During the past decades, research in developmental psychology came along with remarkable reports about the cognitive capabilities of very young children. Numerous studies seem to reveal that young infants master certain tasks targeting these capabilities at a far earlier age than previously assumed. The results are readily interpreted in the same way as with older children: that is, in terms of early or even innate cognitive competence. We will henceforth call them cognitively rich interpretations. In contrast, cognitively lean interpretations of the same findings – that is, interpretations in terms of basic psychological or even physiological principles – attract much less attention, if uttered at all.
Newborns' apparent imitation of others' facial and manual gestures (Meltzoff & Moore Reference Meltzoff and Moore1977) has been a prime example for this trend. On the one hand, it has been interpreted as a process of active matching between others' and one's own actions (Meltzoff & Moore Reference Meltzoff and Moore1977), a primitive form of self-consciousness (Gallagher Reference Gallagher2000), and even as the early basis of intersubjectivity, communication, and social cognition (Meltzoff & Moore Reference Meltzoff, Moore, Nadel and Butterworth1999a). During the past 20 years, these cognitively rich interpretations have attracted much interest, despite a substantial number of failed attempts in replicating the original effects (e.g., Anisfeld Reference Anisfeld1996; Koepke et al. Reference Koepke, Hamm and Legerstee1983). On the other hand, cognitively lean re-interpretations of neonatal imitation – for instance, assuming simple attentional processes or experimental artifacts – were brought forward (e.g., Anisfeld Reference Anisfeld1991; Heyes & Watson Reference Heyes and Watson1981). They were, however, surprisingly rare and are only slowly being perceived by a broader audience (e.g., Oostenbroek et al. Reference Oostenbroek, Slaughter, Nielsen and Suddendorf2013). To our mind, the work by K&A in analysing the physiological mechanisms of neonatal behaviour encourages a general debate about cognitively rich versus lean interpretations. It should be taken as a wake-up call for a stronger consideration of the cognitive and neurophysiological basics of infant behaviour, and it should foster re-interpretations of results for a wide variety of phenomena in early childhood development.
In this commentary, we present three further examples from the field of infants' perception of, and learning from, others' behaviour for which cognitively rich interpretations have been proposed, but for which cognitively lean ones exist as well: rational imitation, theory of mind, and natural pedagogy.
Gergely et al. (Reference Gergely, Bekkering and Király2002) reported selective imitation of new actions in 14-month-olds. Infants were more likely to imitate if the model performed the action deliberately than if the choice of the action could be justified by external constraints in the model's situation. The authors concluded that 14-month-olds are capable to assess the model's situation and her actions under rational aspects. This concept of “rational imitation” in infancy has henceforth received much notice and has provoked many follow-up studies which pick up the paradigm and the cognitively rich interpretation. Again, only few attempts have been made to find cognitively lean interpretations for this phenomenon (e.g., Paulus et al. Reference Paulus, Hunnius, Vissers and Bekkering2011). Our own research showed that selective imitation might be simply elicited by attentional processes (Beisert et al. Reference Beisert, Zmyj, Liepelt, Jung, Prinz and Daum2012). Infants who were habituated to the external constraints in the model's situation imitated just as often as infants in the situation without constraints. We concluded that infants in the original study were distracted by the unusual external constraints and simply did not focus their attention on the relevant aspects of the model's action.
Remarkable results about young infants' perception of others' actions have also been reported by Onishi and Baillargeon (Reference Onishi and Baillargeon2005). For instance, 15-month-olds looked longer if an agent searched for an object not at the place where she had actually placed it before, but at a place where the object, invisibly for the agent yet visibly for the infant, had just moved. According to the authors, 15-month-olds already understand that people may have false beliefs about reality and expect them to act according to these. The conclusion that infants acquire a theory of mind far earlier than previously assumed has attracted wide attention and motivated further studies along this line with still younger infants (e.g., Southgate & Vernetti Reference Southgate and Vernetti2014). On the contrary, Heyes (Reference Heyes2014) presented a cognitively lean account for the results of this and similar experiments. She argued that the sequence of events which the infants witness comes along with changes of physical characteristics like colours, shapes, and movements in the test stimuli. Infants' perception might thus just be modulated by the level of novelty in the test stimuli compared to events encoded earlier in the sequence.
Finally, Csibra and Gergely (Reference Csibra and Gergely2009) proposed that infants are innately biased for learning from benevolent social interaction partners. According to Csibra and Gergely's natural pedagogy theory, ostensive signals inform infants that they should be attentive and prepared for subsequent learning. Ostensive signals are social stimuli which directly address the infant. For instance, the authors showed that 6-month-olds follow the gaze of an adult only if it is preceded by ostensive signals like mutual eye contact or infant-directed speech, but not if it is preceded by nonsocial signals. Recent research, however, indicates that a more basal attentional account is sufficient to substitute this cognitively rich interpretation (Szufnarowska et al. Reference Szufnarowska, Rohlfing, Fawcett and Gredebäck2014). Salient social stimuli – which were not directed at the infants and, therefore, had no ostensive character – enhanced gaze-following as well as ostensive stimuli. It thus seems that social stimuli, irrespective of whether they are directed at the infant or not, are per se attention-grabbing and thus enhance the infants' alertness in the observation of subsequent actions.
A key principle in science is Ockham's razor: in short, the selection of the simplest among multiple competing theories. As depicted by K&A, cognitively lean interpretations are not necessarily simpler than cognitively rich ones as regards the mechanisms they describe. Still, they are simpler as regards the background assumptions which are behind these mechanisms: basal psychological or physiological principles instead of the prerequisite of early or even innate cognitive competence. Roughly half a century after the cognitive revolution, we would like to argue towards turning the tide in developmental psychology and contemplating cognitively lean in place of cognitively rich interpretations of infant behaviour.
During the past decades, research in developmental psychology came along with remarkable reports about the cognitive capabilities of very young children. Numerous studies seem to reveal that young infants master certain tasks targeting these capabilities at a far earlier age than previously assumed. The results are readily interpreted in the same way as with older children: that is, in terms of early or even innate cognitive competence. We will henceforth call them cognitively rich interpretations. In contrast, cognitively lean interpretations of the same findings – that is, interpretations in terms of basic psychological or even physiological principles – attract much less attention, if uttered at all.
Newborns' apparent imitation of others' facial and manual gestures (Meltzoff & Moore Reference Meltzoff and Moore1977) has been a prime example for this trend. On the one hand, it has been interpreted as a process of active matching between others' and one's own actions (Meltzoff & Moore Reference Meltzoff and Moore1977), a primitive form of self-consciousness (Gallagher Reference Gallagher2000), and even as the early basis of intersubjectivity, communication, and social cognition (Meltzoff & Moore Reference Meltzoff, Moore, Nadel and Butterworth1999a). During the past 20 years, these cognitively rich interpretations have attracted much interest, despite a substantial number of failed attempts in replicating the original effects (e.g., Anisfeld Reference Anisfeld1996; Koepke et al. Reference Koepke, Hamm and Legerstee1983). On the other hand, cognitively lean re-interpretations of neonatal imitation – for instance, assuming simple attentional processes or experimental artifacts – were brought forward (e.g., Anisfeld Reference Anisfeld1991; Heyes & Watson Reference Heyes and Watson1981). They were, however, surprisingly rare and are only slowly being perceived by a broader audience (e.g., Oostenbroek et al. Reference Oostenbroek, Slaughter, Nielsen and Suddendorf2013). To our mind, the work by K&A in analysing the physiological mechanisms of neonatal behaviour encourages a general debate about cognitively rich versus lean interpretations. It should be taken as a wake-up call for a stronger consideration of the cognitive and neurophysiological basics of infant behaviour, and it should foster re-interpretations of results for a wide variety of phenomena in early childhood development.
In this commentary, we present three further examples from the field of infants' perception of, and learning from, others' behaviour for which cognitively rich interpretations have been proposed, but for which cognitively lean ones exist as well: rational imitation, theory of mind, and natural pedagogy.
Gergely et al. (Reference Gergely, Bekkering and Király2002) reported selective imitation of new actions in 14-month-olds. Infants were more likely to imitate if the model performed the action deliberately than if the choice of the action could be justified by external constraints in the model's situation. The authors concluded that 14-month-olds are capable to assess the model's situation and her actions under rational aspects. This concept of “rational imitation” in infancy has henceforth received much notice and has provoked many follow-up studies which pick up the paradigm and the cognitively rich interpretation. Again, only few attempts have been made to find cognitively lean interpretations for this phenomenon (e.g., Paulus et al. Reference Paulus, Hunnius, Vissers and Bekkering2011). Our own research showed that selective imitation might be simply elicited by attentional processes (Beisert et al. Reference Beisert, Zmyj, Liepelt, Jung, Prinz and Daum2012). Infants who were habituated to the external constraints in the model's situation imitated just as often as infants in the situation without constraints. We concluded that infants in the original study were distracted by the unusual external constraints and simply did not focus their attention on the relevant aspects of the model's action.
Remarkable results about young infants' perception of others' actions have also been reported by Onishi and Baillargeon (Reference Onishi and Baillargeon2005). For instance, 15-month-olds looked longer if an agent searched for an object not at the place where she had actually placed it before, but at a place where the object, invisibly for the agent yet visibly for the infant, had just moved. According to the authors, 15-month-olds already understand that people may have false beliefs about reality and expect them to act according to these. The conclusion that infants acquire a theory of mind far earlier than previously assumed has attracted wide attention and motivated further studies along this line with still younger infants (e.g., Southgate & Vernetti Reference Southgate and Vernetti2014). On the contrary, Heyes (Reference Heyes2014) presented a cognitively lean account for the results of this and similar experiments. She argued that the sequence of events which the infants witness comes along with changes of physical characteristics like colours, shapes, and movements in the test stimuli. Infants' perception might thus just be modulated by the level of novelty in the test stimuli compared to events encoded earlier in the sequence.
Finally, Csibra and Gergely (Reference Csibra and Gergely2009) proposed that infants are innately biased for learning from benevolent social interaction partners. According to Csibra and Gergely's natural pedagogy theory, ostensive signals inform infants that they should be attentive and prepared for subsequent learning. Ostensive signals are social stimuli which directly address the infant. For instance, the authors showed that 6-month-olds follow the gaze of an adult only if it is preceded by ostensive signals like mutual eye contact or infant-directed speech, but not if it is preceded by nonsocial signals. Recent research, however, indicates that a more basal attentional account is sufficient to substitute this cognitively rich interpretation (Szufnarowska et al. Reference Szufnarowska, Rohlfing, Fawcett and Gredebäck2014). Salient social stimuli – which were not directed at the infants and, therefore, had no ostensive character – enhanced gaze-following as well as ostensive stimuli. It thus seems that social stimuli, irrespective of whether they are directed at the infant or not, are per se attention-grabbing and thus enhance the infants' alertness in the observation of subsequent actions.
A key principle in science is Ockham's razor: in short, the selection of the simplest among multiple competing theories. As depicted by K&A, cognitively lean interpretations are not necessarily simpler than cognitively rich ones as regards the mechanisms they describe. Still, they are simpler as regards the background assumptions which are behind these mechanisms: basal psychological or physiological principles instead of the prerequisite of early or even innate cognitive competence. Roughly half a century after the cognitive revolution, we would like to argue towards turning the tide in developmental psychology and contemplating cognitively lean in place of cognitively rich interpretations of infant behaviour.