To master a culture’s categorical systems, growing minds must learn both to identify abstract similarities that bind two individual instances into a common category (e.g., recognizing what makes Scooby and Snoopy both “dogs”) and to differentiate instances that belong to different categories (e.g., calling Garfield a “cat” rather than a “dog”). Research has charted how children learn to “lump” and “split” categories in the cognitive literature, but we know much less about how emotion categories develop. Here, we administered laboratory tasks assessing affective abstraction (i.e., the ability to match two images that induce similar affective experiences) and emotion differentiation (i.e., how specifically one categorizes one’s emotions as distinct) to a cross-sectional sample of participants in the USA aged 10-25 (50.62% women). Affective abstraction scores were largely uncorrelated with age, but the ability to match cross-category stimuli of the same valence (e.g., a negative-affect-inducing house and a negative-affect-inducing animal) increased with age. Interestingly, affective abstraction and emotion differentiation scores were positively associated, especially in youth participants. Thus, abstraction and differentiation are not merely opposites of each other (i.e., it’s not the case that people either tend to chronically “lump” emotion categories or chronically “split” emotion categories). Instead, there may be shared cognitive processes that allow individuals to construct emotion categories that are both nuanced and general enough to conform to societal norms, and this covariance is most pronounced in youth. These results shed new light on the development of emotion categorization.

The hypothalamus plays an important role in the regulation of the bodys metabolic state and behaviors related to survival. Despite its importance however, many questions exist regarding the intrinsic and extrinsic connections of the hypothalamus in humans, especially its relationship with the cortex. As a heterogeneous structure, it is possible that the hypothalamus is composed of different subregions, which have their own distinct relationships with the cortex. Previous work on functional connectivity in the human hypothalamus have either treated it as a unitary structure or relied on methodological approaches that are limited in modeling its intrinsic functional architecture. Here, we used resting state data from ultrahigh field 7 Tesla fMRI and a data driven analytical approach to identify functional subregions of the human hypothalamus. Our approach identified four functional hypothalamic subregions based on intrinsic functional connectivity, which in turn showed distinct patterns of functional connectivity with cortex. Overall, all hypothalamic subregions showed stronger connectivity with a cortical network, Cortical Network 1 composed primarily of frontal, midline, and limbic cortical areas and weaker connectivity with a second cortical network composed largely of posterior sensorimotor regions, Cortical Network 2. Of the hypothalamic subregions, the anterior hypothalamus showed the strongest connection to Cortical Network 1, while a more ventral subregion containing the anterior hypothalamus extending to the tuberal region showed the weakest connectivity. The findings support the use of ultrahigh field, high resolution imaging in providing a more incisive investigation of the human hypothalamus that respects its complex internal structure and extrinsic functional architecture.

This paper uses a generative neural network architecture combining unsupervised (generative) and supervised (discriminative) models with a model comparison strategy to evaluate assumptions about the mappings between brain states and behavior. Most modeling in cognitive neuroscience publications assume a one-to-one brain-behavior relationship that is linear, but never test these assumptions or the consequences of violating them. We systematically varied these assumptions using simulations of four ground-truth brain-behavior mappings that involve progressively more complex relationships, ranging from one-to-one linear mappings to many-to-one nonlinear mappings. We then applied our Variational AutoEncoder-Classifier framework to the simulations to show how it accurately captured diverse brain-behavior mappings,provided evidence regarding which assumptions are supported by the data, and illustrated the problems that arise when assumptions are violated. This integrated approach offers a reliable foundation for cognitive neuroscience to effectively model complex neural and behavioral processes, allowing more justified conclusions about the nature of brain-behavior mappings.

The brain continuously anticipates the energetic needs of the body and prepares to meet those needs before they arise, called allostasis. In support of allostasis, the brain continually models the sensory state of the body, called interoception. We replicated and extended a large-scale system supporting allostasis and interoception in the human brain using ultra-high precision 7 Tesla functional magnetic resonance imaging (fMRI) (N = 90), improving the precision of subgenual and pregenual anterior cingulate topography combined with extensive brainstem nuclei mapping. We observed over 90% of the anatomical connections published in tract-tracing studies in non-human animals. The system also included regions of dense intrinsic connectivity broadly throughout the system, some of which were identified previously as part of the backbone of neural communication across the brain. These results strengthen previous evidence for a whole-brain system supporting the modeling and regulation of the internal milieu of the body.

Research in affective science includes over one hundred thousand articles, the vast majority of which have been published in only the past two decades. The size and rapid growth of this field have led to unique challenges for the twenty-first-century scientist including how to develop both breadth and depth of scholarship, curb siloing and promote integrative and interdisciplinary framework, and represent and monitor the field in its entirety. Here, we help address these issues by compactly mapping out this enormous field using citation network analysis (CNA). We generated a citation matrix of over 100,000 publications and over 1 million citations since the seminal works on emotion by Charles Darwin (1872) and William James (1884). Using graph theory metric and content analysis of titles and abstracts, we identified and characterized the contents of 69 research communities, their most influential articles, and their interconnectedness with each other. We further identified potential “missed connections” between communities that share similar content but do not have strong citation-based connections. In doing so, we establish the first, low-dimensional representation, or field-wide map, of a substantial portion of the affective sciences literature. This panoramic view of the field provides affective and non-affective scientists alike with the means to rapidly survey dozens of major research communities and topics in the field, guide scholarship development, and identify gaps and connections for developing an integrative science.

When investigating the brain, bodily, or behavioral correlates of emotional experience, researchers often present participants with stimuli that are assumed to reliably and exclusively evoke an instance of one, and only one, emotion category across participants (e.g., a fear stimulus, a joy stimulus, and so on). These assumptions are driven by a typological view. Here, we tested the extent to which they are met. Across three studies (total N = 453), participants reported their experiences as they viewed silent video clips or static images that were curated from published studies and from online search engines. Two different response formats were used. Overall, the proportion of stimulus-evoked emotion experiences that met even lenient benchmarks for validity and reliability for labeling a stimulus as pertaining to a single emotion category label was exceedingly low. Furthermore, participants frequently used more than one label for a given instance. The findings suggest that typological assumptions, and the nomothetic approach they align with, rely on assumptions that are rarely, if ever, met in stimulus-evoked paradigms. Correspondingly, the use of group-averaged normative ratings masks tremendous variation that is potentially meaningful. An overreliance on these norms may lead to conclusions that emotions are organized as discrete categories, yet these theory-laden conclusions may have limited generalizability regarding the emotional experiences of individual people during these tasks. Rather, emotional experiences evoked by visual stimuli are multifaceted (i.e., involve multiple labels per instance) and vary tremendously across individuals. Future work may benefit from multifaceted measurement of emotion and idiographic, data-driven modeling approaches. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

While much research in social cognitive neuroscience has focused on which brain regions are engaged when processing social content, it remains unclear what these areas are doing in terms of underlying mechanisms. Here, we approached this question using predictive processing theory, which suggests that the brain instantiates a generative model of its sensory environment. Using a novel animated shapes fMRI task, we observed a functional double dissociation between brain regions that were engaged when forming a prediction for agentic movement - which involved the premotor cortex and the lateral parietal cortex, previously implicated in action observation and mirroring - from those associated with processing feedback for updating abstract priors to guide predictions - which involved the dorsomedial and ventrolateral prefrontal cortex, the temporoparietal area, and the lateral temporal cortex, previously associated with mentalizing/theory of mind. We observed parallel functional dissociations in the cerebellar areas affiliated with these networks. These findings suggest new insights into how brain regions associated with action observation/mirroring and mentalizing/theory of mind play complementary roles in supporting facets of a predictive processing architecture underlying social cognition. ### Competing Interest Statement The authors have declared no competing interest. NSF, 2241938

Affective abstraction refers to how people conceptualize affective states in terms of category-level representations that generalize across specific situations (e.g., "fear" as evoked by heights, predators, and haunted houses). Here, we develop a novel task for assessing affective abstraction and test its relations with trait alexithymia, depression, and autism spectrum quotient. In a preregistered online study, participants completed a set of tasks in which they matched a cue image with one of two probe images based on similarity of affective experience. In a discrete emotion version of the task, the cue and target probe matched on a discrete emotion category while controlling for valence. In a valence version of the task, the cue and target probe matched on valence (i.e., pleasantness or unpleasantness). We further varied the degree of abstraction such that some judgments crossed semantic categories (e.g., a house cue with animal probes). Accuracy, as indexed by the proportion of choices that accorded with norms, predicted trait measures of alexithymia, depression, and autism quotient with medium effect sizes. We conducted an integrative data analysis by including data from three other (nonpreregistered) samples (N = 435) and found substantial moderation by sampling population (Amazon Mechanical Turk, college students) and partial moderation by gender identity. Additional constraints on generalization include that our sample included predominantly White American adults between the ages of 23 and 64. These results provide preliminary support for the notion that affective abstraction may reflect a transdiagnostic psychological process of broad relevance to individual differences in affective processing. (PsycInfo Database Record (c) 2025 APA, all rights reserved).

A recently published article by van Heijst et al. attempted to reconcile two research approaches in the science of emotion-basic emotion theory and the theory of constructed emotion-by suggesting that the former explains emotions as bioregulatory states of the body whereas the latter explains feelings that arise from those state changes. This bifurcation of emotion into objective physical states and subjective feelings involves three misleading simplifications that fundamentally misrepresent the theory of constructed emotion and prevent progress in the science of emotion. In this article we identify these misleading simplifications and the resulting factual errors, empirical oversights, and evolutionary oversimplifications. We then discuss why such errors will continue to arise until scientists realize that the two theories are intrinsically irreconcilable. They rest on incommensurate assumptions and require different methods of evaluation. Only by directly considering these differences will these research silos in the science of emotion finally dissolve, speeding the accumulation of trustworthy scientific knowledge about emotion that is usable in the real world.

Is there a universal mapping of physiology to emotion, or do these mappings vary substantially by person or situation? Psychologists, philosophers, and neuroscientists have debated this question for decades. Most previous studies have focused on differentiating emotions on the basis of accompanying autonomic responses using analytical approaches that often assume within-category homogeneity. In the present study, we took an alternative approach to this question. We determined the extent to which the relationship between subjective experience and autonomic reactivity generalizes across, or depends upon, the individual and situation for instances of a single emotion category, specifically, fear. Electrodermal activity and cardiac activity-two autonomic measures that are often assumed to show robust relationships with instances of fear-were recorded while participants reported fear experience in response to dozens of fear-evoking videos related to three distinct situations: spiders, heights, and social encounters. We formally translated assumptions from diverse theoretical models into a common framework for model comparison analyses. Results exceedingly favored a model that assumed situation-dependency in the relationship between fear experience and autonomic reactivity, with subject variance also significant but constrained by situation. Models that assumed generalization across situations and/or individuals performed much worse by comparison. These results call into question the assumption of generalizability of autonomic-subjective mappings across instances of fear, as required in translational research from nonhuman animals to humans, and advance a situated approach to understanding the autonomic correlates of fear experience. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

People place value on emotion categories that inform which emotions to cultivate and which to regulate in life. Here, we examined how people's beliefs about emotion categories varied along three valence-related dimensions: evaluation (good, bad), hedonic feeling (pleasure, displeasure), and desirability (want to feel, do not want to feel). In Studies 1A and 1B, we found that evaluative (good/bad) and hedonic (pleasant/unpleasant) ratings were distinct for certain emotions including lust, anger, shame, fear, and guilt. In Study 2, we found that emotion valuation depended on cultural background in a sample of Asian Americans and Caucasian Americans. Overall, Asian American participants evaluated certain emotions (including, but not limited to, anger, sadness, guilt, and shame) more positively than Caucasian American participants, and this difference was more pronounced on the evaluative rating dimension. Finally, in Study 3, we examined how evaluative and hedonic dimensions further relate with the desire to experience certain emotions and the emotions that people believe they feel in everyday life. Our findings support a model in which evaluative and hedonic dimensions of emotion valuation predict desired emotional states, which in turn predicts beliefs about the reported frequency of emotions experienced in everyday life.

Despite decades of preclinical investigation, there remains limited understanding of the etiology and biological underpinnings of anxiety disorders. Sensitivity to potential threat is characteristic of anxiety-like behavior in humans and rodents, but traditional rodent behavioral tasks aimed to assess threat responsiveness lack translational value, especially with regard to emotionally valenced stimuli. Therefore, development of novel preclinical approaches to serve as analogues to patient assessments is needed. In humans, the fearful face task is widely used to test responsiveness to socially communicated threat signals. In rats, ultrasonic vocalizations (USVs) are analogous social cues associated with positive or negative affective states that can elicit behavioral changes in the receiver. It is therefore likely that when rats hear aversive alarm call USVs (22 kHz), they evoke translatable changes in brain activity comparable with the fearful face task. We used functional magnetic resonance imaging in male and female rats to assess changes in BOLD activity induced by exposure to aversive 22 kHz alarm calls emitted in response to threatening stimuli, prosocial (55 kHz) USVs emitted in response to appetitive stimuli, or a computer-generated 22 kHz tone. Results show patterns of regional activation that are specific to each USV stimulus. Notably, limbic regions clinically relevant to psychiatric disorders (e.g., amygdala, bed nucleus of the stria terminalis) are preferentially activated by either aversive 22 kHz or appetitive 55 kHz USVs. These results support the use of USV playback as a promising translational tool to investigate affective processing under conditions of distal threat in preclinical rat models.

The periaqueductal gray (PAG) is a small midbrain structure that surrounds the cerebral aqueduct, regulates brain–body communication, and is often studied for its role in “fight-or-flight” and “freezing” responses to threat. We used ultra-high field 7-Tesla fMRI to resolve the PAG in humans and distinguish it from the cerebral aqueduct, examining its in vivo function in humans during a working memory task (N = 87). Relative to baseline fixation, both mild and moderate task-elicited cognitive demands elicited bilateral BOLD increases in ventrolateral PAG (vlPAG), a region previously observed to show increased activity during anticipated painful threat in both non-human and human animals. The present task posed only the most minimal (if any) “threat”. The mild-demand condition involved a task easier than remembering a phone number, elicited a heart rate decrease relative to baseline, yet nonetheless elicited a bilateral vlPAG response. Across PAG voxels, BOLD signal intensity correlated with changes in physiological reactivity (relative to baseline) and showed some evidence of spatial organization along the rostral–caudal axis. These findings suggest that the PAG may have a broader role in coordinating brain—body communication during a minimally to moderately demanding task, even in the absence of threat.

Although emotion words such as “anger”, “disgust”, “happiness”, or “pride” are often thought of as mere labels, increasing evidence points to language as being important for emotion perception and experience. Emotion words may be particularly important for facilitating access to the emotion concepts. Indeed, deficits in semantic processing or impaired access to emotion words interfere with emotion perception. Yet, it is unclear what these behavioral findings mean for affective neuroscience. Thus, we examined the brain areas that support processing of emotion words using representational similarity analysis of fMRI data (N = 25). In the task, participants saw 10 emotion words (e.g., “anger”, “happiness”) while in the scanner. Participants rated each word based on its valence on a continuous scale ranging from 0 (Pleasant/Good) to 1 (Unpleasant/Bad) scale to ensure they were processing the words. Our results revealed that a diverse range of brain areas including prefrontal, midline cortical, and sensorimotor regions contained information about emotion words. Notably, our results overlapped with many regions implicated in decoding emotion experience by prior studies. Our results raise questions about what processes are being supported by these regions during emotion experience.

The extent to which neural representations of fear experience depend on or generalize across the situational context has remained unclear. We systematically manipulated variation within and across three distinct fear-evocative situations including fear of heights, spiders, and social threats. Participants (n = 21; 10 females and 11 males) viewed ∼20 s clips depicting spiders, heights, or social encounters and rated fear after each video. Searchlight multivoxel pattern analysis was used to identify whether and which brain regions carry information that predicts fear experience and the degree to which the fear-predictive neural codes in these areas depend on or generalize across the situations. The overwhelming majority of brain regions carrying information about fear did so in a situation-dependent manner. These findings suggest that local neural representations of fear experience are unlikely to involve a singular pattern but rather a collection of multiple heterogeneous brain states.

Thirty years of neuroimaging reveal the set of brain regions consistently associated with pleasant and unpleasant affect in humans-or the neural reference space for valence. Yet some of humans' most potent affective states occur in the context of other humans. Prior work has yet to differentiate how the neural reference space for valence varies as a product of the sociality of affective stimuli. To address this question, we meta-analyzed across 614 social and non-social affective neuroimaging contrasts, summarizing the brain regions that are consistently activated for social and non-social affective information. We demonstrate that across the literature, social and non-social affective stimuli yield overlapping activations within regions associated with visceromotor control, including the amygdala, hypothalamus, anterior cingulate cortex and insula. However, we find that social processing differs from non-social affective processing in that it involves additional cortical activations in the medial prefrontal and posterior cingulum that have been associated with mentalizing and prediction. A Bayesian classifier was able to differentiate unpleasant from pleasant affect, but not social from non-social affective states. Moreover, it was not able to classify unpleasantness from pleasantness at the highest levels of sociality. These findings suggest that highly social scenarios may be equally salient to humans, regardless of their valence.