1. Infant imitation is selective and specific

K&A agree with Meltzoff and Moore's findings that if you poke out your tongue at neonates, they will do so in response. However, K&A do not account for our parallel finding that if you show other control gestures – gestures by the same person, at the same distance, moving at the same rate – infants do not respond with tongue protrusion. Experiments making this direct comparison provide a consensus that adult tongue protrusion is the strongest elicitor of the infant tongue-protrusion response (reviews by Anisfeld Reference Anisfeld1991; Meltzoff & Moore Reference Meltzoff and Moore1997; Reference Meltzoff, Moore, Slater and Muir1999b; Nagy et al. Reference Nagy, Pilling, Orvos and Molnar2013; Ray & Heyes Reference Ray and Heyes2011; Simpson et al. Reference Simpson, Murray, Paukner and Ferrari2014a). Motor stereotypies alone cannot account for this selectivity.

2. Infants imitate manual movements as well as facial movements

K&A's theory of the aerodigestive system does not explain the imitation of manual movements. Echoing the first point on selectivity, neonates imitate different types of manual movements (Meltzoff & Moore Reference Meltzoff and Moore1977; Nagy et al. Reference Nagy, Pal and Orvos2014; Vinter Reference Vinter1986). Piaget's (Reference Piaget, Attegno and Hodgson1962) classic book showed accurate imitation of hand movements in 1- to 4-month-olds. K&A “predict” that both Piaget's and the newer findings should not occur in infants this young.

3. Infants imitate from memory

K&A acknowledge that infants often respond after a delay of a few seconds in experiments of imitation. However, they do not discuss how their theory would handle findings based on long-term memory. Meltzoff and Moore (Reference Meltzoff and Moore1994) showed young infants' tongue protrusions on one day and then presented them with a passive face on the next. Infants produced significantly more tongue protrusions when they saw this passive face after a 24-hour delay than did control infants (who did not see the tongue protrusions the day before). This suggests that young infants have a memory for facial actions. Within the first year of life, infants also imitate object-directed acts based on memory (e.g., Klein & Meltzoff Reference Klein and Meltzoff1999; Meltzoff Reference Meltzoff1988). A comprehensive theory of infant imitation needs to explain imitation from memory.

4. Infants can vary their tongue movements prior to 4 months of age

K&A argue that infants have no directed control of their tongues before 4 months of age. All tongue movements are purported to be the stereotypic thrust/retraction involved in suckling. Three empirical findings suggest that infants' competence exceeds this. First, infants imitate different kinds of tongue movements (Meltzoff & Moore Reference Meltzoff and Moore1994). Second, prior to 4 months of age, infants spontaneously produce diverse cooing sounds, which require tongue movements markedly different from suckling and tongue stereotypies. Third, infants imitate speech sounds by 20 weeks of age as documented by phonetic transcriptions and spectrographic analyses (Kuhl & Meltzoff Reference Kuhl and Meltzoff1996). The tongue shapes and positions used for imitating sounds like “ah” and “uu” differ from each other and from tongue stereotypies. Different sounds elicit different lip and tongue movements.

5. Infant imitation does not fit the timetable of motor stereotypies

As Piaget (Reference Piaget, Attegno and Hodgson1962) reported, vocal and manual imitation are expanding, not contracting, at the age that K&A predict a “drop out” of infant imitation based on the motor stereotypies literature. Furthermore, even though infants may not systematically imitate tongue protrusions under one type of experimental design during this time, they do succeed when the design is changed. This suggests a constraint on performance, not a lack of competence. Meltzoff and Moore (Reference Meltzoff and Moore1992) argued that by 12 weeks of age infants have learned expectancies about face-to-face interactions, and they remember past encounters with a person. When a person switches from demonstrating one facial gesture to another, there is response carryover. Accordingly, we conducted a study with a new design: Two distinctly different adults were used to demonstrate gestures to infants, and each modeled a different gesture. In this case, we observed significant imitation of each adult's gesture. Our active intermodal mapping (AIM) theory proposes that infant facial imitation fits within a larger social context involving face processing and the individuation of people (Meltzoff & Moore Reference Meltzoff and Moore1992; Reference Meltzoff and Moore1997; Reference Meltzoff, Moore, Slater and Muir1999b). Although agreeing that the sensorimotor context is important, I maintain that infant memory and the social context must also be considered when explaining developmental change in imitation.

6. The role of prenatal movements

One central aspect of Meltzoff and Moore's (Reference Meltzoff and Moore1997) AIM theory is that infants' prenatal body movements contribute to postnatal imitation. We argued that such activity is a mechanism by which infants build up an “act space” of possible movements of the lips, tongue, hands, and other body parts. Having produced the movements prenatally, infants are in a better position to make sense of the visual stimulus produced by others and map it to their own productions. Meltzoff and Moore (Reference Meltzoff and Moore1997) provided a detailed model for how infants link the perception and production of matching acts. We described the “metric of equivalence” that underwrites infants' mapping of seen body transformations to those they feel themselves produce. Studies using infant neuroscience techniques, such as electroencephalography (EEG), are providing new insights into this process, including the role of somatosensation and infant neural body maps (Saby et al. Reference Saby, Meltzoff and Marshall2013; Reference Saby, Meltzoff and Marshall2015).

The target article concludes by saying that developmental psychologists habitually adopt a simple black/white nature versus nurture distinction. The field of developmental science, however, left behind that dichotomy more than 40 years ago, and has integrated perceptual learning, motor development, neuroscience findings, and both prenatal and postnatal experience into theories of infant development (Marshall & Meltzoff Reference Marshall and Meltzoff2014; Reference Marshall and Meltzoff2015; Meltzoff & Moore Reference Meltzoff and Moore1997; and many others). That is the larger context in which a comprehensive theory of infant imitation should be situated.