The body's part

A pencil between the teeth

Charles Darwin noticed in 1872 that a feeling and its outward sign seem to feed each other. “The free expression by outward signs of an emotion intensifies it,” he wrote in The Expression of the Emotions in Man and Animals, and holding the signs back softens it. Twelve years later the American psychologist William James turned the idea on its head: we do not cry because we are sad, he argued, we feel sorry because we cry, afraid because we tremble. If that is even partly true, then the body is not only a messenger for the feelings. It is part of how they are made, and you might be able to change a feeling by changing the body.

In 1988 Fritz Strack, Leonard Martin, and Sabine Stepper at the universities of Mannheim and Illinois designed a clever test. They told 92 undergraduates that they were studying how people who had lost the use of their hands do tasks with their mouths. Each student held a felt-tipped marker in the teeth, in the lips only, or in the non-dominant hand while rating four Far Side cartoons from 0 to 9 for funniness. Holding a pen in the teeth, though they did not know it, forces the face into something like a smile. Holding it in the lips forces a pout. The students with the “smile” rated the cartoons as funnier, 5.14 to 4.32, a difference of 0.82 points. It became one of the best-known results in psychology.

Then it was tested again. In 2016 a registered replication report, in which 17 laboratories in eight countries all followed the same preregistered plan, tested 1,894 people. The difference between smile and pout was 0.03 points, with a margin of error that ran from −0.11 to +0.16. None of the 17 labs came near the original 0.82. Strack replied that the new protocol put a video camera on each participant, which makes people self-conscious, and that most of the labs drew on psychology-class students who might have known the hypothesis, and in one reanalysis the few labs with other kinds of participants did find a small effect. A 2018 study found the effect only when no camera was filming, but the larger project described below tested the pen again and found nothing. The original pen-in-the-teeth effect, as stated, did not replicate.

The idea itself is more resilient. In 2022 the Many Smiles Collaboration, a preregistered “adversarial collaboration” in which Strack and other supporters of the idea helped design the tests, ran three different smile tasks in 26 research groups in 19 countries. Of 3,878 people tested, 1,504 remained after exclusions (nearly half guessed the purpose, and others failed attention checks). Deliberately making a happy face, by copying actors’ faces or pulling the lips back toward the ears, raised reported happiness by about 0.3 points on a scale, about the same as looking at a mildly pleasant picture. The pen-in-mouth task, the original, showed essentially nothing (0.04). The fair summary: a deliberate smile can lift mood a little. The famous cartoon effect does not exist. “Smile your way to happiness” is too much to claim.

Botulinum toxin, sold as Botox, offers an indirect test. It blocks the release of the chemical acetylcholine at the nerve endings that tell a muscle to contract, so the muscle goes weak for a few months. Injected between the eyebrows, it makes frowning hard. Small trials have asked whether it relieves depression. In 2012 a study of 30 patients found that one treatment cut depression scores by 47 percent over six weeks, against 9 percent after salt water. A study with 74 patients analyzed in 2014 found that 52 percent responded, against 15 percent. Meta-analyses of the few trials report large effects. But the trials are small, most of the patients were women, many patients guessed which treatment they got, and in most studies, whether the wrinkles faded did not track whether the mood improved, which is not what a pure “face-to-mood” explanation predicts. A study of healthy women in 2010 found something different from cheering up: Botox in the frown lines dulled their reaction to mildly pleasant video clips. The honest summary: an early, intriguing result, not a treatment, and the mechanism is unknown.

Venting feeds the fire

A popular idea says that anger is a pressure, and the safe thing is to let it out: hit a pillow, pound a punching bag, throw darts at a photograph of your boss. The idea has an ancient name, “catharsis,” borrowed from Aristotle’s account of tragedy, and it is still sold today, in “rage rooms” where customers pay to smash plates and old televisions. The body does matter here, but the popular prescription is backwards.

In 2002 the psychologist Brad Bushman angered a large group of students and then divided them. Some hit a punching bag while thinking about the person who had insulted them. Some hit it while thinking about getting fit, which was meant as a distraction. Others sat quietly. Afterward each could blast the evaluator with loud noise. The students who punched the bag while ruminating on the insult were the angriest and the most aggressive. Distraction came next. The group who simply sat was the least angry. Doing nothing worked better than venting. Bushman and his colleagues also found that people who read a persuasive article saying venting is good for you were more eager to hit the bag, and more aggressive afterward. And in a 2001 study a fake “mood-freezing” pill, which the participants were told would hold their mood fixed, eliminated that extra aggression, which suggests people who believe in catharsis aggress partly in the hope of feeling better.

We found no controlled trial of a commercial rage room, so its effects are untested, though the logic of the punching-bag studies is not encouraging. What does help? Evidence is better for stepping back and reinterpreting. In 2012 Dominik Mischkowski, Ethan Kross, and Bushman found that people who described an insult from a distant, observer’s point of view showed fewer aggressive thoughts and less aggression. In another study that year Fabiansson and Denson found that reappraisal and distraction both lowered anger at first, but only reappraisal kept it down, and it led people to make fairer offers in a negotiation. Slow breathing, which we turn to next, is a plausible helper, although it has not been tested head to head against anger in the same way.

Heart, breath, and sleep

Whether or not the body causes feelings, it certainly reports them, and people differ in how well they listen. Psychologists call the sense of the body’s inner state interoception. A classic test, devised by Rainer Schandry in 1981, asks you to silently count your heartbeats over a set time, while a monitor counts them too. People who do well on the task, in a study by Stephen Wiens and colleagues, tended to rate emotional film clips as more intense. In a 2004 brain-imaging study Hugo Critchley and colleagues found that activity in the right anterior insula, a fold of cortex that maps the body’s inner state, predicted how accurately people sensed their heartbeats. London traders who were better at it, in a 2016 study, also made more profit and lasted longer in the market. The skill is real, but heartbeat counting is a shaky measure: in a 2018 study of 572 people, counted beats matched actual beats only weakly, and earlier work by Ring and Brener showed that counts follow what people believe about their heart rate. “Trust your gut” often means, in this literature, “notice your heart,” and the measure is not trustworthy enough for much more.

Breathing is the one body signal you can change directly. In a 2023 trial at Stanford, led by Melis Yilmaz Balban, 108 volunteers were assigned to do one of three breathing exercises, or mindfulness meditation, for five minutes a day for 28 days, at home and online. One exercise was “cyclic sighing”: breathe in slowly through the nose, take a second small sip of air at the top to fill the lungs fully, then let all the air out slowly, ideally through the mouth. Compared with mindfulness meditation, the breathing groups had a bigger improvement in daily mood and a bigger drop in breathing rate, and cyclic sighing did best. The groups were small, the volunteers chose to take part, there was no do-nothing group, and many people did only about two thirds of the sessions. A 2023 meta-analysis of 12 trials with 785 adults found that breathing exercises lowered stress by a small to moderate amount (about −0.35 standard deviations), with similar effects on anxiety and depression, and cautioned against hype. You will read that slow breathing “activates the vagus nerve.” That phrase is a metaphor, not a measured mechanism, and the idea that heart-rate variability is a simple gauge of “vagal tone” is itself disputed.

The body also reports the opposite of calm, and the clearest case is sleep. In 2007 Seung-Schik Yoo, Matthew Walker, and colleagues at the University of California, Berkeley, scanned volunteers shown disturbing pictures, either after a normal night or after a night without sleep. According to a review by the same group, the amygdala’s response was about 60 percent stronger after a night without sleep, and its connection to the prefrontal region that normally calms it was weaker. A 2024 meta-analysis of 154 studies with 5,717 people found that lost sleep reliably lowered positive feelings and raised anxiety. The effect on negative mood was less clear. Walker’s popular book about sleep contains errors that critics have documented, so the primary papers are the safer source.

The body in our judgments

If the body colors feeling, does it also tilt judgment? The most famous claim says that it does, and more strongly than you would believe. In 2011 Shai Danziger, Jonathan Levav, and Liora Avnaim-Pesso reported on 1,112 parole rulings by eight Israeli judges on two parole boards over 50 days. The judges heard 14 to 35 cases a day, with two food breaks that cut the day into three sessions. At the start of each session about 65 percent of requests were granted. The share fell toward zero by the end, and bounced back to about 65 percent after each break. It looked as if hunger turned judges harsh, and the finding went around the world.

But the study never measured hunger, only the timing of breaks. And the order of cases was not random. Keren Weinshall-Margel and John Shapard noted that prisoners without lawyers were heard last in each session and had a much lower chance of success. In their own sample of 227 hearings, 15 percent of unrepresented prisoners won against 35 percent of those with lawyers. Andreas Glöckner showed in 2016 that a perfectly rational judge who tries to avoid starting a long case just before a break, and favorable cases take longer to decide, produces a similar sawtooth. In 2017 Daniël Lakens calculated that the effect, taken literally, was as large as the difference in height between men and women, which is too big to be believable for a psychological nudge. The pattern is in the data. That it is a hunger effect is not established.

“Hangry” is a gentler claim and fares a little better. In 2014 Bushman and colleagues followed 107 married couples for 21 days. Each evening the spouses stuck as many as 51 pins into a voodoo doll of their partner, and the days with lower blood sugar went with more pins. In 2022 Viren Swami and colleagues prompted 64 people five times a day for three weeks and gathered more than 9,000 responses. People who rated themselves as hungrier also rated themselves as angrier and more irritable, and less pleased, even after allowing for age, sex, weight, and diet. An experiment by Jennifer MacCormack and Kristen Lindquist in 2019 added a twist: hunger turned into negative judgment of a researcher only when the participants did not realize they were reading their own state as emotion. That is misattribution again, the lesson of Lesson 2. What did not hold up is the idea that willpower is fueled by sugar. A preregistered test of “ego depletion” in 2016, by 23 labs with 2,141 people, found an effect of 0.04, which is close to zero, and a 2021 test in 36 labs with 3,531 people found 0.06.

Many of the most striking claims of the “embodied” kind failed the same test. In 2008 Lawrence Williams and John Bargh reported that 41 students who briefly held a cup of hot coffee in an elevator rated a stranger as having a warmer personality than those who held iced coffee. Replications of the experiments by Dermot Lynott and colleagues in 2014, with 861 people, and by Christopher Chabris and colleagues in 2019 found nothing. A clipboard’s weight was said to make people judge issues as more important, and a bitter drink to make moral judgments harsher; Matthew Eskine’s bitter-drink study in 2011 used 57 people, and a multi-site replication in 2020 with 1,137 people found little to no support. A pooled analysis of 25 weight and importance studies found a strong sign of publication bias and, once corrected, no effect. A 2022 preprint that repeated 49 “social priming” studies in 70 close replications found that 94 percent of the new effects were smaller than the originals, and only 17 percent were significant in the predicted direction.

The gut deserves a separate word, because a true fact has been inflated into several false ones. About 90 percent of the body’s serotonin is made in the gut, mainly to help move food along. The gut and the brain do talk, along the vagus nerve, most of whose fibers carry signals from the body up to the brain, and a large 2019 study of 1,054 people in Belgium found that two kinds of gut bacteria, Coprococcus and Dialister, were depleted in people with depression. That is a correlation. Trials of probiotics for mood are small and early: a pilot in 2023 with 49 adults was a feasibility study, and a review of 14 human trials found no significant drop in anxiety. The gut is an organ of the body that shares signals with the brain. It does not make your decisions.