Facial expressions can influence emotions

The argument

Research into the relationship between facial expressions and emotions has shown that controlling or preventing facial expressions dampens the experience of the associated emotions, even when other factors are the same. That last qualification is what makes the finding significant. If expressions were merely the outward sign of an emotion generated elsewhere, suppressing the expression would conceal the feeling without altering it. What the research reports instead is a change in the experience itself, which implies that the expression is part of what produces the emotion rather than a readout of it. The causal arrow, on this account, runs in both directions. These findings have provided support to the facial feedback hypothesis, which was first articulated by Charles Darwin during his studies of human and animal emotions. Darwin wrote about the effect of facial expressions to either intensify or soften emotions during emotional experiences — a moderating role, in which the expression adjusts the strength of a feeling already under way. Later, psychologist William James argued for a stronger version: that facial expressions were a primary and independent physiological driver of emotions, generating the emotional state rather than modulating it. The two positions differ in how much work the face is doing, but both treat it as a cause rather than a consequence. A further line of evidence comes from cases where the machinery works differently. The significance of facial feedback to emotions has been emphasized by studies of people with autism spectrum disorders, where difficulties in perceiving the emotions of others have been associated with difficulties in identifying facial expressions. The association is what the hypothesis would predict: if emotional processing is bound up with facial expression, then differences in reading expressions should track differences in reading emotion, and they do.

Premises

Counter-arguments

The facial feedback hypothesis is genuinely contested. A large multi-laboratory registered replication (Wagenmakers and colleagues, 2016) failed to reproduce the classic 'pen-in-teeth' facilitation of amusement, and subsequent meta-analyses report at most small, moderator-dependent effects. Critics argue facial feedback may modestly modulate emotion rather than causally drive it, and that the correlation in autism reflects broader social-cognitive and perceptual differences rather than facial feedback specifically. The evidence supports a weak influence, not the strong claim that expressions are a primary, independent driver of emotional experience.

Rejecting the premises

[Rejecting P1] A prominent multi-lab registered replication failed to reproduce the facial-feedback effect, and meta-analyses find only small, moderator-dependent influences, so 'dampens/intensifies' overstates a fragile finding. [Rejecting P2] The autism correlation is better explained by broader social-cognitive and perceptual differences than by facial feedback specifically, so it does not establish causation. [Rejecting C] The current evidence supports at most a weak modulatory role, not the strong claim that facial expressions are a primary causal driver of emotion.