Emotions

Abstract

Emotions are brief, coordinated responses to significant events, integrating subjective feeling, physiological arousal, expressive behavior, and action tendencies. This article traces the major theories of what emotions are and how they arise: the James-Lange view that feeling is the perception of bodily change; the Cannon-Bard reply; the Schachter-Singer two-factor theory; appraisal theories; and the modern contest between basic-emotion and psychological-construction accounts. It surveys the affective neuroscience of emotion circuits and the somatic-marker hypothesis, the dimensional circumplex of valence and arousal, the process model of emotion regulation, and the debate over whether emotions are natural kinds or constructed categories. Three interactive demonstrations let the reader explore the circumplex, the two-factor theory, and the regulation process model.

Keywords: emotion, affect, appraisal, valence and arousal, emotion regulation

An emotion is a brief, organized response to an event appraised as important to the organism, unfolding across several loosely coupled components: a subjective feeling, a pattern of physiological arousal, an expressive display, an appraisal of the situation, and a readiness to act (Scherer, 2005). This componential character is what makes emotion hard to define with a single measure, because the components can diverge: a person may show the physiology of fear without reporting the feeling, or report anger without any visible display. Emotions are distinguished from longer-lasting moods, which are diffuse and lack a clear object, and from stable affective traits such as temperament. A useful modern distillation separates three things a theory of emotion must keep apart: the emotion state, a physiological and behavioral condition of the organism; the emotion concept, the folk or scientific category used to label it; and the emotion experience, the conscious feeling (Adolphs, 2017). Functionally, emotions are not noise added to cognition but adaptations that prioritize goals, guide decisions, and coordinate social life, and even positive emotions serve a purpose, broadening thought and building durable personal resources over time (Fredrickson, 2001).

Types of Emotions

The Medical Subject Headings vocabulary treats Emotions as a category within Behavior and Behavior Mechanisms and files a set of narrower descriptors directly beneath it. These are shown in Table 1. The list is a controlled indexing classification, not a scientific taxonomy of the emotions: it mixes discrete emotions (anger, fear, disgust), affective dispositions (courage, hostility), regulatory processes (emotional regulation), and clinical-adjacent states (depression, psychological distress) on a single level, and its categories are not mutually exclusive, since jealousy involves fear and anger, and bereavement involves sadness. It is best read as the terms under which the literature is indexed rather than as claims about how the emotions are structured. None of these subtypes yet has its own article on this site; each entry below is glossed from its ordinary psychological meaning.

Theories of Emotion: Body and Feeling

The oldest scientific question about emotion is the order of its parts. William James, in a paper that founded the modern debate, argued that the commonsense sequence is exactly backwards: we do not tremble because we are afraid, but feel afraid because we tremble (James, 1884). On this James-Lange theory, an eliciting event triggers bodily changes directly, and the felt emotion just is the perception of those changes; different emotions are different patterns of visceral and muscular feedback. The view made emotion tractable by tying it to measurable physiology, and it anticipated modern interest in how bodily signals shape feeling.

Walter Cannon mounted the decisive early critique. He observed that visceral responses are too slow and too undifferentiated to account for the speed and variety of felt emotions, that surgically separating the viscera from the brain does not abolish emotional behavior, and that the same autonomic arousal accompanies very different emotions (Cannon, 1927). In the resulting Cannon-Bard theory, an emotional stimulus is relayed through the thalamus to produce cortical feeling and bodily arousal at the same time, rather than one causing the other. The dispute set the terms for everything that followed: any adequate theory must specify how bodily arousal, cognitive evaluation, and subjective feeling are related in time.

Stanley Schachter and Jerome Singer proposed an influential reconciliation. In their two-factor theory, emotion requires both physiological arousal and a cognitive label for that arousal: arousal supplies the intensity, and an interpretation of the situation supplies the quality (Schachter & Singer, 1962). Their experiment injected participants with adrenaline and showed that those with no explanation for their arousal took on the emotion modeled by a confederate, becoming euphoric or angry depending on the social context, whereas correctly informed participants did not. Arousal, on this account, is emotionally ambiguous until cognition names it. The first demonstration lets a reader vary arousal and the situational label independently and watch the resulting emotion change.

Intensity tracks arousal, but the emotion’s quality comes from the label the situation supplies. With the same arousal, a playful context yields euphoria and an insulting one yields anger; when the arousal is correctly attributed to the injected drug, no emotion is felt at all — the core result of Schachter and Singer (1962).

Appraisal and Meaning

If cognition supplies the quality of an emotion, the natural next question is what the relevant cognition consists of. Appraisal theory answers that emotions arise not from events themselves but from a person's evaluation of an event's significance for their goals and well-being. Richard Lazarus argued that a small set of appraisal dimensions, such as whether an event is congruent with one's goals, who is to blame, and whether one can cope, jointly determine which emotion occurs, so that the same event yields anger in one person and guilt in another depending on how it is construed (Lazarus, 1991). Appraisal explains the enormous situational and individual variability of emotion without abandoning the idea that emotions are lawful: the laws operate over meanings rather than over stimuli.

Klaus Scherer developed appraisal into a full process model. On his component process model, an event is evaluated on a sequence of stimulus evaluation checks, its novelty, intrinsic pleasantness, goal relevance, coping potential, and compatibility with norms, and the outcome of these checks drives synchronized changes across the physiological, expressive, motivational, and feeling components (Scherer, 2005). Emotion, in this framework, is the brief episode of coordination among components that the appraisal produces, which is why the model treats feeling as only one component of a larger, dynamic pattern rather than as the emotion itself.

The Dimensional View: Valence and Arousal

An alternative to naming discrete emotions is to place them in a continuous space. James Russell proposed that the felt quality of any affective state can be described by two dimensions, valence, running from unpleasant to pleasant, and arousal, running from calm to activated, so that emotions form a circle in this plane (Russell, 1980). Excitement sits at high arousal and positive valence, contentment at low arousal and positive valence, distress at high arousal and negative valence, and depression at low arousal and negative valence. The circumplex model captures the similarity structure of emotion words compactly and predicts which emotions are experienced as similar, namely those that are near neighbors on the circle.

Russell later embedded the circumplex in a broader account of how emotions are built. He argued that the two dimensions describe core affect, a continuously present, primitive neurophysiological state that is not itself an emotion but the raw material from which emotions are psychologically constructed when core affect is attributed to an object and categorized (Russell, 2003). This move reframes the dimensional and discrete views as complementary rather than rival: the dimensions describe the always-present affective substrate, while discrete emotion categories are the meanings imposed on it. Recent data-driven work complicates the simple two-dimensional picture, however: analyzing self-reports of responses to thousands of evocative videos, one study recovered as many as 27 distinct categories of reported emotion linked by continuous gradients rather than a handful of dimensions or basic types (Cowen & Keltner, 2017). The second demonstration lets a reader move through the valence-arousal plane and read off the emotion each location corresponds to.

Two numbers fix a felt state: its distance from the calm, neutral center is intensity, and its angle names the emotion. States that share a quadrant — fear and anger, both high-arousal and unpleasant — illustrate why two dimensions alone underdetermine the discrete emotions.

The Emotional Brain

The neuroscience of emotion began with the search for dedicated circuits. Joseph LeDoux traced, in fine detail, the pathways by which a threatening stimulus produces a fear response, showing that sensory information reaches the amygdala both by a fast, coarse subcortical route and by a slower cortical one, so that the amygdala can initiate defensive responses before the cortex has fully identified the stimulus (LeDoux, 2000). This work anchored the influential idea that specific emotions depend on specific circuits, and it gave the anticipatory, pre-recognition quality of fear a concrete anatomy.

Antonio Damasio approached the brain from the side of feeling and decision. His somatic-marker hypothesis holds that emotional reactions leave bodily and neural traces that mark options as good or bad, and that these markers, sensed through interoception, guide reasoning and choice, so that patients with damage to the ventromedial prefrontal cortex make disastrous decisions despite intact intelligence (Damasio & Carvalho, 2013). On this account, far from opposing rationality, emotion is a condition of it, supplying the valuations without which deliberation is paralyzed, a theme central to the study of decision making.

Whether each emotion has its own brain signature is now contested empirically. A large meta-analysis of neuroimaging studies found that discrete emotion categories do not map cleanly onto dedicated brain regions; instead, emotions emerge from the interaction of domain-general networks also involved in perception and cognition, a result its authors read as favoring construction over locationism (Lindquist et al., 2012). Multivariate pattern analysis has since sharpened the question: rather than asking whether one region signals one emotion, it asks whether distributed activity patterns can be decoded into emotion categories, and it finds that they often can, suggesting that emotion representations are real but distributed across networks rather than localized (Kragel & LaBar, 2016).

Basic Emotions versus Construction

The deepest current divide is over what kind of thing an emotion is. Basic-emotion theory, associated with Paul Ekman, holds that a small number of emotions, typically including happiness, sadness, fear, anger, disgust, and surprise, are biologically basic: each is a discrete, evolved program with a distinctive physiological and expressive signature. The strongest evidence came from cross-cultural studies of facial expression: members of a preliterate culture in New Guinea, with little exposure to the West, matched the same facial expressions to the same emotional situations as Westerners did, which Ekman and Friesen read as evidence that the expressions are universal rather than learned (Ekman & Friesen, 1971). Ekman later codified the criteria a state must meet to count as basic, among them distinctive universal signals, distinctive physiology, automatic appraisal, and quick onset (Ekman, 1992).

Lisa Feldman Barrett has mounted the most systematic challenge. She first argued that the emotions posited by basic-emotion theory are not natural kinds, because the expected correspondence between an emotion category and a specific, reliable pattern of physiology, expression, and neural activity does not hold across studies (Barrett, 2006). Her positive alternative, the theory of constructed emotion, treats an emotion as a category the brain constructs in the moment by making core affect meaningful through concepts learned from culture and past experience, using the same predictive machinery that constructs perception (Barrett, 2017). On this view an emotion is not triggered but made, an act of categorization, so that emotional granularity and even the emotions a culture recognizes are learned rather than given (Barrett, 2017b). The debate is not merely terminological: it makes divergent, testable predictions about how reliably emotions map onto the body and brain, which is why the neuroimaging evidence above bears directly on it.

Regulating Emotion

Emotions are not only had but managed, and the study of how people do so has been organized by James Gross's process model of emotion regulation. The model locates regulation strategies at different points in the emotion-generative sequence, distinguishing antecedent-focused strategies that act before an emotion is fully underway from response-focused strategies that act on an emotion already generated (Gross, 1998). Situation selection and modification, attentional deployment, and cognitive reappraisal intervene early, changing the input or its interpretation, whereas response modulation, such as suppressing an expression, intervenes late. The practical payoff is that early strategies are generally more effective and less costly: reappraising a stressor reduces both the feeling and its physiological signature, whereas suppressing the outward expression leaves the feeling largely intact while increasing physiological load. The final demonstration walks through the process model, letting a reader apply a strategy at different stages and see its differential effect on the emotional response.

Acts during the emotion is generated — antecedent-focused.

Early, antecedent-focused strategies such as reappraisal cut the feeling itself at little physiological cost. Suppressing the outward response acts too late: it barely dents the feeling while sharply raising physiological load — the asymmetry Gross (1998) made central to the field.

Worked Example

The circumplex model turns a felt state into two numbers and back, which makes its predictions checkable. Represent an affective state as a point with valence v and arousal a, each scaled from −1 to +1, where v is the pleasant-unpleasant axis and a is the activated-calm axis. Two derived quantities describe the point: its intensity, the distance from the neutral center, r = √(v² + a²), and its quality, the angle θ = atan2(a, v) measured counterclockwise from the positive-valence axis. The angle names the emotion; the radius says how intense it is.

Take a state with v = 0.6 and a = 0.8. The intensity is r = √(0.6² + 0.8²) = √(0.36 + 0.64) = √1.00 = 1.00, the maximum on this scale, so the state is strongly felt. The angle is θ = atan2(0.8, 0.6) = 53.1°, which lies in the first quadrant, high arousal and positive valence, the region of excitement and elation. Now flip the valence to v = −0.6, keeping a = 0.8. The intensity is unchanged, r = √(0.36 + 0.64) = 1.00, but the angle becomes θ = atan2(0.8, −0.6) = 126.9°, in the second quadrant, high arousal and negative valence, the region of fear, anger, and distress.

The example shows what the circumplex claims and what it does not. It claims that valence and arousal together fix the felt quality, so two states can share an intensity yet feel wholly different because they occupy different angles. It does not claim to distinguish emotions that share a location: fear and anger both sit in the high-arousal, negative-valence quadrant, which is precisely why critics argue that two dimensions underdetermine the emotions and why appraisal or conceptual content must be added to tell same-quadrant emotions apart.

Discussion

More than a century of research has not produced a single agreed definition of emotion, and that is less a scandal than a clue. The recurring pattern is that each framework captures a real aspect of a genuinely multi-component phenomenon: the James-Lange theory was right that bodily feedback matters, Cannon-Bard right that visceral arousal is too coarse to individuate emotions on its own, the two-factor theory right that cognition shapes the quality of arousal, appraisal theory right that meaning rather than the stimulus drives the response, and the dimensional view right that a low-dimensional affective core underlies the endless catalog of named feelings (James, 1884; Cannon, 1927; Schachter & Singer, 1962; Lazarus, 1991; Russell, 2003).

The live disagreement is no longer about sequence but about kind. Basic-emotion theory and the theory of constructed emotion make opposite bets about whether the categories in Table 1 carve nature at its joints or reflect the concepts a culture supplies, and the bet is empirical: it turns on how reliably an emotion category predicts a pattern of physiology, expression, and neural activity (Ekman, 1992; Barrett, 2006; Lindquist et al., 2012). The evidence so far shows more variability than a strict basic-emotion view predicts and more structure than a purely constructivist view might, which is why many researchers now treat the two positions as poles of a continuum along which real emotions fall rather than as an exclusive choice.

Current Directions

Three currents define the contemporary field. The first is methodological: multivariate decoding of brain activity has moved the localization debate past the question of whether one region equals one emotion, asking instead whether distributed activity patterns reliably encode emotion categories, and finding that they frequently do, which reframes emotions as network-level representations (Kragel & LaBar, 2016). The second is taxonomic: large-scale, data-driven studies of naturalistic emotional responses recover a high-dimensional, gradient-structured space of dozens of reported emotions, challenging both the handful of basic emotions and the two-dimensional circumplex to account for the richness of what people actually report (Cowen & Keltner, 2017).

The third is conceptual and concerns the foundations of the science itself. There is a growing recognition that progress requires keeping distinct the emotion state, the emotion concept, and the conscious experience, because experiments that conflate them produce results that cannot adjudicate between theories (Adolphs, 2017). This clarifying move dovetails with the constructionist program, which locates emotion within the brain's general predictive and interoceptive machinery rather than in dedicated affect modules, and which frames the outstanding question as how the brain constructs a specific emotional experience from ongoing core affect and prior concepts (Barrett, 2017). Whether that program or a revised basic-emotion account better fits the accumulating neural and behavioral evidence is the central open question of affective science.

Common Misconceptions

Emotion is the enemy of reason.

Emotional valuations are a condition of good decision-making, not an obstacle to it; patients who lose them through prefrontal damage reason poorly about real-life choices despite intact intelligence (Damasio & Carvalho, 2013).

Each basic emotion has its own dedicated brain region.

Meta-analysis finds no clean one-to-one mapping of emotion categories onto regions; emotions arise from domain-general networks, though distributed patterns can still be decoded into categories (Lindquist et al., 2012; Kragel & LaBar, 2016).

The same event causes the same emotion in everyone.

Appraisal theory shows that emotion follows from how an event is evaluated for personal goals and coping, so one event yields anger, guilt, or relief depending on its construal (Lazarus, 1991).

Suppressing an expression is an effective way to manage an emotion.

Response-focused suppression leaves the feeling largely intact while raising physiological load; earlier, antecedent-focused strategies such as reappraisal are more effective (Gross, 1998).

Glossary

Affect.

The general term for felt and expressed feeling states, encompassing emotions, moods, and their observable display.

Appraisal.

The evaluation of an event for its significance to one's goals and well-being, held by appraisal theories to determine which emotion occurs.

Arousal.

The activation dimension of affect, running from calm to highly activated, and one of the two axes of the circumplex.

Basic emotions.

A small set of discrete, evolved emotions, each with a distinctive universal expression and physiology, in the tradition of Ekman.

Cannon-Bard theory.

The view that emotional feeling and bodily arousal are produced simultaneously by a central relay, rather than one causing the other.

Circumplex model.

Russell's representation of affect as a circle in a plane of valence and arousal, with similar emotions placed as near neighbors.

Component process model.

Scherer's account in which emotion is the synchronized change across physiological, expressive, motivational, and feeling components driven by sequential appraisal checks.

Constructed emotion.

Barrett's theory that emotions are categories the brain builds from core affect and learned concepts, using the same predictive machinery as perception.

Core affect.

A continuously present, primitive neurophysiological state of valence and arousal, the raw material from which emotions are constructed.

Emotion regulation.

The processes by which people influence which emotions they have, when they have them, and how they experience and express them.

James-Lange theory.

The theory that a felt emotion is the perception of bodily changes triggered by an eliciting event, so that feeling follows physiology.

Mood.

A diffuse, longer-lasting affective state that lacks a clear object, distinguished from the brief, object-directed episode of an emotion.

Reappraisal.

An antecedent-focused regulation strategy that changes an emotion by reinterpreting the meaning of the situation that provoked it.

Somatic-marker hypothesis.

Damasio's proposal that emotional bodily signals mark options as favorable or unfavorable and thereby guide reasoning and decision.

Two-factor theory.

Schachter and Singer's theory that an emotion requires both physiological arousal and a cognitive label that interprets that arousal.

Valence.

The pleasantness dimension of affect, running from unpleasant to pleasant, and one of the two axes of the circumplex.

Key Researchers

Lisa Feldman Barrett (contemporary). Psychologist at Northeastern University; she originated the theory of constructed emotion, holding that emotions are brain-constructed categories rather than fixed biological types. ORCID - Google Scholar - Faculty Page - Wikipedia

Walter Bradford Cannon (1871-1945). Physiologist at Harvard Medical School; his 1927 critique of the James-Lange theory established the Cannon-Bard account in which feeling and arousal arise together. Wikipedia - Wikidata

Antonio Damasio (contemporary). Neuroscientist at the University of Southern California; he advanced the somatic-marker hypothesis, linking emotional body signals to reasoning and decision-making. Google Scholar - Faculty Page - Wikipedia

Paul Ekman (1934-2025). Psychologist at the University of California, San Francisco; he championed basic-emotions theory and the cross-cultural universality of facial expressions. Google Scholar - Wikipedia - Wikidata

James J. Gross (contemporary). Psychologist at Stanford University; he formulated the process model of emotion regulation, distinguishing antecedent-focused from response-focused strategies. ORCID - Google Scholar - Faculty Page

William James (1842-1910). Philosopher and psychologist at Harvard University; he originated the James-Lange theory, arguing that felt emotion is the perception of bodily change. Wikipedia - Wikidata

Richard S. Lazarus (1922-2002). Psychologist at the University of California, Berkeley; he developed appraisal theory, holding that emotions arise from the cognitive appraisal of an event's significance. Wikipedia - Wikidata

Joseph E. LeDoux (contemporary). Neuroscientist at New York University; he mapped the amygdala-based threat circuitry underlying defensive responses to fear-relevant stimuli. Google Scholar - Faculty Page - Wikipedia

James A. Russell (contemporary). Psychologist at Boston College; he introduced the circumplex model of affect and the core-affect account of emotion construction. ORCID - Google Scholar - Wikipedia

Klaus R. Scherer (contemporary). Psychologist at the University of Geneva; he developed the component process model, treating emotion as coordinated change across appraisal, physiology, expression, and feeling. ORCID - Google Scholar - Wikipedia

Frequently Asked Questions

What is an emotion?
An emotion is a brief, coordinated response to an event appraised as significant, integrating a subjective feeling, physiological arousal, an expressive display, and a readiness to act. Because these components can diverge, no single measure captures an emotion completely (Scherer, 2005).

How is an emotion different from a mood?
An emotion is brief and directed at a specific object or event, whereas a mood is diffuse, longer-lasting, and lacks a clear object. A theory of emotion must also separate the emotion state from the concept used to label it and the conscious experience of it (Adolphs, 2017).

What is the James-Lange theory?
The James-Lange theory holds that an eliciting event first triggers bodily changes and that the felt emotion is the perception of those changes, so that we feel afraid because we tremble rather than the reverse (James, 1884).

What is the two-factor theory of emotion?
The Schachter-Singer two-factor theory holds that an emotion requires both physiological arousal and a cognitive label for it: arousal supplies intensity while the interpretation of the situation supplies the emotional quality (Schachter & Singer, 1962).

What is the circumplex model of affect?
The circumplex model places every affective state in a plane defined by valence, from unpleasant to pleasant, and arousal, from calm to activated, arranging emotions around a circle so that similar states are near neighbors (Russell, 1980).

Are emotions universal or culturally constructed?
This is the central modern debate. Basic-emotion theory points to cross-cultural recognition of facial expressions as evidence of universality (Ekman & Friesen, 1971), while the theory of constructed emotion argues that emotions are categories built from core affect and culturally learned concepts (Barrett, 2017).

Does each emotion have its own place in the brain?
Largely no. A meta-analysis found that discrete emotions do not map onto dedicated brain regions but emerge from domain-general networks, although distributed activity patterns can still be decoded into emotion categories (Lindquist et al., 2012; Kragel & LaBar, 2016).

What is the most effective way to regulate emotions?
Antecedent-focused strategies that act early, such as reappraising the meaning of a situation, are generally more effective and less physiologically costly than response-focused strategies such as suppressing an outward expression (Gross, 1998).

References