The Mere Co-Presence: Synchronization of Autonomic Signals and Emotional Responses across Co-Present Individuals Not Engaged in Direct Interaction

Abstract

Existing evidence suggests that in social contexts individuals become coupled in their emotions and behaviors. Furthermore, recent biological studies demonstrate that the physiological signals of interacting individuals become coupled as well, exhibiting temporally synchronized response patterns. However, it is yet unknown whether people can shape each other's responses without the direct, face-to-face interaction. Here we investigated whether the convergence of physiological and emotional states can occur among "merely co-present" individuals, without direct interactional exchanges. To this end, we measured continuous autonomic signals and collected emotional responses of participants who watched emotional movies together, seated side-by-side. We found that the autonomic signals of co-present participants were idiosyncratically synchronized and that the degree of this synchronization was correlated with the convergence of their emotional responses. These findings suggest that moment-to-moment emotional transmissions, resulting in shared emotional experiences, can occur in the absence of direct communication and are mediated by autonomic synchronization.

Fig 1. Analysis approach.

The effect of co-presence on autonomic synchronization was assessed by comparing experimental and control inter-subject correlation (ISC) of participants’ autonomic signals. Experimental ISC was calculated for participants who viewed the movie together, in the same group. Control ISC was calculated for participants who viewed the movie not together, in different groups.

Fig 2. Average autonomic response time-courses elicited by positive and negative movies.

Figure presents normalized EDA (blue) and HR (red) response time-courses (2HZ), averaged across participants. Shaded areas represent standard errors.

Fig 3. The effect of co-presence on autonomic ISC.

A. Control (H₀) distributions (not co-present) and the experimental means (co-present, dashed lines), for both autonomic measures (ISC_EDA, ISC_HR) and for both emotional movies. H₀ distributions represent movie-driven ISC of ANS signals. As can be seen in the figure, the experimental group-wise ISCs fall in the right tail of the control distributions, indicating the positive effect of co-presence on synchronization of participants’ autonomic activity. B. Time shifts in ANS synchronization: we assessed whether the ANS signals of co-present individuals were more tightly linked in time. Figure presents the control (H₀) distributions and the experimental means (dashed lines) of the time lags in which two ANS signals exhibited maximal ISC. As can be seen in the figure, the time shifts in ANS synchronization of the co-present participants were significantly smaller in three out of four ANS measures. Lag_EDA in the positive movie showed a trend towards significance.

Fig 4. Association of autonomic synchronization and emotional convergence.

Correlation between the extent of autonomic synchronization, as indexed by ISC_composite scores, and the degree of emotional convergence (small delta scores signify higher convergence). Results are shown separately for positive and negative emotional movies.