Let’s Dance TogetherSynchrony, Shared Intentionality and Cooperation

Why does marching in step with strangers make you more likely to help them? Not metaphorically — measurably, with real money on the line. Three experiments later, the answer turns out to be more specific than anyone assumed, and it changes how we think about why humans ever started dancing together in the first place. Collective music and dance appear in every culture on record. Reddish, Fischer, and Bulbulia note that such performances have long been hypothesized to bind people together and increase cooperation — and prior experiments supported that idea. However, those experiments left a mechanism unexplained. When bodies move in time, something shifts socially. The question is what, exactly, is doing the work. To answer it, the authors separated two phenomena that usually travel together. Synchrony is the low-level matching of behavior — moving or sounding at the same frequency or phase as another person. Shared intentionality is something higher-order: the explicit sharing of a goal to produce that joint action. In spontaneous conversation, both can arise together without anyone trying. In communal ritual, participants actively collaborate to stay in time. Reddish and colleagues wanted to know whether cooperation springs from the body matching alone or whether it requires the mind to be aiming at the same target. Experiment one set up four conditions. In the shared-goal condition, participants were explicitly instructed to work together to move in time with each other. In the synchrony condition, they moved to the same rhythmic beat — producing synchrony as a byproduct without anyone intending it as a joint project. In the asynchrony condition, they moved to different rhythms. In a passive control, there was no movement at all. After six minutes of movement, cooperation was measured with a public goods game — each person chose how much real money to contribute to a shared pot, from zero to five dollars. The shared-goal condition won. Mean contributions were three dollars and eighty-eight cents in the shared-goal condition, three dollars and fifty-three cents in the synchrony condition, two dollars and eighty-seven cents in the asynchrony condition, and three dollars and twenty-two cents in the passive control. The difference between shared-goal and asynchrony was statistically significant, with an effect size of r equals point three two. A linear trend across all four conditions was also significant, consistent with the prediction that cooperation increases as you move from asynchrony to passive to synchrony to intentional synchrony. Self-report measures of interdependent self-construal — how much participants experienced themselves as part of the group rather than as separate individuals — showed the same pattern, with the shared-goal condition scoring highest by a wide margin. There was one important caveat. In the shared-goal condition, fifty-five percent of participants chose the maximum contribution of five dollars. That ceiling effect likely suppressed the measured gap between conditions. The true advantage of shared intentionality may have been larger than the numbers show. Experiment two moved the manipulation out of movement and into voice. Twenty-seven volunteers in groups of three read aloud a list of emotionally neutral one-syllable words for six minutes. In the synchrony condition, they were told to speak in time with each other. In the sequential condition, they were told to speak out of phase — reading different columns so the three voices would be offset. Crucially, both conditions involved shared intentionality: everyone was deliberately trying to coordinate. The only difference was whether the goal was unison or offset. This let the authors isolate synchrony's contribution when shared intentionality is held constant. The result was clean. Participants who vocalized in synchrony were nearly six times more likely to choose the cooperative option in a stag-hunt coordination game — sixty-two percent versus twenty-one percent chose cooperation, and the odds ratio was five point eight seven. Perceived synchrony ratings confirmed the manipulation worked: the synchrony group averaged five point five on a seven-point scale, while the sequential group averaged four point four eight. That gap in perception translated directly into behavior. Self-report measures of trust, similarity, and interdependent self-construal did not differ significantly between conditions, but trust did correlate with cooperative choice at r equals point seven five. Synchrony pushed behavior even when self-report scales couldn't detect why. Experiment three ran the full factorial test. Eighty-six participants were assigned to one of six conditions crossing three levels of movement — in-phase synchrony, frequency-locked sequential movement, or out-of-time asynchrony — with two goal conditions: a group goal to coordinate together or an individual goal. Foot-pedal timing data gave an objective synchrony measure. Cooperation was again measured with the stag-hunt game. The interaction between movement and goal was significant: a chi-square of seven point ninety-four with two degrees of freedom, and a p-value of point zero two. That interaction was entirely driven by the group-goal conditions. When participants shared a goal, in-phase synchrony produced dramatically more cooperation than either sequential or asynchronous movement — ninety-three percent of participants in the synchrony-plus-group-goal cell chose the cooperative option, compared with forty-three to sixty-two percent in the other cells. The odds of cooperating after synchrony were thirteen times higher than after sequential movement and seventeen point thirty-three times higher than after asynchrony. Without a group goal, synchrony didn't pull away from the other conditions at all. Social unity — a composite of perceived cooperation, similarity, and group cohesion — was higher after synchrony than after asynchrony, with a Cohen's d of zero point eighty-five, and it correlated with cooperative choice at r equals point three two. Trust, by contrast, showed no significant effects across movement or goal conditions. That's worth pausing on. Trust is the mechanism most people would reach for when explaining why moving together makes you more cooperative. These data say it's not the proximate driver — or at least not in the timeframe of these experiments. The authors tested their proposed mechanism with a path analysis. The model they proposed goes like this: movement synchrony and a shared group goal lead to perceived synchrony, which leads to perceived evidence of successful cooperation, which produces social unity and trust, which then predicts cooperative behavior. That model fit the data well — a chi-square of thirty-two point zero six with twenty-four degrees of freedom, a comparative fit index of point ninety-four, and a root mean square error of approximation of point zero six. A reversed model, in which cooperative behavior produced the psychological states rather than the other way around, fit poorly: a chi-square of eighty-nine point twenty-eight, a comparative fit index of point seventy-three, and a root mean square error of approximation of point one nine. The directional claim held up. What this model says, in plain terms, is that synchrony functions as feedback. When a group shares the goal of moving together and perceives that they are succeeding, each iteration of coordinated action is evidence that the group is working. That perceptual signal reinforces cooperative tendencies in real time. It is not that synchrony makes you like people more and you therefore cooperate. It is that synchrony, when framed as a joint project, gives the group continuous proof that it can coordinate — and that proof compounds. This is meaningfully different from simpler accounts in which motor mimicry blurs the self with others and cooperation follows automatically. The blurring may occur, but without a shared goal directing attention to the group's success, the cooperative effect is weaker and less reliable. Reddish, Fischer, and Bulbulia are honest about the limits of their setup. Lab tasks lasted six minutes; real ritual lasts hours. Participants were strangers; real communities share identity, dress, and cultural meaning. These differences matter for generalization. The authors are not claiming to have reproduced a communal ceremony in a university basement. Within those limits, what they establish is a specific causal structure. Synchrony combined with shared intentionality produces the strongest cooperative response, and the reinforcement model provides the best available account of why. The cautious cultural implication follows: collective music and dance, conserved across human societies, may persist in part because they reliably generate this feedback loop. Groups that maintained coordinated ritual practices would have had access to a technology for producing cooperation on demand — and the collectivist values that sustain those practices may themselves be part of the mechanism, not merely its context. That is a long reach from a public goods game and a set of foot pedals. But the path between them, at least, has been mapped. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.