Guillaume Duchenne de Boulogne was a nineteenth-century French neurologist with an interest in which muscles produce which expressions, and a method that would not survive an ethics committee. He applied electrical probes to the faces of his subjects — including a man with a facial nerve condition that left him insensitive to pain — and photographed the results, stimulating individual muscles to see what each contributed.
He found that a smile of genuine enjoyment involves two muscles. The zygomaticus major pulls the corners of the mouth up and back. The orbicularis oculi, the ring of muscle around the eye socket, tightens, raising the cheek, narrowing the eye aperture, and producing crow's feet at the outer corner. Duchenne's observation, published in 1862, was that the first muscle obeys the will but the second obeys only the sweet emotions of the soul, and that a smile lacking it unmasks a false friend.
Ekman revived the distinction a century later, and it is now standard to call a smile involving both muscles a Duchenne smile.
The claim that made it interesting is that the outer portion of orbicularis oculi is not reliably under voluntary control. Most people cannot contract it deliberately without also doing something else with their face, which is why posed smiles in photographs so often look wrong in a way that is hard to name. Estimates of how many people can produce the marker on demand vary, and actors and practiced performers are better at it than average, but the general asymmetry holds: the mouth is easy to command and the eye ring is not.
A 2020 study found an ingenious way to test this from the opposite direction. Researchers examined smiles in people who had received botulinum toxin injections to the orbicularis oculi — a common cosmetic treatment for crow's feet, which paralyzes exactly the muscle in question. The chemical intervention removed the marker while leaving the emotion intact, and observers' judgments of the smiles changed accordingly.
What all this supports is narrower than the popular version. The Duchenne marker is a reasonably good indicator that a smile is accompanied by positive affect. It is not a lie detector, and the absence of the marker is much weaker evidence than its presence: people smile politely, socially, nervously and out of embarrassment, and none of those are deceptions. A social smile with an inert upper face is the ordinary currency of professional life, produced thousands of times a day by people concealing nothing more sinister than boredom.
Where it earns its place is as a calibration tool for warmth specifically. Somebody who is performing enthusiasm — for you, for a product, for a plan — is producing an emotional display on demand, and the eye ring is the part of that display they cannot easily supply.
This is why photographers say things designed to make you laugh rather than telling you to smile, why the practiced sales smile reads as slightly wrong, and why diplomatic photographs are studied for it. The mouth reports policy. The eyes report state.
See Attack: manufacturing warmth on camera