Excessive self-blaming emotions are commonly observed in anxiety disorders, with qualitatively similar symptomatology reported in subclinical populations. Interpretation of moral information requires assessing the social conceptual information, a process overseen by the superior anterior temporal lobe (sATL). Feelings of self-blame evoke interactions of sATL and socio-affective regions, and previous research shows that subclinical anxiety modulates the organisation of the self-blame circuitry. This study aimed to extend these findings by exploring links of trait-anxiety with (i) self-blaming emotions and associated behaviours in an experimental task, and (ii) self-blame-dependent neural activity and connectivity, as observed during reliving of autobiographical guilt memories. We also explored the role of resting-state fMRI in linking these phenomena. Increased anxiety was linked to stronger self-blaming emotions, and more pronounced self-attacking and hiding. When experiencing negative emotions about themselves (i.e. shame and self-anger), anxious individuals were also less likely to disengage from self-focused thoughts. These behavioural findings were paralleled by enhanced self-blame-related connectivity between the left sATL and bilateral posterior subgenual cingulate cortex. Distinct patterns of activity and connectivity within the ATL-related circuitry were furthermore linked to individual differences in intensity of the self-blaming emotions and approach-avoidance motivation towards the guilt memories. As such, the results of the current study link stronger self-blaming emotions in anxious individuals with specific maladaptive patterns of behaviour. Furthermore, the work provides robust evidence for the important role of ATL-related circuitry in self-blame processing, supporting its broader involvement in social conceptual processing and its alterations in subclinical anxiety.

Abstract The medial prefrontal cortex is central to learning flexible internal models across diverse domains, yet the functional specialization enabling this remains unknown. We tested whether medial prefrontal specialization is representational (encoding domain-specific features) or computational (implementing domain-general computations). During fMRI, participants learned probabilistic features of virtual environments representing spatial, social, and sequential domain knowledge. Although each domain used different features, they shared the same feature-to-latent state mapping, matching their computational demands. The medial PFC showed no domain-specific feature representations. Instead, its neural patterns revealed a triad of specialized yet domain-invariant computations. Ventromedial PFC patterns reflected probabilistic inference, abstracting hidden probability distributions from observations and tracking trial-wise posterior task state changes within a low-dimensional latent space. Anteromedial PFC organized task states along orthogonal axes, tracking directional shifts within states and switches between different states, suggesting a global task coordinate system. Dorsomedial PFC patterns represented task dynamics, using predictive surprise to monitor validity of the current internal model and switch task policies. These results suggest a principled architecture in medial PFC where three general-purpose computations jointly enable learning world-models across diverse environments.

Objectives: Coherent speech remains focused on the topic at hand, avoiding irrelevant or tangential comments. Discourse coherence often declines with age and this has been linked to declines in cognitive control processes that regulate activation of knowledge and ideas. The present study tested whether external topic reminders could improve coherence by reducing reliance on internal control processes. Methods: 77 young and older adults produced spoken discourse in response to topic prompts. In the baseline condition, the topic prompt was removed when the participant began speaking. This was compared with conditions where a reminder was presented on screen, either throughout their response or only during the last 30 seconds. Results: Both age groups produced more coherent, topic-focused responses when they had access to a topic reminder. Late reminders had an additional proactive impact on participants’ speech, increasing its coherence even before the reminder appeared. This proactive effect was short-lived in the older group but more sustained in the young group. Discussion: When speaking, people must maintain the current discourse goal (i.e., intended topic), in order to prioritise relevant information and inhibit irrelevant ideas. Topic reminders boost activation of the discourse goal, reducing the need to maintain this through internal control. Furthermore, people proactively strengthened goal activation when they anticipated a later reminder, but older people were less effective at maintaining this throughout their response. Together, these results suggest that internal maintenance of discourse goals is less effective in older people, though they do benefit from external cues that support topic maintenance.

Narrative comprehension involves creating a mental representation of the events of the story: a "situation model". Maintaining a situation model is thought to be supported by the Default Mode Network (DMN), but recent work suggests that the semantic system, and specifically the ventrolateral anterior temporal lobe (ATL), may play a role in reflecting on and restructuring the situation model via internally-driven or endogenous semantic processing. The present study used fMRI to investigate how ATL and DMN brain regions respond under varying exogenous, or input-driven, and endogenous processing demands when reading social and non-social stories. We studied neural responses to three types of situation model manipulation: 1) add - incorporating new information into the situation model, 2) use - using the information in the situation model to support comprehension of narrative language input, and 3) reconfigure - restructuring the situation model. Relative to add, the use and reconfigure manipulations tended to elicit greater activation in regions of the DMN, including the dorsomedial prefrontal cortex, posterior cingulate cortex and precuneus, as well as the bilateral superior, middle and inferior ATL. Relative to non-social stories, add and use manipulations in social stories engaged the left anterior middle and superior temporal gyri and inferior parietal lobule (IPL), whereas reconfigure manipulations engaged the right superior, middle and inferior frontal gyri and IPL. The present results inform a developing framework for coordination between the ATL and DMN during narrative comprehension.

Semantic cognition (use of acquired world knowledge to guide behaviour) is critical in everyday life. Semantic cognition is often assumed to be preserved in later life, compensating for functional declines in other cognitive domains. However, aging research rarely considers age-related effects on non-verbal knowledge or on semantic control processes that regulate how knowledge is activated and used. We addressed this by conducting the most detailed assessment of semantic cognition across the adult lifespan to date, involving 537 UK adults aged between 20 and 91. Verbal semantic knowledge increased linearly across adulthood while non-verbal knowledge reached a plateau at age 60. In contrast, controlled semantic processing showed age-related decline, particularly in the ability to inhibit task-irrelevant semantic knowledge. These results indicate that semantic cognition is not uniformly preserved in old age: though older people know more than young people, they are less able to use their knowledge flexibly in novel situations.

How do people use semantic knowledge to guide decision-making? Semantic decisions can be affected both by the intrinsic qualities of the concepts being considered and the relationships between them. Age-related change in these factors is rarely studied; yet this can provide valuable insights into how semantic cognition develops across the lifespan. We investigated the factors influencing semantic decisions in 537 adults aged 20–91 years. Participants made two types of verbal semantic decision: global relatedness judgements and feature-specific judgements based on object colour. We predicted accuracy and response time from age, the intrinsic properties of the item being judged, and the strengths of target and distractor relationships on each trial. Effects of intrinsic properties varied markedly with age. Younger adults made more errors when faced with less frequent and less concrete items, but these effects diminished across adulthood, suggesting that greater experience can reinforce the representation of uncommon and abstract concepts. In contrast, higher semantic diversity impaired performance disproportionately in older people, suggesting that older adults experienced greater difficulty when concepts activated many different associations. Strong but task-irrelevant semantic distractors also led to more errors, particularly among older adults. These findings indicate that the determinants of semantic decision-making shift systematically across adulthood. Younger people have difficulty dealing with infrequent, abstract concepts with which they have limited experience; while advancing age is associated with greater difficulty resolving competition among simultaneously active aspects of meaning. Our data support the view that semantic knowledge accumulates throughout adulthood while goal-directed selection of knowledge declines.

Prompted monologues – where people speak at length in response to a topic prompt – offer a controlled but ecologically valid way to study discourse production. However, there is little systematic understanding of how speaker characteristics, topic properties, and experimental context influence discourse of this kind. The Edinburgh Discourse Corpus (EDC) addresses these limitations, bringing together 7,261 spoken monologues collected across nine studies varying in experimental settings, instructions, and task demands. Participants produced speech for between 50 and 90s at a time, in response to a diverse set of prompts probing knowledge and opinions about a range of topics. The corpus contains responses from 462 participants and comprises over a million words of transcribed speech. We calculated 23 measures that capture lexical, semantic, syntactic and discourse-level properties and used Principal Components Analysis to reduce these to seven interpretable components. These speech components were shaped by characteristics of the speaker, prompt, and experimental context. Older adults exhibited reduced speech coherence, positivity, and emotional intensity, but also drew on a more sophisticated vocabulary. Structured prompts elicited more lexically-constrained content while unstructured, open-ended prompts elicited more emotionally-charged language. Experimental context, including setting, task demands, and instruction types, also affected speech, highlighting the role of contextual and pragmatic factors on discourse. These findings suggest that discourse properties are sensitive to a wide range of speaker, prompt and contextual influences, with implications for the design of future discourse elicitation studies. The full EDC, including all transcripts and speech measures, is available as an open-access resource.

Creative ideas often stem from a deliberate process involving the exploration and exploitation of our existing knowledge. This study explores the cognitive mechanisms behind creative problem-solving, focusing on the phases of idea generation and evaluation in the Remote Associates Test (RAT). Generating novel ideas relies on an effective, goal-directed search through semantic memory, while evaluation requires the careful assessment of potential ideas for their appropriateness. These phases are analogous to semantic control processes, which manage the search, manipulation and selection of knowledge in line with specific goals. The current work investigates how individual differences in these cognitive processes relate to performance in divergent and convergent thinking tasks, by deconstructing the standard RAT to isolate generation and evaluation processes. Further, we examined the role of retrieval strategies, specifically clustering (grouping related concepts) and switching (moving between different areas of semantic space), in creative thought. Our findings indicate that the ability to generate ideas under multiple constraints predicts fluency in divergent thinking tasks, while strong evaluation skills may inhibit originality. Furthermore, the analysis of clustering and switching strategies reveals that effective semantic search involves a balance between deep exploration of specific conceptual areas and broad exploration across different semantic clusters. These findings suggest that multiple strategies can lead to creative success, underscoring the importance of flexible cognitive strategies in navigating the semantic space during creative problem-solving.

Conceptual knowledge-about objects, events, and social behaviour-is represented within the semantic system, but it is unclear if different conceptual categories engage the same portions of the system. This is perhaps most relevant for event-based, or thematic, knowledge and social knowledge which is acquired through social experiences. The present study investigated neural specialisation for social concepts by examining whether distinct semantic regions or hubs represent taxonomic versus thematic relations and social versus non-social relations. Specialisation was examined in two groups with different social experiences: autistic and non-autistic adults. There were minimal behavioural and no neural differences between groups, suggesting that differences in social experiences between autistic and non-autistic people may be better understood at the interactional level. In whole-brain analyses across both groups, taxonomic relations engaged the semantic control network to a greater extent than thematic relations did, and an overlapping portion of the rostroventral area of left angular gyrus was engaged by both thematic (relative to taxonomic) and social (relative to non-social) relations. Region of interest analyses revealed a more complex pattern within bilateral angular gyri. The results suggest that angular gyrus represents conceptual knowledge in a graded fashion, including specialisation for thematic and social relations.

Reliance on internal predictive models of the world is central to many theories of human cognition. Yet it is unknown whether humans acquire multiple separate internal models, each evolved for a specific domain, or maintain a globally unified representation. Using fMRI during naturalistic experiences (movie watching and narrative listening), we show that three topographically distinct midline prefrontal cortical regions perform distinct predictive operations. The ventromedial PFC updates contextual predictions (States), the anteromedial PFC governs reference frame shifts for social predictions (Agents), and the dorsomedial PFC predicts transitions across the abstract state spaces (Actions). Prediction-error-driven neural transitions in these regions, indicative of model updates, coincided with subjective belief changes in a domain-specific manner. We find these parallel top-down predictions are unified and selectively integrated with visual sensory streams in the Precuneus, shaping participants' ongoing experience. Results generalized across sensory modalities and content, suggesting humans recruit abstract, modular predictive models for both vision and language. Our results highlight a key feature of human world modeling: fragmenting information into abstract domains before global integration.

Older people’s speech is often less globally coherent than young people’s, i.e., less relevant to the topic at hand. This has sometimes been attributed to age-related change in communication goals, yet little is known about how different age groups modify their discourse style when they have different goals. Here, young and older adults produced discourse on a range of topics, first without specific instructions and then when instructed to be as entertaining or as focused as possible. When aiming to entertain, both age groups spoke more quickly and were less globally coherent, particularly in the later stages of responses. This later decline in coherence was more pronounced in the older group. These results indicate that deviations from the topic of discussion can stem from a desire to entertain one’s audience. Thus, a shift in communication goals is a viable possible explanation for age-related declines in coherence.

Creative thinking is a complex, higher-order ability that draws on multiple cognitive systems. However, the contribution of specific semantic control processes to creativity remains unclear. The current study had two goals: First, we investigated how individual differences in semantic knowledge and control contribute to divergent and convergent styles of creative thinking, beyond the involvement of domain-general executive functions. Second, we explored whether there were age-related differences in semantic and executive abilities, and if these differences influenced the ability to think creatively. Specifically, we examined the role of the two components of semantic control: controlled retrieval and semantic selection. In our study, 63 younger adults and 64 older adults completed semantic, executive, and creative thinking measures. Younger adults demonstrated better executive functioning, while older adults exhibited superior semantic knowledge, controlled retrieval, and convergent thinking abilities. Crucially, there were no age differences across several divergent thinking metrics: automated originality scoring, human ratings, or uniqueness. Regression analyses indicated that semantic knowledge and updating executive ability influenced convergent thinking abilities across both age groups. In contrast, semantic control abilities were predictive of divergent thinking skills, but only in the younger group. Our results emphasize the key role of the semantic system in creative thought, and, critically, indicate that divergent and convergent thinking may rely on different aspects of semantic cognition. Moreover, the recruitment of these abilities varies across the lifespan, in line with increased knowledge reserves and declines in executive control seen in older adults.

Allele-specific expression (ASE) outlier detection is a powerful tool for identifying genes affected by large effect rare genetic regulatory variants but suffers from data sparsity and noisy signal in low-count genes. Genome phasing can be utilized to aggregate ASE signal along haplotypes to alleviate both sparsity and noise. Yet statistical tools for utilizing haplotype-level ASE data for rare variant interpretation are lacking. Here, we present ANEVA-h, to quantify the amount of genetic variation in gene expression from haplotype-level ASE data in a population, enabling more accurate and comprehensive detection of regulatory effects. We apply ANEVA-h to GTEx project data, along with a compatible dosage outlier test, to show an over 2-fold increase in the number of testable genes, reduction of spurious outlier calls, and improved enrichment for rare high-impact variants. In clinical cohorts of neuromuscular and congenital heart disease, it enhances gene prioritization and identifies candidate diagnoses missed by DROP-MAE and ANEVA. Finally, we analyze globally diverse populations to characterize the impact of ancestry background in reference and the test population. We provide tools and data necessary to facilitate integration of haplotype level ASE outlier testing in rare variant interpretation pipelines.

Word embeddings derived from large language corpora have been successfully used in cognitive science and artificial intelligence to represent linguistic meaning. However, there is continued debate as to how well they encode useful information about the perceptual qualities of concepts. This debate is critical to identifying the scope of embodiment in human semantics. If perceptual object properties can be inferred from word embeddings derived from language alone, this suggests that language provides a useful adjunct to direct perceptual experience for acquiring this kind of conceptual knowledge. Previous research has shown mixed performance when embeddings are used to predict perceptual qualities. Here, we tested if we could improve performance by leveraging the ability of Transformer-based language models to represent word meaning in context. To this end, we conducted two experiments. Our first experiment investigated noun representations. We generated decontextualised (‘charcoal’) and contextualised (‘the brightness of charcoal’) Word2Vec and BERT embeddings for a large set of concepts and compared their ability to predict human ratings of the concepts’ brightness. We repeated this procedure to also probe for the shape of those concepts. In general, we found very good prediction performance for shape, and more modest performance for brightness. The addition of context did not improve perceptual prediction performance. In Experiment 2, we investigated representations of adjective-noun pairs. Perceptual prediction performance was generally found to be good, with the non-additive nature of adjective brightness reflected in the word embeddings. We also found that the addition of context had a limited impact on how well perceptual features could be predicted. We frame these results against current work on the interpretability of language models and debates surrounding embodiment in human conceptual processing.

Effective communication involves a delicate balance between generating novel, engaging content and maintaining a coherent narrative. The neural mechanisms underlying this balance between coherence and creativity in discourse production remain unexplored. The aim of the current study was to investigate the relationship between coherence and creativity in spontaneous speech, with a specific focus on the interaction among three key neural networks: the Default Mode Network, Multiple-Demand Network, and the Semantic Control Network. To this end, we conducted a two-part analysis. At the behavioural level, we analysed speech samples produced in response to topic cues, computing measures of global coherence (indexing the degree of connectedness to the main topic) and Divergent Semantic Integration (DSI; reflecting the diversity of ideas incorporated in the narrative). Coherence and divergence in speech were negatively correlated, suggesting a trade-off between maintaining a coherent narrative structure and incorporating creative elements. At the neural level, higher global coherence was associated with greater activation in the Multiple-Demand Network, emphasising its role in organising and sustaining logical flow in discourse production. In contrast, functional connectivity analyses demonstrated that higher DSI was related to greater coupling between the Default Mode and Multiple-Demand Networks, suggesting that creative speech relies on a dynamic interplay between associative and executive processes. These results provide new insights into the cognitive and neural processes underpinning spontaneous speech production, highlighting the complex interplay between different brain networks in managing competing demands of being coherent and creative.

Producing coherent discourse requires us to regulate the content of our speech and avoid interference from discourse-irrelevant concepts that become active in semantic memory. The inhibitory deficit hypothesis proposes that coherence declines in later life are due to a reduced ability to inhibit these irrelevant ideas. However, existing evidence in support of this view is correlational. We performed an experimental test of the hypothesis by asking young (18-25) and older (70-90) participants to produce discourse on a range of topics while attending to two types of visual distractors: images of meaningful concepts and meaningless abstract patterns. The overall global coherence of responses was lower when participants were distracted (cf. no distraction) but this effect was not larger for meaningful distractors. Participants also spoke more slowly under distraction. These effects did not differ between age groups. Critically, however, in the meaningful distractor condition, responses diverged from the original topic more quickly than in the other conditions. This effect was only present in older participants. These results suggest two underlying effects at play. First, performing a concurrent task has a general effect on the speed and coherence of discourse, which in this study was age-invariant. Second, for older people, tasks that activate a series of irrelevant semantic representations have an additional cumulative effect on discourse content, causing it to deviate off topic more rapidly. Our results support the inhibitory deficit hypothesis and suggest that older people can improve their coherence by avoiding semantically-laden environmental distractors like TV or radio programmes.

Identifying the brain regions that process concrete and abstract concepts is key to understanding the neural architecture of thought, memory and language. We review current theories of concreteness effects and test their neural predictions in a meta-analysis of 72 neuroimaging studies (1400 participants). Our analysis includes more than twice as many studies as previous meta-analyses, allowing for a more sensitive mapping of these effects across the brain. We also conducted a quantitative assessment of the degree to which concreteness effects aligned with a range of large-scale functional brain networks. Our results suggest that concrete and abstract concepts vary both in the information-processing modalities they engage and in the demands they place on cognitive control processes. Abstract concepts preferentially activated networks for social cognition (particularly for sentences), language and semantic control (particularly when presented as single words). Concrete concepts preferentially activated action processing regions when presented in sentences, though we found no evidence that they activated visual networks. Specialisation for both concept types was present in different parts of the default mode network (DMN), with effects dissociating along a social-spatial axis. Concrete concepts generated greater activation in a medial temporal DMN component, implicated in constructing mental models of spatial contexts and scenes. In contrast, abstract concepts showed greater activation in frontotemporal DMN regions involved in social and language processing. These results align with prior claims that generating models of situations and events is a core DMN function and indicate specialisation within DMN for different aspects of these models.

Understanding narratives requires at least transient access to the semantic system, to decode incoming content, and prolonged access to the default mode network to maintain and manipulate the narrative model. Subregions within the integrative semantic hubs in bilateral anterior temporal lobe appear differentially sensitive to the need to rapidly decode external input (exogenous processing) versus reflecting on context stored in the narrative model (endogenous processing). The latter is most consistently reported in the middle temporal gyrus portion of the hub, suggesting that this region serves as a critical hinge point, dynamically interacting with the default mode network to facilitate endogenous processing. The present study investigated this by characterizing the functional connectivity profiles of anterior temporal lobe subregions during movie-viewing and examining content-evoked changes in these profiles. Compared to other anterior temporal lobe subregions, middle temporal gyrus was more functionally connected to the default mode network, and these connections were strengthened during moments with limited incoming information, providing viewers with a chance to reflect on the content. Rather than being functionally distinct networks, the semantic and default mode systems dynamically interact to facilitate reflection or endogenous semantic processing. Future work should further characterize how neural systems dynamically shift from integrated to segregated states in response to everyday processing demands.