In education and early childcare, the media, popular ...

Does research in affective neuroscience justify responsive early childcare? Dr Bruce Maxwell Université du Québec à Trois-Rivières 3351 boul. des Forges, R2061 TROIS-RIVIÈRES G9A 5H7 QUÉBEC CANADA [email protected]

Dr Eric Racine I.R.C.M.

1

In education and early childcare, the media, popular science and some neuroscientists have heralded the promises of ―brain-based‖ interventions and the revolutionary promises of more scientific and objective understandings of learning and development (OECD, 2002; Gura, 2005; Racine et al. 2006; Meltzoff et al., 2009). Although the hope for more enlightened education and child care practices is laudable, the eagerness to translate basic neuroscience research to the educational setting raises important ethical questions, especially given the epistemological weight that neuroscience explanations presently carry in comparison to more traditional psychological, behavior or even folk understandings (Frazzetto & Anker, 2009; Racine & Illes, 2009). Given the strong perennial interest in developmental science among the public and in the media (Nelson & Thompson, 2000), it is no surprise that a body of literature has emerged in recent years that attempts to draw practical conclusions about childcare from the new field of affective and social neuroscience. The purpose of this literature, as characterized by one author, is to inform caregivers ―about the impact of different ways of parenting on a child‘s brain‖ (Sunderland 2006, p. 6). Along with frequent child-caregiver physical contact, nursing on demand, late weaning and cosleeping, ―responsiveness‖ or ―responsive parenting‖1 is an abiding element in a package of early childcare practices almost universally advocated in this literature on the grounds that it has a positive impact on the development of the infant‘s ―social brain‖ (Fonagy et al., 2002; Schore, 2003; Gerhardt, 2004; Sunderland, 2006; Newton, 2008; Leach, 2010). The aim of this paper is to describe and examine critically one claim that has emerged in the translational research on education and childcare: namely, that ―responsive childcare‖ practices significantly benefit children‘s social development and may be a protective factor against developmental problems such as autism, anxiety disorder, attention deficit and hyperactivity disorder (ADHD), and depression. We begin by presenting three examples of the argument advocating responsive parenting on the grounds that it is favourable to the development of a healthy stress response system. Drawing on Shonkoff‘s (2000) schema for classifying the practical significance of scientific information for child policy, we then examine the evidence in favour of this claim. The 1

―Responsive childcare‖ describes prompt, consistent, and sensitive reactions to an infant‘s distress on the part of its caregiver. Viewed as a social process, responsive childcare comprises three elements described in a review published in the Bulletin of the World Health Organization. First, caregivers observe signs of distress or discomfort. Second, they interpret those signs by accurately assigning a proximate cause. Third, they act rapidly, reliably, and appropriately for the sake of addressing the infant‘s immediate needs (Eschel et al., 2006). In this schema, ―appropriateness‖ is the most manifestly value-laden concept and demands special attention. Some common ways that parents and other caregivers respond to infant distress, while likely to be regarded as appropriate by those caregivers themselves, would not be considered appropriate interventions in the sense that the term ―responsive childcare‖ is typically used in the parenting ethics literature. For example, many new parents employ the extinction method of sleep training (i.e., allowing babies to cry themselves to sleep) because they believe that it will help their child learn to fall asleep quickly and without adult intervention. Similarly, some caregivers adhere to a rigid feeding schedule and will delay feeding even though they believe the infant is showing signs of hunger. A parent might do this out of concern for ―spoiling‖ an infant by letting his desires dictate the household schedule or because the parent believes that babies ―need structure‖. From the perspective of the caregivers in these examples both practices are appropriate to the extent that they are consistent with their beliefs about good parenting. From the perspective of responsive childcare as it is normally conceived, by contrast, neither practice is appropriate. Responsive childcare implies prioritizing the immediate relief of infant distress and the infant‘s present comfort in childcare choices.

2

conclusion of our brief overview of this evidence is that, while the conclusiveness and applicability of the key evidence is questionable, the posited link between responsive early childcare and the neurological health of the stress response system remains a reasonable hypothesis that merits further investigation. Notwithstanding this assessment, we argue in the following section that parenting science writers and their publishers have an ethical responsibility to stress the tentativeness of this link. To support this argument, we point to the potential for causing undue distress in families with children affected by severe neuro-psychological troubles and to the need to restore developmental science‘s credibility as a source of knowledge about how early childcare practices may contribute to or frustrate children‘s health and welfare.

The neuroscience argument for responsive early childcare The accounts of the supposedly positive effects of responsive childcare on neurological development to be found in the new translational literature on the neuroscience of parenting are, in terms of their content and scientific rigour, as varied as can be expected from an intellectually diverse group of writers approaching a wide array of scientific research. Whether aimed at professional audience or a non-specialist audience of new parents, a basic argument recurs in the translational literature on the neuroscience of parenting. Simple enough to seem to offer clear practical guidance for infant care, the argument advances that responsive childcare, by helping to maintain low levels of stress hormones in the infant‘s brain, is conducive to the ―normal‖, ―healthy‖ or ―optimal‖ development of the neurological systems involved in stress regulation.2 The argument is discernable, for instance, in Sonia Lupien and colleagues‘ scholarly review article on the deleterious effects of stress on the brain published in Nature Reviews Neuroscience (2009). Humans, they state, give birth to relatively immature young and most of neuroendocrine maturation, including the stress response system, occurs ex utero (2009, p. 438).The authors claim, next, that because of human babies‘ extreme physical vulnerability, they may have a species-specific predisposition to 2

The prior question of what is meant by a ―well regulated stress response system‖ tends to be glossed over in the neuroscience of parenting literature with the assumption that ―psychosocial problems‖ themselves constitute evidence of a disregulated stress response system. This is a potentially problematic conceptualization of healthy stress regulation if only because it would seem to exclude that very troubling emotional experiences (depression or acute anxiety for instance) might, in certain circumstances, be considered appropriate, rational, or ―adaptive‖. Instead of tackling this problem, suffice it for use to state what we mean by this term. Based on our understanding of the phenomenon of brain plasticity, we consider the view that there is a universally ―healthy‖ or ―normal‖ stress response system for humans to be scientifically unsound. The stress response system, like other neurological systems, undergoes structural changes in response to its environment such that the ―adaptiveness‖ of a neurological system is always a function of the environment in which it has to operate. For example, a stressful environment of early life will likely favour the emergence of a stress response system that is adaptive in a stressful life environment but which may be maladaptive in low-stress environments. Hence, in our view, an environment of early care that leads to a ―healthy‖ or ―normal‖ stress response system is one that leads to a stress response system that is well suited to the social environment in which the child will have to live. In light of the complexity of human cultures and the rapidity of societal change, significant mismatches between childrearing practices and the dominant ethos of adult life are possible.

3

anticipate the threat posed by being left unattended by an adult and to become psychologically stressed in the absence of constant care (2009, p. 436). The stress response system is particularly susceptible to the effects of environmental stress through at least the first year of life and possibly well into childhood (2009, p. 437). Lupien and colleagues conclude by throwing down the gauntlet to move to a phase of translational research on responsive early childcare. ―After more than 30 years of research on the negative effects of stress on the brain,‖ they write, ―it is now time to turn our attention to the potential positive impact of early intervention on brain development‖ (2009, p. 442). One intervention that they identify as being an especially salient means of limiting the activation of the stress response in infants is ―sensitive and supporting care‖ (2009, p. 436). A simpler form of the neuroscience and responsive parenting argument appears in the popular parenting writer Penelope Leach‘s latest book, the Essential first year (2010). Without quoting specific studies, Leach claims that neuroscience research has shown that when babies are left alone to cry for long periods the level of the stress hormone cortisol in their brains rises and that repeated exposure to high levels of cortisol permanently alters the neurological systems involved in stress management. Parents, she urges, must understand that babies cry when they are left alone for a reason. Their nervous systems are immature and, because they cannot yet comfort themselves, they need their caregivers‘ help. Responsive childcare and more specifically the avoidance of the extinction method of sleep training, Leach concludes, favours the development of a nervous system which reacts appropriately to life‘s stresses and may protect children against the risk of neuropsychological conditions such as anxiety and depression later in life. A final example of the neuroscience argument for responsive early childcare can be found in Sue Gerhardt‘s book, Why love matters (2004). Gerhardt‘s analysis also focuses on the evidence from developmental neuroscience on the ―corrosive‖ effects of cortisol on infants‘ brains. Summarizing her interpretation of this evidence, Gerhardt writes: ―[e]ssentially, the stress response system is affected by how much early stress it has to deal with, and how well the system is helped to recover. It seems that what you put in to the system is what you get out – a well-resourced and well-regulated infant becomes a child and adult who can regulate himself or herself well, whilst a poorly resourced and poorly regulated infant becomes a child who cannot regulate herself well‖ (2004, p. 66). By permanently structuring the neurological pathways underlying stress regulation, responsive early childcare becomes, for Gerhard as for Leach, a protective factor against anxiety and depression.

Epistemological limitations of evidence cited: conclusiveness and applicability A claim frequently voiced in the neuroscience of parenting literature is that the scientific evidence on the effects of stress on the developing brain justifies policy initiatives, public advocacy work, and the implementation of educational interventions aimed at promoting responsive parenting (Dawson et al., 2000; Barrett, 2006). To what extent are the 4

scientific claims that ground such assertions justified? Before addressing this difficult question let us emphasize, in anticipation of an easy mistunderstanding of our intentions, that we are not concerned with challenging the well established links in the cognitivebehavioural literature between responsive caregiving and various aspects of child health and development, including language, cognitive and psychosocial development (Eschel et al., 2006; Nelson et al., 2007). What we are concerned with, rather, is the strength of the existing evidence in support of a specific claim that is central to the neuroscience argument for responsive parenting: namely, that responsive parenting is a causative factor in the development of well regulated stress response system. The translation of research on human development practice and policy is a highly complex and often politically charged affair. Jack Shonkoff has argued that, in this context, effective translation of the results of research on child development to the realm of policy and practice requires that careful attention be paid to the distinction between established knowledge, reasonable hypotheses, and unwarranted or irresponsible assertions in the field of investigation (Shonkoff, 2000, p. 183). It is the very nature of scientific knowledge about human development, Shonkoff reminds us, to be provisional and limited. The base of established knowledge is in constant evolution. Mirroring calls for translational prudence in the neuroethics literature (Wolpe et al., 2005; Illes & Racine, 2005; Fenton et al., 2009; McEwen, 2002.), Shonkoff advocates a conservative approach to policy not only because it enhances developmental science‘s potential contribution to the advancement of human health and well being but also because, without it, public trust in psychological science‘s role in developing sound child and family policy risks being undermined. ―When pronouncements go beyond the knowledge base, or when unrealistically ambitious impacts are predicted,‖ Shonkoff writes, ―accountability becomes a high stakes enterprise. Achievable promises and a nondefensive acknowledgement of the limits of science are critical safeguards that preserve credibility, strengthen the quest for greater understanding, and ultimately strengthen our ability to enhance human health and well-being‖ (Shonkoff, 2000, p. 187). Due to the nature of the research typically adduced, the work of discerning its practical significance in terms of questions that may be meaningful to ordinary parents and caregivers poses formidable interpretive challenges for both non-experts/interpreters and the experts/scientists doing the research themselves. Furthermore, while metaanalyses like Lupien and colleagues‘ (2009) and Dawson and colleagues‘ (2000) indicate that stress affects various measures of development and on brain systems involved in cognition and associated with mental health, the heterogeneous research sources on which claims about the effects of early stress on the stress response system invariably paint a highly complex picture. Of course, a comprehensive analysis of this research is well beyond the scope of this paper. Our more modest aim in this section is to briefly present two observations about the evidence cited that appear to raise doubts about its conclusiveness and its applicability. Our first observation is that the evidence for three claims that are key to the neuroscience argument for responsive parenting are inconclusive: (1) that chronic high cortisol levels affects behavioral outcomes, (2) that the more stressed the child becomes 5

the higher his corisol levels rise, (3) that the effects of cortisol on the stress-regulatory system is largely irreversible. For example, Lupien and colleagues (2009) cite research which shows that the levels of the stress hormone glucocorticoid of all toddlers in all-day daycare progressively rise during the day. However, increases vary not only according to how ―sensitive‖ the available care is but also according to toddlers‘ individual emotional and behavioural profile. Despite these findings, they assert that ―there is no evidence that the elevated glucocorticoid levels associated with being in day care affect development‖ (2009, p. 436). That is to say, In contrast, several studies on children exposed to severe deprivation or neglect experience have reported hypocortisolism—i.e. chronically low levels of glucocorticoids in chronically stressed children. Finally, and contrary to Gerhardt and Leach‘s more popular versions of the neuroscience and responsive parenting argument, Lupien and colleagues point out that what is known about the longitudinal effects of this neuro-physiological reaction to situations of chronic high stress indicates that the effects of environmental stress on the stress response system are not permanent. ―Sensitive and supportive care‖, they write, ―normalizes [these children‘s] their basal glucocorticoid level after only 10 weeks‖ (2009, p. 437). To sum up, some kinds of experiences associated with high cortisol levels (i.e., attending all-day daycare) appear not to be a risk factor for the emergence of developmental problems, some stressful experiences—especially very stressful ones—do not always lead to raised cortisol levels, and finally there is evidence to suggest that the effects of stressful experiences on stress-regulatory systems are not permanent but may be reversible. Our second observation about the epistemological limitations of the evidence cited in support of the neuroscience argument for responsive parenting concerns it applicability to the context of most families. The biological research concerned with investigating the influence of stress on specific neuro-endocrinal pathways is experimentally contrived and highly technical. A significant portion of it is also animalbased owing to necessary ethical restrictions on manipulating children‘s social environments for the sake of hypothesis testing about the effects of different external stressors on neuroendocrinal mechanisms. The specific animal research that has been key to understanding the effects of stress on the functioning of neuro-endocrinal pathways associated with the mammalian stress-response system (esp. the hypothalamus-pituitaryadrenal or HPA axis) typically involves subjecting newborn rodents and primates to forms of severe maternal deprivation: prolonged or ―repeated, unpredictable separation from the mother, unpredictable maternal feedings, or spontaneous maternal abusive behaviour‖ (Lupien et al. 2009, p. 436). That is to say, hypotheses about the effects of stress on the developing brain are in this body of research addressed negatively, with the key neuroscience research tending to look not at the beneficial effects of responsive infant care but the negative effects of abuse and neglect. Typical relevant human studies identify potentially stressful conditions (all-day day-care, being cared for by a clinically depressed mother, etc.), measure the effects of these conditions on cortisol levels, and extrapolate the developmental effects of high cortisol on the children‘s activity of the HPA axis or other pathway. The underlying biological model of the effects of exposure to

6

high levels of cortisol on the body‘s stress-regulatory system is, again, derived from animal studies (Lupien et al. 2009, p. 436-437).3 In our reading of the situation, two related limitations stand out from this sketch of the evidence. The first is that responsive parenting is not a factor studied in the research linking severe abuse and neglect to poor developmental outcomes. When the implications of this research for parenting practice and policy are considered, the responsive parenting construct plays the role of an interpretive lens through which to assess the research‘s significance. This use of the responsive early childcare construct may be sound if, as interpreters like Sunderland (2006), Gerhard (2004), Newton (2008) and Lupien (2009) seem to assume, the stress induced by being exposed to unresponsive parenting were comparable in kind and in degree to that induced in situations of extreme abuse and neglect. Second, even if one brackets the complexities identified above about generalizing across types of stressful situations, the operative comparison between the circumstance of children or animals exposed to extreme conditions of deprivation, neglect and abuse and those of most human infants seems doubtful. At issue is how well conclusions about the effects of social abuse and neglect on the mammalian stress response system apply to the situations of human infants exposed to varying degrees of ―unresponsive‖ within a normal family environment that is otherwise reasonably predictable, caring, safe and healthy. In our view, the relevance of the evidence drawn from animal studies significantly drops when one considers that the likely audience of the parenting neuroscience literature are ―normal‖ families in this sense. A noteworthy exception to the typical animal studies cited in support of the neuroscience argument for responsive early childcare is Meaney and colleagues‘ research programme on the epigenetic effects of ―high-caring environments‖ on the stress response systems of rat pups (cf. Francis et al., 1999; McGowan et al., 2009).4 What is unique about this research is that responsive early childcare—or at least a closely analogous construct—does appear as a variable in the experimental design. In effect, ―responsive early childcare‖ is tacitly operationalized in terms of naturally occurring 3

A difficult question in this literature is how ―negative effects‖ are operationalized; it is usually done in a multifactorial way that identifies measure that do not easily map onto meaningful concepts for nonscientists. Parents/care givers are interested in knowing socially meaningful effects like ―secure‖, ―calm‖, ―happy‖, etc. Whereas researchers try to assess ―pathophysiological changes‖ by measuring the effects on things like growth rates (weight and height), cognitive performance (learning impairments), ―reduced hippocampal volume‖, ―behavioural disturbances‖, hormone levels in parts of the brain regarded as significant for stress regulation, and ―basal HPA axis activity‖, slower shutdown of the HPA response to acute stress (a sign of a disregulated stress response system). 4 The experiments in question involved isolating female rats with a high frequency of maternal licking and grooming and arched-back nursing and showing that the offspring of these rats had different stress responses as indicated by differential expression of genes involved in stress response (e.g., increase in hippocampal glucocorticoid reception messenger RNA). By doing so, the authors found, first, that whenever a female rat with high frequency of ―caring maternal‖ behaviours took care of pups (biologically hers or sham-adopted animals), the pups were less fearful under conditions of novelty. The study‘s second conclusion, all the more impressive because it stands in tension with a common understanding of genetics, was that the environment of early care mediates hereditary changes in the way genes in brain regions that regulate stress reactivity are expressed. In other words, the effects of ―caring‖ behaviours on brain development—i.e., lower stress reactivity—may be genetically inheritable.

7

variations in patterns of maternal licking and grooming and arched-back nursing on the part of female rats. It is this feature of these protocols that renders interpretive comparisons with human families more suggestive and helps to explain why the Meaney studies are so widely cited by authors making the neuroscience and responsive early childcare argument. In conclusion, considering the limitations of the current research on the effects of environmental stressors on the development of the human stress-regulatory system and of its applicability to the circumstances of most families, we conclude that the posited link between responsive early childcare and more effective responses to environmental stress is at best a reasonable hypothesis in Shonkoff‘s (2000) sense. It warrants further research but, because it oversteps the research on which it is purportedly based, is not at present adequate grounds on which to prescribe the package of care practices comprising ―responsive parenting‖. To be clear, our conclusion stands independently of the body of behavioural-cognitive research on responsive parenting that claims to demonstrate links between responsive parenting and various positive developmental outcomes. One may object that, if this latter body of evidence is conclusive, then our conclusion is inconsequential from the perspective of parental decision making. The details of the scientific argument matter less, perhaps, than the finding that responsive early childcare is grounded in some established knowledge about child development. In reply to this objection, the next and final section of this paper advances that there are foreseeable negative impact of the particular neuroscience justification for responsive parenting on families with children suffering from developmental disorders and on the public trust in developmental science‘s potential contribution to health and well being.

Ethical issues: overestimating the benefits of responsive parenting and restoring public trust in developmental science Although good intentions are undoubtedly behind efforts to advance policy and practice recommendations based on neuroscience research about the effects of stress on brain development, the circumstances of parenting science—and especially those of parenting neuroscience—generate particular obligations for scientists and science writers working in this area. Interpreting the research in a scientifically sound way is only part of the interpretive challenge of drawing reasonable practical conclusions. Given growing evidence on the ―seductive allure of neuroscience explanations‖, and against the background assumption among the reading public that developmental science is often manipulated to promote scientists‘ and science writers‘ personal views about good parenting, translational work in parenting neuroscience demands extra caution in their handling of the relevant scientific information. The rhetorical appeal of neuroscience-based explanations has been observed both in studies of media coverage of neuroscience (Racine et al., 2005; Racine et al., 2010) and in research in social psychology on how non-experts evaluate the evidential strength of neuroscience explanations (Weisenberg et al., 2008; McCabe & Castel, 2008). 8

Weisberg and colleagues‘ (2008) study in particular is cautionary. This study revealed that explanations with superfluous neuroscience information tend to be regarded as being more credible than explanations based only on behavioral-cognitive information. Furthermore, when bad explanations included neuroscience information, non-experts found it more difficult to perceive the weaknesses of those explanations, and were thus more likely to judge a bad explanation as convincing.5 As John Bruer (1998) and others have stressed, the air of scientific credibility that attends neuroscience justifications for educational interventions and family policy is not innocuous hype. It can distract attention from more credible accounts of social problems and end up diverting resources away from more effective strategies for promoting child welfare and positive development (Bruer, 1998; Chugani, 1998; Hinton, 2002). A flagrant example of the impact of the seductive allure of neuroscience explanations on child policy is was the former Georigia governor Zell Miller‘s provision of several million dollars in the 1998 state budget to fund the purchase and delivery of a recording of classical music to for families of all newborn children. The legislation, now widely considered an embarrassment, was based on a hasty interpretation of neuroscience research on benefits of exposure to classical music on cognitive development, or the socalled ―Mozart effect‖ (Rauscher et al., 1993). Against the background of the apparent seductiveness neuroscience explanations, neuroessentialism, and hype for neuropolicy, claims about the potential ―toxicity‖ of stress the developing brain due to insufficiently responsive early childcare are apt to create unwarranted anxiety in parents unless these claims are presented very tentatively. As we have seen, the neuroscience of parenting literature repeatedly advances that responsive parenting constitutes a protective factor against disabling psychodevelopmental disorders like autism-spectrum disorders, anxiety, depression, and ADHD. Faced with such claims, it is imminently foreseeable that parents, typically unaware of the complexity involved in reliably assessing the effect of stress on the developing brain, will overestimate the developmental benefits of responsive parenting. Especially pernicious potential misunderstanding of the link between responsive parenting and healthy development is the overestimation of the developmental risks of ―unresponsive‖ early childcare practices like the extinction method of sleep training and the eschewal of infant-parent co-sleeping. The reaction of well intentioned but misled parents with children suffering from severe developmental disorder is, again, predictable enough: they 5

Another phenomenon that illustrates that neuroscience information sometimes distorts non-experts‘ perceptions of scientific explanations is neuroessentialism, a reductionistis view about the relationship between the brain and the self. In neuroessentialist interpretations, the self becomes equated with or reduced to the brain (Roskies, 2002; Racine et al., 2005). The particular form of neuro-essentialism that is perceptible in the neuroscience and responsive parenting argument is a distorted a view of parental ethics which threatens to reduce early childcare to early ―braincare‖. The neuroscience of parenting literature, that is, seems to tacitly encourage a kind of ―parental neuroessentialism‖ according to which the parent‘s role is to act in the interest of the healthy development of the child‘s brain rather than in the interest of the child as an experiencing self. One particularly telling instance of statement of parental neuroessentialism appears in Margot Sunderland‘s book, when she advises parents that, ―[i]n your ‗brain sculpting‘ role as a parent, you can influence the establishing of strong communicating pathways between your child‘s anterior cingulated and his higher brain. This will enhance his ability to both feel and reflect on another person‘s distress‖ (2006, p. 228).

9

may come to believe that if only they had been more ―responsive‖ their child would today be healthy and normal. This is not to deny that the findings about the effects of stress on brain development suggest important avenues of inquiry that, if pursued, might in future demonstrate that certain parenting practice associated with responsive early childcare do offer long-term benefits for children‘s psycho-social well being. Our view, as advanced in the last section, is rather that the ―unresponsive‖ practices some parents routinely use in caring for infants (the extinction method of sleep training, rigid care scheduling, etc.) is a significant factor in the emergence of some common psychological disorders is at present speculative. Hence, pending more specific knowledge about the effects of the associated stress experiences on brain development, statements such as Penelope Leach‘s to the effect that ―the kind of anxiety induced by the extinction method of sleep training may relate to anxiety right through adult life‖ (2010) are alarmist and, to this extent, constitute ethically irresponsible use of neuroscience information. We turn back now to Shonkoff‘s (2000) plea for a conservative approach to drawing policy and practice implications from scientific information in the interest of public trust. The genre of popular parenting science writing is widely perceived as having helped undermine public trust in the capacity of developmental science to inform parenting practices. Where public trust needs to be restored, we believe, the responsibility falls on science writers to engage with the public forthrightly—not for the sake of intellectual integrity alone but to better promote children‘s health and well being. The genre of translational literature on parenting science is reputed for misusing scientific evidence to lend a false air of legitimacy to what is ultimately authors‘ ―common sense‖ or personal views about how to raise children properly (Barrett, 2006). One of the most poignant examples of this accusation comes in a memorable passage in the most polemical chapter of Jerome Kagan‘s notoriously combative book, Three Seductive Ideas (1998). Here, Kagan argues that the doctrine of infant determinism—i.e., the common belief that an absence of stable, attentive care by loving parents imperils children‘s future well being—is based on little more than an ―ethical judgement as to which maternal behaviours [...] are considered the most virtuous‖ (1998, p. 101). Babies, he concedes, are undoubtedly sensitive to stress. There is no question that their physical and emotional well-being lies at the mercy of caregivers. Common conviction holds, however, that experiences of stress and low quality care in infancy put children at risk for behaviour problems, neuropsychological disorders and frustrating relationships in adulthood. In light of what psychological research has shown about the influence of temperament on personality and human resiliency to even the most aversive childhood experiences, Kagan concludes that any parent who seriously believes that ―their infant‘s future is controlled in a nontrivial way by the events in the first hours, weeks, and months of life‖ (1998, p. 86), have been taken in by scientifically irresponsible psychologists. Kagan‘s critique draws attention forcefully to the way that a scientist‘s prior beliefs about ―good parenting‖ can bias the choice of theoretical frameworks, hypothesis generation, and the process of data collection and interpretation. This is not, of course, to suggest that value-free developmental science is desirable or even possible. Although social scientists, consciously or inadvertently, often attempt to conceal the value10

ladenness of their own work by couching constructs in neutral-sounding descriptors like ―emotionally balanced‖, ―resilient to stress‖, ―secure‖, ―optimal‖ or ―health‖, they can and do bring a personal perspective to their work. The personal perspective has a negative effect on inquiry, however, when it fuels distortions of the scientific enterprise and, most notably, the confirmation bias: ―cherry picking‖ evidence that seems to support one‘s ethical view or accepting supporting evidence without sufficient scrutiny while disregarding or underestimating the import of countervailing evidence. An instance of the likely influence of the personal perspective on the neuroscience argument for responsive parenting comes, in fact, from the very body of evidence to which Kagan is refers in his critique of the doctrine of infant determinism, namely the large literature in psychology on temperament as a factor in personal development. As we saw above, infant determinism is a basic presupposition of the neuroscience of argument for responsive parenting, a theoretical orientation that would place it in the mainstream of theories of child development in modern psychology (Kagan, 1998). Starting in the 1960s, a key scientific development outside psychology—i.e., the discovery of DNA the subsequent rise of genetics—disrupted the broad consensus that building blocks for future personality development are laid in early life experience. Perceiving the potential implications of genetics for understanding personality development, psychologists quickly mounted a campaign against Attachment Theory which was at that time was the dominant school of infant determinism in academic psychology (Karen, 1998). And so it was that the field of behaviour genetics was born, with a meta-theoretical framework inspired by biological science and coinciding with increased public awareness of the influence of genes on health outcomes. Under the aegis of Jerome Kagan‘s longitudinal studies on temperament and Thomas Bouchard‘s (1990) research programme on identical twins, a mass of striking and highly publicized evidence emerged that the impact of early childcare on developmental outcomes was overwhelmed by genetically rooted individual temperaments (cf. overviews in Karen 1998, p. 293ff. and Barrett, 2006, pp. 129ff.).6 Temperament research continues as an active research area in developmental psychology (Barrett 2006) and the general absence of attempts to integrate knowledge about the effects of stress on neurological development with the knowledge on the role of temperament on stress resiliency should be regarded as an additional epistemological limitation of the neuroscience argument for responsive parenting. It is common knowledge that popular writings on childcare for an audience of professionals or new parents typically do little to dispel the public impression that developmental science simply reflects to one or another psychologist‘ or science writer‘s personal prejudices. Less well appreciated, perhaps, are the potentially harmful consequences of the propagation of imprudent interpretations of child development research on children‘s health and well being. In our view, a danger of the scepticism and cynicism towards the possibility of reliable or well established knowledge in child 6

The central issue of the so-called temperament debate of the 1980s, as perplexing as it was emotionally charged, seemed to be the validity of attachment theory‘s key claim that the quality of attachment between an infant and its primary caregiver, as this construct is defined by attachment theory, is the primary determinant of a person‘s personality, as defined by temperamentalists (Kagan, 1998; Karen, 1998; Barrett 2006).

11

development engendered by parenting-advice literature is that these attitudes push the discourse on good parenting into the private sphere by encouraging the subjectivization and personalization of beliefs about parenting. Closing off to public and responsible scientific scrutiny the huge grey zone of traditional human childcare practices that lies between abuse and neglect does no service to children. Since parenting science writers are one of the key points of contact between science and the public, they can render an important public service by helping to restore this trust. With respect to writings on the neuroscience of parenting, this means exploiting the current public fascination with brain science to engage audiences intelligently and honestly about role of neuroscience research in the ongoing task of reviewing our childcare habits and understanding their effects on children. Conclusion This paper presented and examined critically the claim, recurrent in scientific and popular writings on early care and brain development, that ―responsive parenting‖ in infancy offers children significant developmental benefits and may protect them against future neurological troubles such as autism, anxiety disorder, attention deficit and hyperactivity disorder, and depression. Addressing the question of whether research in affective neuroscience justifies responsive early childcare, we tried to show, demands sensitivity to the complex interaction between the epistemological underpinnings of neuroscience research in this area, the practical and policy consequences of early translation of research, and the role that values and perceptions play in informing decisions made by neuroscientists, policy-makers, and parents in this translational effort.

References Barrett, H. 2006. Attachment and the perils of parenting. London: National Family and Parenting Institute. Bouchard, Thomas J., David T. Lykken, Matthew McGue, Nancy L. Segal, & Auke Tellegen. 1990. Science, 250 (4978): 223-229. Bruer, J.T. 1998. The brain and child development: time for some critical thinking. Public Health Reports, 113 (5): 388-398. Chugani, H.T. 1998. Neuroscience and public policy. Public Health Reports, 113(i6): 480. Dawson, G., S.B. Ashman, & L.J. Carver. (2000). The role of early experience in shaping behavioral and brain development and its implications for social policy. Development and Psychopathology, 12: 695-712. Eschel, N., et al., 2006. Responsive parenting: interventions and outcomes. World Health Organization Bulletin, 83 (12): 991-998. Fenton, A., L. Meynell, & F. Baylis. 2009. Responsibility and speculation: On possible applications of pediatric fMRI. Am J Bioeth, 9 (1): W1-2. Fonagy, P., G. Gergely, E. Jurist, & M. Target. 2002. Affect-regulation, mentalization, and the development of the self. New York: Other Press.

12

Francis, D, J. Diorio, Dong Liu, & M. J. Meaney. 1999. Nongenomic Transmission across generations of maternal behavior and stress responses in the rat. Science, 286 (5442): 1155-1158. Frazzetto, G. & S. Anker. 2009. Neuroculture. Nat Rev Neurosci, 10 (11): 815-21. Gerhardt, S. 2004. Why love matters. London: Routledge. Gura, T. 2005. Educational research: Big plans for little brains. Nature, 435 (7046): 11561158. Hinton, V.J. 2002. Ethics of neuroimaging in pediatric development. Brain and Cognition, 50 (3): 455-468. Illes, J. & E. Racine. 2005. Imaging or imagining? A neuroethics challenge informed by genetics. American Journal of Bioethics, 5 (2): 5-18. Kagan, J. 1998. Three seductive ideas. Cambridge, MA: Harvard University Press. Karen, R. 1998. Becoming attached. New York: Oxford University Press. Leach, P. 2010. The essential first year. New York: DK Publishing. Lupien, S., B. S. McEwen, M. R. Gunnmar, & C. Heim. 2009. Effects of stress throughout the lifespan on the brain, behaviour and cognition. Nature, 10: 434445. McCabe, D.P. & A.D. Castel. 2008. Seeing is believing: the effect of brain images on judgments of scientific reasoning. Cognition, 107 (1): 343-52. McEwen, B. 2002. The End of Stress as we Know it. Washington, DC: Dana Press. McGowan, P. O. et al. 2009. Nature Neuroscience 12, 3: 342-348. Meltzoff, A.N., et al. 2009. Foundations for a new science of learning. Science, 325 (5938): 284-8. Nelson, C.A. et al. 2007. Cognitive recovery in socially deprived young children: the Budapest Early Intervention Project. Science, 318: 1937-1940. Newton, R.P. 2008. The attachment connection. Oakland, CA: New Harbinger Publications. OECD. 2002. Understanding the brain: Toward a new learning science, Vol. 1. Paris: OECD Publications. 115. Racine, E., O. Bar-Ilan, & J. Illes. 2005. fMRI in the public eye. Nature Reviews Neuroscience. 6 (2): 159-164. Racine, E., O. Bar-Ilan, & J. Illes. 2006, Brain imaging: A decade of coverage in the print media. Science Communication, 28 (1): 122-142. Racine, E. & J. Illes. 2009. "Emergentism" at the crossroads of philosophy, neurotechnology, and the enhancement debate, in Handbook of Philosophy and Neuroscience, J. Bickle (ed.). New York: Oxford University Press. Racine, E., et al. 2010. Contemporary neuroscience in the media. Soc Sci Med, 71 (4): 725-33. Rauscher, F., G. Shaw, & K. Ky, 1993. Music and spatial task performance. Nature, 365: 611. Roskies, A. 2002. Neuroethics for the new millenium. Neuron, 35 (1): 21-3. Schore, A. 2003. Affect dysregulation and disorders of the self. New York: Norton. Shonkoff, J.P. 2000. Science, policy, and practice. Child Development, 71 (1): 181-187. Sunderland, M. 2006. The science of parenting. New York: DK Publishing. Thompson, R. & C.A. Nelson. 2001. Developmental science and the media. American Psychologist, 56 (1): 5-15. 13

Weisberg, D.S., et al. 2008. The seductive allure of neuroscience explanations. J Cogn Neurosci, 20 (3): 470-7. Wolpe, P.R., K.R. Foster, & D.D. Langleben. 2005. Emerging neurotechnologies for liedetection: Promises and perils. American Journal of Bioethics, 5 (2): 39-49.

14