Contributions of brain’s language network to the behavioral language performance

Contributions of brain’s language network to the behavioral language performance | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Contributions of brain’s language network to the behavioral language performance Hoorieh Darvishi¹˒², Parham Zargar², Reza Rajimehr² This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8370100/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 15 You are reading this latest preprint version Abstract The neural basis of individual differences in language ability remains incompletely understood. While previous research has identified brain regions associated with language processing, few studies have systematically examined whether higher language performance corresponds to greater neural engagement across cortical and cerebellar language network. Using data from the Human Connectome Project (HCP), we investigated neural activation patterns in 70 healthy adults stratified into high-performing and low-performing groups (35 subjects per group) based on composite scores from language tests, while controlling for social cognition to isolate language-specific effects. Functional MRI data from the story versus math task were analyzed to compare activation patterns between groups. The high-performing group exhibited significantly greater activation across distributed language regions in both hemispheres, with particularly robust effects in left frontal and temporal areas. Brain-behavior correlation analyses across the full sample confirmed that activation strength within these regions predicted language ability. Cerebellar analysis revealed greater recruitment of right posterior regions in the high-performing group, extending the pattern of enhanced activation beyond cortical networks. These findings demonstrate that superior language ability is associated with more robust engagement of a distributed cerebro-cerebellar language network, suggesting that individual differences in language skills are reflected in the magnitude of neural resource recruitment during language processing. Biological sciences/Neuroscience Biological sciences/Psychology Social science/Psychology Language network Cerebellum Functional MRI Language performance Brain-behavior coupling Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction The human language network (LN) is a complex, left-lateralized system comprising anatomically distinct but functionally interconnected regions within the frontal, temporal, and parietal lobes ( 1 ). This network collaboratively supports the core components of language processing: production, comprehension, and the integration of linguistic information ( 2 , 3 ). Understanding how these anatomical components contribute to language processing requires examining the specialized roles of each region. Language production and executive control are primarily mediated by the frontal lobe. The inferior frontal gyrus (IFG), or Broca's area, is a key hub, with its subregions performing specialized roles. The pars opercularis (po) is central to syntactic processing and phonological working memory ( 4 – 6 ), while the pars triangularis (pt) manages semantic retrieval and lexical selection ( 7 , 8 ). The pars orbitalis (porb) also contributes by integrating abstract semantic content ( 9 , 10 ). Complementing the IFG, the precentral gyrus (PreCG) and supplementary motor area (SMA) govern articulatory planning and speech initiation ( 11 , 12 ). Higher-level cognitive support is provided by the dorsolateral prefrontal cortex (DLPFC), which applies executive control over working memory and conflict resolution during complex language tasks ( 13 – 15 ). Language comprehension is primarily anchored in the temporal lobe. The superior temporal gyrus (STG), particularly its posterior part which includes Wernicke’s area ( 16 ), is crucial for processing phonological information and enabling auditory comprehension ( 17 , 18 ). The left posterior middle temporal gyrus (pMTG) works in concert with other regions, such as the angular gyrus (AG), to support lexical-semantic processing ( 19 ). At a higher level, the anterior temporal lobe (ATL) serves as a hub for integrating semantic and conceptual information into coherent meaning ( 20 ). The inferior parietal lobule (IPL) holds a pivotal position, functioning as a major integration center for inputs originating from other cortical lobes. This multifaceted region, encompassing the AG and the supramarginal gyrus (SMG) ( 21 , 22 ), operates as a multimodal hub critical for the integration of phonological, semantic, and syntactic information ( 23 , 24 ). Furthermore, the IPL contributes substantially to verbal working memory ( 25 ) and is considered essential for the comprehension of complex discourse ( 26 , 27 ). Beyond the classic cortical areas, the cerebellum has emerged as a key modulator of language processing. This contribution is primarily localized to the right lateral posterior hemisphere, specifically Crus I, Crus II, and Lobule VI, which form reciprocal loops with cerebral language regions ( 28 ). Crus I is engaged in syntactic processing and verbal working memory, supporting linguistic sequencing and temporal prediction during sentence comprehension ( 28 , 29 ). Crus II contributes to semantic retrieval and verbal fluency, with damage to this area linked to deficits in lexical access and semantic control ( 30 ). Finally, Lobule VI is tied to phonological processing and articulatory planning, coordinating with motor cortical areas to ensure fluent speech production ( 31 ). Together, these cerebellar regions also support broader cognitive functions essential for efficient language use, including error monitoring, attention, and prediction ( 32 , 33 ). Having outlined the key anatomical components of the cortical and cerebellar LN, it is crucial to consider their functional significance. The following section explores the direct relationship between brain activity within this network and behavioral performance on language tasks. The relationship between brain activation in the language-related regions and behavioral performance remains an active area of investigation ( 34 ). Evidence suggests that higher activation within the canonical cortical and cerebellar LN can be associated with better language task performance ( 35 ). Neuroimaging studies have shown that, as individuals perform better in top-down syntactic prediction rather than in basic syntactic composition, activity in the left IFG and the posterior superior temporal sulcus (pSTS) increases ( 36 ). Additionally, a study found that stronger functional connectivity within the left-hemisphere (LH) LN was positively correlated with verbal fluency scores in patients with drug-resistant epilepsy ( 37 ). Furthermore, increased functional connectivity within the LN supports language performance in healthy aging despite gray matter loss ( 38 ). A study found that stronger right-hemisphere (RH) functional connectivity supported executive aspects of language in older adults ( 39 ). In addition, another investigation demonstrated that stronger functional connectivity within the LN correlates positively with verbal fluency scores ( 40 ). Finally, higher RH activity during language tasks is associated with better language performance, especially under higher cognitive demands ( 41 , 42 ). Despite these findings, relatively few studies have directly quantified the relationship between the magnitude of neural activation and behavioral language performance across both cortical and cerebellar regions. Consequently, it remains unclear whether individuals with higher language scores consistently exhibit greater activation in these areas, or if alternative mechanisms, such as neural efficiency, contribute to this outcome. In the present investigation, we sought to address this ambiguity by examining whether higher behavioral performance on language tests correlates with greater magnitude of neural activation within the core cortical and cerebellar LN components. To mitigate the possibility that lower performance reflects task artifacts or non-linguistic processing deficits rather than genuine language-related difficulties, we ensured that both groups differing in language ability exhibited robust performance on a validated, independent social cognitive task. Employing functional neuroimaging in conjunction with detailed behavioral assessments, we investigated brain–behavior coupling to clarify the mechanistic contribution of neural activity patterns to inter-individual variability in language ability. Results To investigate the neural correlates of varying language abilities, participants were first selected based on their performance on outside-scanner language tests: a picture vocabulary test and a reading recognition test (PicVocab_AgeAdj and ReadEng_AgeAdj). Figure 1 A shows the distributions of these language test scores across the full sample, revealing substantial inter-individual variability. We then stratified the cohort into distinct high- and low-performing groups based on a composite language score derived from these two measures. We defined the groups according to the mean and standard deviation of the composite score: the high-performing group (Group H) included individuals scoring more than one standard deviation above the mean, and the low-performing group (Group L) included those scoring more than one standard deviation below the mean. A critical step in our design was to isolate the effects of language ability from other general cognitive functions. To achieve this, we used performance on a social cognition task as a matching variable. Only participants who demonstrated high performance on the social cognition task were included for selection into either Group H or Group L (Fig. 1 B). This procedure yielded two groups that were widely divergent in language abilities but matched for social cognition, ensuring that potential confounds from general cognitive deficits were minimized. The final sample consisted of 35 subjects in Group H and 35 subjects in Group L. Cortical activation differences To evaluate the neural correlates of the performance groups, we calculated effect size maps (Cohen's d) for the story versus math contrast within each group separately. Group H exhibited robust activation across the canonical LN, with large effect sizes observed in temporal and frontal regions, as well as medial parietal cortex (Fig. 2 A). In contrast, Group L showed activation in a similar set of regions but with markedly smaller effect sizes throughout the entire network (Fig. 2 B). A visual comparison of the two maps confirms that the magnitude of activation was substantially greater in Group H than in Group L. To statistically validate and localize these findings, we performed a comparison between the groups across functionally defined regions of interest (ROIs) in both hemispheres. Figure 3 A displays the anatomical locations of these ROIs, which include seven regions in the LH and five in the RH: posterior temporal (PT), middle temporal (MT), anterior temporal (AT), inferior frontal (IF), ventromedial prefrontal cortex (VMPFC), dorsomedial prefrontal cortex (DMPFC), and medial parietal (MP). The functional definition of these ROIs was established using a rigorous procedure. To delineate the anatomical boundaries and avoid circular analysis, we generated a Cohen's d map for an independent validation sample (excluding the 70 subjects used in the main analysis) based on the story versus math contrast. The resulting activation map was then thresholded at d = 0.8 to define the ROIs shown in Fig. 3 A. Complementing the effect size comparison, we performed a voxel-wise statistical analysis to identify regions showing significant group differences (Group H > Group L). We utilized Threshold-Free Cluster Enhancement (TFCE), with the resulting statistical map displaying significance scores on a logarithmic scale (0–2.02). This analysis confirmed that Group H recruited LN to a significantly greater extent than Group L, particularly within the LH (Fig. 3 B). Quantitative region-of-interest (ROI) analysis To quantify these activation patterns, we conducted a targeted region-of-interest (ROI) analysis, extracting the mean activation for each group across both hemispheres (Fig. 4 ). Statistical comparison using independent samples t-tests with false discovery rate (FDR) correction for multiple comparisons revealed that Group H exhibited significantly greater activation than Group L across 11 of the 12 ROIs examined. In the LH, Group H showed significantly stronger activation across six of seven ROIs. The most robust effects emerged in DMPFC_Left (t = 6.56, FDR-corrected p = 1.19×10⁻⁸), IF_Left (t = 5.04, FDR-corrected p = 8.47×10⁻⁶), and AT_Left (t = 4.84, FDR-corrected p = 1.38×10⁻⁵). Significant group differences were also observed in MT_Left (t = 4.66, FDR-corrected p = 1.74×10⁻⁵), PT_Left (t = 4.55, FDR-corrected p = 2.37×10⁻⁵), and MP_Left (t = 2.74, FDR-corrected p = 0.0092). Only VMPFC_Left showed no significant group difference (t = 1.01, FDR-corrected p = 0.315). In the RH, Group H demonstrated significantly greater activation across all five ROIs examined. The strongest effects were found in IF_Right (t = 4.76, FDR-corrected p = 1.47×10⁻⁵), followed by PT_Right (t = 2.96, FDR-corrected p = 0.0054) and AT_Right (t = 3.53, FDR-corrected p = 9.86×10⁻⁴). Significant differences also emerged in VMPFC_Right (t = 2.28, FDR-corrected p = 0.0288) and MP_Right (t = 2.04, FDR-corrected p = 0.0472). In summary, this quantitative ROI analysis demonstrates that superior language performance is associated with significantly greater neural engagement across distributed language-related regions in both hemispheres, with the strongest effects observed in DMPFC_Left, and robust activation differences spanning frontal, temporal, and parietal areas bilaterally. Brain–behavior correlation analysis To examine how neural activation within the LN relates to inter-individual variation in linguistic ability, we performed ROI-wise brain–behavior correlations across twelve defined cortical regions spanning both hemispheres (Fig. 5 ). For each subject, mean β-estimates (story vs math) were extracted from each ROI and related to standardized language ability scores (Mean_z), derived from PicVocab_AgeAdj and ReadEng_AgeAdj measures. This structure allowed us to quantify whether activation strength in subcomponents of the LN predicts behavioral performance across participants. Across LH ROIs, we observed reliable positive brain–behavior correlations, most prominently in DMPFC_Left (r = .27, p = 8.68×10⁻¹⁹), followed by IF_Left (r = .23, p = 1.00×10⁻¹³), MT_Left (r = .22, p = 1.70×10⁻¹³), PT_Left (r = .21, p = 1.77×10⁻¹¹), AT_Left (r = .19, p = 9.69×10⁻¹⁰), MP_Left (r = .15, p = 5.72×10⁻⁷), and VMPFC_Left (r = .13, p = 2.37×10⁻⁵). A similar but weaker organization was present in the RH, where the strongest effects appeared in PT_Right (r = .19, p = 1.45×10⁻⁹), followed by AT_Right (r = .18, p = 9.60×10⁻⁹), IF_Right (r = .16, p = 1.89×10⁻⁷), MP_Right (r = .15, p = 5.35×10⁻⁷), and VMPFC_Right (r = .10, p = .0012). All correlations were positive in direction, indicating that higher language ability was consistently associated with stronger β responses across distributed LN subregions in both hemispheres. Cerebellar activation analysis Finally, we examined the contribution of the cerebellum. First, a probabilistic map ( 43 ) was used to identify cerebellar regions consistently activated during language tasks (Fig. 6 A). This map confirmed that activation was most probable in the right posterior cerebellum, an area encompassing Crus I, Crus II, and Lobule VI, with additional extension into Lobule VIIIA, which together form part of the cerebello-cortical loops engaged during language processing. To examine group-specific engagement of these cerebellar regions, we calculated Cohen's d effect size maps for the story versus math contrast in each group separately. In the current study, both Group H (Fig. 6 B) and Group L (Fig. 6 C) showed activations specifically in Crus I and Crus II, though the magnitude of activation differed between groups. Group H showed robust, positive effect sizes throughout Crus I and Crus II, confirming their strong recruitment. In contrast, Group L exhibited weaker activation. This pattern suggests that cerebellar engagement during language processing scales with behavioral performance, paralleling the cortical findings. Discussion Our results reveal a clear and multi-faceted pattern of neural differences linked to language ability. Our behavioral selection successfully isolated two groups differing in language ability but matched for social cognition (Fig. 1 ). Across the cerebrum, voxel-wise and ROI-based analyses consistently demonstrated that Group H exhibited significantly greater activation in canonical language regions compared to Group L (Figs. 2 – 4 ). Critically, brain-behavior correlations confirmed that activation strength within these regions predicted language ability across the entire sample, with the strongest relationships observed in LH frontal and temporal areas (Fig. 5 ). This pattern extended to the cerebellum, where our analysis revealed stronger recruitment of language-related subregions, specifically Crus I and Crus II, in Group H relative to Group L (Fig. 6 ). This comprehensive evidence demonstrates that superior language ability is associated with more robust and widespread engagement of the entire cerebro-cerebellar language network. After controlling for social cognition, Group H exhibited significantly greater activation across distributed language regions spanning LH PT, MT, AT, IF, DMPFC, and MP, as well as RH PT, AT, IF, VMPFC, and MP. This pattern was strongly left-lateralized but also included key regions in the RH, demonstrating that language ability relies on coordinated activity across both hemispheres and extends beyond cortical networks to include critical cerebellar contributions. The pronounced LH dominance in Group H aligns with extensive evidence establishing the left cortical network as the core system for language processing. The heightened activation we observed in LH IF and temporal regions corresponds directly to their known roles in syntactic processing and semantic integration, respectively, as confirmed by meta-analyses ( 44 , 45 ). Furthermore, the increased RH activation in Group H, particularly in IF_Right and VMPFC_Right, likely reflects greater recruitment of executive and supportive language functions needed for higher-level performance. This supports findings that the RH is engaged during more complex language demands ( 42 , 44 , 45 ). In the cerebellum, the greater activation in Crus I and Crus II within Group H highlights the importance of the cerebro-cerebellar loop in language processing. These findings are consistent with established functional topography: Crus I and Crus II are tied to cognitive aspects of language including syntactic and semantic processing ( 28 – 31 ). The attenuated activation in Group L may therefore represent less efficient engagement of this crucial modulatory circuit, directly contributing to their performance differences. This aligns with evidence that cerebellar integrity is critical for fluent language execution ( 30 – 32 ). Our results also contribute to the broader debate on neural mechanisms underlying individual differences in language ability. The clear pattern of greater activation associated with better performance in our cohort aligns with models where superior ability is linked to enhanced recruitment of neural resources ( 35 – 41 , 46 ). This pattern in healthy individuals provides a valuable baseline for understanding clinical conditions; for instance, language network alterations in patients with epilepsy have been shown to correlate with performance differences similar to those we observed ( 34 – 37 ). Similarly, our findings resonate with studies demonstrating that stronger functional connectivity within the LN correlates with better verbal fluency, suggesting that the activation patterns we observe during tasks reflect underlying network architecture ( 40 ). Limitations Several limitations of this study highlight avenues for future research. Our cross-sectional design, based on a sample from the HCP, establishes a strong association but cannot determine the causal relationship between brain activation and language ability. Future longitudinal studies are needed to track how these brain-behavior patterns develop over time and whether training-induced improvements in language performance correspond to changes in neural engagement. Although we controlled for social cognition, other unmeasured variables could have influenced the results. Additionally, our analysis focused primarily on activation magnitude; future work incorporating functional connectivity measures could provide deeper insight into the network dynamics supporting high-level language performance. Finally, while our cerebellar analysis revealed group differences in Crus I and Crus II, a more fine-grained parcellation of cerebellar subregions and their specific contributions to different aspects of language processing warrants further investigation. Methods Participants In this study, we used the HCP S1200 dataset of healthy adults aged 22–35 ( https://www.humanconnectome.org/study/hcp-young-adult/document/1200-subjects-data-release ). Subjects were recruited from Washington University (St. Louis, MO) and the surrounding area. The HCP data were acquired using protocols approved by the Washington University institutional review board, and written informed consent was obtained from all subjects. From the 1,206 subjects in the release, 157 subjects lacked complete functional data for the story versus math contrast and were excluded. The remaining 1,049 subjects had complete functional data for both the language task and the social cognition task. From this cohort of 1,049 subjects, we employed an extreme groups approach to identify individuals differing substantially in language ability. For each individual, age-adjusted scores from the NIH toolbox picture vocabulary test (PicVocab_AgeAdj) and the NIH toolbox oral reading recognition test (ReadEng_AgeAdj)—both administered outside the scanner as part of the standard HCP behavioral battery—were first standardized (z-scored) against the full sample of 1,049 participants. These two z-scores were then averaged to create a composite language score (Mean_z) for each participant. To isolate language ability from general cognitive function, we used behavioral performance on the Theory of Mind task (ToM_AgeAdj score) as a matching criterion. Only participants who demonstrated high performance on this social cognition measure were eligible for group selection. From this subset, the low-performing group (Group L) comprised 35 individuals with composite language scores more than one standard deviation below the mean, while the high-performing group (Group H) comprised 35 individuals with scores more than one standard deviation above the mean (Total N = 70). This procedure ensured that both groups demonstrated intact social cognitive abilities while differing substantially in language ability. Social cognition performance calculation Since the HCP dataset does not provide a pre-calculated accuracy score for the ToM task, we computed accuracy using participant responses across the six possible response categories. These categories reflect how participants judged each video clip: A = correctly identifying Random trials as Random, B = correctly identifying Mental (ToM) trials as Mental, C = incorrectly judging Random trials as Mental, D = responding "Unsure" to Random trials, E = incorrectly judging Mental trials as Random, and F = responding "Unsure" to Mental trials. Accuracy was calculated using the formula: Accuracy = (A + B) - (C + D + E + F), which rewards correct responses (A and B) while penalizing all incorrect or unsure responses (C, D, E, and F). This accuracy score was used to identify participants with high performance on the social cognition task. Task paradigm Language tests Language abilities were assessed using two standardized measures from the NIH toolbox. Vocabulary knowledge was measured with the picture vocabulary test (age-adjusted scale score; PicVocab_AgeAdj), a computerized adaptive test of general vocabulary knowledge that indexes crystallized verbal ability. Participants listened to an audio recording of a word and selected, from four photographic images, the picture that best matched the word’s meaning. Reading ability was assessed with the oral reading recognition test (age-adjusted scale score; ReadEng_AgeAdj), a computerized adaptive measure of reading decoding skill and crystallized ability in which participants read and pronounce letters and words as accurately as possible. For both measures, age-adjusted scores were normed using age-appropriate bands of the toolbox norming sample (18–29 or 30–35), where a score of 100 indicates national-average performance and 115 or 85 indicate performance ± 1 SD relative to the participant’s age band; higher scores indicate better ability. Language task (story vs. math) Functional data in this study were based on the HCP language processing task ( 47 ). The language task consisted of two runs (run duration = 3:57 min:sec). In each run, 4 blocks of a story task were interleaved with 4 blocks of a math task. The lengths of blocks varied, and the average duration of blocks was approximately 30 s. In the story blocks, participants were presented with brief auditory stories (5–9 sentences) adapted from Aesop's fables, followed by a 2-alternative forced-choice question that asked participants about the topic of the story. For example, after a story about an eagle that saves a man who had done him a favor, participants were asked "Was that about revenge or reciprocity?" Participants pressed a button to select either the first or the second choice. The math task also included trials that were presented auditorily. In these trials, participants completed a series of simple arithmetic (addition and subtraction) operations (e.g., "Fourteen plus twelve"), followed by "equals" and then two choices (e.g., "twenty-nine or twenty-six"). Participants pressed a button to select either the first or the second answer. The math task was adaptive to maintain a similar level of difficulty across participants. The math condition served as a control that matched the story condition for auditory input, motor response demands, and block duration, while minimizing language-specific semantic processing. Social cognition task (Theory of Mind) The ToM task assessed social cognition by presenting participants with short video clips (20 seconds each) depicting simple geometric shapes—such as squares, circles, and triangles—either engaging in meaningful interactions or moving in a random, non-social manner ( 47 ). These animations were adapted from classic ToM stimuli in which social interactions are inferred purely from the motion dynamics of the shapes. Following each clip, participants judged whether the sequence portrayed a mental interaction—that is, behavior suggestive of intentions, beliefs, or emotional states—not sure, or no interaction, indicating random movement without apparent social content. The task consisted of two runs (each lasting 3 minutes and 27 seconds), with each run containing five video blocks: in one run, two Mental and three Random blocks were presented, whereas in the other run the order was three Mental and two Random blocks. Each video block was separated by a 15-second fixation period. This paradigm reliably engages neural systems supporting ToM by requiring participants to interpret animate-like social behavior from abstract visual stimuli. Data acquisition The HCP MRI data acquisition has previously been described in detail ( 48 – 50 ). Images were acquired using a customized 3T Siemens 'Connectom' Skyra scanner having a 100 mT/m SC72 gradient insert and a standard Siemens 32-channel RF-receive head coil. At least one 3D T1-weighted MPRAGE image and one 3D T2-weighted SPACE image were acquired at 0.7 mm isotropic resolution. Whole-brain task fMRI data were acquired using a multi-band EPI sequence with parameters of TR = 720 ms, TE = 33.1 ms, flip angle = 52°, 2 mm isotropic voxels, 72 slices, and multi-band acceleration factor of 8. Spin echo field maps were acquired during both structural and fMRI scanning sessions to enable accurate cross-modal registration of structural and functional images in each subject. Analysis of structural data Structural images (T1-weighted and T2-weighted) were used for extracting subcortical gray matter structures and reconstructing cortical surfaces in each subject. Volume data were transformed from native space into MNI space using a nonlinear volume-based registration. For accurate cross-subject registration of cortical surfaces, a multimodal surface matching (MSM) algorithm was used. The MSM algorithm had two versions: 'MSMSulc' (non-rigid surface alignment based on folding patterns) and 'MSMAll' (optimized alignment of cortical areas using sulcal depth maps plus features from other modalities including myelin maps, resting-state network maps, and visuotopic connectivity maps). Data in our work were based on MSMAll registration. After surface and volume registration, cortical vertices were combined with subcortical gray matter voxels to form the standard 'CIFTI grayordinates' space (91,282 vertices/voxels with 2 mm cortical vertex spacing and 2 mm isotropic subcortical voxels). Analysis of fMRI data Functional images were minimally preprocessed using the HCP pipelines ( 50 ). Preprocessing included correction for spatial distortions due to gradient nonlinearity and B0 field inhomogeneity, fieldmap-based unwarping of EPI images, motion correction, brain-boundary-based registration of EPI to structural T1-weighted scans, non-linear registration to MNI space, and grand-mean intensity normalization. Data from the cortical gray matter ribbon were projected onto the surface and then onto the standard grayordinates space. Data were minimally smoothed by a 2 mm FWHM Gaussian kernel in the grayordinates space. Thus, smoothing was constrained to the cortical surface mesh in each hemisphere. The preprocessed functional time-series were entered into a general linear model (GLM) to estimate functional activities in each vertex/voxel in each run. Two regressors/predictors were included in the GLM design of the language task: story and math. Each predictor covered the duration of a block (~ 30 s). All regressors were convolved with a canonical hemodynamic response function and its temporal derivatives. The time-series were temporally filtered with a Gaussian-weighted linear high-pass filter with a cutoff of 200 s, to remove low-frequency drifts/fluctuations presumably unrelated to the task design. The time-series were also prewhitened to remove temporal autocorrelations in the fMRI signal. For the story versus math contrast, the contrast of parameter estimate (COPE) was computed based on beta values of the GLM. Fixed-effects analyses were conducted to estimate the average effects across runs within each subject. The resulting subject-level contrast maps were used for all subsequent group-level analyses. Statistical analysis Group-level effect size maps To visualize and quantify activation patterns within each performance group, we calculated Cohen's d effect size maps for the story versus math contrast separately for Group H and Group L. For each group, Cohen's d was computed at each vertex/voxel as the mean COPE across subjects divided by the standard deviation of COPE values. These effect size maps enabled direct comparison of activation magnitude between groups while accounting for within-group variability. Voxel-wise group comparison To identify regions showing significant differences in activation between Group H and Group L, we performed voxel-wise statistical analysis using Threshold-Free Cluster Enhancement (TFCE) as implemented in FSL's randomise tool ( 51 ). TFCE enhances cluster-like structures in statistical maps without requiring an arbitrary cluster-forming threshold. We tested the contrast Group H > Group L using 5,000 permutations to generate a null distribution. The resulting statistical maps were family-wise error (FWE) corrected at p < 0.05. For visualization, significance scores were displayed on a logarithmic scale ranging from 0 to 2.02. Region-of-interest (ROI) definition To enable precise quantification of group differences across anatomically defined language regions, we established functional ROIs using an independent validation approach to avoid circular analysis. From the full cohort of 1,049 subjects with complete language task data, we excluded the 70 subjects used in our main analysis (35 in Group H and 35 in Group L), leaving an independent validation sample of 979 subjects. Using this independent sample, we generated a group-level Cohen's d effect size map for the story versus math contrast. This map was thresholded at d = 0.8 to identify regions showing robust language-related activation. The resulting activation clusters were then manually parcellated into functionally defined ROIs corresponding to seven regions in the LH and five regions in the RH: posterior temporal (PT), middle temporal (MT), anterior temporal (AT), inferior frontal (IF), ventromedial prefrontal cortex (VMPFC), dorsomedial prefrontal cortex (DMPFC), and medial parietal (MP). ROI-based group comparison For each of the 12 functionally defined ROIs, we extracted the mean COPE value (story vs. math) for each subject in both Group H and Group L. Independent samples t-tests were conducted to compare mean activation between groups for each ROI. To control for multiple comparisons across the 12 ROIs, we applied false discovery rate (FDR) correction using the Benjamini-Hochberg procedure (62). FDR-corrected p-values < 0.05 were considered statistically significant. Effect sizes (Cohen's d) were calculated for each ROI comparison as the difference in group means divided by the pooled standard deviation. Brain-behavior correlation analysis To examine the relationship between neural activation and individual differences in language ability across the full range of performance, we conducted brain-behavior correlation analyses using all 1,049 subjects with complete language task data. For each subject, mean β-estimates (story vs. math) were extracted from each of the 12 ROIs. These activation values were correlated with the composite language ability score (Mean_z, derived from z-scored PicVocab_AgeAdj and ReadEng_AgeAdj measures) using Pearson correlation. Statistical significance was assessed using two-tailed tests, with p-values reported for each ROI without correction for multiple comparisons, as these analyses were exploratory and aimed at characterizing the continuous brain-behavior relationship across the entire sample rather than testing a specific hypothesis about individual ROIs. Cerebellar analysis Cerebellar activation patterns were analyzed separately due to the distinct functional organization of the cerebellum. We first identified language-relevant cerebellar regions using a probabilistic atlas map. This probabilistic map, based on coordinates from multiple neuroimaging studies, indicated that language-related cerebellar activation is most probable in the right posterior cerebellum, specifically Crus I, Crus II, Lobule VI, and Lobule VIIIA. For our cohort, we calculated Cohen's d effect size maps for the story versus math contrast separately for Group H and Group L in cerebellar space. Visual inspection of these maps confirmed that both groups showed activation specifically in Crus I and Crus II, consistent with the probabilistic atlas but not extending to Lobule VI and Lobule VIIIA in our sample. We qualitatively compared activation magnitude between groups by examining the distribution and intensity of positive Cohen's d values within these cerebellar regions. Declarations Author’s contributions R.R. conceived the idea and designed the analyses; H.D. and P.Z. performed data analysis and prepared figures; P.Z. wrote the manuscript; R.R. critically revised the manuscript. All authors approved the final version of the manuscript. Data availability statement This study used publicly available data from the Human Connectome Project (HCP). Access to the HCP dataset can be obtained through the official HCP data portal. The analysis codes developed for this study are available upon reasonable request. Conflict of interests statement The authors have no conflicts of interest to declare. Acknowledgments This research was supported by IPM (Institute for Research in Fundamental Sciences). Funding Declaration Funding : This research was funded by IPM (Institute for Research in Fundamental Sciences). References Schroën, J. A. M. et al. Causal evidence for a coordinated temporal interplay within the language network. Proc. Natl. Acad. Sci. 120 (47), e2306279120 (2023). Binder, J. R. et al. Human Brain Language Areas Identified by Functional Magnetic Resonance Imaging. J. Neurosci. 17 (1), 353–362 (1997). Lukic, S. et al. Common and distinct neural substrates of sentence production and comprehension. NeuroImage 224 , 117374 (2021). Hopkins, W. D. et al. A comprehensive analysis of variability in the sulci that define the inferior frontal gyrus in the chimpanzee ( Pan troglodytes ) brain. Am. J. Biol. Anthropol. 179 (1), 31–47 (2022 Sept). Friederici, A. D. The Brain Basis of Language Processing: From Structure to Function. Physiol. Rev. 91 (4), 1357–1392 (2011). Fedorenko, E., Behr, M. K. & Kanwisher, N. Functional specificity for high-level linguistic processing in the human brain. Proc. Natl. Acad. Sci. 2011 Sept 27 ;108(39):16428–16433 . Konstantopoulos, K. & Giakoumettis, D. Neuroimaging in neurogenic communication disorders 718 (Academic, 2023). Hagoort, P. M. U. C. & Memory Unification, Control) and beyond. Front Psychol [Internet]. [cited 2023 Dec 10];4. Available from: http://journal.frontiersin.org/article/ (2013). 10.3389/fpsyg.2013.00416/abstract Belyk, M., Brown, S., Lim, J. & Kotz, S. A. Convergence of semantics and emotional expression within the IFG pars orbitalis. NeuroImage 156 , 240–248 (2017). Price, C. J. A review and synthesis of the first 20years of PET and fMRI studies of heard speech, spoken language and reading. NeuroImage 62 (2), 816–847 (2012). Bullock, L., Forseth, K. J., Woolnough, O., Rollo, P. S. & Tandon, N. Supplementary motor area in speech initiation: A large-scale intracranial EEG evaluation of stereotyped word articulation. iScience 28 (1), 111531 (2025). Liu, J. R., Zhao, L., Hullett, P. W. & Chang, E. F. Speech sequencing in the human precentral gyrus. Nat. Hum. Behav. 2025 July 16 ;9(11):2327–2344 . Klaus, J. & Schutter, D. J. L. G. The Role of Left Dorsolateral Prefrontal Cortex in Language Processing. Neuroscience 377 , 197–205 (2018). Hertrich, I., Dietrich, S., Blum, C. & Ackermann, H. The Role of the Dorsolateral Prefrontal Cortex for Speech and Language Processing. Front. Hum. Neurosci. 15 , 645209 (2021). Yu, L. et al. A left-lateralized dorsolateral prefrontal network for naming. Cell. Rep. 44 (5), 115677 (2025). Shen, W., Yuan, Y., Liu, C. & Luo, J. The roles of the temporal lobe in creative insight: an integrated review. Think. Reason . 23 (4), 321–375 (2017). Wani, P. D. From Sound to Meaning: Navigating Wernicke’s Area in Language Processing. Cureus [Internet]. Sept 21 [cited 2025 Nov 25]; (2024). Available from: https://www.cureus.com/articles/283534-from-sound-to-meaning-navigating-wernickes-area-in-language-processing Binder, J. R. Current Controversies on Wernicke’s Area and its Role in Language. Curr. Neurol. Neurosci. Rep. 17 (8), 58 (2017). Zhang, X. et al. The neural correlates of sub-lexical semantics and its integration with the lexical meaning in reading Chinese characters. J. Neurolinguistics . 69 , 101176 (2024). Jung, J., Williams, S. & Lambon Ralph, M. A. The role of GABA in semantic memory and its neuroplasticity. eLife. June 3 ;12:RP91771. (2025). Zhang, W., Xiang, M. & Wang, S. The role of left angular gyrus in the representation of linguistic composition relations. Hum. Brain Mapp. 43 (7), 2204–2217 (2022). Stoeckel, C., Gough, P. M., Watkins, K. E. & Devlin, J. T. Supramarginal gyrus involvement in visual word recognition. Cortex 45 (9), 1091–1096 (2009). Martin, R. C., Yue, Q., Zahn, R. & Lu, Y. The role of variation in phonological and semantic working memory capacities in sentence comprehension: neural evidence from healthy and brain-damaged individuals. Cogn. Affect. Behav. Neurosci. 25 (1), 240–262 (2025). Graves, W. W. et al. Correspondence between cognitive and neural representations for phonology, orthography, and semantics in supramarginal compared to angular gyrus. Brain Struct. Funct. 228 (1), 255–271 (2023). Jonides, J. et al. The Role of Parietal Cortex in Verbal Working Memory. J. Neurosci. 18(13) (1), 5026–5034 (1998 July). Barbey, A. K., Colom, R. & Grafman, J. Neural mechanisms of discourse comprehension: a human lesion study. Brain 137 (1), 277–287 (2014). Glanz, O., Hader, M., Schulze-Bonhage, A., Auer, P. & Ball, T. A Study of Word Complexity Under Conditions of Non-experimental, Natural Overt Speech Production Using ECoG. Front. Hum. Neurosci. 15 , 711886 (2022). Nakatani, H., Nakamura, Y. & Okanoya, K. Respective Involvement of the Right Cerebellar Crus I and II in Syntactic and Semantic Processing for Comprehension of Language. Cerebellum 22 (4), 739–755 (2022). Stoodley, C. J., Valera, E. M. & Schmahmann, J. D. Functional topography of the cerebellum for motor and cognitive tasks: An fMRI study. NeuroImage 59 (2), 1560–1570 (2012). Silveri, M. C. Contribution of the Cerebellum and the Basal Ganglia to Language Production: Speech, Word Fluency, and Sentence Construction—Evidence from Pathology. Cerebellum 20 (2), 282–294 (2021). Simonyan, K. & Fuertinger, S. Speech networks at rest and in action: interactions between functional brain networks controlling speech production. J. Neurophysiol. 113 (7), 2967–2978 (2015). Mannarelli, D. et al. Cerebellum’s Contribution to Attention, Executive Functions and Timing: Psychophysiological Evidence from Event-Related Potentials. Brain Sci. 13 (12), 1683 (2023). Peterburs, J. & Desmond, J. E. The role of the human cerebellum in performance monitoring. Curr. Opin. Neurobiol. 40 , 38–44 (2016). Van Ettinger-Veenstra, H., McAllister, A., Lundberg, P., Karlsson, T. & Engström, M. Higher Language Ability is Related to Angular Gyrus Activation Increase During Semantic Processing, Independent of Sentence Incongruency. Front Hum Neurosci [Internet]. 2016 Mar 11 [cited 2025 Nov 25];10. Available from: http://journal.frontiersin.org/Article/ 10.3389/fnhum.2016.00110/abstract Weber, S., Hausmann, M., Kane, P. & Weis, S. The relationship between language ability and brain activity across language processes and modalities. Neuropsychologia 146 , 107536 (2020 Sept). Matchin, W., Hammerly, C. & Lau, E. The role of the IFG and pSTS in syntactic prediction: Evidence from a parametric study of hierarchical structure in fMRI. Cortex 88 , 106–123 (2017). Pravatà, E. et al. Functional Connectivity MR Imaging of the Language Network in Patients with Drug-Resistant Epilepsy. Am. J. Neuroradiol. 32 (3), 532–540 (2011). Pistono, A. et al. Increased functional connectivity supports language performance in healthy aging despite gray matter loss. Neurobiol. Aging . 98 , 52–62 (2021). Gertel, V. H., Zhang, H. & Diaz, M. T. Stronger right hemisphere functional connectivity supports executive aspects of language in older adults. Brain Lang. 206 , 104771 (2020 July). Miró-Padilla, A., Bueichekú, E., Ventura-Campos, N., Palomar-García, M. Á. & Ávila, C. Functional connectivity in resting state as a phonemic fluency ability measure. Neuropsychologia 97 , 98–103 (2017). Yue, Q. et al. Brain Modularity Mediates the Relation between Task Complexity and Performance. J. Cogn. Neurosci. 29(9) (1), 1532–1546 (2017 Sept). Van Ettinger-Veenstra, H. M. et al. Right-hemispheric brain activation correlates to language performance. NeuroImage 49 (4), 3481–3488 (2010). Casto, C. et al. The cerebellar components of the human language network [Internet]. Neuroscience; [cited 2025 Nov 29]. Available from: http://biorxiv.org/lookup/doi/ (2025). 10.1101/2025.04.14.645351 Turker, S., Kuhnke, P., Eickhoff, S. B., Caspers, S. & Hartwigsen, G. Cortical, subcortical, and cerebellar contributions to language processing: A meta-analytic review of 403 neuroimaging experiments. Psychol. Bull. 149 (11–12), 699–723 (2023). Rodd, J. M., Vitello, S., Woollams, A. M. & Adank, P. Localising semantic and syntactic processing in spoken and written language comprehension: An Activation Likelihood Estimation meta-analysis. Brain Lang. 141 , 89–102 (2015). Warren, J. E., Crinion, J. T., Lambon Ralph, M. A. & Wise, R. J. S. Anterior temporal lobe connectivity correlates with functional outcome after aphasic stroke. Brain 132 (12), 3428–3442 (2009). Barch, D. M. et al. Function in the human connectome: Task-fMRI and individual differences in behavior. NeuroImage 80 , 169–189 (2013). Uğurbil, K. et al. Pushing spatial and temporal resolution for functional and diffusion MRI in the Human Connectome Project. NeuroImage 80 , 80–104 (2013). Smith, S. M. et al. Resting-state fMRI in the Human Connectome Project. NeuroImage 80 , 144–168 (2013). Glasser, M. F. et al. The minimal preprocessing pipelines for the Human Connectome Project. NeuroImage 80 , 105–124 (2013). Spisák, T. et al. Probabilistic TFCE: A generalized combination of cluster size and voxel intensity to increase statistical power. NeuroImage 185 , 12–26 (2019). Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 13 Apr, 2026 Reviews received at journal 03 Apr, 2026 Reviewers agreed at journal 19 Mar, 2026 Reviews received at journal 10 Mar, 2026 Reviewers agreed at journal 25 Feb, 2026 Reviewers agreed at journal 23 Feb, 2026 Reviews received at journal 17 Feb, 2026 Reviewers agreed at journal 11 Feb, 2026 Reviewers agreed at journal 04 Feb, 2026 Reviewers agreed at journal 04 Feb, 2026 Reviewers invited by journal 04 Feb, 2026 Editor invited by journal 22 Dec, 2025 Editor assigned by journal 18 Dec, 2025 Submission checks completed at journal 18 Dec, 2025 First submitted to journal 15 Dec, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8370100","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":586720313,"identity":"28dd65ff-ad65-413c-9d1e-267f753e0738","order_by":0,"name":"Hoorieh Darvishi¹˒²","email":"","orcid":"","institution":"Institute for Cognitive Science Studies","correspondingAuthor":false,"prefix":"","firstName":"Hoorieh","middleName":"","lastName":"Darvishi¹˒²","suffix":""},{"id":586720314,"identity":"9d22c27c-449a-4688-b34c-5c880a8f088d","order_by":1,"name":"Parham Zargar²","email":"","orcid":"","institution":"Institute for Research in Fundamental Sciences","correspondingAuthor":false,"prefix":"","firstName":"Parham","middleName":"","lastName":"Zargar²","suffix":""},{"id":586720315,"identity":"298d697b-01ea-4d58-9a7c-c0b3a113c517","order_by":2,"name":"Reza Rajimehr²","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAs0lEQVRIiWNgGAWjYHCChAMMDAcY+JlBbANStEg2k6AFBA4wGBwgVq15+4GHh25U3Mk3Ps57dANDwT3CWmTOJCQczjnzzHLbYb60GwwGxYS1SDAAteS2HTYwO8xjBtSSQIQW/gdALf8OGxg3E61FAmRLw2EDA2bitQBtyTl22EAC5JcE4hyWk/w5p+awAX//2WM3PvwhQgsDAw9MFQ8wVonRwMDAfgChZRSMglEwCkYBNgAAap483mkCNOYAAAAASUVORK5CYII=","orcid":"","institution":"Institute for Research in Fundamental Sciences","correspondingAuthor":true,"prefix":"","firstName":"Reza","middleName":"","lastName":"Rajimehr²","suffix":""}],"badges":[],"createdAt":"2025-12-15 21:53:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8370100/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8370100/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102212076,"identity":"9f6232c6-5cf5-4f7d-9c4d-a2ed5facaa01","added_by":"auto","created_at":"2026-02-09 12:32:00","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":991455,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBehavioral distributions and selection of language-performance groups.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Distributions of outside-scanner language scores across the full cohort (N = 1206). Histograms show performance on the Picture Vocabulary Test (top), the Oral Reading Recognition Test (middle), and the composite language score derived from the standardized average of the two measures (bottom). These distributions illustrate broad inter-individual variability and provided the basis for defining high- and low-performing subgroups.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, Relationship between language performance and social cognition accuracy (ToM task). Scatterplots display ToM accuracy as a function of vocabulary performance (top), reading performance (middle), and the composite language score (bottom). Vertical lines in the bottom panel indicate ±1 s.d. from the composite-score mean, which were used to identify potential high-performing and low-performing individuals. Although more subjects fell above the +1 s.d. threshold, only 35 subjects with the highest performance (red) were selected to match the group size of the low-performing group, ensuring balanced sampling across groups. All selected participants met the criterion for high social-cognition performance, yielding two groups that diverged in language ability while remaining matched on social cognition.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8370100/v1/90d29c6d62166030a0791296.png"},{"id":102296887,"identity":"efd8f98c-9e40-4e9c-964a-cd963ad9a795","added_by":"auto","created_at":"2026-02-10 10:22:30","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1039372,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGroup-level effect size maps (Cohen’s d) for the story \u0026gt; math contrast in high- and low-performing language groups.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Cohen’s d maps reveal strong engagement of the cortical language network, with pronounced effect sizes in left frontal, temporal, and medial parietal regions. Warmer colors indicate larger positive story-evoked responses, reflecting robust recruitment of language-selective areas in high-performing individuals. Lateral (L), medial (M), left-hemisphere (LH), and right-hemisphere (RH) surface views are accompanied by flat-patch projections (bottom), facilitating visualization of distributed activation patterns across the cortical sheet.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, The overall spatial pattern of activation is broadly similar to Group H; however, effect sizes are substantially attenuated across nearly all language-related regions, including frontal, temporal, and parietal cortices. This reduction indicates weaker recruitment of the language network during linguistic processing in low-performing individuals. As in Panel A, surface views and corresponding flat-patch maps (bottom) are displayed using a symmetric color scale (–2.08 to 1.7), enabling direct comparison between groups and demonstrating that higher behavioral language ability is associated with stronger and more extensive cortical activation.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8370100/v1/a646d3ae195882cad01ae00e.png"},{"id":102212077,"identity":"1be5a91b-4fdd-4469-a12f-2f936130549a","added_by":"auto","created_at":"2026-02-09 12:32:00","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":609731,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDefinition of cortical ROIs and voxel-wise group comparison\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Language-related regions of interest (ROIs) were derived from an independent validation sample to avoid circularity. A Cohen’s \u003cem\u003ed\u003c/em\u003e map for the story \u0026gt; math contrast (excluding the 70 subjects used in the main analysis) was thresholded at \u003cem\u003ed\u003c/em\u003e = 0.8, and the resulting activation clusters were manually parcellated into seven ROIs in the left hemisphere (PT, MT, AT, IF, VMPFC, DMPFC, MP) and five homologous ROIs in the right hemisphere. The top panels show the 3D cortical surface projections, and the lower panels depict the same ROIs on flat-patch cortical representations to highlight spatial boundaries.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, Threshold-Free Cluster Enhancement (TFCE) analysis (5,000 permutations) identified regions where activation was significantly greater in the high-performing group (H \u0026gt; L). Statistical values are displayed on a log-scaled significance map, with ROI outlines (cyan) overlaid to show correspondence between TFCE-detected clusters and the independently defined language ROIs. Lower panels again present flat-patch cortical maps to illustrate the full spatial extent of these group differences.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8370100/v1/62d78a8b3612db6bcc0b1e03.png"},{"id":102297130,"identity":"db0c100f-0e31-4d88-9e43-2f8ba31a3c07","added_by":"auto","created_at":"2026-02-10 10:25:59","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":539353,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGroup differences in story \u0026gt; math activation across cortical language ROIs.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBar plots show mean story \u0026gt; math activation (COPE values) for the high-performing (Group H, red) and low-performing (Group L, blue) language groups in left-hemisphere (left panel) and right-hemisphere (right panel) regions of interest (ROIs). Error bars indicate ±1 s.e.m. Independent-samples \u003cem\u003et\u003c/em\u003e-tests with false discovery rate (FDR) correction across the 12 ROIs revealed significantly greater activation for Group H in 11 of 12 regions, with left VMPFC as the only non-significant ROI. The strongest group differences occurred in DMPFC_Left, IF_Left, and AT_Left, and in IF_Right, with robust effects spanning frontal, temporal, and medial parietal cortices bilaterally. Asterisks denote FDR-corrected significance levels (p* \u0026lt; 0.05; p** \u0026lt; 0.01; p*** \u0026lt; 0.001).\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8370100/v1/a9a021eb2e04eb5b1dffd2cd.png"},{"id":102297133,"identity":"0746857d-0354-4692-9336-51a292224834","added_by":"auto","created_at":"2026-02-10 10:26:01","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1215906,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBrain–behavior correlations across cortical language-network ROIs.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eScatterplots show the relationship between mean activation (β-estimates for the story \u0026gt; math contrast) and standardized language ability scores across all participants (N = 1,049) for each of the twelve functionally defined cortical ROIs. Each panel depicts individual data points (blue) and the best-fitting linear regression line (red). Across regions, stronger activation was consistently associated with higher language performance, yielding positive correlations in both hemispheres. The left-hemisphere ROIs exhibited the strongest associations, particularly in the dorsomedial prefrontal cortex (DMPFC_Left), inferior frontal gyrus (IF_Left), and middle and posterior temporal regions (MT_Left, PT_Left). Right-hemisphere effects were similar in direction but weaker in magnitude. Reported statistics (r and p values) correspond to Pearson correlations computed across the full cohort. Together, these results demonstrate a robust, graded coupling between activation strength within the language network and individual differences in linguistic ability.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8370100/v1/372a5a85386a905825377a7d.png"},{"id":102297139,"identity":"63f1c773-99ba-402c-8b1c-44d6a525790b","added_by":"auto","created_at":"2026-02-10 10:26:02","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":683801,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCerebellar contributions to language processing in high- and low-performing groups.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Probabilistic map of language-related cerebellar activations. The flat-patch representation shows the probability, across participants, that each cerebellar vertex belongs to an individually defined language-responsive region. This probabilistic overlap map identifies a consistent language-associated territory in the right posterior cerebellum spanning Crus I, Crus II, Lobule VI, and Lobule VIIIA regions repeatedly implicated in higher-level linguistic processing. Warmer colors indicate higher overlap probability.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb–c\u003c/strong\u003e, Group-level cerebellar effect-size maps for the story \u0026gt; math contrast. Cohen’s d maps are shown for the high-performing group (Group H) (B) and the low-performing group (Group L) (C), displayed on matched cerebellar flat patches. Both groups exhibit activation centered in Crus I and Crus II; however, Group H demonstrates substantially stronger positive effect sizes, indicating more robust recruitment of cerebellar language regions. This mirrors the cortical findings and suggests that the magnitude of cerebellar engagement scales with individual differences in language ability.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8370100/v1/b7475178bce605523727affd.png"},{"id":102300448,"identity":"ffae49e7-7109-46ec-9961-2ebcc5dde686","added_by":"auto","created_at":"2026-02-10 11:14:43","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5786853,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8370100/v1/9e147f0e-303f-4a33-a1b2-759f9c9ca1a4.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Contributions of brain’s language network to the behavioral language performance","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe human language network (LN) is a complex, left-lateralized system comprising anatomically distinct but functionally interconnected regions within the frontal, temporal, and parietal lobes (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). This network collaboratively supports the core components of language processing: production, comprehension, and the integration of linguistic information (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). Understanding how these anatomical components contribute to language processing requires examining the specialized roles of each region.\u003c/p\u003e \u003cp\u003eLanguage production and executive control are primarily mediated by the frontal lobe. The inferior frontal gyrus (IFG), or Broca's area, is a key hub, with its subregions performing specialized roles. The pars opercularis (po) is central to syntactic processing and phonological working memory (\u003cspan additionalcitationids=\"CR5\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e), while the pars triangularis (pt) manages semantic retrieval and lexical selection (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). The pars orbitalis (porb) also contributes by integrating abstract semantic content (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). Complementing the IFG, the precentral gyrus (PreCG) and supplementary motor area (SMA) govern articulatory planning and speech initiation (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). Higher-level cognitive support is provided by the dorsolateral prefrontal cortex (DLPFC), which applies executive control over working memory and conflict resolution during complex language tasks (\u003cspan additionalcitationids=\"CR14\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eLanguage comprehension is primarily anchored in the temporal lobe. The superior temporal gyrus (STG), particularly its posterior part which includes Wernicke\u0026rsquo;s area (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e), is crucial for processing phonological information and enabling auditory comprehension (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e). The left posterior middle temporal gyrus (pMTG) works in concert with other regions, such as the angular gyrus (AG), to support lexical-semantic processing (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e). At a higher level, the anterior temporal lobe (ATL) serves as a hub for integrating semantic and conceptual information into coherent meaning (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe inferior parietal lobule (IPL) holds a pivotal position, functioning as a major integration center for inputs originating from other cortical lobes. This multifaceted region, encompassing the AG and the supramarginal gyrus (SMG) (\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e), operates as a multimodal hub critical for the integration of phonological, semantic, and syntactic information (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e). Furthermore, the IPL contributes substantially to verbal working memory (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e) and is considered essential for the comprehension of complex discourse (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eBeyond the classic cortical areas, the cerebellum has emerged as a key modulator of language processing. This contribution is primarily localized to the right lateral posterior hemisphere, specifically Crus I, Crus II, and Lobule VI, which form reciprocal loops with cerebral language regions (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e). Crus I is engaged in syntactic processing and verbal working memory, supporting linguistic sequencing and temporal prediction during sentence comprehension (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e). Crus II contributes to semantic retrieval and verbal fluency, with damage to this area linked to deficits in lexical access and semantic control (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e). Finally, Lobule VI is tied to phonological processing and articulatory planning, coordinating with motor cortical areas to ensure fluent speech production (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e). Together, these cerebellar regions also support broader cognitive functions essential for efficient language use, including error monitoring, attention, and prediction (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eHaving outlined the key anatomical components of the cortical and cerebellar LN, it is crucial to consider their functional significance. The following section explores the direct relationship between brain activity within this network and behavioral performance on language tasks.\u003c/p\u003e \u003cp\u003eThe relationship between brain activation in the language-related regions and behavioral performance remains an active area of investigation (\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e). Evidence suggests that higher activation within the canonical cortical and cerebellar LN can be associated with better language task performance (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e). Neuroimaging studies have shown that, as individuals perform better in top-down syntactic prediction rather than in basic syntactic composition, activity in the left IFG and the posterior superior temporal sulcus (pSTS) increases (\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e). Additionally, a study found that stronger functional connectivity within the left-hemisphere (LH) LN was positively correlated with verbal fluency scores in patients with drug-resistant epilepsy (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e). Furthermore, increased functional connectivity within the LN supports language performance in healthy aging despite gray matter loss (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e). A study found that stronger right-hemisphere (RH) functional connectivity supported executive aspects of language in older adults (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e). In addition, another investigation demonstrated that stronger functional connectivity within the LN correlates positively with verbal fluency scores (\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e). Finally, higher RH activity during language tasks is associated with better language performance, especially under higher cognitive demands (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDespite these findings, relatively few studies have directly quantified the relationship between the magnitude of neural activation and behavioral language performance across both cortical and cerebellar regions. Consequently, it remains unclear whether individuals with higher language scores consistently exhibit greater activation in these areas, or if alternative mechanisms, such as neural efficiency, contribute to this outcome.\u003c/p\u003e \u003cp\u003eIn the present investigation, we sought to address this ambiguity by examining whether higher behavioral performance on language tests correlates with greater magnitude of neural activation within the core cortical and cerebellar LN components. To mitigate the possibility that lower performance reflects task artifacts or non-linguistic processing deficits rather than genuine language-related difficulties, we ensured that both groups differing in language ability exhibited robust performance on a validated, independent social cognitive task. Employing functional neuroimaging in conjunction with detailed behavioral assessments, we investigated brain\u0026ndash;behavior coupling to clarify the mechanistic contribution of neural activity patterns to inter-individual variability in language ability.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eTo investigate the neural correlates of varying language abilities, participants were first selected based on their performance on outside-scanner language tests: a picture vocabulary test and a reading recognition test (PicVocab_AgeAdj and ReadEng_AgeAdj). Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA shows the distributions of these language test scores across the full sample, revealing substantial inter-individual variability. We then stratified the cohort into distinct high- and low-performing groups based on a composite language score derived from these two measures. We defined the groups according to the mean and standard deviation of the composite score: the high-performing group (Group H) included individuals scoring more than one standard deviation above the mean, and the low-performing group (Group L) included those scoring more than one standard deviation below the mean.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eA critical step in our design was to isolate the effects of language ability from other general cognitive functions. To achieve this, we used performance on a social cognition task as a matching variable. Only participants who demonstrated high performance on the social cognition task were included for selection into either Group H or Group L (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB). This procedure yielded two groups that were widely divergent in language abilities but matched for social cognition, ensuring that potential confounds from general cognitive deficits were minimized. The final sample consisted of 35 subjects in Group H and 35 subjects in Group L.\u003c/p\u003e \u003cp\u003eCortical activation differences\u003c/p\u003e \u003cp\u003eTo evaluate the neural correlates of the performance groups, we calculated effect size maps (Cohen's d) for the story versus math contrast within each group separately. Group H exhibited robust activation across the canonical LN, with large effect sizes observed in temporal and frontal regions, as well as medial parietal cortex (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). In contrast, Group L showed activation in a similar set of regions but with markedly smaller effect sizes throughout the entire network (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). A visual comparison of the two maps confirms that the magnitude of activation was substantially greater in Group H than in Group L.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo statistically validate and localize these findings, we performed a comparison between the groups across functionally defined regions of interest (ROIs) in both hemispheres. Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA displays the anatomical locations of these ROIs, which include seven regions in the LH and five in the RH: posterior temporal (PT), middle temporal (MT), anterior temporal (AT), inferior frontal (IF), ventromedial prefrontal cortex (VMPFC), dorsomedial prefrontal cortex (DMPFC), and medial parietal (MP).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe functional definition of these ROIs was established using a rigorous procedure. To delineate the anatomical boundaries and avoid circular analysis, we generated a Cohen's d map for an independent validation sample (excluding the 70 subjects used in the main analysis) based on the story versus math contrast. The resulting activation map was then thresholded at d\u0026thinsp;=\u0026thinsp;0.8 to define the ROIs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA.\u003c/p\u003e \u003cp\u003eComplementing the effect size comparison, we performed a voxel-wise statistical analysis to identify regions showing significant group differences (Group H\u0026thinsp;\u0026gt;\u0026thinsp;Group L). We utilized Threshold-Free Cluster Enhancement (TFCE), with the resulting statistical map displaying significance scores on a logarithmic scale (0\u0026ndash;2.02). This analysis confirmed that Group H recruited LN to a significantly greater extent than Group L, particularly within the LH (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003eQuantitative region-of-interest (ROI) analysis\u003c/p\u003e \u003cp\u003eTo quantify these activation patterns, we conducted a targeted region-of-interest (ROI) analysis, extracting the mean activation for each group across both hemispheres (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Statistical comparison using independent samples t-tests with false discovery rate (FDR) correction for multiple comparisons revealed that Group H exhibited significantly greater activation than Group L across 11 of the 12 ROIs examined.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn the LH, Group H showed significantly stronger activation across six of seven ROIs. The most robust effects emerged in DMPFC_Left (t\u0026thinsp;=\u0026thinsp;6.56, FDR-corrected p\u0026thinsp;=\u0026thinsp;1.19\u0026times;10⁻⁸), IF_Left (t\u0026thinsp;=\u0026thinsp;5.04, FDR-corrected p\u0026thinsp;=\u0026thinsp;8.47\u0026times;10⁻⁶), and AT_Left (t\u0026thinsp;=\u0026thinsp;4.84, FDR-corrected p\u0026thinsp;=\u0026thinsp;1.38\u0026times;10⁻⁵). Significant group differences were also observed in MT_Left (t\u0026thinsp;=\u0026thinsp;4.66, FDR-corrected p\u0026thinsp;=\u0026thinsp;1.74\u0026times;10⁻⁵), PT_Left (t\u0026thinsp;=\u0026thinsp;4.55, FDR-corrected p\u0026thinsp;=\u0026thinsp;2.37\u0026times;10⁻⁵), and MP_Left (t\u0026thinsp;=\u0026thinsp;2.74, FDR-corrected p\u0026thinsp;=\u0026thinsp;0.0092). Only VMPFC_Left showed no significant group difference (t\u0026thinsp;=\u0026thinsp;1.01, FDR-corrected p\u0026thinsp;=\u0026thinsp;0.315).\u003c/p\u003e \u003cp\u003eIn the RH, Group H demonstrated significantly greater activation across all five ROIs examined. The strongest effects were found in IF_Right (t\u0026thinsp;=\u0026thinsp;4.76, FDR-corrected p\u0026thinsp;=\u0026thinsp;1.47\u0026times;10⁻⁵), followed by PT_Right (t\u0026thinsp;=\u0026thinsp;2.96, FDR-corrected p\u0026thinsp;=\u0026thinsp;0.0054) and AT_Right (t\u0026thinsp;=\u0026thinsp;3.53, FDR-corrected p\u0026thinsp;=\u0026thinsp;9.86\u0026times;10⁻⁴). Significant differences also emerged in VMPFC_Right (t\u0026thinsp;=\u0026thinsp;2.28, FDR-corrected p\u0026thinsp;=\u0026thinsp;0.0288) and MP_Right (t\u0026thinsp;=\u0026thinsp;2.04, FDR-corrected p\u0026thinsp;=\u0026thinsp;0.0472).\u003c/p\u003e \u003cp\u003eIn summary, this quantitative ROI analysis demonstrates that superior language performance is associated with significantly greater neural engagement across distributed language-related regions in both hemispheres, with the strongest effects observed in DMPFC_Left, and robust activation differences spanning frontal, temporal, and parietal areas bilaterally.\u003c/p\u003e \u003cp\u003eBrain\u0026ndash;behavior correlation analysis\u003c/p\u003e \u003cp\u003eTo examine how neural activation within the LN relates to inter-individual variation in linguistic ability, we performed ROI-wise brain\u0026ndash;behavior correlations across twelve defined cortical regions spanning both hemispheres (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). For each subject, mean β-estimates (story vs math) were extracted from each ROI and related to standardized language ability scores (Mean_z), derived from PicVocab_AgeAdj and ReadEng_AgeAdj measures. This structure allowed us to quantify whether activation strength in subcomponents of the LN predicts behavioral performance across participants.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e Across LH ROIs, we observed reliable positive brain\u0026ndash;behavior correlations, most prominently in DMPFC_Left (r\u0026thinsp;=\u0026thinsp;.27, p\u0026thinsp;=\u0026thinsp;8.68\u0026times;10⁻\u0026sup1;⁹), followed by IF_Left (r\u0026thinsp;=\u0026thinsp;.23, p\u0026thinsp;=\u0026thinsp;1.00\u0026times;10⁻\u0026sup1;\u0026sup3;), MT_Left (r\u0026thinsp;=\u0026thinsp;.22, p\u0026thinsp;=\u0026thinsp;1.70\u0026times;10⁻\u0026sup1;\u0026sup3;), PT_Left (r\u0026thinsp;=\u0026thinsp;.21, p\u0026thinsp;=\u0026thinsp;1.77\u0026times;10⁻\u0026sup1;\u0026sup1;), AT_Left (r\u0026thinsp;=\u0026thinsp;.19, p\u0026thinsp;=\u0026thinsp;9.69\u0026times;10⁻\u0026sup1;⁰), MP_Left (r\u0026thinsp;=\u0026thinsp;.15, p\u0026thinsp;=\u0026thinsp;5.72\u0026times;10⁻⁷), and VMPFC_Left (r\u0026thinsp;=\u0026thinsp;.13, p\u0026thinsp;=\u0026thinsp;2.37\u0026times;10⁻⁵). A similar but weaker organization was present in the RH, where the strongest effects appeared in PT_Right (r\u0026thinsp;=\u0026thinsp;.19, p\u0026thinsp;=\u0026thinsp;1.45\u0026times;10⁻⁹), followed by AT_Right (r\u0026thinsp;=\u0026thinsp;.18, p\u0026thinsp;=\u0026thinsp;9.60\u0026times;10⁻⁹), IF_Right (r\u0026thinsp;=\u0026thinsp;.16, p\u0026thinsp;=\u0026thinsp;1.89\u0026times;10⁻⁷), MP_Right (r\u0026thinsp;=\u0026thinsp;.15, p\u0026thinsp;=\u0026thinsp;5.35\u0026times;10⁻⁷), and VMPFC_Right (r\u0026thinsp;=\u0026thinsp;.10, p\u0026thinsp;=\u0026thinsp;.0012). All correlations were positive in direction, indicating that higher language ability was consistently associated with stronger β responses across distributed LN subregions in both hemispheres.\u003c/p\u003e \u003cp\u003eCerebellar activation analysis\u003c/p\u003e \u003cp\u003eFinally, we examined the contribution of the cerebellum. First, a probabilistic map (\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e) was used to identify cerebellar regions consistently activated during language tasks (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA). This map confirmed that activation was most probable in the right posterior cerebellum, an area encompassing Crus I, Crus II, and Lobule VI, with additional extension into Lobule VIIIA, which together form part of the cerebello-cortical loops engaged during language processing.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo examine group-specific engagement of these cerebellar regions, we calculated Cohen's d effect size maps for the story versus math contrast in each group separately. In the current study, both Group H (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB) and Group L (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC) showed activations specifically in Crus I and Crus II, though the magnitude of activation differed between groups. Group H showed robust, positive effect sizes throughout Crus I and Crus II, confirming their strong recruitment. In contrast, Group L exhibited weaker activation. This pattern suggests that cerebellar engagement during language processing scales with behavioral performance, paralleling the cortical findings.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur results reveal a clear and multi-faceted pattern of neural differences linked to language ability. Our behavioral selection successfully isolated two groups differing in language ability but matched for social cognition (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Across the cerebrum, voxel-wise and ROI-based analyses consistently demonstrated that Group H exhibited significantly greater activation in canonical language regions compared to Group L (Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e–\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Critically, brain-behavior correlations confirmed that activation strength within these regions predicted language ability across the entire sample, with the strongest relationships observed in LH frontal and temporal areas (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). This pattern extended to the cerebellum, where our analysis revealed stronger recruitment of language-related subregions, specifically Crus I and Crus II, in Group H relative to Group L (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThis comprehensive evidence demonstrates that superior language ability is associated with more robust and widespread engagement of the entire cerebro-cerebellar language network. After controlling for social cognition, Group H exhibited significantly greater activation across distributed language regions spanning LH PT, MT, AT, IF, DMPFC, and MP, as well as RH PT, AT, IF, VMPFC, and MP. This pattern was strongly left-lateralized but also included key regions in the RH, demonstrating that language ability relies on coordinated activity across both hemispheres and extends beyond cortical networks to include critical cerebellar contributions.\u003c/p\u003e \u003cp\u003eThe pronounced LH dominance in Group H aligns with extensive evidence establishing the left cortical network as the core system for language processing. The heightened activation we observed in LH IF and temporal regions corresponds directly to their known roles in syntactic processing and semantic integration, respectively, as confirmed by meta-analyses (\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e). Furthermore, the increased RH activation in Group H, particularly in IF_Right and VMPFC_Right, likely reflects greater recruitment of executive and supportive language functions needed for higher-level performance. This supports findings that the RH is engaged during more complex language demands (\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn the cerebellum, the greater activation in Crus I and Crus II within Group H highlights the importance of the cerebro-cerebellar loop in language processing. These findings are consistent with established functional topography: Crus I and Crus II are tied to cognitive aspects of language including syntactic and semantic processing (\u003cspan additionalcitationids=\"CR29 CR30\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e–\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e). The attenuated activation in Group L may therefore represent less efficient engagement of this crucial modulatory circuit, directly contributing to their performance differences. This aligns with evidence that cerebellar integrity is critical for fluent language execution (\u003cspan additionalcitationids=\"CR31\" citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e–\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOur results also contribute to the broader debate on neural mechanisms underlying individual differences in language ability. The clear pattern of greater activation associated with better performance in our cohort aligns with models where superior ability is linked to enhanced recruitment of neural resources (\u003cspan additionalcitationids=\"CR36 CR37 CR38 CR39 CR40\" citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e–\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e). This pattern in healthy individuals provides a valuable baseline for understanding clinical conditions; for instance, language network alterations in patients with epilepsy have been shown to correlate with performance differences similar to those we observed (\u003cspan additionalcitationids=\"CR35 CR36\" citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e–\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e). Similarly, our findings resonate with studies demonstrating that stronger functional connectivity within the LN correlates with better verbal fluency, suggesting that the activation patterns we observe during tasks reflect underlying network architecture (\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eLimitations\u003c/p\u003e \u003cp\u003eSeveral limitations of this study highlight avenues for future research. Our cross-sectional design, based on a sample from the HCP, establishes a strong association but cannot determine the causal relationship between brain activation and language ability. Future longitudinal studies are needed to track how these brain-behavior patterns develop over time and whether training-induced improvements in language performance correspond to changes in neural engagement. Although we controlled for social cognition, other unmeasured variables could have influenced the results. Additionally, our analysis focused primarily on activation magnitude; future work incorporating functional connectivity measures could provide deeper insight into the network dynamics supporting high-level language performance. Finally, while our cerebellar analysis revealed group differences in Crus I and Crus II, a more fine-grained parcellation of cerebellar subregions and their specific contributions to different aspects of language processing warrants further investigation.\u003c/p\u003e "},{"header":"Methods","content":"\u003cp\u003eParticipants\u003c/p\u003e\u003cp\u003eIn this study, we used the HCP S1200 dataset of healthy adults aged 22–35 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.humanconnectome.org/study/hcp-young-adult/document/1200-subjects-data-release\u003c/span\u003e\u003cspan address=\"https://www.humanconnectome.org/study/hcp-young-adult/document/1200-subjects-data-release\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Subjects were recruited from Washington University (St. Louis, MO) and the surrounding area. The HCP data were acquired using protocols approved by the Washington University institutional review board, and written informed consent was obtained from all subjects. From the 1,206 subjects in the release, 157 subjects lacked complete functional data for the story versus math contrast and were excluded. The remaining 1,049 subjects had complete functional data for both the language task and the social cognition task.\u003c/p\u003e\u003cp\u003eFrom this cohort of 1,049 subjects, we employed an extreme groups approach to identify individuals differing substantially in language ability. For each individual, age-adjusted scores from the NIH toolbox picture vocabulary test (PicVocab_AgeAdj) and the NIH toolbox oral reading recognition test (ReadEng_AgeAdj)—both administered outside the scanner as part of the standard HCP behavioral battery—were first standardized (z-scored) against the full sample of 1,049 participants. These two z-scores were then averaged to create a composite language score (Mean_z) for each participant.\u003c/p\u003e\u003cp\u003eTo isolate language ability from general cognitive function, we used behavioral performance on the Theory of Mind task (ToM_AgeAdj score) as a matching criterion. Only participants who demonstrated high performance on this social cognition measure were eligible for group selection. From this subset, the low-performing group (Group L) comprised 35 individuals with composite language scores more than one standard deviation below the mean, while the high-performing group (Group H) comprised 35 individuals with scores more than one standard deviation above the mean (Total N = 70). This procedure ensured that both groups demonstrated intact social cognitive abilities while differing substantially in language ability.\u003c/p\u003e\u003cp\u003eSocial cognition performance calculation\u003c/p\u003e\u003cp\u003eSince the HCP dataset does not provide a pre-calculated accuracy score for the ToM task, we computed accuracy using participant responses across the six possible response categories. These categories reflect how participants judged each video clip: A = correctly identifying Random trials as Random, B = correctly identifying Mental (ToM) trials as Mental, C = incorrectly judging Random trials as Mental, D = responding \"Unsure\" to Random trials, E = incorrectly judging Mental trials as Random, and F = responding \"Unsure\" to Mental trials. Accuracy was calculated using the formula: Accuracy = (A + B) - (C + D + E + F), which rewards correct responses (A and B) while penalizing all incorrect or unsure responses (C, D, E, and F). This accuracy score was used to identify participants with high performance on the social cognition task.\u003c/p\u003e\u003cp\u003eTask paradigm\u003c/p\u003e\u003cp\u003eLanguage tests\u003c/p\u003e\u003cp\u003eLanguage abilities were assessed using two standardized measures from the NIH toolbox. Vocabulary knowledge was measured with the picture vocabulary test (age-adjusted scale score; PicVocab_AgeAdj), a computerized adaptive test of general vocabulary knowledge that indexes crystallized verbal ability. Participants listened to an audio recording of a word and selected, from four photographic images, the picture that best matched the word’s meaning. Reading ability was assessed with the oral reading recognition test (age-adjusted scale score; ReadEng_AgeAdj), a computerized adaptive measure of reading decoding skill and crystallized ability in which participants read and pronounce letters and words as accurately as possible. For both measures, age-adjusted scores were normed using age-appropriate bands of the toolbox norming sample (18–29 or 30–35), where a score of 100 indicates national-average performance and 115 or 85 indicate performance ± 1 SD relative to the participant’s age band; higher scores indicate better ability.\u003c/p\u003e\u003cp\u003eLanguage task (story vs. math)\u003c/p\u003e\u003cp\u003eFunctional data in this study were based on the HCP language processing task (\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e). The language task consisted of two runs (run duration = 3:57 min:sec). In each run, 4 blocks of a story task were interleaved with 4 blocks of a math task. The lengths of blocks varied, and the average duration of blocks was approximately 30 s.\u003c/p\u003e\u003cp\u003eIn the story blocks, participants were presented with brief auditory stories (5–9 sentences) adapted from Aesop's fables, followed by a 2-alternative forced-choice question that asked participants about the topic of the story. For example, after a story about an eagle that saves a man who had done him a favor, participants were asked \"Was that about revenge or reciprocity?\" Participants pressed a button to select either the first or the second choice. The math task also included trials that were presented auditorily. In these trials, participants completed a series of simple arithmetic (addition and subtraction) operations (e.g., \"Fourteen plus twelve\"), followed by \"equals\" and then two choices (e.g., \"twenty-nine or twenty-six\"). Participants pressed a button to select either the first or the second answer. The math task was adaptive to maintain a similar level of difficulty across participants. The math condition served as a control that matched the story condition for auditory input, motor response demands, and block duration, while minimizing language-specific semantic processing.\u003c/p\u003e\u003cp\u003eSocial cognition task (Theory of Mind)\u003c/p\u003e\u003cp\u003eThe ToM task assessed social cognition by presenting participants with short video clips (20 seconds each) depicting simple geometric shapes—such as squares, circles, and triangles—either engaging in meaningful interactions or moving in a random, non-social manner (\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e). These animations were adapted from classic ToM stimuli in which social interactions are inferred purely from the motion dynamics of the shapes. Following each clip, participants judged whether the sequence portrayed a mental interaction—that is, behavior suggestive of intentions, beliefs, or emotional states—not sure, or no interaction, indicating random movement without apparent social content. The task consisted of two runs (each lasting 3 minutes and 27 seconds), with each run containing five video blocks: in one run, two Mental and three Random blocks were presented, whereas in the other run the order was three Mental and two Random blocks. Each video block was separated by a 15-second fixation period. This paradigm reliably engages neural systems supporting ToM by requiring participants to interpret animate-like social behavior from abstract visual stimuli.\u003c/p\u003e\u003cp\u003eData acquisition\u003c/p\u003e\u003cp\u003eThe HCP MRI data acquisition has previously been described in detail (\u003cspan additionalcitationids=\"CR49\" citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e–\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e). Images were acquired using a customized 3T Siemens 'Connectom' Skyra scanner having a 100 mT/m SC72 gradient insert and a standard Siemens 32-channel RF-receive head coil. At least one 3D T1-weighted MPRAGE image and one 3D T2-weighted SPACE image were acquired at 0.7 mm isotropic resolution. Whole-brain task fMRI data were acquired using a multi-band EPI sequence with parameters of TR = 720 ms, TE = 33.1 ms, flip angle = 52°, 2 mm isotropic voxels, 72 slices, and multi-band acceleration factor of 8. Spin echo field maps were acquired during both structural and fMRI scanning sessions to enable accurate cross-modal registration of structural and functional images in each subject.\u003c/p\u003e\u003cp\u003eAnalysis of structural data\u003c/p\u003e\u003cp\u003eStructural images (T1-weighted and T2-weighted) were used for extracting subcortical gray matter structures and reconstructing cortical surfaces in each subject. Volume data were transformed from native space into MNI space using a nonlinear volume-based registration. For accurate cross-subject registration of cortical surfaces, a multimodal surface matching (MSM) algorithm was used. The MSM algorithm had two versions: 'MSMSulc' (non-rigid surface alignment based on folding patterns) and 'MSMAll' (optimized alignment of cortical areas using sulcal depth maps plus features from other modalities including myelin maps, resting-state network maps, and visuotopic connectivity maps). Data in our work were based on MSMAll registration. After surface and volume registration, cortical vertices were combined with subcortical gray matter voxels to form the standard 'CIFTI grayordinates' space (91,282 vertices/voxels with 2 mm cortical vertex spacing and 2 mm isotropic subcortical voxels).\u003c/p\u003e\u003cp\u003eAnalysis of fMRI data\u003c/p\u003e\u003cp\u003eFunctional images were minimally preprocessed using the HCP pipelines (\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e). Preprocessing included correction for spatial distortions due to gradient nonlinearity and B0 field inhomogeneity, fieldmap-based unwarping of EPI images, motion correction, brain-boundary-based registration of EPI to structural T1-weighted scans, non-linear registration to MNI space, and grand-mean intensity normalization. Data from the cortical gray matter ribbon were projected onto the surface and then onto the standard grayordinates space. Data were minimally smoothed by a 2 mm FWHM Gaussian kernel in the grayordinates space. Thus, smoothing was constrained to the cortical surface mesh in each hemisphere.\u003c/p\u003e\u003cp\u003eThe preprocessed functional time-series were entered into a general linear model (GLM) to estimate functional activities in each vertex/voxel in each run. Two regressors/predictors were included in the GLM design of the language task: story and math. Each predictor covered the duration of a block (~ 30 s). All regressors were convolved with a canonical hemodynamic response function and its temporal derivatives. The time-series were temporally filtered with a Gaussian-weighted linear high-pass filter with a cutoff of 200 s, to remove low-frequency drifts/fluctuations presumably unrelated to the task design. The time-series were also prewhitened to remove temporal autocorrelations in the fMRI signal.\u003c/p\u003e\u003cp\u003eFor the story versus math contrast, the contrast of parameter estimate (COPE) was computed based on beta values of the GLM. Fixed-effects analyses were conducted to estimate the average effects across runs within each subject. The resulting subject-level contrast maps were used for all subsequent group-level analyses.\u003c/p\u003e\u003ch2\u003eStatistical analysis\u003c/h2\u003e\u003cp\u003eGroup-level effect size maps\u003c/p\u003e\u003cp\u003eTo visualize and quantify activation patterns within each performance group, we calculated Cohen's d effect size maps for the story versus math contrast separately for Group H and Group L. For each group, Cohen's d was computed at each vertex/voxel as the mean COPE across subjects divided by the standard deviation of COPE values. These effect size maps enabled direct comparison of activation magnitude between groups while accounting for within-group variability.\u003c/p\u003e\u003cp\u003eVoxel-wise group comparison\u003c/p\u003e\u003cp\u003eTo identify regions showing significant differences in activation between Group H and Group L, we performed voxel-wise statistical analysis using Threshold-Free Cluster Enhancement (TFCE) as implemented in FSL's randomise tool (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e). TFCE enhances cluster-like structures in statistical maps without requiring an arbitrary cluster-forming threshold. We tested the contrast Group H \u0026gt; Group L using 5,000 permutations to generate a null distribution. The resulting statistical maps were family-wise error (FWE) corrected at p \u0026lt; 0.05. For visualization, significance scores were displayed on a logarithmic scale ranging from 0 to 2.02.\u003c/p\u003e\u003cp\u003eRegion-of-interest (ROI) definition\u003c/p\u003e\u003cp\u003eTo enable precise quantification of group differences across anatomically defined language regions, we established functional ROIs using an independent validation approach to avoid circular analysis. From the full cohort of 1,049 subjects with complete language task data, we excluded the 70 subjects used in our main analysis (35 in Group H and 35 in Group L), leaving an independent validation sample of 979 subjects.\u003c/p\u003e\u003cp\u003eUsing this independent sample, we generated a group-level Cohen's d effect size map for the story versus math contrast. This map was thresholded at d = 0.8 to identify regions showing robust language-related activation. The resulting activation clusters were then manually parcellated into functionally defined ROIs corresponding to seven regions in the LH and five regions in the RH: posterior temporal (PT), middle temporal (MT), anterior temporal (AT), inferior frontal (IF), ventromedial prefrontal cortex (VMPFC), dorsomedial prefrontal cortex (DMPFC), and medial parietal (MP).\u003c/p\u003e\u003cp\u003eROI-based group comparison\u003c/p\u003e\u003cp\u003eFor each of the 12 functionally defined ROIs, we extracted the mean COPE value (story vs. math) for each subject in both Group H and Group L. Independent samples t-tests were conducted to compare mean activation between groups for each ROI. To control for multiple comparisons across the 12 ROIs, we applied false discovery rate (FDR) correction using the Benjamini-Hochberg procedure (62). FDR-corrected p-values \u0026lt; 0.05 were considered statistically significant. Effect sizes (Cohen's d) were calculated for each ROI comparison as the difference in group means divided by the pooled standard deviation.\u003c/p\u003e\u003cp\u003eBrain-behavior correlation analysis\u003c/p\u003e\u003cp\u003eTo examine the relationship between neural activation and individual differences in language ability across the full range of performance, we conducted brain-behavior correlation analyses using all 1,049 subjects with complete language task data. For each subject, mean β-estimates (story vs. math) were extracted from each of the 12 ROIs. These activation values were correlated with the composite language ability score (Mean_z, derived from z-scored PicVocab_AgeAdj and ReadEng_AgeAdj measures) using Pearson correlation. Statistical significance was assessed using two-tailed tests, with p-values reported for each ROI without correction for multiple comparisons, as these analyses were exploratory and aimed at characterizing the continuous brain-behavior relationship across the entire sample rather than testing a specific hypothesis about individual ROIs.\u003c/p\u003e\u003cp\u003eCerebellar analysis\u003c/p\u003e\u003cp\u003eCerebellar activation patterns were analyzed separately due to the distinct functional organization of the cerebellum. We first identified language-relevant cerebellar regions using a probabilistic atlas map. This probabilistic map, based on coordinates from multiple neuroimaging studies, indicated that language-related cerebellar activation is most probable in the right posterior cerebellum, specifically Crus I, Crus II, Lobule VI, and Lobule VIIIA.\u003c/p\u003e\u003cp\u003eFor our cohort, we calculated Cohen's d effect size maps for the story versus math contrast separately for Group H and Group L in cerebellar space. Visual inspection of these maps confirmed that both groups showed activation specifically in Crus I and Crus II, consistent with the probabilistic atlas but not extending to Lobule VI and Lobule VIIIA in our sample. We qualitatively compared activation magnitude between groups by examining the distribution and intensity of positive Cohen's d values within these cerebellar regions.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eAuthor\u0026rsquo;s contributions\u003c/p\u003e\n\n\u003cp\u003eR.R. conceived the idea and designed the analyses; H.D. and P.Z. performed data analysis and prepared figures; P.Z. wrote the manuscript; R.R. critically revised the manuscript. All authors approved the final version of the manuscript.\u003c/p\u003e\n\n\u003cp\u003eData availability statement\u003cspan dir=\"RTL\"\u003e \u003c/span\u003e\u003c/p\u003e\n\n\u003cp\u003eThis study used publicly available data from the Human Connectome Project (HCP). Access to the HCP dataset can be obtained through the official HCP data portal. The analysis codes developed for this study are available upon reasonable request.\u003c/p\u003e\n\n\u003cp\u003eConflict of interests statement\u003c/p\u003e\n\n\u003cp\u003eThe authors have no conflicts of interest to declare.\u003c/p\u003e\n\n\u003cp\u003eAcknowledgments\u003c/p\u003e\n\n\u003cp\u003eThis research was supported by IPM (Institute for Research in Fundamental Sciences).\u003c/p\u003e\n\n\u003cp\u003eFunding Declaration\u003c/p\u003e\n\n\u003cp\u003eFunding : This research was funded by IPM (Institute for Research in Fundamental Sciences).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSchro\u0026euml;n, J. A. M. et al. Causal evidence for a coordinated temporal interplay within the language network. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e \u003cb\u003e120\u003c/b\u003e (47), e2306279120 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBinder, J. R. et al. Human Brain Language Areas Identified by Functional Magnetic Resonance Imaging. \u003cem\u003eJ. Neurosci.\u003c/em\u003e \u003cb\u003e17\u003c/b\u003e (1), 353\u0026ndash;362 (1997).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLukic, S. et al. Common and distinct neural substrates of sentence production and comprehension. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e224\u003c/b\u003e, 117374 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHopkins, W. D. et al. A comprehensive analysis of variability in the sulci that define the inferior frontal gyrus in the chimpanzee (\u003cem\u003ePan troglodytes\u003c/em\u003e) brain. \u003cem\u003eAm. J. Biol. Anthropol.\u003c/em\u003e \u003cb\u003e179\u003c/b\u003e (1), 31\u0026ndash;47 (2022 Sept).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFriederici, A. D. The Brain Basis of Language Processing: From Structure to Function. \u003cem\u003ePhysiol. Rev.\u003c/em\u003e \u003cb\u003e91\u003c/b\u003e (4), 1357\u0026ndash;1392 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFedorenko, E., Behr, M. K. \u0026amp; Kanwisher, N. Functional specificity for high-level linguistic processing in the human brain. \u003cem\u003eProc. Natl. Acad. Sci. 2011 Sept\u003c/em\u003e \u003cb\u003e27\u003c/b\u003e;108(39):16428\u0026ndash;16433 .\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKonstantopoulos, K. \u0026amp; Giakoumettis, D. \u003cem\u003eNeuroimaging in neurogenic communication disorders\u003c/em\u003e 718 (Academic, 2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHagoort, P. M. U. C. \u0026amp; Memory Unification, Control) and beyond. Front Psychol [Internet]. [cited 2023 Dec 10];4. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://journal.frontiersin.org/article/\u003c/span\u003e\u003cspan address=\"http://journal.frontiersin.org/article/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fpsyg.2013.00416/abstract\u003c/span\u003e\u003cspan address=\"10.3389/fpsyg.2013.00416/abstract\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBelyk, M., Brown, S., Lim, J. \u0026amp; Kotz, S. A. Convergence of semantics and emotional expression within the IFG pars orbitalis. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e156\u003c/b\u003e, 240\u0026ndash;248 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrice, C. J. A review and synthesis of the first 20years of PET and fMRI studies of heard speech, spoken language and reading. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e62\u003c/b\u003e (2), 816\u0026ndash;847 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBullock, L., Forseth, K. J., Woolnough, O., Rollo, P. S. \u0026amp; Tandon, N. Supplementary motor area in speech initiation: A large-scale intracranial EEG evaluation of stereotyped word articulation. \u003cem\u003eiScience\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e (1), 111531 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, J. R., Zhao, L., Hullett, P. W. \u0026amp; Chang, E. F. Speech sequencing in the human precentral gyrus. \u003cem\u003eNat. Hum. Behav. 2025 July\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e;9(11):2327\u0026ndash;2344 .\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlaus, J. \u0026amp; Schutter, D. J. L. G. The Role of Left Dorsolateral Prefrontal Cortex in Language Processing. \u003cem\u003eNeuroscience\u003c/em\u003e \u003cb\u003e377\u003c/b\u003e, 197\u0026ndash;205 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHertrich, I., Dietrich, S., Blum, C. \u0026amp; Ackermann, H. The Role of the Dorsolateral Prefrontal Cortex for Speech and Language Processing. \u003cem\u003eFront. Hum. Neurosci.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 645209 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu, L. et al. A left-lateralized dorsolateral prefrontal network for naming. \u003cem\u003eCell. Rep.\u003c/em\u003e \u003cb\u003e44\u003c/b\u003e (5), 115677 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShen, W., Yuan, Y., Liu, C. \u0026amp; Luo, J. The roles of the temporal lobe in creative insight: an integrated review. \u003cem\u003eThink. Reason\u003c/em\u003e. \u003cb\u003e23\u003c/b\u003e (4), 321\u0026ndash;375 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWani, P. D. From Sound to Meaning: Navigating Wernicke\u0026rsquo;s Area in Language Processing. Cureus [Internet]. Sept 21 [cited 2025 Nov 25]; (2024). Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cureus.com/articles/283534-from-sound-to-meaning-navigating-wernickes-area-in-language-processing\u003c/span\u003e\u003cspan address=\"https://www.cureus.com/articles/283534-from-sound-to-meaning-navigating-wernickes-area-in-language-processing\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBinder, J. R. Current Controversies on Wernicke\u0026rsquo;s Area and its Role in Language. \u003cem\u003eCurr. Neurol. Neurosci. Rep.\u003c/em\u003e \u003cb\u003e17\u003c/b\u003e (8), 58 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, X. et al. The neural correlates of sub-lexical semantics and its integration with the lexical meaning in reading Chinese characters. \u003cem\u003eJ. Neurolinguistics\u003c/em\u003e. \u003cb\u003e69\u003c/b\u003e, 101176 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJung, J., Williams, S. \u0026amp; Lambon Ralph, M. A. The role of GABA in semantic memory and its neuroplasticity. eLife. June \u003cb\u003e3\u003c/b\u003e;12:RP91771. (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, W., Xiang, M. \u0026amp; Wang, S. The role of left angular gyrus in the representation of linguistic composition relations. \u003cem\u003eHum. Brain Mapp.\u003c/em\u003e \u003cb\u003e43\u003c/b\u003e (7), 2204\u0026ndash;2217 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStoeckel, C., Gough, P. M., Watkins, K. E. \u0026amp; Devlin, J. T. Supramarginal gyrus involvement in visual word recognition. \u003cem\u003eCortex\u003c/em\u003e \u003cb\u003e45\u003c/b\u003e (9), 1091\u0026ndash;1096 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartin, R. C., Yue, Q., Zahn, R. \u0026amp; Lu, Y. The role of variation in phonological and semantic working memory capacities in sentence comprehension: neural evidence from healthy and brain-damaged individuals. \u003cem\u003eCogn. Affect. Behav. Neurosci.\u003c/em\u003e \u003cb\u003e25\u003c/b\u003e (1), 240\u0026ndash;262 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGraves, W. W. et al. Correspondence between cognitive and neural representations for phonology, orthography, and semantics in supramarginal compared to angular gyrus. \u003cem\u003eBrain Struct. Funct.\u003c/em\u003e \u003cb\u003e228\u003c/b\u003e (1), 255\u0026ndash;271 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJonides, J. et al. The Role of Parietal Cortex in Verbal Working Memory. \u003cem\u003eJ. Neurosci.\u003c/em\u003e \u003cb\u003e18(13)\u003c/b\u003e (1), 5026\u0026ndash;5034 (1998 July).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarbey, A. K., Colom, R. \u0026amp; Grafman, J. Neural mechanisms of discourse comprehension: a human lesion study. \u003cem\u003eBrain\u003c/em\u003e \u003cb\u003e137\u003c/b\u003e (1), 277\u0026ndash;287 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGlanz, O., Hader, M., Schulze-Bonhage, A., Auer, P. \u0026amp; Ball, T. A Study of Word Complexity Under Conditions of Non-experimental, Natural Overt Speech Production Using ECoG. \u003cem\u003eFront. Hum. Neurosci.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 711886 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNakatani, H., Nakamura, Y. \u0026amp; Okanoya, K. Respective Involvement of the Right Cerebellar Crus I and II in Syntactic and Semantic Processing for Comprehension of Language. \u003cem\u003eCerebellum\u003c/em\u003e \u003cb\u003e22\u003c/b\u003e (4), 739\u0026ndash;755 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStoodley, C. J., Valera, E. M. \u0026amp; Schmahmann, J. D. Functional topography of the cerebellum for motor and cognitive tasks: An fMRI study. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e59\u003c/b\u003e (2), 1560\u0026ndash;1570 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSilveri, M. C. Contribution of the Cerebellum and the Basal Ganglia to Language Production: Speech, Word Fluency, and Sentence Construction\u0026mdash;Evidence from Pathology. \u003cem\u003eCerebellum\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e (2), 282\u0026ndash;294 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSimonyan, K. \u0026amp; Fuertinger, S. Speech networks at rest and in action: interactions between functional brain networks controlling speech production. \u003cem\u003eJ. Neurophysiol.\u003c/em\u003e \u003cb\u003e113\u003c/b\u003e (7), 2967\u0026ndash;2978 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMannarelli, D. et al. Cerebellum\u0026rsquo;s Contribution to Attention, Executive Functions and Timing: Psychophysiological Evidence from Event-Related Potentials. \u003cem\u003eBrain Sci.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e (12), 1683 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeterburs, J. \u0026amp; Desmond, J. E. The role of the human cerebellum in performance monitoring. \u003cem\u003eCurr. Opin. Neurobiol.\u003c/em\u003e \u003cb\u003e40\u003c/b\u003e, 38\u0026ndash;44 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Ettinger-Veenstra, H., McAllister, A., Lundberg, P., Karlsson, T. \u0026amp; Engstr\u0026ouml;m, M. Higher Language Ability is Related to Angular Gyrus Activation Increase During Semantic Processing, Independent of Sentence Incongruency. Front Hum Neurosci [Internet]. 2016 Mar 11 [cited 2025 Nov 25];10. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://journal.frontiersin.org/Article/\u003c/span\u003e\u003cspan address=\"http://journal.frontiersin.org/Article/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fnhum.2016.00110/abstract\u003c/span\u003e\u003cspan address=\"10.3389/fnhum.2016.00110/abstract\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeber, S., Hausmann, M., Kane, P. \u0026amp; Weis, S. The relationship between language ability and brain activity across language processes and modalities. \u003cem\u003eNeuropsychologia\u003c/em\u003e \u003cb\u003e146\u003c/b\u003e, 107536 (2020 Sept).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMatchin, W., Hammerly, C. \u0026amp; Lau, E. The role of the IFG and pSTS in syntactic prediction: Evidence from a parametric study of hierarchical structure in fMRI. \u003cem\u003eCortex\u003c/em\u003e \u003cb\u003e88\u003c/b\u003e, 106\u0026ndash;123 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePravat\u0026agrave;, E. et al. Functional Connectivity MR Imaging of the Language Network in Patients with Drug-Resistant Epilepsy. \u003cem\u003eAm. J. Neuroradiol.\u003c/em\u003e \u003cb\u003e32\u003c/b\u003e (3), 532\u0026ndash;540 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePistono, A. et al. Increased functional connectivity supports language performance in healthy aging despite gray matter loss. \u003cem\u003eNeurobiol. Aging\u003c/em\u003e. \u003cb\u003e98\u003c/b\u003e, 52\u0026ndash;62 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGertel, V. H., Zhang, H. \u0026amp; Diaz, M. T. Stronger right hemisphere functional connectivity supports executive aspects of language in older adults. \u003cem\u003eBrain Lang.\u003c/em\u003e \u003cb\u003e206\u003c/b\u003e, 104771 (2020 July).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMir\u0026oacute;-Padilla, A., Bueichek\u0026uacute;, E., Ventura-Campos, N., Palomar-Garc\u0026iacute;a, M. \u0026Aacute;. \u0026amp; \u0026Aacute;vila, C. Functional connectivity in resting state as a phonemic fluency ability measure. \u003cem\u003eNeuropsychologia\u003c/em\u003e \u003cb\u003e97\u003c/b\u003e, 98\u0026ndash;103 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYue, Q. et al. Brain Modularity Mediates the Relation between Task Complexity and Performance. \u003cem\u003eJ. Cogn. Neurosci.\u003c/em\u003e \u003cb\u003e29(9)\u003c/b\u003e (1), 1532\u0026ndash;1546 (2017 Sept).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Ettinger-Veenstra, H. M. et al. Right-hemispheric brain activation correlates to language performance. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e49\u003c/b\u003e (4), 3481\u0026ndash;3488 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCasto, C. et al. The cerebellar components of the human language network [Internet]. Neuroscience; [cited 2025 Nov 29]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://biorxiv.org/lookup/doi/\u003c/span\u003e\u003cspan address=\"http://biorxiv.org/lookup/doi/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1101/2025.04.14.645351\u003c/span\u003e\u003cspan address=\"10.1101/2025.04.14.645351\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTurker, S., Kuhnke, P., Eickhoff, S. B., Caspers, S. \u0026amp; Hartwigsen, G. Cortical, subcortical, and cerebellar contributions to language processing: A meta-analytic review of 403 neuroimaging experiments. \u003cem\u003ePsychol. Bull.\u003c/em\u003e \u003cb\u003e149\u003c/b\u003e (11\u0026ndash;12), 699\u0026ndash;723 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRodd, J. M., Vitello, S., Woollams, A. M. \u0026amp; Adank, P. Localising semantic and syntactic processing in spoken and written language comprehension: An Activation Likelihood Estimation meta-analysis. \u003cem\u003eBrain Lang.\u003c/em\u003e \u003cb\u003e141\u003c/b\u003e, 89\u0026ndash;102 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWarren, J. E., Crinion, J. T., Lambon Ralph, M. A. \u0026amp; Wise, R. J. S. Anterior temporal lobe connectivity correlates with functional outcome after aphasic stroke. \u003cem\u003eBrain\u003c/em\u003e \u003cb\u003e132\u003c/b\u003e (12), 3428\u0026ndash;3442 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarch, D. M. et al. Function in the human connectome: Task-fMRI and individual differences in behavior. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e80\u003c/b\u003e, 169\u0026ndash;189 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUğurbil, K. et al. Pushing spatial and temporal resolution for functional and diffusion MRI in the Human Connectome Project. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e80\u003c/b\u003e, 80\u0026ndash;104 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith, S. M. et al. Resting-state fMRI in the Human Connectome Project. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e80\u003c/b\u003e, 144\u0026ndash;168 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGlasser, M. F. et al. The minimal preprocessing pipelines for the Human Connectome Project. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e80\u003c/b\u003e, 105\u0026ndash;124 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSpis\u0026aacute;k, T. et al. Probabilistic TFCE: A generalized combination of cluster size and voxel intensity to increase statistical power. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e185\u003c/b\u003e, 12\u0026ndash;26 (2019).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Language network, Cerebellum, Functional MRI, Language performance, Brain-behavior coupling","lastPublishedDoi":"10.21203/rs.3.rs-8370100/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8370100/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe neural basis of individual differences in language ability remains incompletely understood. While previous research has identified brain regions associated with language processing, few studies have systematically examined whether higher language performance corresponds to greater neural engagement across cortical and cerebellar language network. Using data from the Human Connectome Project (HCP), we investigated neural activation patterns in 70 healthy adults stratified into high-performing and low-performing groups (35 subjects per group) based on composite scores from language tests, while controlling for social cognition to isolate language-specific effects. Functional MRI data from the story versus math task were analyzed to compare activation patterns between groups. The high-performing group exhibited significantly greater activation across distributed language regions in both hemispheres, with particularly robust effects in left frontal and temporal areas. Brain-behavior correlation analyses across the full sample confirmed that activation strength within these regions predicted language ability. Cerebellar analysis revealed greater recruitment of right posterior regions in the high-performing group, extending the pattern of enhanced activation beyond cortical networks. These findings demonstrate that superior language ability is associated with more robust engagement of a distributed cerebro-cerebellar language network, suggesting that individual differences in language skills are reflected in the magnitude of neural resource recruitment during language processing.\u003c/p\u003e","manuscriptTitle":"Contributions of brain’s language network to the behavioral language performance","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-09 12:31:55","doi":"10.21203/rs.3.rs-8370100/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-13T06:29:11+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-03T09:55:45+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"147688065906175433725066072839108793603","date":"2026-03-19T23:06:00+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-10T11:13:24+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"94668835936651015968810710733730483287","date":"2026-02-25T13:41:40+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"266829917234413936187257628734361597629","date":"2026-02-23T08:27:52+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-17T18:26:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"243731851590799982638567091359635443450","date":"2026-02-11T22:20:46+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"147688065906175433725066072839108793603","date":"2026-02-04T23:35:40+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"239803789029365747591331727330794963424","date":"2026-02-04T15:36:00+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-02-04T11:52:03+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-12-23T04:45:47+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-12-18T09:49:12+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-12-18T09:46:30+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-12-15T21:35:11+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"8c41e79e-d4e7-4b72-bacb-7ae76017755d","owner":[],"postedDate":"February 9th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[{"id":62424921,"name":"Biological sciences/Neuroscience"},{"id":62424922,"name":"Biological sciences/Psychology"},{"id":62424923,"name":"Social science/Psychology"}],"tags":[],"updatedAt":"2026-04-13T06:39:53+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-09 12:31:55","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8370100","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8370100","identity":"rs-8370100","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}