Keven & Akins (K&A) re-analyze Meltzoff and Moore's (Reference Meltzoff and Moore1977) results, re-interpreting the so-called NI of tongue protrusion and retraction (TP/R) gestures by the infant from an exhaustive description of motor control acquisition. They adopt a neurodevelopmental perspective that shares some crucial arguments with the neuroconstructivist model (Mareschal et al. Reference Mareschal, Johnson, Sirois, Spratling, Thomas and Westermann2007): developmental explanation; neural development as activity and context-dependent; multiplicity of contexts; chronotropy; lack of univocal correspondence between behaviors and cognitive processes; and interdependence among different levels (i.e., genetic, neural, cognitive, behavioral, environmental) both in psychological development and for the explanation of this development.

K&A's work constitutes a bright and parsimonious redefinition of TP/R gestures as developmental stereotypes (i.e., subcortically controlled, arousal dependent, and faded once infants acquire motor control), focusing their analysis on neural and behavioral levels. We strongly support the contention that NI behaviors neither fulfill the criteria for imitation nor that they are precursors for genuine imitation. The purpose of this commentary will be twofold: (a) to support this assumption from a cognitive level of analysis; and (b) to add some evidences emphasizing the relevance of dyadic interaction in the development of imitation.

Scientific evidence seems to show that the cognitive system of the newborn would not be prepared yet to accomplish all of the tasks involved in imitation. Their perceptive and attentional capacities (Volpe Reference Volpe2008), and face processing and intersensory processing abilities (Bahrick et al. Reference Bahrick, Lickliter and Flom2004; Johnson et al. Reference Johnson, Senju and Tomalski2015; Lewkowicz Reference Lewkowicz2014; Morton & Johnson Reference Morton and Johnson1991), among others, are too weak yet. In the following weeks, organism-environment interactions will shape the development of social orienting, and infants will soon be able to respond contingently to social stimuli. There is just a “chronotropic coincidence” between the progressive improvement in social orienting abilities and the fading of NI. We share K&A's disagreement with the social hypothesis view on the explanation of this drop-out “from a change in performance not competence, as the later emergence of sophisticated imitation makes clear” (sect. 2, para. 5). NI and early social orienting abilities do not correspond respectively to a subcortical and cortical control of a unique behavior with the same cognitive and social function, but they follow distinct developmental lines. Imitation would be found only in the developmental trajectory of social orientation.

The beginning of imitation behaviors has been established by 6–8 months (Oostenbroek et al. Reference Oostenbroek, Suddendorf, Nielsen, Redshaw, Kennedy-Costantini, Davis, Clark and Slaughter2016). However, in our view, there would be earlier specific types of behaviors, similar in their appearance to NI, but again with a totally different function and origin, that would be truly incipient forms of matching behaviors, and precursors of genuine intentional and intended (Carpenter & Tomasello Reference Carpenter, Tomasello, Chater and Hurley2005) imitation. Those were called mimicry behaviors, and have been scarcely considered in infants imitation studies (Moody & McIntosh Reference Moody, McIntosh, Rogers and Williams2006). Mimicry behaviors emerge in typical development when NI behaviors are disappearing, and while the infant is acquiring cortical motor control but before the development of cognitive functions involved in mature imitation. They are still automatic, non-intentional, and non-goal directed, but they imply the copy of a model and are triggered by a specific stimulus, not by general arousal. Even when involuntary, mimicry behaviors do also form a substrate for the directed behaviors to follow, and they have a clear social function (Carpenter et al. Reference Carpenter, Uebel and Tomasello2013; Moody & McIntosh Reference Moody, McIntosh, Rogers and Williams2006).

A clear example of those mimicry behaviors is social smiling. The social smile emerges around 8 weeks, coinciding with the 2-month shift: when infants show improved head control and gaze direction, increased alertness and sustained attention, and an increasing ability also to explore features of the face (Wörmann et al. Reference Wörmann, Holodynski, Kärtner and Keller2012).

Consequently, significant changes in dyadic interaction between adult and infant arise during this second month (Lavelli & Fogel Reference Lavelli and Fogel2005). Another set of evidence shows that, when mother-infant dyads from two different contexts regarding the frequency of face-to-face interactions are compared, no differences are found at 6 weeks. However, at 12 weeks, the frequency of smiling behaviors is higher for the group of dyads in the more interactive context (Wörmann et al. Reference Wörmann, Holodynski, Kärtner and Keller2012).

Thus, social interaction is needed for mimicry behaviors to develop, and they influence and are influenced by variables associated with social contact. Mimicry has been called the “social glue” (Lakin et al. Reference Lakin, Jefferis, Cheng and Chartrand2003): It promotes affiliation, empathy, and pro-social behaviors, both in adults (Chartrand & Lakin Reference Chartrand and Lakin2013; Duffy & Chartrand Reference Duffy and Chartrand2015) and young children (Carpenter et al. Reference Carpenter, Uebel and Tomasello2013). In infant-adult dyads, from one side, mothers interpret an infant's imitation as an invitation to respond contingently (Wörmann et al. Reference Wörmann, Holodynski, Kärtner and Keller2012). From the other, infants whose mothers show more imitative behaviors are more able to perceive contingency in interactions (Legerstee & Varghese Reference Legerstee and Varghese2001; Soussignan et al. Reference Soussignan, Nadel, Canet and Gerardin2006).

Interaction provides the dyad with opportunities of repetition (Thelen Reference Thelen1981b), and this repetition (with certain characteristics, not perfectly contingent …) will progressively enable a mutual adjustment between infant and adult. From very early on, the presence and frequency of imitative behaviors is an indicator of positive interactions, regarding the sensitive responsiveness of the mother, the attentiveness to caregiver shown by the infant, and the degree of mutuality shared by the dyad (Wan et al. Reference Wan, Green, Elsabbagh, Johnson, Charman and Plummer2013).

Mimicry is a precursor of intentional imitation, but it does not fade when the dyad develops those more sophisticated forms of imitation. Some preliminary results from a longitudinal study we are carrying out on the developmental trajectory of imitation in infants show that mimicry behaviors increase in frequency from 9 to 15 months old. Mimicry behaviors would be acting as an enhancer of social interaction through development.

Earlier behaviors such as reflex smile or TP/R gestures are crucial in human development, and are part of an infant's repertoire of motor behaviors that will be used to imitate. However, our hypothesis, as is K&A's, would be that its function is far from sharing the developmental trajectory of imitation. Imitation behaviors will develop in an interactive context, and mimicry will emerge in the beginning of that developmental trajectory, and will continue to promote interaction through its whole course.