The role of sensorimotor experience in the development of mimicry in infancy

Abstract During social interactions we often have an automatic and unconscious tendency to copy or ‘mimic’ others’ actions. The dominant view on the neural basis of mimicry appeals to an automatic coupling between perception and action. It has been suggested that this coupling is formed through associative learning during correlated sensorimotor experience. Although studies with adult participants have provided support for this hypothesis, little is known about the role of sensorimotor experience in supporting the development of perceptual‐motor couplings, and consequently mimicry behaviour, in infancy. Here we investigated whether the extent to which an observed action elicits mimicry depends on the opportunity an infant has had to develop perceptual‐motor couplings for this action through correlated sensorimotor experience. We found that mothers’ tendency to imitate their 4‐month‐olds’ facial expressions during a parent‐child interaction session was related to infants’ facial mimicry as measured by electromyography. Maternal facial imitation was not related to infants’ mimicry of hand actions, and instead we found preliminary evidence that infants’ tendency to look at their own hands may be related to their tendency to mimic hand actions. These results are consistent with the idea that mimicry is supported by perceptual‐motor couplings that are formed through correlated sensorimotor experience obtained by observing one's own actions and imitative social partners.


S.2. Individual muscle activations
In the paper we reported an effect of maternal imitation group on the mimicry scores in the direct gaze condition. Here we demonstrate that the results are identical when we analyse the individual muscle activations instead. A repeated measures analysis on the EMG activity with Muscle region (Frontalis vs. Masseter), and Action type (Eyebrow vs. Mouth) as within-subjects factors, and maternal imitation (high vs. low) as between-subjects factor demonstrated a significant interaction between Muscle region, Action, and Maternal imitation, F (1, 25) = 6.617, p = .016, ηp 2 =.209. We followed-up on the significant three-way interaction by performing separate repeated measures analyses for the high and low maternal imitation group with Muscle region (Frontalis vs. Masseter), and Action type (Eyebrow vs. Mouth) as within-subjects factors. These analyses demonstrated that there only was a significant interaction between Muscle and Action type in the high maternal imitation group, F (1, 13) = 10.335, p = .007, ηp 2 =.443, but not in the low maternal imitation group, F (1, 12) = .224, p = .645, ηp 2 =.018. Infants in the high maternal imitation group showed significantly more frontalis region activation than masseter region activation during the observation of

S.3. Averted gaze condition
In the paper we focused on the relationship between sensorimotor experience and mimicry of actions accompanied by direct gaze, as this was the only condition in which we previously found evidence for mimicry (de Klerk et al., 2018). For completeness, we report the results for the equivalent analyses performed on the averted gaze conditions here.

S.3.1. Facial mimicry
Correlational analyses demonstrated that there was no significant relationship between maternal facial imitation and infant mimicry of facial actions accompanied by

S.4 Overt hand mimicry coding
Previously we did not find evidence for mimicry of hand actions (de Klerk et al., 2018). Although this does not preclude the possibility that there is nevertheless meaningful variability in the EMG measure of hand mimicry, it is also possible that the hand EMG data was not reliable enough to detect an effect (for example because the electrodes were too big to measure from specific hand muscles). Therefore, to obtain an additional index of hand mimicry we also coded the videos for overt mimicry of hand actions. Videos were coded offline and all trials in which the infant performed hand actions such as hand opening and closing, or finger movements, were given a code of 1. Trials in which the infant did not see at least two thirds of the action were excluded from analysis. Additionally, trials were excluded if the infant's hands were not visible in the video or if the infant was holding onto something.
The coder was unaware of which trial type the infant was observing. We calculated a hand mimicry score by taking the probability that the infant performed a hand action when they observed a hand trial, and subtracting the probability that the infant performed a hand action when they observed a facial action trial. The resulting measure represents the infant's tendency to specifically perform hand actions when observing hand trials in the Hand_Direct and Hand_Averted trials. For three of the infants who were included in the hand EMG analyses their hands were not visible for a sufficient number of trials to code for overt hand mimicry. Thus the analyses on the overt hand mimicry were based on 20 infants.

S.4.1 Overt hand mimicry results
The results of the analyses on the overt hand mimicry scores replicated those on the EMG measures of hand mimicry. There was no significant relationship between the proportion of time the infants spent looking at their own hands during the PCI and their overt mimicry of hand actions accompanied by direct or averted gaze, all p's > .756.
However, again when we created a grouping variable based on a median split of the proportion of time the infants spent looking at their own hands, we found group differences in overt hand mimicry in the direct gaze condition between those infants who showed more versus less interest in their own hands during the PCI. An ANOVA on the overt hand mimicry scores in the direct gaze condition with 'hand interest' group (High vs. Low) as between subjects' factor showed a marginally significant effect of group, F (1, 18) = 3.847, p = .065, ηp 2 = .176. There was no effect of hand interest group on overt mimicry of hand actions accompanied by averted gaze, F (1, 18) = .002, p = .966, ηp 2 = .00. As can be seen in Supplementary Figure 5, infants in the high hand interest group (N=10) showed a greater tendency to overtly mimic hand actions accompanied by direct gaze compared to infants in the low hand interest group (N=10).
Again, there was no effect of maternal facial imitation group on overt hand mimicry, F (1, 18) = .783, p = .388, ηp 2 = .042, and no correlation between maternal facial imitation and infant overt hand mimicry, r (18) = -.037, p =.878 (lower 95% CI= -.361, upper 95% CI = .270). The fact that these results are the same as those obtained using the hand EMG measure suggests that this latter measure provided a reliable index of mimicry, and provides converging evidence for the idea that infants' interest in their own hands seems to be related to their tendency to mimic others' hand actions.