Keven & Akins (K&A) proposal challenges decades of received wisdom concerning infant cognitive development. In addition, their case is important because it contains potentially significant methodological and philosophical lessons for cognitive science. Beyond a set of brief remarks concerning its relevance to nativist theorizing in developmental psychology, K&A leave us to guess what other lessons we ought to take away from their contribution. Below, I offer three partially related, but independent, proposals. First, their case suggests why a target behavior ought to be investigated in an ecologically valid way, especially before a hypothesis concerning its functional role has been formulated. Second, it demonstrates how attention to biology is not a peripheral side issue to explanations of psychological phenomena. Finally, their case highlights why alternative explanations, especially those that rely on simpler mechanisms to explain adaptive behavior, have to be ruled out prior to attributing cognitively sophisticated abilities to animals, including humans.
Meltzoff and Moore's central goal was to show that infant imitative competence is present in neonatal infants (1977). In addition, they attempted to explain its presence through postulating a cognitively sophisticated matching process (so-called “active intermodal matching” or AIM) that facilitates imitation via the recognition of the model as “like me” (Meltzoff & Moore Reference Meltzoff, Moore and Rochat1995; Reference Meltzoff and Moore1997; Meltzoff Reference Meltzoff2007). However, their model presumes that the target behavior for analysis (i.e., TP/R) has been correctly identified as imitation behavior. K&A give us good reason to suspect it has not. Why not? One reason is that Meltzoff and Moore investigated TP/R independently of how human infant behavior is environmentally embedded. That is, they failed to attend to the conditions under which TP/R emerges, specifically under its typical pre- and postnatal environmental conditions. Without doing so, it is unclear how they could have properly formulated a hypothesis concerning the appropriate stimulus situation responsible for evoking TP/R. In other words, Meltzoff and Moore did not establish the ecological validity of their experimental design prior to attributing a functional role to TP/R. One of the central lessons from ethology is that, in order to correctly identify the biologically significant role of some behavior, it must be investigated under natural ecological conditions (see Tinbergen Reference Tinbergen1963). In my view, K&A's case against Meltzoff and Moore depends on the notion of ecological validity, but it does so without an explicit discussion of the wider significance of this principle to developmental psychology. Without such a discussion, it remains unclear what broader implications ecological validity (as a methodological precept) has for the science of human cognitive development.
In addition to illustrating the danger of investigating behavior in an ecologically invalid manner, the K&A case against Meltzoff and Moore highlights the risk of attempting to explain the presence and structure of psychological processes in the absence of details concerning how those processes are embodied. Whereas it is true that Meltzoff and Moore (Reference Meltzoff and Moore1997) have postulated a mechanism for imitation in infants, they treat it as if it could be investigated solely in psychological terms, that is, as if psychological explanation is autonomous from biological explanation. This view, articulated and defended by Jerry Fodor (Reference Fodor1974; Reference Fodor1997), holds that psychology and biology are distinct domains of inquiry. This has been a staple of classical cognitive science, and seems to be a view shared by Meltzoff and Moore in practice, if not in principle. In contrast, one of the central claims of the embodied turn in cognitive science is that cognitive capacities are constituted by and “emerge from the recurrent sensorimotor patterns that enable action to be perceptually guided” (Varela et al. Reference Varela, Thompson and Rosch1991, p. 173). This appears to be the context within which K&A's hypothesis ought to be understood, and it looks unlikely that it could have been formulated if they, like Meltzoff and Moore, attempted to explain the presence of TP/R without considering how it unfolds through biological development. Although there is substantial disagreement among proponents of embodied cognitive science as to what its philosophical commitments are, or ought to be (see, e.g., Chemero Reference Chemero2011; Kiverstein & Clark Reference Kiverstein and Clark2009; Shapiro Reference Shapiro2010) there is broad agreement that attention to bodily properties/processes is not dispensable or secondary to characterizing/explaining psychological phenomena. Do K&A agree that their case demonstrates why psychological explanation is not autonomous from biological explanation?
Finally, K&A's hypothesis, if correct, is an example of how what was considered to be evidence of a high-level cognitive process (i.e., imitation) is, in fact, better explained through an appeal to so-called “low-level” mechanisms (i.e., rhythmically stereotyped motor behavior). This is analogous to what has been called a “killjoy explanation” in comparative psychology (Dennett Reference Dennett1983; Shettleworth Reference Shettleworth2010). Such an explanation is “killjoy” in the present context because it tempers the view that cognitively sophisticated mechanisms are required to explain human behavior. It is an appropriate description, here, because Meltzoff and Moore hypothesized that the arousal of TP/R behavior in infants is evidence of a complex cognitive process (i.e., AIM) that is then invoked to explain how infants solve a complex cognitive task (i.e., the correspondence problem). The issue is that if TP/R arousal is the result of the spontaneous activity of a subsystem implicated in aerodigestive sensorimotor development, as K&A argue, then it does not require positing a cognitively sophisticated mechanism to explain. Are K&A committed to the view that developmental psychologists ought to consider killjoy explanations in problem areas other than imitation?
In conclusion, K&A's hypothesis is important because it calls into question decades of theorizing predicated on the conclusion that neonatal infants possess cognitively sophisticated mechanisms for imitation. In addition, K&A's hypothesis is significant even if it fails to completely account for the functional role of TP/R in development, because it shows how attention to biological details matters for the formulation and testing of hypotheses concerning the development of psychological phenomena. In any case, K&A's hypothesis demonstrates how attributing functional significance to behavior is risky if done in an ecologically invalid way, without attention to how it unfolds as a result of embodied processes, and if a sophisticated cognitive mechanism is presumed to be responsible for its presence, when in fact it is not.
Keven & Akins (K&A) proposal challenges decades of received wisdom concerning infant cognitive development. In addition, their case is important because it contains potentially significant methodological and philosophical lessons for cognitive science. Beyond a set of brief remarks concerning its relevance to nativist theorizing in developmental psychology, K&A leave us to guess what other lessons we ought to take away from their contribution. Below, I offer three partially related, but independent, proposals. First, their case suggests why a target behavior ought to be investigated in an ecologically valid way, especially before a hypothesis concerning its functional role has been formulated. Second, it demonstrates how attention to biology is not a peripheral side issue to explanations of psychological phenomena. Finally, their case highlights why alternative explanations, especially those that rely on simpler mechanisms to explain adaptive behavior, have to be ruled out prior to attributing cognitively sophisticated abilities to animals, including humans.
Meltzoff and Moore's central goal was to show that infant imitative competence is present in neonatal infants (1977). In addition, they attempted to explain its presence through postulating a cognitively sophisticated matching process (so-called “active intermodal matching” or AIM) that facilitates imitation via the recognition of the model as “like me” (Meltzoff & Moore Reference Meltzoff, Moore and Rochat1995; Reference Meltzoff and Moore1997; Meltzoff Reference Meltzoff2007). However, their model presumes that the target behavior for analysis (i.e., TP/R) has been correctly identified as imitation behavior. K&A give us good reason to suspect it has not. Why not? One reason is that Meltzoff and Moore investigated TP/R independently of how human infant behavior is environmentally embedded. That is, they failed to attend to the conditions under which TP/R emerges, specifically under its typical pre- and postnatal environmental conditions. Without doing so, it is unclear how they could have properly formulated a hypothesis concerning the appropriate stimulus situation responsible for evoking TP/R. In other words, Meltzoff and Moore did not establish the ecological validity of their experimental design prior to attributing a functional role to TP/R. One of the central lessons from ethology is that, in order to correctly identify the biologically significant role of some behavior, it must be investigated under natural ecological conditions (see Tinbergen Reference Tinbergen1963). In my view, K&A's case against Meltzoff and Moore depends on the notion of ecological validity, but it does so without an explicit discussion of the wider significance of this principle to developmental psychology. Without such a discussion, it remains unclear what broader implications ecological validity (as a methodological precept) has for the science of human cognitive development.
In addition to illustrating the danger of investigating behavior in an ecologically invalid manner, the K&A case against Meltzoff and Moore highlights the risk of attempting to explain the presence and structure of psychological processes in the absence of details concerning how those processes are embodied. Whereas it is true that Meltzoff and Moore (Reference Meltzoff and Moore1997) have postulated a mechanism for imitation in infants, they treat it as if it could be investigated solely in psychological terms, that is, as if psychological explanation is autonomous from biological explanation. This view, articulated and defended by Jerry Fodor (Reference Fodor1974; Reference Fodor1997), holds that psychology and biology are distinct domains of inquiry. This has been a staple of classical cognitive science, and seems to be a view shared by Meltzoff and Moore in practice, if not in principle. In contrast, one of the central claims of the embodied turn in cognitive science is that cognitive capacities are constituted by and “emerge from the recurrent sensorimotor patterns that enable action to be perceptually guided” (Varela et al. Reference Varela, Thompson and Rosch1991, p. 173). This appears to be the context within which K&A's hypothesis ought to be understood, and it looks unlikely that it could have been formulated if they, like Meltzoff and Moore, attempted to explain the presence of TP/R without considering how it unfolds through biological development. Although there is substantial disagreement among proponents of embodied cognitive science as to what its philosophical commitments are, or ought to be (see, e.g., Chemero Reference Chemero2011; Kiverstein & Clark Reference Kiverstein and Clark2009; Shapiro Reference Shapiro2010) there is broad agreement that attention to bodily properties/processes is not dispensable or secondary to characterizing/explaining psychological phenomena. Do K&A agree that their case demonstrates why psychological explanation is not autonomous from biological explanation?
Finally, K&A's hypothesis, if correct, is an example of how what was considered to be evidence of a high-level cognitive process (i.e., imitation) is, in fact, better explained through an appeal to so-called “low-level” mechanisms (i.e., rhythmically stereotyped motor behavior). This is analogous to what has been called a “killjoy explanation” in comparative psychology (Dennett Reference Dennett1983; Shettleworth Reference Shettleworth2010). Such an explanation is “killjoy” in the present context because it tempers the view that cognitively sophisticated mechanisms are required to explain human behavior. It is an appropriate description, here, because Meltzoff and Moore hypothesized that the arousal of TP/R behavior in infants is evidence of a complex cognitive process (i.e., AIM) that is then invoked to explain how infants solve a complex cognitive task (i.e., the correspondence problem). The issue is that if TP/R arousal is the result of the spontaneous activity of a subsystem implicated in aerodigestive sensorimotor development, as K&A argue, then it does not require positing a cognitively sophisticated mechanism to explain. Are K&A committed to the view that developmental psychologists ought to consider killjoy explanations in problem areas other than imitation?
In conclusion, K&A's hypothesis is important because it calls into question decades of theorizing predicated on the conclusion that neonatal infants possess cognitively sophisticated mechanisms for imitation. In addition, K&A's hypothesis is significant even if it fails to completely account for the functional role of TP/R in development, because it shows how attention to biological details matters for the formulation and testing of hypotheses concerning the development of psychological phenomena. In any case, K&A's hypothesis demonstrates how attributing functional significance to behavior is risky if done in an ecologically invalid way, without attention to how it unfolds as a result of embodied processes, and if a sophisticated cognitive mechanism is presumed to be responsible for its presence, when in fact it is not.