Do Newborns Have the Ability to Imitate? We can now rule out some long-standing explanations for why the effect might be difficult to detect, only some research groups observe it, the published literature is biased

Do Newborns Have the Ability to Imitate? Virginia Slaughter. Trends in Cognitive Sciences, Vol 25, Issue 5, pp 377-387, March 13, 2021. https://doi.org/10.1016/j.tics.2021.02.006

Highlights

Although many assume that newborn infants imitate others, new data and analyses suggest it is not a reliable effect.

Meta-analysis of human neonatal imitation studies revealed an overall positive effect which was not moderated by major methodological variations, but did vary by research group.

Future studies of newborn imitation should adopt modern procedures to eliminate potential biases.

Researchers should also test models of how imitation could be learned. Associative Sequence Learning proposes that coincident experience of producing and perceiving body gestures over the first year of life, creates mirror systems to support imitation.

Neonatal imitation is widely accepted as fact and cited as evidence of an inborn mirror neuron system that underpins human social behaviour, even though its existence has been debated for decades. The possibility that newborns do not imitate was reinvigorated recently by powerful longitudinal data and novel analyses. Although the evidence is still mixed, recent research progresses the debate by ruling out some long-standing explanations for why the effect might be difficult to detect, by showing that only some research groups observe it, and by revealing indications that the published literature is biased. Further advances will be made with updated testing procedures and reporting standards, and investigation of new research questions such as how infants could learn to imitate.

Concluding Remarks

When considering all the evidence, it is hard to maintain the conviction that newborns imitate. However, there is still no definitive answer to the question. To find an answer, several things need to happen. Researchers should replace the 40-year-old methodology that has nurtured this controversy with modern approaches to data collection, analysis and reporting. It is also time to ask new questions: are newborns physically capable of imitating (Box 2)? To what extent do adults imitate babies in everyday interactions? Can imitation be trained in the first months of life? It also would be helpful if authors writing about newborns, imitation or mirror neurons, acknowledged the ongoing controversy rather than treating neonatal imitation as a fact.

Box 2

Do Newborns Have Voluntary Control of Oral–Facial Movements?

Putting aside the controversial evidence for and against newborn imitation, Keven and Akins [

57.

] considered whether newborn infants’ sensory–motor and brain development enables imitation. Their detailed analysis drawing on modern accounts of early neuromotor development suggested that imitation of mouth gestures, and tongue protrusion in particular, is beyond the capacity of neonates whose nervous systems are still adapting to postpartum life. It takes months for the newborn brain to coordinate breathing with sucking and swallowing liquids. As these functions mature, newborns engage in repetitive oral activity, including a lot of mouth opening and closing, and tongue protrusion and retraction. These patterned behaviours are involuntary, driven by subcortical brain mechanisms, and increase when newborns are aroused.

Keven and Akins’ analysis suggests that voluntary production of oral–facial movements in response to a matching model is physically impossible in newborns, because these gestures are generated by the brain’s subcortex in the first months of life. This would mean that so-called imitation in the newborn period is simply coincidental, since neonates’ involuntary mouth movements increase in response to the arousing sight of an experimenter’s animated face. This conclusion has been offered previously, based on observations that newborns increase their tongue protrusions in response to various arousing stimuli including light displays and orchestral music [

6.

]. Keven and Akins’ analysis also addresses the claim that newborn imitation fades out around 2–3 months of age: at that stage, the ‘wiring up’ of sucking, swallowing, and respiration functions are complete, so infants no longer involuntarily produce oral–facial movements in response to arousing stimuli.

Keven and Akins’ analysis demonstrates the value of asking different questions in relation to newborn imitation, rather than just focusing on ‘do they, or don’t they?’ However, their analysis was itself controversial, as evident in the open peer commentaries accompanying their article. For instance, some commentators dismissed the analysis as irrelevant to newborn imitation, relying on claims that a range of facial and manual gestures, in addition to tongue protrusion and mouth opening, are imitated by neonates [

58.

,

59.

]. Others accepted that newborns’ oral–facial movements promote maturation of sucking and swallowing but argued that they simultaneously function as communicative signals in face-to-face interactions with caregivers [

60.

,

61.

].

There is no doubt that imitation is a central element of human development. Indeed, there is a vast literature documenting what children imitate, from whom, and under what circumstances [

55.

]. However, we still do not know when or how this ubiquitous behaviour emerges, which means that we do not truly understand it (see Outstanding Questions). If imitation turns out to be learned rather than inborn, this would not diminish the theoretical significance of imitation for infant–parent bonding, social learning or later-developing interpersonal skills. Rather, it would highlight the brain’s proclivity to create connections between ourselves and others, from the first months of life.

Outstanding Questions

Is there evidence of imitation when newborns are tested with objective measures such as EMG?

Does video modelling of gestures genuinely increase newborns’ imitative response as suggested by the meta-analysis? If so, why?

What accounts for the meta-analytic finding that neonatal imitation varies by research group?

How frequently do infants experience observation–action correspondences during a typical day and is this variable related to production of imitation?

Are different types of observation–action correspondence (e.g., self-observation, mirror exposure, and caregiver mimicry) related to different forms of imitation in infancy?

Can imitation be promoted in infants with observation-action training as predicted by ASL?