Keven & Akins (K&A) conclude correctly that orofacial stereotypies are critical to the development of aerodigestion in neonates and are unlikely to co-occur with imitative behavior. Much has been made of facial imitation at birth. Neonates project their tongues in reaction to scientists doing the same, suggesting that proper imitation starts at birth. Fetuses have been protruding their tongues long before birth and postnatally. One needs, therefore, to address automatic imitation (mimicry) rather than proper imitation. The purpose of tongue-protrusion relates more to food being too hot or unpalatable. For the neonate, tongue-protrusion optimizes food rejection, a process commenced in the womb. Tongue protrusion (generally accompanied by other forms of grimacing and back extending) is an effective means of rejecting food and signaling the caregiver that something is wrong. If neonates and infants also increase tongue protrusions when adults simulate a happy face or finger point, then it is not imitation, but likely excitement at seeing an adult perform an interesting task. Imitation is not an innate behavior but tongue-protrusion is, and their combined response to environmental cues postpartum is learned in a baby's first months. Infants learn to imitate others based on observing others imitate them. Neonates are not born with the ability to copy others; they acquire that skill during the first months of life (cf. Leisman et al. Reference Leisman, Macahdo, Melillo and Mualem2012; Melillo & Leisman Reference Melillo and Leisman2009).

The widely accepted view that newborn infants imitate lacks supporting evidence. Existing data suggest that infants do not imitate until their second year; imitation of different kinds emerges at different ages. The evidence is consistent with a dynamic systems account in which imitation ability is not an inherited, specialized module but is the emergent product of a system of social, cognitive, and motor components, each with its own developmental history.

The necessity of an internal description or representation of a motor response helps explain why imitation of orofacial gestures is such a good candidate for imitation via a mirror system. It is well-known that fetuses perform mouth-opening and closing and tongue protrusion (Longo Reference Longo, Kosobud and Bertenthal2008; Prechtl Reference Prechtl, Wade and Whiting1986). These gesticulations are part of the neonate's behavioral repertoire at birth. Neuroanatomical evidence shows that the corticobulbar tract is already myelinated, innervating mouth and tongue (Sarnat Reference Sarnat2003), enabling automatic elicitation of the observed response. Whereas some claim that these behaviors are reflexive and would not necessitate an internal representation, Lepage & Théoret (Reference Lepage and Théoret2007) noted that imitative behaviors are automatic rather than reflexive and demonstrate that orofacial gestures follow visual and auditory stimuli. A reflexive response, however, would not be elicited by more than one modality. Support for neonatal imitation suggests that infants are more likely to match actions after each has been presented over time (~40 sec), rather than instantly (Anisfeld Reference Anisfeld1991). This finding is more consistent with a mirror system where activation is expected to build up gradually over time as the gesture is modeled, as opposed to explanations claiming that the behavior is merely reflexive. The reflex would be present at birth, but gesture modeling would be built over time.

Data intimate that infant imitation is nonreflexive and possesses a developmental course similar to many primitive reflexes, increasing until approximately 2 months of age, declining and virtually disappearing by 5 months of age (Fontaine Reference Fontaine1984). During this same timeframe, primitive reflexes are gradually inhibited (McGraw Reference McGraw1943), signifying that similar cortical inhibitory processes may suppress spontaneous imitation. As automatic prompting of orofacial gestures becomes suppressed with age, imitation does not vanish. Instead, it becomes subject to voluntary control. This result is seen in adults with frontal lobe lesions with impaired inhibitory control, who display compulsive imitation (Lhermitte et al. Reference Lhermitte, Pillon and Serdaru1986).

One nativistic explanation implies that imitation is a unitary competency – a dedicated behavior having evolved as a unit (Ferrari et al. Reference Ferrari, Rozzi and Fogassi2006a). It has been suggested that neonatal imitation is an inherited, specialized neurological mechanism for imitative behavior in human infants and adults (e.g., Iacoboni & Dapretto Reference Iacoboni and Dapretto2006). A second account of the origins of imitation (cf. Gottlieb Reference Gottlieb2007) suggests that the capacity to match others' behaviors is not present at birth but emerges during the second year continuing to mature over time, with no heritable specialized mechanism. Instead, imitation arises from the infant's attainment of cognitive, motor, and social skills. Differing explanations of the development of imitation are conceivable because the literature provides inconsistent reports that can be variously interpreted.

Neonatal imitation is key in developmental cognition because it asserts a core nativist position for the origins of knowledge. Neonatal imitation is achievable only if infants receive significant awareness of their bodies or agency. It is imperative to ask whether support for this claim is compelling. In newborn imitation experiments, infants are typically exposed to two different behaviors. In a majority of cases, one is tongue-protrusion, the other being mouth-opening. Infants outside of imitation experiments normally produce both behaviors. The literature reports that newborns selectively increase their production of each behavior over baseline after seeing that particular behavior modeled (e.g., Meltzoff & Moore Reference Meltzoff and Moore1983).

Anisfeld (cf. Reference Anisfeld, Hurley and Chater2005) assessed numerous studies of neonatal imitation, which had tongue protrusion as a focal behavior. Only tongue protrusion was consistently matched by newborns in different studies. If neonates imitate only one behavior, then matching may be a consequence of arousal and not of imitation. Anisfeld (Reference Anisfeld1996) and Jones (Reference Jones1996) reported that infants increased rates of tongue protrusion when interested or aroused by stimuli (e.g., flashing colored lights) not resembling human tongue protrusion. Humphrey (Reference Humphrey, Reese and Lipsitt1970) showed that neonates manifested tongue protrusion in response to palm-touching, approaching and receding pens, small balls, and short segments of The Barber of Seville overture (Jones Reference Jones2006a). The results corresponded with patterns in imitation experiments, suggesting that all stimuli are arousing; tongue protrusion is a common response of neonates to numerous arousing stimuli in different sensory modalities.

Therefore, it is likely that newborns' matching of tongue protrusion is not imitation but a manifestation of the infant's concentration in or arousal by the modeler's exhibition of the same behavior. The support for the nativist assertion that newborn infants imitate is not compelling, and we should proceed on the assumption that they do not.