Until more compelling evidence emerges, however, a propensity to imitate from birth should no longer be considered an established phenomenon

Re-evaluating the neonatal imitation hypothesis. Janine Oostenbroek et al. Developmental Science, DOI: 10.1111/desc.12720

We appreciate Meltzoff et al.’s (2017) detailed commentary on our study of neonatal imitation (Oostenbroek et al., 2016) and the opportunity to clarify critical aspects of the findings. First, however, we wish to highlight that our data were collected as part of a larger longitudinal investigation, in which we set out to examine potential relations between neonatal imitation and later emerging aspects of social cognition (Suddendorf, Oostenbroek, Nielsen, & Slaughter, 2013). Our aim was not to test the veracity of the phenomenon of neonatal imitation, but to assess whether early individual differences predict later imitative and other socio-cognitive capacities, which would support the view of neonatal imitation as the foundation of social cognition (Meltzoff, 2002). When coding was finalized, it became clear that, to our surprise, the data challenged the very existence of the phenomenon whose trajectory and consequences we sought to chart.

Meltzoff et al. (2017) claim that methodological problems biased our results towards null effects. We assert that there is no compelling reason to think that any of these critiques undermine our challenge to the field. For instance, they argue that too many stimuli were used in a within-subjects design, which may have led to neonatal fatigue and disengagement. If that were the case, however, then initial trials should have elicited imitation but not later ones. We found no such order effects in any of our analyses (see Supplementary Material in Oostenbroek et al., 2016). Meltzoff et al. rightly note that infants cannot be expected to imitate behaviors that they are incapable of producing. Our data show, however, that with the exception of the ‘eee’ vocalizations and sad faces, infants produced all target behaviors either occasionally or regularly. [...]

The literature on neonatal imitation has a long history of post-hoc theorizing. In response to early failures to replicate the effect, for example, it was stipulated that neonates should be unfamiliar with the model’s face to maintain interest during the experiment (Meltzoff & Moore, 1983a). Later, however, it was argued that infants imitate familiar faces to communicate recognition (Meltzoff, 2005; Meltzoff & Moore, 1994), and that ‘imitation is not modified by which person serves as the model’ (Meltzoff & Moore, 1992, p. 492). Now, Meltzoff et al. again argue that the model should be unfamiliar, even suggesting that this is the ‘key’ to eliciting neonatal imitation. Setting aside the obvious contradictions here, it seems untenable to claim that the ‘engine and mechanism for the growth of social cognition’ (Meltzoff, 2002, p. 7) would manifest itself only with unfamiliar people.