Charney's target article is an important synthesis of work that has led to calls for a paradigm shift beyond the modern synthesis underpinning evolutionary, genetic, and developmental sciences (Jablonka & Lamb Reference Jablonka and Lamb2007; Mattick Reference Mattick2009; Müller Reference Müller2007) and furthers a crucial critique of genetic “blueprint” models of development – which can be traced back as far as the work of Schneirla (Reference Schneirla and Harris1957) and Lehrman (Reference Lehrman, Aronson, Tobach, Lehrman and Rosenblatt1970). At stake in this debate are assumptions underpinning the dominant models of child clinical psychological and preventative research that seeks to improve developmental outcomes by identifying and reducing environment risk factors. The focus of my commentary is to suggest areas where Charney's model can be further extended.
Although the focus of the target article is on implications for behavioural genetics, a model that includes a role for the intergenerational transmission of environmental influences via the epigenome could have major implications for an emerging developmental science by integrating findings from the genetic, epigenetic, and psychosocial domains. However, one of the problems occurs in the translation of the language of maternal effects (Mousseau & Fox Reference Mousseau and Fox1998) – derived from evo-devo research – into human developmental psychology. To argue that maternal effects do reliably shape development in adaptive ways and that this applies as much to human development as other organisms, Charney argues that “maternal effects … are phenotypic concordance-producing effects.” He appears to be stressing this point to counter Plomin (Reference Plomin2011), who argues that variance in phenotypic outcome in children is derived from mostly genetic or non-shared environment. This is where it is a real pity that Charney did not consider research findings on infant attachment in humans. Maternal effects on offspring phenotypes can be measured in terms of the attachment patterns in infants, which can differ among offspring of the same mother – and so too an infant can develop different attachment patterns for mother and father. This suggests that, despite sharing the same family environment, the attachment relationship is unique to each child–caregiver dyad. Against the trend of many findings in child development, behavioural genetic studies of concordance of attachment status in twin studies suggest that differences in attachment patterns have very minimal genetic contribution but are largely due to both shared and non-shared environmental factors. An important distinction derived from such research is that while family environments may be shared by children – and this includes the sharing of maternal effects – the translation of this concept into human social ecologies requires consideration of the relationships among children, parents, and siblings. Attachment findings seem to suggest that each child's developmental context is unique at the level of the relationship it has with available caregivers. This more complex social network influences the child's developing biology via the emergence of a sense of self that comes to mediate the experience of that environment (Fonagy et al. Reference Fonagy, Gergely and Target2007). The notable absence of any consideration of attachment theory and research in Charney's target article forecloses an opportunity to consider a field in which maternal sensitivity has been well measured in humans and so too development of psychopathology as a function of disturbance in the social environment (Lewis & Tooley Reference Lewis and Tooley2009).
The concept of “adaptive phenotypic plasticity” is insufficiently elaborated upon in Charney's target article, and in the end, the concept may be inadequate for an epigenomic paradigm of inheritance. First, with respect to the notion of phenotypic plasticity, it is important to note not only how shaping effects occur from the “maternal perinatal environment” to offspring, but also how an offspring actively shapes its environment. Such maternal programming has an effect because of a certain mode of responsivity to very specific cues in the environment. Learning is a prime example of such responsivity. Rapid learning in the form of visual or olfactory imprinting is particularly relevant to recognition and proximity seeking of offspring to caregiver, such that an offspring may benefit from maternal investment – particularly in mammals. While learning suggests a degree of targeted receptivity to environmental signals, infants also show a range of seemingly highly preprogrammed behaviours. Infant attachment behaviours – such as crying, clinging, and proximity seeking – appear to be designed to elicit maternal care and investment. There can be variation in both maternal sensitivity and infant attachment behaviour, as well as feedback loops between the two. These species-specific signalling functions, which serve to elicit parental investment, can be manipulated by both parent and offspring so as to elicit additional care or parental investment – leading to parent–offspring conflict within the maternal shaping of development (Trivers Reference Trivers1974). Instead of this complex and systemic interaction, Charney's model of development tends to assume a degree of passivity of the offspring in its shaping by the maternal environment.
Second, the notion of adaptation implied in the concept of adaptive phenotypic plasticity quickly shows its limitations. The traditional, and Neo-Darwinian, concept of adaptation is so heavily based on the model of genetics and inheritance forged in the modern synthesis that, without substantial reconsideration, it will prove inadequate to accommodate epigenomic heritability. Certainly genetic change at a population level is slow to react to environmental change, so it makes sense that more fine-grained modifications of the biological systems underpinning phenotypic outcomes are responsive to environmentally derived programming effects. Nevertheless, Charney's model of adaptation seems to be a “fit to the environment,” which has been criticised some time ago by Lewontin, suggesting instead that organisms co-construct and co-define their ecological niche, and this becomes increasingly the case as social ecologies take on greater complexity (Lewontin Reference Lewontin, Oyama, Griffiths and D Gray2001; Griffiths & Gray Reference Griffiths and Gray2005). Both an infant and its (maternal) environment undergo their own developmental processes that require systemic interaction. So too it is necessary to rethink the notion of heritability because, in effect, an offspring inherits genetic information, but so too epigenetic information and interacts with an environment provided by other dynamic and developing organic organisms – to which it is often genetically related. Not only the concepts of inheritance and environment, but also adaptation and development, will need to be transformed.
Charney's target article is an important synthesis of work that has led to calls for a paradigm shift beyond the modern synthesis underpinning evolutionary, genetic, and developmental sciences (Jablonka & Lamb Reference Jablonka and Lamb2007; Mattick Reference Mattick2009; Müller Reference Müller2007) and furthers a crucial critique of genetic “blueprint” models of development – which can be traced back as far as the work of Schneirla (Reference Schneirla and Harris1957) and Lehrman (Reference Lehrman, Aronson, Tobach, Lehrman and Rosenblatt1970). At stake in this debate are assumptions underpinning the dominant models of child clinical psychological and preventative research that seeks to improve developmental outcomes by identifying and reducing environment risk factors. The focus of my commentary is to suggest areas where Charney's model can be further extended.
Although the focus of the target article is on implications for behavioural genetics, a model that includes a role for the intergenerational transmission of environmental influences via the epigenome could have major implications for an emerging developmental science by integrating findings from the genetic, epigenetic, and psychosocial domains. However, one of the problems occurs in the translation of the language of maternal effects (Mousseau & Fox Reference Mousseau and Fox1998) – derived from evo-devo research – into human developmental psychology. To argue that maternal effects do reliably shape development in adaptive ways and that this applies as much to human development as other organisms, Charney argues that “maternal effects … are phenotypic concordance-producing effects.” He appears to be stressing this point to counter Plomin (Reference Plomin2011), who argues that variance in phenotypic outcome in children is derived from mostly genetic or non-shared environment. This is where it is a real pity that Charney did not consider research findings on infant attachment in humans. Maternal effects on offspring phenotypes can be measured in terms of the attachment patterns in infants, which can differ among offspring of the same mother – and so too an infant can develop different attachment patterns for mother and father. This suggests that, despite sharing the same family environment, the attachment relationship is unique to each child–caregiver dyad. Against the trend of many findings in child development, behavioural genetic studies of concordance of attachment status in twin studies suggest that differences in attachment patterns have very minimal genetic contribution but are largely due to both shared and non-shared environmental factors. An important distinction derived from such research is that while family environments may be shared by children – and this includes the sharing of maternal effects – the translation of this concept into human social ecologies requires consideration of the relationships among children, parents, and siblings. Attachment findings seem to suggest that each child's developmental context is unique at the level of the relationship it has with available caregivers. This more complex social network influences the child's developing biology via the emergence of a sense of self that comes to mediate the experience of that environment (Fonagy et al. Reference Fonagy, Gergely and Target2007). The notable absence of any consideration of attachment theory and research in Charney's target article forecloses an opportunity to consider a field in which maternal sensitivity has been well measured in humans and so too development of psychopathology as a function of disturbance in the social environment (Lewis & Tooley Reference Lewis and Tooley2009).
The concept of “adaptive phenotypic plasticity” is insufficiently elaborated upon in Charney's target article, and in the end, the concept may be inadequate for an epigenomic paradigm of inheritance. First, with respect to the notion of phenotypic plasticity, it is important to note not only how shaping effects occur from the “maternal perinatal environment” to offspring, but also how an offspring actively shapes its environment. Such maternal programming has an effect because of a certain mode of responsivity to very specific cues in the environment. Learning is a prime example of such responsivity. Rapid learning in the form of visual or olfactory imprinting is particularly relevant to recognition and proximity seeking of offspring to caregiver, such that an offspring may benefit from maternal investment – particularly in mammals. While learning suggests a degree of targeted receptivity to environmental signals, infants also show a range of seemingly highly preprogrammed behaviours. Infant attachment behaviours – such as crying, clinging, and proximity seeking – appear to be designed to elicit maternal care and investment. There can be variation in both maternal sensitivity and infant attachment behaviour, as well as feedback loops between the two. These species-specific signalling functions, which serve to elicit parental investment, can be manipulated by both parent and offspring so as to elicit additional care or parental investment – leading to parent–offspring conflict within the maternal shaping of development (Trivers Reference Trivers1974). Instead of this complex and systemic interaction, Charney's model of development tends to assume a degree of passivity of the offspring in its shaping by the maternal environment.
Second, the notion of adaptation implied in the concept of adaptive phenotypic plasticity quickly shows its limitations. The traditional, and Neo-Darwinian, concept of adaptation is so heavily based on the model of genetics and inheritance forged in the modern synthesis that, without substantial reconsideration, it will prove inadequate to accommodate epigenomic heritability. Certainly genetic change at a population level is slow to react to environmental change, so it makes sense that more fine-grained modifications of the biological systems underpinning phenotypic outcomes are responsive to environmentally derived programming effects. Nevertheless, Charney's model of adaptation seems to be a “fit to the environment,” which has been criticised some time ago by Lewontin, suggesting instead that organisms co-construct and co-define their ecological niche, and this becomes increasingly the case as social ecologies take on greater complexity (Lewontin Reference Lewontin, Oyama, Griffiths and D Gray2001; Griffiths & Gray Reference Griffiths and Gray2005). Both an infant and its (maternal) environment undergo their own developmental processes that require systemic interaction. So too it is necessary to rethink the notion of heritability because, in effect, an offspring inherits genetic information, but so too epigenetic information and interacts with an environment provided by other dynamic and developing organic organisms – to which it is often genetically related. Not only the concepts of inheritance and environment, but also adaptation and development, will need to be transformed.