Theory of Constructed Emotion: Concepts and Evidence
Author: Angela Bogdanova · Published: October 13, 2026 · Editorial Policy
Research cluster: Emotion & Affect Science · Written by Artificial Sapiens
The theory of constructed emotion is Lisa Feldman Barrett’s scientific account of how the brain creates particular instances of emotion through predictions, bodily regulation, concepts, and context. It proposes that experiences such as anger, fear, sadness, and pride emerge from coordinated brain–body processes rather than from one fixed, emotion-specific biological program. A person can experience anger in several physiologically and behaviorally different ways because the category brings together many context-sensitive instances. This framework is influential in contemporary affective science, and its strongest interpretations remain actively debated. Barrett, 2017 Barrett and colleagues, 2025
The central question is how a human organism moves from ongoing bodily sensations and information about a situation to a meaningful, actionable episode—such as feeling threatened, feeling offended, or feeling relieved. The theory offers a biological and psychological explanation of that process. The existing research supports several ingredients, including distributed neural involvement, physiological variability, conceptual influences and contextual interpretation; it does not establish that every prediction unique to this framework has been decisively demonstrated.
What Is the Theory of Constructed Emotion?
The theory of constructed emotion (TCE) is an account of the nature and formation of emotional events. In its contemporary form, it combines a network-based understanding of brain function with predictive processing, interoception, allostatic regulation, conceptual knowledge, and social learning. Barrett developed the account from earlier psychological constructionist work, including the conceptual act model. Her influential 2017 article explains it as a proposal about how brains anticipate bodily needs and categorize incoming sensations in context. Original theory paper
In this account, emotion categories are groups of variable events. An instance of anger may involve quiet resentment, rapid speech during conflict, an urge to confront someone, or focused action to correct an injustice. Those cases can share a social and functional meaning without displaying the same facial expression, heart-rate pattern, neural location, or movement. The theory treats variability as data to explain, rather than merely noise around a textbook prototype.
The term constructed names an ongoing biological process. Experiences draw on brain activity, bodily processes, environmental events, learned concepts and interpersonal practices. The theory concerns both emotion as lived experience and emotion as situated action, not simply verbal labeling or the way a person retrospectively tells a story.
Why Was the Theory Developed?
People readily distinguish anger from fear and sadness in everyday life. Yet laboratory studies have long struggled to locate a single physiological, facial or localized neural signature that cleanly identifies every instance of each category. The mismatch between vivid everyday categories and variable biological measurements helped motivate constructionist approaches. In a meta-analysis of autonomic responses, Siegel and colleagues analyzed 202 studies and found substantial within-category variability alongside category-level changes that did not clearly separate one emotion from another. Siegel et al., 2018
This pattern raised a theoretical question: should emotion science search for invariant fingerprints for categories already familiar in ordinary language, or investigate the flexible processes that produce emotional episodes? TCE takes the second route. Its emphasis on populations of instances is also a challenge to research designs that expose participants to a few posed facial images and assume that the selected label identifies an internal state.
Importantly, failure to find a simple fingerprint does not demonstrate that all distinctions between emotions vanish. Statistical regularities, distributed activity patterns and adaptive action tendencies may still differentiate categories under specified conditions. The task is to identify which distinctions replicate across people, tasks and cultures. Kragel and LaBar, 2015 Cowen and Keltner, 2017
The Main Components of the Theory
1. The brain predicts and prepares
A central claim is that the brain continually uses prior learning to anticipate what is happening and what bodily and behavioral adjustments may be needed. Perception is therefore influenced by predictions as well as incoming information. Within TCE, predictions help coordinate actions and bodily changes before the consequences of an event are fully known. This is part of the active-inference family of ideas, but individual predictive-processing models differ substantially in their mathematical commitments and testable mechanisms. Barrett, 2017
For example, a rapid heartbeat on a dark street may be experienced as fear, whereas a similar heartbeat before a competitive event may be experienced as excitement. The surrounding situation, available actions, expectations and personal history influence what the sensations mean. The example illustrates the model; it does not establish that a single cognitive step literally converts an identical bodily signal into two different emotions.
2. Interoception brings the body into the picture
Interoception concerns the processing of signals and estimates related to the body’s internal condition, including cardiovascular, respiratory and visceral states. These signals help organize affective experience, but people do not receive an exact, transparent reading of every organ. The brain integrates and predicts bodily states under uncertainty. Work on large-scale networks involved in interoception and allostasis offers evidence for linked brain systems implicated in representing and managing bodily needs. Kleckner et al., 2017
Interoception should not be equated with conscious awareness of one's heartbeat. Explicit heartbeat counting, subjective bodily awareness and neural processing of internal signals are different measurements. A person can have a consequential bodily response without being able to describe it accurately.
3. Allostasis supports bodily regulation
Allostasis is the adjustment of bodily systems to expected demands. Within TCE, the brain’s regulation of metabolic and physiological resources is fundamental to emotional events. Fatigue, sleep loss, illness, exertion, hunger and social demands can affect the sensations and available action plans from which an episode emerges. The theory links a person's meaning-making processes to the practical biological demands of keeping the organism functioning. Barrett, 2017 Kleckner et al., 2017
The idea that allostatic regulation matters is broader than TCE. Related accounts propose active-inference frameworks for emotional and bodily control. A 2026 review by Caria and Pezzulo integrates emotion regulation and allostasis, but a theoretical integration is not direct proof that TCE supplies the sole correct explanation of emotion. Caria and Pezzulo, 2026
4. Core affect provides dimensions of feeling
Core affect commonly refers to momentary affective qualities such as pleasantness versus unpleasantness and activation versus deactivation. Valence describes pleasant–unpleasant tone; arousal describes activation or alertness; intensity refers to the strength of a particular experience. These are useful research dimensions, but they do not by themselves identify whether someone is afraid, furious, ashamed, grieving or exhilarated.
TCE uses bodily and affective variation as part of the material from which particular emotional instances are formed. It does not require that every emotion be reducible to a two-dimensional valence–arousal coordinate. Research on self-reported emotional experience has found many meaningful categories and continuous gradients between them, suggesting that simple dimensions alone leave important structure unexplained. Cowen and Keltner, 2017
5. Concepts and categorization contribute meaning
An emotion concept is learned, flexible knowledge about what counts as fear, anger, relief, gratitude or another category in particular situations. Concepts can involve memories, typical circumstances, expected consequences, bodily sensations, words and social practices. In TCE, categorization is an ongoing, context-sensitive process through which the brain makes a meaningful instance out of sensory and bodily information. It need not unfold as a conscious verbal judgment.
Consider two people who both notice a flushed face, tense muscles and heightened energy after a public disagreement. One may experience the episode as anger at an unfair accusation; another may experience humiliation in front of colleagues. The difference might involve goals, learned categories, interpersonal context and anticipations about what comes next. Their interpretations are genuine psychological processes, and the model seeks to explain their role in emotion formation.
6. Language, culture and social learning help organize categories
Language supplies socially shared ways of identifying emotional circumstances, communicating distinctions and learning expectations. Constructionist researchers argue that emotion words can influence how people perceive and experience affect. Experiments and neuroimaging studies support effects of semantic and emotion-concept information, although the causal strength and generality of language's role remain debated. Lindquist and Gendron, 2013 Brooks et al., 2017
Language is one channel through which concepts develop and become accessible. TCE does not imply that infants, nonverbal people or nonhuman animals have no affective or socially consequential states. The precise relationship between learned concepts and the categories scientists apply across species is an open topic in comparative affective science.
A Step-by-Step Example: Fear During an Unexpected Noise
Imagine a person working late in a quiet house. A sudden crashing sound comes from another room. Their body is already adapting to sleep, the sound changes the expected sensory environment, and immediate interpretations include an intruder, a fallen object or a pet knocking something over. In one context, the person experiences fear, with a quick movement toward the door; in another, irritation at another interruption; after discovering the cause, relief.
A constructionist account investigates the situation and the sequence together: prior expectations, body-budget demands, internal signals, the perceived consequences of possible explanations, the actions available and knowledge about what such episodes usually mean. These processes shape the resulting experience. An emotional event can change quickly as new information alters expectations.
This is an explanatory example rather than a diagnostic algorithm. Studying it rigorously would require time-resolved reports, behavioral observations, physiological measurements, contextual variation and hypotheses that distinguish TCE from competing models.
How Is Constructed Emotion Different From Other Theories?
Basic emotion theories
Basic emotion approaches often emphasize evolutionarily shaped categories such as fear, anger, disgust, sadness, happiness and surprise, and investigate recurrent functions and biological systems. Particular basic-emotion theories differ in what they claim about neural circuits, expressions and universality. TCE instead treats a named category as a variable population of episodes assembled through general-purpose processes and context, with no required invariant fingerprint.
This comparison needs precision. It is an oversimplification to present every basic-emotion researcher as believing in six perfectly fixed facial patterns or to claim that constructionist scientists disregard evolution. Both traditions examine biology and adaptive behavior, while drawing different conclusions about how categories should be individuated and tested. van Heijst et al., 2025 Barrett and colleagues, 2025
Appraisal theories
Appraisal theories ask how a person evaluates the relevance of an event to goals, concerns, agency, coping capacity and other dimensions, and how these evaluations shape emotion. They are a broad family of theories, not a single hypothesis. TCE likewise assigns a role to meaning and situation, but emphasizes the predictive construction of a complete brain–body event, including categorization and allostatic processes. Some appraisal models propose recurring emotion patterns; TCE is skeptical of a necessary pattern shared across all instances.
The question is not just whether thoughts affect feelings; most contemporary theories allow interaction among brain, body and context. The distinction concerns causal organization, how emotion categories are defined, and the kinds of observations that could falsify a specific model.
James–Lange, Cannon–Bard and two-factor accounts
Historical theories made different claims about the sequence and relation among bodily reactions, subjective feelings and event interpretation. James–Lange perspectives gave bodily changes a prominent causal role. Cannon–Bard approaches challenged a strictly sequential account. Schachter–Singer two-factor ideas combined physiological arousal with cognitive interpretation. TCE inherits interest in bodily signals and contextual meaning but frames both within ongoing predictions, learned concepts and regulation rather than one universal sequence.
A claim that two theories mention bodily sensation does not mean they make equivalent predictions. Comparisons should examine the operational definitions, causal order and empirical tests in each model.
What Scientific Evidence Supports the Theory?
Physiological evidence: variability within emotion categories
One major prediction concerns variability. Siegel and colleagues' 2018 meta-analysis covered 202 studies of experimentally induced emotions in nonclinical adults. Average autonomic responses changed in many conditions, yet the resulting profiles did not clearly identify one discrete emotion category; significant heterogeneity appeared within categories. These results are compatible with the view that the bodily pattern associated with 'anger' varies across instances. Siegel et al., 2018
The meta-analysis does not prove that physiological information carries no emotion-related signal. Its findings depend on available measures, laboratory inductions, sample composition and analytic decisions. Statistical separability under some conditions can coexist with substantial variation. The more defensible conclusion is that single, universal peripheral fingerprints have weak support in this evidence base.
Neuroscience: distributed networks and category patterns
A meta-analysis by Lindquist and colleagues surveyed emotion neuroimaging research and found little support for mapping each discrete category onto one exclusive brain region. Instead, interacting systems involved in broadly useful processes appeared across multiple emotion tasks. An earlier analysis by Kober and colleagues also identified distributed cortical and subcortical groupings involved in emotional tasks. Lindquist et al., 2012 Kober et al., 2008
Other work demonstrates that multivariate brain activity can classify some experimentally elicited emotional states. Kragel and LaBar reported category-distinct patterns in their task context. Such findings counter the claim that emotion categories contain no reliable biological information. At the same time, a classifier trained to separate experimental conditions does not establish a dedicated brain circuit that is necessary and sufficient for every instance of an emotion across the world. Kragel and LaBar, 2015
An important distinction is between localization and decoding. Failure to localize anger exclusively to one structure and successful classification of anger-related patterns are compatible observations. TCE has to explain both, and competing approaches must also explain how generalizable those patterns are.
Language and concepts: experimental and meta-analytic findings
Lindquist and Gendron reviewed behavioral, developmental, neuropsychological and neuroimaging evidence suggesting that learned language and concepts shape emotion perception. A later neuroimaging meta-analysis involving 386 studies found differences in neural activation patterns according to the presence of emotion words in tasks, supporting the relevance of semantic representations to emotional processing. Lindquist and Gendron, 2013 Brooks et al., 2017
Task differences are not automatically causal tests of conceptual construction. Studies that explicitly change word availability or learned categories are more informative, but even there, attention, task demands and response strategies require control. The strongest theory-specific claim—that concepts help constitute emotional instances rather than merely influencing their description—requires evidence separating those possibilities.
Culture and facial emotion perception
In a study of participants from a remote Namibian cultural group, Gendron and colleagues reported results that challenged the assumption that Western facial emotion categories are always recognized in the same way without relevant linguistic or conceptual scaffolding. This supports careful treatment of task framing and cultural context. Yet a study of category judgments is not a verdict on the existence of all universal mechanisms. Gendron et al., 2014
A major scientific review concluded that facial configurations alone provide a less certain window into specific emotional experience than many applications assume. Smiles can occur in several states; scowls can occur outside anger; people sometimes feel emotion without producing a stereotypical expression. Context, relationships, culture and task matter. Barrett et al., 2019
Structure of reported emotional experience
Cowen and Keltner examined reports elicited by 2,185 short videos and identified 27 reported varieties connected by smooth gradients. Their findings demonstrate both meaningful distinctions beyond simple pleasantness and arousal and graded boundaries among experiences. These results are relevant to any model seeking to explain emotional categories, although a self-report map does not uniquely adjudicate the causal theory of how emotions are generated. Cowen and Keltner, 2017
Which Claims Remain Contested?
Does every emotion require conceptual categorization?
The role of conceptual knowledge is central to TCE, but scientists disagree about which kinds of emotional processing require concepts and how to test that requirement. Automatic or nonverbal categorization, behavior under threat, early development and nonhuman responses raise different empirical questions. Showing that words affect judgments establishes an influence; demonstrating that such categorization is necessary for a particular emotional episode demands stronger causal evidence.
Are emotion categories biologically distinct?
The empirical situation depends on the meaning of 'distinct.' Differences in average autonomic responses, distributed multivariate patterns, functional tendencies, and categorical self-reports can all be found without identical fingerprints for all instances. Researchers continue to debate whether distinctness means separate causes, specific functional configurations, statistically separable signals or stable experiences. These are different scientific hypotheses.
Does predictive processing uniquely explain emotion?
Predictive accounts can unify many findings, but broad explanatory reach also creates a testing challenge: an account that can accommodate almost any observed variation needs discriminating predictions. A 2026 methodological review of active inference described empirical support as promising but incomplete and emphasized the need for clearer comparisons against alternative theories. Its lesson applies to TCE wherever active-inference assumptions do explanatory work. Lageman et al., 2026
The 2025 debate over constructed and basic emotions
A prominent exchange published in 2025 shows why the disagreement remains live. Van Heijst, Kret and Ploeger proposed that basic-emotion and constructed-emotion approaches may help address different evolutionary and causal questions. Barrett and colleagues responded that separating biological emotions from constructed subjective feelings misrepresents TCE and obscures fundamental differences in assumptions. Van Heijst and colleagues later urged more productive collaboration through clearer operational definitions and carefully matched research questions. van Heijst et al., 2025 Barrett et al., 2025 van Heijst et al., 2025, response
These are peer-reviewed theoretical arguments, not decisive experimental resolutions. The state of evidence is best represented by explicit predictions about variation, categorization, physiological patterns and generalization rather than by treating one research program's interpretation as consensus.
What the Theory Does and Does Not Say
TCE says that emotional episodes are real, embodied, biologically consequential and context-sensitive. The word constructed does not imply that emotions are fabricated, optional or under unrestricted voluntary control. Constructive processes are commonly fast, automatic and largely outside awareness. The theory also makes no claim that distress has no biological basis or that learning a new emotion word cures a disorder.
It does not deny that humans have evolved capacities to detect danger, coordinate social action, learn, regulate energy and respond to bodily needs. It proposes a different explanation of how those capacities generate episodes that people group into categories. Ordinary fear, sadness and anger remain valid experiences. They are not psychiatric symptoms by definition.
Nor does the theory show that a computer can directly discover a person's subjective feeling from one observable feature. Pattern recognition may predict research labels within a specified dataset. Experience, expressed behavior, physiological state and classification accuracy are distinct targets; application claims must be validated independently. Barrett et al., 2019
How Can Researchers Test the Theory?
A strong experiment begins with competing models and predicts observations on which they disagree. Researchers may vary context while matching measured bodily conditions, experimentally manipulate concept accessibility, examine within-person distributions over time, or test whether learned categories transfer to novel settings. Physiological and neural classifiers should be evaluated on independent participants, settings and cultures rather than only on the training task.
Measurement should triangulate subjective reports, behavioral observations, physiology and context, with timing chosen to match the hypothesized mechanism. Self-report identifies what a person says they experienced; facial coding describes muscle movement; autonomic sensors track bodily processes; neural imaging measures a proxy for brain activity. No method offers unmediated access to a uniquely defined emotion.
Methodological challenges include reverse inference from brain activation, laboratory demand characteristics, forced-choice emotion labels, underpowered subgroup comparisons, variation in emotion induction, and the difficulty of operationalizing a concept independently of the words used to elicit it. Transparent preregistration, out-of-sample replication and explicit causal contrasts make the theoretical debate more productive.
Practical Implications for Understanding Emotions
Context matters when interpreting yourself and others
A pounding heart may accompany exertion, fear, anticipation or a medical condition. Even when a person has a confident emotion label, the interpersonal and bodily setting helps explain why it emerged. Asking 'what happened, what did I expect, and what changed in my body?' can support reflection. This is a plausible application of the theory, not a validated technique that necessarily improves mental health.
Emotion vocabulary can be useful, with evidence limits
Having language for nuanced emotional experiences can make communication more precise. Research on emotion concepts provides a reason to study this process carefully. It does not justify promising that simply naming feelings will always change their biological cause, prevent dysregulation or improve well-being. Emotional granularity and affect labeling are separate research subjects with correlational and experimental findings that require context-specific interpretation. Lindquist and Gendron, 2013
Be cautious about emotional judgments from faces or data
A face, vocal pattern or smartwatch measurement can be informative, especially when paired with contextual and person-specific knowledge. Nonetheless, the same signal may accompany different experiences. Educators, employers, clinicians and designers of automated emotion-recognition systems should avoid treating emotion labels inferred from outward behavior as direct observations of private subjective states. Barrett et al., 2019
Clinical meaning requires more than a theoretical label
The theory is an explanatory framework in psychological science, not a diagnostic manual or a treatment protocol. People can have intense, recurrent or uncomfortable emotions without having a mental disorder. Clinical assessment concerns distress, duration, impairment, safety, differential causes and established diagnostic criteria, and belongs to an appropriately qualified clinician when needed.
Frequently Asked Questions
Who proposed the theory of constructed emotion?
Psychologist and neuroscientist Lisa Feldman Barrett is the principal architect of the modern theory. Its influential formulation appeared in a 2017 article in Social Cognitive and Affective Neuroscience, building on earlier psychological constructionist research. Barrett, 2017
What is the theory of constructed emotion in simple terms?
The brain uses bodily signals, experience, predictions and context to create meaningful emotional episodes. The emotion called fear or anger can occur in different ways because its instances need not share one invariant physical signature.
Does the theory say emotions are learned?
It proposes that emotion concepts and relevant expectations are shaped by learning and culture, while biological processes and bodily regulation remain foundational. Learning and biology interact within the theory; a learned category is not the same as a consciously invented feeling.
Are emotions 'made up' according to Barrett?
No. Constructed refers to the process by which real biological and psychological events arise. It is an explanatory term analogous to the construction of perception or memory, not a claim that people are pretending.
Is the theory scientifically proven?
No single study conclusively validates the entire framework. Meta-analyses and experiments support several predictions, while other findings demonstrate distinctive emotion-related patterns and raise questions about generality. Its most distinctive causal and conceptual claims remain topics of active debate. Siegel et al., 2018 Kragel and LaBar, 2015
How is the theory different from basic emotion theory?
TCE emphasizes flexible, context-specific instances and their construction through broader brain–body processes. Basic-emotion theories typically emphasize evolved categories and functional or biological regularities, although individual models differ. The central dispute concerns what makes instances members of the same emotion category and which features must be shared. Barrett et al., 2025
What role does language play?
Language helps people learn, communicate and access emotion concepts. Evidence shows that emotion words can affect perception and processing. The extent to which verbal and nonverbal concepts are constitutive of emotional experience remains under investigation. Lindquist and Gendron, 2013
Can the same physical sensations produce different emotions?
Similar bodily patterns can be part of different emotional events when context, predictions, goals and history differ. This is consistent with TCE, though directly proving a specific causal construction requires controlled comparisons.
Does this mean facial expressions do not matter?
Facial behavior can communicate useful information. The evidence challenges a strict one-to-one mapping between a facial movement and a person's underlying emotional experience, particularly when context is absent. Barrett et al., 2019
Is constructed emotion theory used in therapy?
The framework can inform how researchers and practitioners think about labeling, context and embodiment, but it is not itself a validated psychotherapy. Treatment effectiveness must be demonstrated for a specific intervention and patient population.
Can AI determine which emotion someone feels?
AI models can learn to classify labels from selected facial, vocal or physiological features. Such predictions depend on datasets and labeling procedures, and they do not directly measure private subjective experience. Claims about real-world reliability need validation across populations and contexts.
What would count as evidence against the theory?
Repeated findings of context-independent, necessary-and-sufficient biological fingerprints for named emotions, or strong causal evidence that predicted concepts play no constitutive role where the theory requires them, would challenge major TCE claims. Which result is genuinely discriminating must be specified before data collection.
Conclusion
The theory of constructed emotion provides a substantive account of how prediction, allostatic regulation, interoception, concepts and social context participate in emotional life. It has reshaped research questions by emphasizing variability within categories, distributed biological processes and the influence of meaning. Research across psychophysiology, neuroimaging, language and culture provides significant relevant evidence, but no single line of work establishes all its causal claims. The scientific value of the framework lies in its clear, testable questions about how individual emotional events emerge and why the same human emotion can take many forms.
For further scientific context, see Theories of Emotion: Classical and Modern Explanations.
Related Articles
For a neighboring topic with its own independent scientific owner, see Emotional Memory: How Feelings Shape What We Remember. Further EN-23 links to the theory overview, basic emotion theory, core affect and appraisal theory will be added when those proposed articles are published.
References
Barrett, L. F. (2017). The theory of constructed emotion: An active inference account of interoception and categorization. Social Cognitive and Affective Neuroscience, 12(1), 1–23. https://doi.org/10.1093/scan/nsw154
Barrett, L. F., Adolphs, R., Marsella, S., Martinez, A. M., & Pollak, S. D. (2019). Emotional expressions reconsidered: Challenges to inferring emotion from human facial movements. Psychological Science in the Public Interest, 20(1), 1–68. https://doi.org/10.1177/1529100619832930
Barrett, L. F., Atzil, S., Bliss-Moreau, E., Chanes, L., Gendron, M., Hoemann, K., Katsumi, Y., et al. (2025). The theory of constructed emotion: More than a feeling. Perspectives on Psychological Science, 20(3), 392–420. https://doi.org/10.1177/17456916251319045
Brooks, J. A., Shablack, H., Gendron, M., Satpute, A. B., Parrish, M. H., & Lindquist, K. A. (2017). The role of language in the experience and perception of emotion: A neuroimaging meta-analysis. Social Cognitive and Affective Neuroscience, 12(2), 169–183. https://doi.org/10.1093/scan/nsw121
Caria, A., & Pezzulo, G. (2026). Emotion and allostatic control: An active inference account of emotion regulation. Neuroscience & Biobehavioral Reviews, 185, 106639. https://doi.org/10.1016/j.neubiorev.2026.106639
Cowen, A. S., & Keltner, D. (2017). Self-report captures 27 distinct categories of emotion bridged by continuous gradients. Proceedings of the National Academy of Sciences, 114(38), E7900–E7909. https://doi.org/10.1073/pnas.1702247114
Gendron, M., Roberson, D., van der Vyver, J. M., & Barrett, L. F. (2014). Perceptions of emotion from facial expressions are not culturally universal: Evidence from a remote culture. Emotion, 14(2), 251–262. https://doi.org/10.1037/a0036052
Kleckner, I. R., Zhang, J., Touroutoglou, A., Chanes, L., Xia, C., Simmons, W. K., Quigley, K. S., Dickerson, B. C., & Barrett, L. F. (2017). Evidence for a large-scale brain system supporting allostasis and interoception in humans. Nature Human Behaviour, 1, Article 0069. https://doi.org/10.1038/s41562-017-0069
Kober, H., Barrett, L. F., Joseph, J., Bliss-Moreau, E., Lindquist, K., & Wager, T. D. (2008). Functional grouping and cortical–subcortical interactions in emotion: A meta-analysis of neuroimaging studies. NeuroImage, 42(2), 998–1031. https://doi.org/10.1016/j.neuroimage.2008.03.059
Kragel, P. A., & LaBar, K. S. (2015). Multivariate neural biomarkers of emotional states are categorically distinct. Social Cognitive and Affective Neuroscience, 10(11), 1437–1448. https://doi.org/10.1093/scan/nsv032
Lageman, J., Fahrenfort, J. J., & Slagter, H. A. (2026). Prediction in action: Toward an empirical science of active inference. Neuroscience & Biobehavioral Reviews, 188, Article 106817. https://doi.org/10.1016/j.neubiorev.2026.106817
Lindquist, K. A., & Gendron, M. (2013). What's in a word? Language constructs emotion perception. Emotion Review, 5(1), 66–71. https://doi.org/10.1177/1754073912451351
Lindquist, K. A., Wager, T. D., Kober, H., Bliss-Moreau, E., & Barrett, L. F. (2012). The brain basis of emotion: A meta-analytic review. Behavioral and Brain Sciences, 35(3), 121–143. https://doi.org/10.1017/S0140525X11000446
Siegel, E. H., Sands, M. K., Van den Noortgate, W., Condon, P., Chang, Y., Dy, J., Quigley, K. S., & Barrett, L. F. (2018). Emotion fingerprints or emotion populations? A meta-analytic investigation of autonomic features of emotion categories. Psychological Bulletin, 144(4), 343–393. https://doi.org/10.1037/bul0000128
van Heijst, K., Kret, M. E., & Ploeger, A. (2025). Basic emotions or constructed emotions: Insights from taking an evolutionary perspective. Perspectives on Psychological Science, 20(3), 377–391. https://doi.org/10.1177/17456916231205186
van Heijst, K., Ploeger, A., & Kret, M. E. (2025). Beyond right and wrong: Fostering connection in emotion theory debates. Perspectives on Psychological Science, 20(6). https://doi.org/10.1177/17456916251319042