In their article, Keven & Akins (K&A) describe newborns' tongue protrusion and retraction (TP/R) as an early emerging orofacial stereotypy that serves to facilitate the activity-dependent development of core aerodigestive functions. This view suggests that TP/R is not an imitative response and emphasizes that development is continuous from the prenatal to the postnatal period (Einspieler et al. Reference Einspieler, Marschik and Prechtl2008; Prechtl Reference Prechtl1984). We agree with this view but would like to comment on the terminology of “stereotypies” used in the target article and discuss additional considerations regarding the learning opportunities initiated by neonatal behavior in a social context.

First, referring to TP/R movements as stereotypies is misleading. This terminology is commonly associated with behaviors that lack variability and may be a sign of, or predictor for, various developmental disorders (Goldman et al. Reference Goldman, Wang, Salgado, Greene, Kim and Rapin2009; Matson et al. Reference Matson, Dempsey and Fodstad2009). In contrast, TP/R movements show considerable variability and have a clear developmental relevance as discussed by K&A. Consequently, TP/R movements might not bear characteristics of motor stereotypies from a neurodevelopmental or clinical perspective. Therefore, we propose referring to TP/R more generally as “movements” or “behaviors” to avoid a potential misunderstanding.

Second and most importantly, although behaviors driven by endogenous central pattern generators (CPGs) are critical for development, the exogenous social consequences following neonatal movements need to be considered as well. Infants do not develop in a vacuum but are born into a rich social environment that responds dynamically to their actions. In fact, maternal engagement is synchronized and strongly influenced by the newborns' behaviors (Feldman Reference Feldman2007). Specific behaviors of the mother (e.g., “motherese,” face-to-face exchanges, affectionate touch) are genetically determined but also dynamically adjust to environmental factors (Keller Reference Keller2003; Leckman et al. Reference Leckman, Feldman, Swain, Eicher, Thompson and Mayes2004). For example, when responding to infant's cooing, mothers not only heighten their fundamental frequency, but also introduce pauses that facilitate learning about turn-taking in communication (e.g., Stevenson et al. Reference Stevenson, Ver Hoeve, Roach and Leavitt1986). These examples demonstrate that the child's (motor) behavior affects how the parent interacts with the child. These changes result in “developmental cascades” (Masten & Cicchetti Reference Masten and Cicchetti2010) where the child's behavior triggers a new or different response in the parent that may then facilitate subsequent development across domains.

Evidence for such developmental cascades also comes from older infants: Mothers offer different verbal feedback to sharing bids from crawling infants than to sharing bids from walking infants (Karasik et al. Reference Karasik, Tamis-LeMonda and Adolph2011; Reference Karasik, Tamis-LeMonda and Adolph2014). Therefore, we believe that a caregiver's observations of neonatal movements (including TP/R and other CPG behaviors) influence the caregiver's responses and interactions with the child. Neonatal behaviors can initiate developmental cascades by triggering a certain response in the caregiver (such as a vocalization or imitation by the caregiver) and consequently serve a social function, even if this was not the “intention” of the child. Put differently, although the newborn may not “imitate” the parent, what matters more is how the parent reacts and responds to the child's behavior.

Finally, we would like to also note that infants (including newborns) are sensitive to contingencies present in their environment and adjust their own behaviors accordingly. For example, newborns adjust their sucking rate (an aerodigestive behavior just like TP/R) in order to elicit a preferred stimulus such as a speech sound or a stimulus that is contingent on their own sucking rate (DeCasper & Carstens Reference DeCasper and Carstens1981; DeCasper & Fifer Reference DeCasper and Fifer1980; Floccia et al. Reference Floccia, Nazzi and Bertoncini2000; Vouloumanos & Werker Reference Vouloumanos and Werker2007). Older infants learn the contingencies between movements of their limbs such as arms or legs and changes in their environment (Needham et al. Reference Needham, Barrett and Peterman2002; Rovee & Rovee Reference Rovee and Rovee1969). In these examples, the infants' initial motor activity (either sucking or limb movements) is not aimed at eliciting a certain outcome. However, the repeated exposure to contingencies between motor behavior and environmental response leads the infant to increase the motor activity in order to re-elicit the response (e.g., kick their legs more to make a mobile hanging over a crib move more). Consequently, it is possible that newborns may be able to also adjust their rate of TP/R behaviors in response to observed behaviors of their parent. Together, these two mechanisms can act as a positive feedback loop: Newborns engage in TP/R, parents notice and respond, newborns increase TP/R behavior, parents respond more, and so on.

Consequently, TP/R behaviors may well have a secondary impact on socio-cognitive development during dynamic and contingent parent-child interactions. One example for how motor activity can affect socio-cognitive development via similar feedback loops and cascades is pointing. The motor pattern underlying pointing (i.e., extending the index finger) is present already before birth (e.g., Einspieler et al. Reference Einspieler, Prayer and Prechtl2012; Marschik et al. Reference Marschik, Prechtl, Prayer, Peyton and Einspieler2013). However, the use of pointing for communicative purposes emerges only around 9 to 14 months of age (Liszkowski et al. Reference Liszkowski, Brown, Callaghan, Takada and de Vos2012). Before the infant uses pointing to communicate, parents inevitably perceive pointing as a communicative signal from the infant and respond to the pointing bid. Over repeated observations of the caregivers' reaction to pointing, infants eventually learn to use pointing communicatively (Bates et al. Reference Bates, Camaioni and Volterra1975; Carpendale & Carpendale Reference Carpendale and Carpendale2010; Vygotsky Reference Vygotsky1978). Following the same logic, TP/R does not need to be intended as imitation by the newborn, but may provide the newborn with opportunities to learn about imitation through observation of the parent's response to this repeating behavior.

In summary, the mechanisms of the developing system are highly interconnected. One behavior, such as TP/R may exist to serve a particular biological purpose. However, during the dynamic exchanges with the environment this behavior will inevitably be used for other purposes as well. In particular, socio-cognitive functions such as communication and language are secondary functions to respiration and orofacial functions for food intake. A comprehensive account of infant development needs to consider that seemingly simple behaviors such as TP/R may serve such a dual purpose: endogenously generated at first and at the same time facilitating socio-cognitive development via social interactions and responses from the environment.