and phonetic encoding. Although both groups showed sensitivity to lexical frequency, older adults exhibited reduced and temporally altered neural sensitivity to lexical frequency during early and intermediate processing, alongside increased influence of phonological distractors. Both groups continued to exhibit lexical-phonological interactions during later phonetic encoding. ERP decoding and temporal generalization analyses revealed relatively preserved word-frequency representations across age groups, whereas syllable-frequency and phonological-facilitation representations emerged later and were less temporally stable in older adults. These findings indicate that healthy aging affects both lexical and phonological processing, with more pronounced alterations in the timing and temporal stability of phonological representations. Age-related differences extended across multiple stages of spoken word production. The pattern of increased sensitivity to phonological competition and reduced stability of phonological representations is consistent with the inhibitory deficit hypothesis and the transmission deficit hypothesis, respectively, suggesting that age-related changes in spoken word production arise from both less efficient phonological transmission and reduced control over competing phonological information.

Speech envelopes convey low-frequency temporal structure, including speech rhythm and prosody, and provide important cues for early speech processing in newborns. However, how the newborn brain processes speech envelopes and how brief exposure to them affects post-stimulation neural activity remain unclear. Using a mixed design with language condition as a between-subjects factor and resting-state time as a within-subjects factor, we assigned 73 newborns from Mandarin-speaking environments to a Mandarin group (n = 25), a French group (n = 23), or an English group (n = 25). Functional near-infrared spectroscopy (fNIRS) was used to record hemodynamic signals during stimulation with the corresponding speech envelopes and during resting-state periods before and after stimulation, allowing us to examine task-related activation patterns and changes in resting-state neural activity. All three speech envelopes elicited neural responses in frontotemporal regions, indicating that the newborn frontotemporal system can process both native-language envelopes and rhythmically distinct nonnative envelopes. Comparisons of the pre- and post-stimulation resting states further revealed a significantly strengthened frontotemporal functional connectivity network in the Mandarin group, without corresponding changes in the amplitude of local spontaneous activity. These findings suggest that the neural effects of brief exposure to speech envelopes may persist into the post-stimulation resting period and are primarily expressed as enhanced functional coordination across brain regions. Such post-stimulation network changes may reflect the newborn brain's brief retention of heard speech and, with continued language exposure, facilitate the gradual stabilization and maturation of early language networks, supporting subsequent speech processing and language learning.

Interactive Specialization hypothesizes that brain regions initially have poorly defined roles and gradually specialize through interactions with the environment and other brain regions during postnatal development. The mechanisms underlying language acquisition are without consensus, split between innate maturation and experience-dependent theories. We analyzed a four-year longitudinal dataset of children aged 5, 7, and 9, which included developmental surveys, psychoeducational tests, fMRI, and diffusion-weighted imaging data. We observed evidence of increased functional integration over time, with region-specific improvements in brain network efficiency and functional activity correlations changing substantially. Notably, the functional coupling between the left and right cerebral white matter regions changed significantly across this age window (z=-2.99,p=0.003), declining from a strong correlation of task-related BOLD activity at age 5 (ρ=0.85) to a substantially weaker correlation by age 9 (ρ=0.21). Within language-associated networks, relay-involved tracts were consistently longer than non-relay tracts (p<0.0001 at each age) and this gap widened between ages 5 and 9, whereas no comparable difference or developmental trend was observed in non-language networks, indicating a network-specific reconfiguration toward a more economical wiring configuration. While findings support the theory of interactive specialization, no direct link was found between daily activities and either language outcomes or network efficiency, and associations with standardized language ability were small and mixed in direction.

Speech production requires orchestration of multiple brain systems that support the unfolding of the spoken message across hierarchical linguistic levels, such as phonemes, syllables, words or phrases. Transitions between linguistic levels are critical for connected speech production, yet the neural dynamics of, for example, syllable-level and word-level transitions remain unknown. In this electroencephalography (EEG) study, we use time-frequency analysis and source localization to determine differences between word-boundary and within-word syllable transitions. To this end, pseudoword pairs comprising a total of six consonant-vowel (CV) syllables with different word-boundary positions were used. Adults produced the pseudoword pairs at the pace of a visual metronome, such that syllables were uttered at matching times independently of their relative positions. Accordingly, target syllables could reference either within-word or between-word syllabic transitions, while other linguistic properties, including articulation, stress patterns, co-articulation or prosody were matched. EEG time-frequency analyses of neural sources revealed sensitivity to hierarchical structure. Sources in left and right inferior frontal lobes, as well as left superior temporal lobe were differentially recruited when producing the same syllables, in the same utterance position, but under different pseudoword boundary contexts. The right inferior frontal source showed extensive time-frequency modulation in word transitions that included elevated event-related synchronization across theta and beta oscillations. Alpha suppression was found in the left superior temporal lobe, possibly related to increased sensory prediction at word onset. Despite efforts to control speech pace and syllable acoustics across conditions, small residual differences in timing and acoustic properties were observed.

Adolescents with hearing loss (HL) often display difficulties in written language processing, yet the contribution of morphosyntax remains uncertain. We examined written German morphosyntax in adolescents with HL, most of whom used cochlear implants, relative to monolingual and multilingual hearing controls. Participants completed a minimal two-word grammaticality-judgment task designed to reduce sentence-level working-memory demands while EEG was recorded at the onset of the second word. Accuracy was near ceiling, but the HL group responded more slowly than both hearing comparison groups, and higher working-memory scores were associated with better accuracy within the HL group. In unified ERP models, incongruent pairs elicited overall N400 and P600 effects without interactions with group or construction. Earlier components showed no reliable HL-specific violation pattern. Under these low working-memory demands, later neural responses to minimal agreement violations were detectable at the sample level, while the HL group showed slower behavioral responding and a within-group association between working memory and accuracy.

Semantic comprehension requires accessing multiple dimensions of meaning, yet the temporal dynamics of these dimensions remain unclear. Previous research has predominantly examined the neural dynamics of lexical semantics as a whole, which is equivalent to investigating the average of neural dynamics across various dimensions of semantic processing. Here, we analyzed openly released electrocorticography (ECoG) recordings from 9 participants listening to naturalistic stories and used time-resolved encoding models to decompose neural responses into six semantic dimensions: vision, motor, space, time, socialness, and emotion. When analyzing all words together, nearly all dimensions showed a common response peaking around 350 ms. However, when neural dynamics of multidimensional semantic access were analyzed separately for concrete and abstract words, a dissociation in access latencies emerged: for concrete words, sensorimotor dimensions reached peak activation earlier than non-sensorimotor dimensions; whereas for abstract words, the social dimension peaked at 225 ms, earlier than other dimensions. These findings demonstrate that the brain dynamically prioritizes semantic dimensions depending on concept type.

Our understanding of the neurobiology of language development is imprecise, particularly for bilingual children in their heritage language, since much research focuses on monolingual children or societal majority languages. In this study, we use a novel computational approach to quantifying linguistic predictions with fNIRS neuroimaging to examine bilingual children's (N = 88, ages 7-12) brain activity as they listen to a naturalistic story in their heritage language of Spanish. The children demonstrated successful neural tracking of word predictability in their heritage language, suggesting they can form rich linguistic representations in Spanish. There were some trends for reading and listening comprehension abilities in English and Spanish to modulate neural tracking of predictability in a Spanish story, but no effects were reliable enough to interpret with confidence. Overall, this first use of a predictability-based computational, naturalistic listening comprehension approach with bilingual children demonstrates the feasibility of this approach and its utility for studying language development.

Chinese character satiation is a phenomenon whereby prolonged or repeated inspection induces a compelling subjective experience of structural disintegration. Yet, objective neurophysiological evidence for the proposed mechanism of orthographic decomposition has remained elusive. The present study employed representational similarity analysis (RSA) on electroencephalography (EEG) data to bridge this gap. We first established distinct neural templates for holistic (whole-character judgment) and decompositional (component-judgment) processing. Subsequently, during a satiation phase involving repeated character exposure, we measured how the neural representation's relative similarity to the decompositional versus holistic templates changed between early (low-repetition) and late (high-repetition) trials. Results revealed a significant positive correlation between this objectively measured neural shift and participants' subjective satiation ratings, localized to 620-675 ms post character onset: participants reporting stronger satiation experience exhibited a more pronounced neural representational shift, moving away from the holistic template and toward the decompositional template. These findings provide neurophysiological support for the orthographic decomposition hypothesis for Chinese character satiation, indicating that individual differences in the subjective feeling of disintegration are associated with a measurable reorganization of the underlying neural representation.