The great American psychologist William James proposed that bodily sensations – a thumping heart, a sweaty palm – aren’t merely a consequence of our emotions, but may actually cause them. In his famous example, when you see a bear and your pulse races and you start running, it’s the running and the racing pulse that makes you feel afraid.
Consistent with James’ theory (and similar ideas put forward even earlier by Charles Darwin), a lot of research has shown that the expression on our face seems not only to reflect, but also to shape how we’re feeling. One of the most well-known and highly cited pieces of research to support the “facial feedback hypothesis” was published in 1988 and involved participants looking at cartoons while holding a pen either between their teeth, forcing them to smile, or between their lips, forcing them to pout. Those in the smile condition said they found the cartoons funnier.
But now an attempt to replicate this modern classic of psychology research, involving 17 labs around the world and a collective subject pool of 1894 students, has failed. “Overall, the results were inconsistent with the original result,” the researchers said.
The replication effort which has been published online in Perpsectives on Psychological Science attempted to stay extremely close to the original 1988 study, but there were a few differences. For example, the instructions to the participants were delivered by video to avoid experimenters inadvertently influencing the participants. And the participants were videoed during the study to ensure that they held the pen correctly in their mouth. As in the original, the aims of the research were disguised as test of motor control and consistent with this cover story, participants first had to perform some tasks with the pen (such as drawing lines between numbers) before looking at the cartoons.
Overall, nine of the participating labs found results that were in the same direction as the original research – participants with a smiling expression caused by holding the pen in their teeth tended to rate the cartoons as funnier than the pouting participants. But the size of the difference was much smaller than in the original research. And when the results from these nine labs were added to the eight who found results in the other direction, the overall outcome was no effect.
The replication researchers, led by E. J. Wagenmakers of the University of Amsterdam, said their results had failed to replicate the original “in a statistically compelling fashion” but they noted that this does not mean the entire facial feedback hypothesis is dead in the water. Many diverse studies have supported the hypothesis, including research involving participants who have undergone botox treatment, which affects their facial muscles.
In a commentary published alongside the replication effort, Fritz Strack, lead author of the original 1988 classic, said that he lauded “the replicators’ effort in this extensive enterprise” but that there were several issues with their methodology that cause him to be concerned with the validity of what he considers to be a surprising outcome (note that Strack proposed that his 1988 study be subjected to a replication attempt and he provided all his original materials to the replication team). Among the issues he raises is that being videoed may have affected the students’ emotional experience through causing them to feel self-conscious.
“… while a first look at the current data seems to suggest that the  SMS facial-feedback study has been convincingly ‘non replicated’,” he writes, “a closer inspection of the replication studies reveals several methodological and statistical issues that need to be considered before drawing further conclusions on the validity of the method, of the model, or of the underlying mechanism.”