Does Smiling Actually Make You Happier?

In 1988, Fritz Strack, Leonard Martin and Sabine Stepper ran an experiment that had to be clever about a simple problem: if you ask people to smile and then ask them how happy they feel, you have all but told them what answer you expect. So instead, they gave people a pen and asked them to hold it in one of two ways. Some held it between their teeth, which forces the same muscles used in a genuine smile into action without anyone using the word "smile". Others held it between their lips, which does close to the opposite, activating a muscle that makes smiling difficult. Everyone then rated how funny they found a set of cartoons, with no idea their face had anything to do with the study.

People holding the pen with their teeth rated the cartoons as funnier than people holding it with their lips. The interpretation was that facial expression is not only an output of emotion - a smile because you feel amused - but also, in a small way, an input. The technical name is the facial feedback hypothesis, and the study became one of the most cited demonstrations of it.

Why would a facial expression feed back into feelings at all? Psychology has floated a couple of candidate explanations without fully settling on one. One idea, self-perception theory, suggests you partly work out your own emotional state from your own behaviour and expression, much as you would infer someone else's mood from watching their face. A second, more specific idea is that physically contracting certain facial muscles sends signals to the brain that are woven into the same circuitry that produces emotional experience, independent of any interpretation at all. Neither has been confirmed as the mechanism, and both may play a small part - it is one of the honest loose ends in this area of research.

A seventeen-lab test says: not so fast

In 2016, a large international collaboration led by Eric-Jan Wagenmakers set out to repeat the exact pen-in-mouth procedure as faithfully as possible, this time across seventeen independent laboratories with a combined sample of close to nineteen hundred people, following a protocol agreed in advance and reviewed in part by Strack himself. Across that much larger and more carefully controlled sample, the effect on how funny people rated the cartoons came out essentially at zero, and not statistically significant. The classic demonstration, run this rigorously, did not show up.

Strack later published his own reflection on the result, raising a fair methodological question: the replication filmed every participant, and knowing you are on camera while holding a pen in an odd way in your mouth is not quite the same experience as doing it unobserved. It is a real possibility, and also a reminder of how hard it is to run the exact same study twice - but it did not overturn the headline result, which stands as a large, well-powered failure to replicate one specific, famous demonstration.

Zooming out to everything else

A single failed replication of one paradigm does not settle whether facial feedback is real in general, because dozens of other studies, using other methods and other emotions, existed too. In 2019, Nicholas Coles, Jeff Larsen and Heather Lench combined the results of decades of that research in a large meta-analysis. Their conclusion was carefully worded and is close to the honest middle ground on this whole topic: facial feedback does have a real effect on emotional experience, but it is small, and it varies a lot depending on which emotion and which method researchers use. Effects tended to be more consistent for positive expressions like smiling than for expressions of anger, fear or disgust.

A bigger, more direct test of smiling specifically

The most thorough recent test came in 2022, when Coles and dozens of collaborators ran the Many Smiles Collaboration: dozens of labs around the world, close to four thousand participants, testing three different ways of getting someone's face into a smile-like configuration, including one method modelled closely on the original pen technique and another that asked people to mimic photographs of smiling faces. Averaged across methods, there was a small but real increase in self-reported happiness, and the effect was strongest for the method that had people actively mimic a genuine smiling face rather than the more artificial pen-in-mouth version.

Where this leaves the idea

Put the pieces together and a coherent, honest story appears. The specific 1988 pen demonstration, taken in isolation, has not held up well under close scrutiny. The broader claim that your facial expression can nudge your emotional experience a little, especially toward positive feelings and especially when the expression resembles a real smile rather than an artificial substitute, has more support across many studies than any single failed replication would suggest - it is just a modest nudge, not a switch you can flip to override a bad day.

What this means for you

Do not expect grinning at a mirror to fix real sadness, and never treat this as a substitute for dealing with what is actually making you unhappy or for professional support if low mood persists. But there is reasonable evidence that a genuine smile, especially one prompted by something that would make you smile anyway, tracks with and can gently reinforce a positive emotion rather than only following it. If you want to understand how you personally experience and regulate positive feelings, or how your particular sense of humour works, that is exactly the territory our positive emotions, emotion regulation and humour style tests are designed to map, in more depth than any single facial expression ever could.