Here’s a small experiment. Pick a feeling you’ve had recently — something more complicated than plain happy or sad. Something that had edges to it, that sat on you for a few hours. Now try to describe it in three words. Then try again in five. Then try again in fifteen.
If you’re like most people, the task gets quickly harder as the required precision increases. The three-word version — sad, frustrated, tired — comes easily. The fifteen-word version — a kind of resigned disappointment, tangled up with some guilt about feeling disappointed, and an edge of restlessness that I recognise from other moments when I should have spoken up and didn’t — takes real effort. And yet it describes what you felt more accurately than the shorter version did.
This small difference — between being able to name feelings with broad categories versus fine detail — has turned out to be one of the more consequential findings in modern emotion research. People who can describe their feelings precisely have, on average, better mental health, more resilient relationships, better responses to stress, and greater ability to regulate their own emotional states. The capacity is called emotional granularity, and it’s been central to the work of an unusual researcher whose broader theory of emotions has upended much of what psychology once believed about what feelings even are.
The researcher who changed the question
The psychologist Lisa Feldman Barrett, at Northeastern University in Boston, has spent several decades challenging what had been, for most of the twentieth century, the dominant scientific theory of emotions. That theory — associated most strongly with the American psychologist Paul Ekman — held that emotions were largely universal, pre-wired, biologically given. Humans in all cultures, Ekman argued from his famous cross-cultural research, recognised the same basic emotions from facial expressions: happiness, sadness, anger, fear, surprise, disgust. These were, in his framework, the fundamental emotions, common to our species, each with its own distinctive physiological signature and expressive pattern.
The theory had enormous influence. It shaped clinical psychology, corporate training programmes, animation and film design, airport security protocols, and popular understanding of feelings for decades. Ekman’s work on microexpressions — the tiny involuntary facial movements supposedly revealing true emotions — became the basis for lie-detection training programmes in law enforcement and, eventually, a television series.
Barrett’s argument, developed across several decades and summarised in her 2017 book How Emotions Are Made, is that almost every piece of the classical theory is wrong or, at minimum, seriously misleading. Emotions, she argues, are not pre-wired universal modules. They are constructed — by your brain, in the moment — from three inputs: raw physiological data from your body, predictions your brain makes about what’s likely happening, and the emotional vocabulary your culture has given you to interpret what you’re feeling.
This is a radical claim. It implies that emotions are, in a deep sense, made rather than discovered. What you feel when your heart races and your mouth goes dry isn’t predetermined to be fear or anger or excitement — it depends, partly, on what your brain’s interpretive framework predicts the feeling should be. A professional basketball player before a big game and a person in a dark alley at night may have nearly identical physiological states, but they experience them as wildly different emotions because their brains are constructing the interpretations from different contexts.
The research behind the theory
Barrett’s theoretical framework is grounded in a large body of empirical research, much of it her own. Among the most striking findings:
Attempts to find specific physiological signatures for each “basic emotion” have largely failed. Despite thousands of studies across decades, researchers haven’t been able to reliably distinguish anger from fear, or disgust from sadness, by physiological measurement alone. The patterns of heart rate, skin conductance, muscle tension and hormonal release that the classical theory predicted should distinguish the basic emotions don’t hold up in careful studies. People in the same emotional state have wildly varying physiological patterns, and people in different emotional states often have almost identical ones.
Cross-cultural evidence for the universal recognition of emotions has turned out to be weaker than Ekman’s original studies suggested. Some recent work, including studies Barrett has conducted in cultures relatively isolated from Western media, has found that the supposed universal facial-expression recognitions don’t hold up consistently. People in some cultures don’t interpret the “universal” anger face as anger, or the “universal” fear face as fear. The universality of the original findings appears to have been substantially driven by cultural contamination — the global spread of Western media and emotion-labelling that effectively trained people worldwide to recognise the Western categories.
And brain-imaging research has failed to find the distinct neural signatures for basic emotions that the classical theory predicted. Different emotions don’t reliably activate different specific brain regions. The same regions are activated across many different emotional experiences, and the neural activity during emotion is better predicted by the specific content of the emotional experience (what you’re experiencing it about) than by the emotional category it would be sorted into.
Barrett’s framework, called the theory of constructed emotion, accommodates all of this. If emotions are constructed in the moment from raw physiological data plus cultural interpretation, then we shouldn’t expect clean universal categories, clean physiological signatures, or clean brain activation patterns. We should expect exactly what we find: a messy, culturally-variable, context-dependent mapping between bodily states and the labels we apply to them.
Emotional granularity — and why it matters
The most practically consequential part of Barrett’s research, for ordinary life, is the work on what she calls emotional granularity — the degree to which a person can distinguish between related emotional states.
Low-granularity people use broad categories. They’re happy or sad or angry or fine. Their emotional vocabulary, when they introspect, is coarse-grained. Whatever they’re feeling gets sorted into one of a small number of bins.
High-granularity people distinguish more finely. They can tell the difference between melancholic and disappointed, between frustrated and resentful, between content and peaceful, between anxious and fearful and uneasy. Their emotional vocabulary is richer, and they use it to track more specific internal states.
What Barrett and colleagues have found, across many studies, is that high-granularity people have meaningfully better mental health outcomes. They’re less likely to develop depression and anxiety disorders. They recover faster from negative experiences. They respond more effectively to stress. They have better relationships, in part because they can communicate their internal states more precisely to others. They’re less likely to use alcohol or drugs to self-medicate, presumably because they’re able to identify what they’re actually feeling and respond to it directly rather than trying to dampen an undifferentiated unpleasantness.
The mechanism, according to Barrett’s framework, has to do with the predictive work the brain is doing. When your brain can accurately identify the specific state you’re in — I’m feeling disappointed because a specific expectation wasn’t met, and the disappointment will pass once I’ve processed that — it can respond to the state appropriately. When your brain can only identify a vague unpleasant state, it doesn’t know what to do with it, and the state lingers or gets acted out in unhelpful ways.
This has a direct practical implication: increasing your emotional vocabulary is, in a measurable sense, good for your mental health. This isn’t vague self-help. It’s one of the better-supported findings in emotion research. Reading novels that describe feelings in fine detail. Learning the vocabulary of mood. Paying attention to the distinctions between closely-related emotional states. Practising naming what you’re feeling with specificity rather than generality. All of these, across studies, correlate with measurable improvements in emotional regulation.
The counter-thread worth hearing
The theory of constructed emotion has been enormously influential, and it’s also not the only serious scientific view. Several prominent researchers continue to defend versions of the classical theory, and it would be misleading to present Barrett’s framework as if it had simply won the debate.
The American psychologist Dacher Keltner at Berkeley, who studied with Ekman, has developed a research programme that argues for a larger set of emotional categories than Ekman’s original six but maintains that emotions have meaningful biological and universal components. Keltner’s research on embarrassment, pride, sympathy, awe and several other states has produced cross-cultural findings that are harder to explain by cultural construction alone.
The Estonian-American neuroscientist Jaak Panksepp, who died in 2017, spent his career documenting what he called the basic emotional operating systems of the mammalian brain — specific neural circuits (SEEKING, RAGE, FEAR, LUST, CARE, PANIC/GRIEF, PLAY) that he argued underlay emotional experience across species. His work on affective neuroscience has produced findings that are hard to reconcile with a purely constructionist view.
So the honest picture is that Barrett’s theory is a significant and well-supported scientific position, but it hasn’t replaced the classical theory in any final way. The field is actively contested. What both sides agree on — and what matters most for ordinary life — is that emotional experience is more complex than folk intuition suggests, that the specific words we use for emotions do real work in shaping our experience, and that developing a richer vocabulary for feelings has measurable practical benefits regardless of which underlying theory turns out to be more correct.
What to do with this
For ordinary life, the most useful takeaway from this research is probably this: invest in your emotional vocabulary. Not for performative reasons. Not to seem more sophisticated. Because it actually helps.
A few specific practices.
When you notice you’re feeling something, try to name it with more precision than bad or off or weird. Are you sad? Anxious? Frustrated? Restless? Resentful? Disappointed? Ashamed? Bored? Each of these is a different state, and each calls for a different response. Forcing yourself to distinguish them — even slowly, even imperfectly — is the practice that builds granularity over time.
Read literature that describes emotions with precision. Good fiction often distinguishes between emotional states with a subtlety that ordinary vocabulary lacks. A book that describes the specific difference between nostalgia and wistfulness, or between grief and mourning and melancholy, gives you language that’s directly applicable to your own internal states. People who read a great deal tend, on average, to have higher emotional granularity, and this is probably one of the mechanisms by which reading supports mental health.
Pay attention to the vocabulary of feelings in cultures and languages different from your own. The Portuguese word saudade names a specific kind of longing for something that may never come back. The Japanese word amae describes a specific emotional state of dependence and acceptance. The Finnish word sisu names a particular kind of determined perseverance. Each of these is a feeling humans everywhere can have, but some cultures have built precise labels for them while others haven’t. Borrowing vocabulary from other traditions expands the range of states you can identify in yourself.
Don’t assume your current emotional vocabulary is adequate just because it’s familiar. The words you currently have for feelings are the ones you happened to learn in your early family life. They’re unlikely to be maximally precise. If you’ve felt, for years, that something is off without being able to say what, the problem may be partly that you don’t yet have the vocabulary for what’s off. Acquiring it is work, but it’s among the most useful work you can do for your own wellbeing.
The question that remains
The deepest thing Barrett’s research suggests, leaving aside the theoretical debates, is that what you feel depends — more than you probably realise — on how you’ve learned to parse what you feel. Give yourself different categories, and you’ll have different experiences. This is not relativism. The bodily states are real. But the emotional experiences you construct out of those states are, to a surprising degree, your own work, shaped by the language and categories your culture has given you and the ones you’ve developed for yourself.
This is a kind of freedom, if you take it seriously. You are not entirely at the mercy of your feelings, because your feelings are partly your own construction. What you build them into is, in a meaningful sense, up to you.
The question worth carrying, especially the next time you find yourself feeling something hard that you can’t quite name:
What are you actually feeling, right now, if you had to describe it with enough precision that someone else could feel it too — and what does having to search for the words tell you about the vocabulary you’ve been using?
Key research referenced: Lisa Feldman Barrett, How Emotions Are Made (2017) and the theory of constructed emotion; Paul Ekman’s foundational research on basic emotions and facial expressions; Dacher Keltner’s work on expanded emotional categories; Jaak Panksepp’s affective neuroscience research on basic emotional circuits; Barrett and colleagues’ research on emotional granularity and mental health outcomes.