MEG reveals distinct neural signatures as pairs improvise in leader–follower roles
Real-time coordination between two people is never just a matter of moving in sync. In this preprint, investigators used magnetoencephalography and hand kinematics to examine how the brain differentiates between leading and following during a joint-improvisation mirror-game task with predefined roles. Sixteen dyads took part, and the design allowed the team to compare leader–follower coordination at the level of cortical activity and movement-linked dynamics rather than relying only on behavioral timing.
The central finding is a split in neural strategy. Followers showed stronger theta-band activity across frontal, cingulate, and sensorimotor regions, along with stronger coherence between movement kinematics and occipital cortical activity. Leaders, by contrast, showed reduced alpha-band power in sensorimotor and premotor regions and stronger beta-band activity in occipital cortices. Those patterns suggest that the two roles recruit different neural resources even while producing the same shared task behavior.
Followers recruited monitoring networks more strongly
Theta-band activity is often associated with control, monitoring, and adjustment of ongoing behavior. In this study, the follower role was linked to a theta cluster spanning frontal, cingulate, and sensorimotor areas, a distribution that fits the behavioral demands of staying aligned with another person’s movement and adapting quickly when the partner changes pace or direction. The pattern also matched the stronger coherence between movement kinematics and occipital cortical activity in followers, suggesting heavier dependence on visual information about motion.
That combination points to a practical division of labor in social movement: followers appear to lean more on online monitoring and sensory updating, especially visual tracking of the partner’s actions. The occipital involvement is particularly relevant because the task depended on continuous observation of the other person’s movement. In a setting like this, the follower does not simply copy. The follower continuously estimates the partner’s next move and recalibrates in real time.
A single sentence captures the shift: followers carried the heavier sensory-monitoring burden.
Leaders showed a different balance of prediction and motor preparation
The leader role was associated with lower alpha power in sensorimotor and premotor regions. In many motor contexts, reduced alpha power is interpreted as greater cortical engagement, and here it aligns with the demands of initiating and shaping movement rather than merely tracking it. Leaders also showed stronger beta-band activity in occipital cortices, a result that is less intuitive at first glance but may reflect elevated demand for predictive visual processing during self-guided coordination.
This pattern suggests that leading in a dyadic improvisation task is not passive or automatic. Instead, it likely requires active action planning, anticipation of the partner’s response, and tight internal control over timing. The premotor and sensorimotor changes fit that interpretation well, since those regions sit near the core of movement preparation and execution. Occipital beta activity may reflect a more specialized visual role too, possibly tied to maintaining a stable perceptual model of the interaction while the leader sets the motion trajectory.
In other words, leadership in this task was not about reduced effort. It appears to have shifted the burden from reactive adjustment toward proactive control.
Frequency-specific dynamics separated the two roles
The study examined delta, theta, alpha, beta, and gamma bands, but the clearest role-related effects emerged in theta, alpha, and beta. That matters because it argues against a single undifferentiated “social brain” explanation for coordination. Instead, the data point to frequency-specific mechanisms that split by role and by function. Theta activity aligned with monitoring in followers, alpha suppression aligned with motor engagement in leaders, and beta differences suggested distinct predictive demands in visual cortex.
This is useful because coordinated interaction is often described in broad terms, yet the brain likely handles it through several layered processes at once. One layer tracks the partner’s movement. Another prepares one’s own next action. A third adjusts the coupling between perception and action so the two bodies remain coordinated. The present findings map those layers onto distinct oscillatory signatures.
The work also reinforces the value of pairing MEG with movement kinematics. Without the kinematic signal, the interpretation would have rested only on cortical oscillations; without the neural data, the behavioral measures would have missed role-specific differences in how the interaction was controlled. Together, the two streams of data gave a clearer view of how real-time coordination is organized.
What the preprint adds, and what it cannot yet settle
Because this is a preprint, the findings remain provisional until peer review and replication. The sample was limited to sixteen dyads, which leaves open questions about how robust the observed patterns are across larger groups, different age ranges, task variants, or more naturalistic forms of social interaction. The mirror-game setup also used predefined leader and follower roles, so the results may not fully capture situations in which leadership shifts dynamically within the same interaction.
Still, the study makes a strong mechanistic case that leader and follower roles are neurally distinct even when outward movement looks tightly coupled. The role separation appeared in both power and coherence measures, and it involved cortical regions tied to motor control, monitoring, and visual processing. For social neuroscience, that is a meaningful advance because it moves beyond asking whether two people are synchronized and toward asking how each participant’s brain supports the interaction in a different way.
The clearest next step is to test whether these spectral signatures hold when role assignment is more fluid, when partners differ in skill, or when the task places heavier cognitive load on one participant. Until then, the most concrete takeaway is narrow but important: in coordinated improvisation, leading and following are not mirror images in the brain.