Core topic 04 of 13
What an emotion is, and why psychologists still argue about it
An emotion is a coordinated response involving physical arousal, attention and interpretation. What researchers disagree about is whether a small set of emotions comes built in with fixed facial signatures, or whether the brain constructs each one on the spot from raw bodily sensation plus learned concepts.
This argument matters more than it sounds. If anger has a fixed facial signature, you can build software to detect it in airports. If anger is constructed differently in different people and cultures, that software is measuring nothing. Several governments have bought the first version.
The two accounts
Paul Ekman's work from the 1960s onward found that people in isolated communities matched photographs of facial expressions to emotion words at above chance rates, which supported a set of universal basic emotions. Lisa Feldman Barrett's theory of constructed emotion argues that what is universal is unpleasant or pleasant arousal, and that everything on top of it is category learning. Her 2019 review with four co-authors examined over a thousand studies and concluded that people scowl when angry about 30 percent of the time, and scowl for plenty of other reasons.
Nobody has won. The practical takeaway is that reading emotion from a face alone is unreliable, which also applies to the claims on body language.
Findings you can rely on
Affect labeling reduces intensity Replicated
Saying what you feel, in words, lowers the intensity of the feeling. Matthew Lieberman's group found lower amygdala activity when people labelled emotional faces, and behavioural studies in exposure therapy have found that people who verbalised their fear during the session showed better outcomes a week later than people who used distraction.
Reappraisal beats suppression Replicated
James Gross has spent thirty years comparing emotion regulation strategies. Changing how you interpret a situation lowers both the felt emotion and the physiological response. Pushing the expression down while feeling it just as strongly raises blood pressure, hurts memory for the conversation, and makes the other person more stressed too.
Emotional granularity predicts outcomes Replicated
People who use emotion words precisely, distinguishing anxious from restless from dreading, regulate better than people whose entire vocabulary is good and bad. In longitudinal studies this predicts lower alcohol use in response to stress and fewer aggressive responses to provocation.
The facial feedback hypothesis Contested
Fritz Strack's 1988 pen-in-the-mouth study found people rated cartoons as funnier when their face was forced into a smile shape. Seventeen labs tried to repeat it in 2016 and failed. A multi-lab study in 2022 with over 3,000 participants found a small effect for deliberately posed smiles. So the claim survives in weakened form.
Crying releases toxins Debunked
William Frey proposed in the 1980s that emotional tears carry stress hormones out of the body. The chemical differences reported were tiny and the research was never replicated at scale. Crying does often precede feeling better, but the evidence points to social support arriving after the crying rather than to any chemical flush.
Where the body comes in
Sensitivity to reward is also what separates people on the extraversion dimension, which is why the same brain systems appear on the introversion and extraversion page. People high in extraversion show stronger positive emotion responses to the same pleasant event.
Interoception, the sense of your own internal state, sets a floor under emotional experience. People who are better at detecting their own heartbeat report more intense emotions. Hunger, poor sleep and illness all shift the raw arousal signal that gets interpreted as an emotion, which is the least mystical explanation available for why everything feels worse at 2am. The sleep and dreams covers what one bad night does to emotional reactivity.
Watch this on the channelEmotion research explained in short videos, including what emotion recognition software gets wrong.
Open Psychology Cue on YouTubeQuestions people ask about this topic
Are there six basic emotions?
It was the dominant view for decades, based on Ekman's cross-cultural facial studies. A 2019 review led by Lisa Feldman Barrett found facial movements vary too much across people and situations to map reliably onto specific emotions.
Why does naming a feeling make it weaker?
Affect labeling reduces amygdala response and self-reported intensity. The effect is modest and the mechanism is debated, but it has held up across several studies and shows up in exposure therapy outcomes.
What is emotional granularity?
The ability to draw fine distinctions between feelings, such as telling irritation apart from resentment. Higher granularity predicts better regulation and lower problem drinking under stress.
Does smiling make you happier?
Slightly, in some conditions. The original pen-in-mouth studies failed a 17-lab replication in 2016, and a larger 2022 study found a small effect for deliberately posed smiles.