back to Bowlby, briefly

John Bowlby

Two names that keep cropping up in the child protection literature (apart from Bowlby) are those of Bruce Perry and Allan Schore. Perry is a recognised expert in child mental health and is especially interested in the effects of trauma. Schore has been described as ‘the American Bowlby’ and as ‘the world’s leading expert in neuropsychoanalysis’. I want to focus on one paper by each of them. I’ve chosen the Perry paper because it tackles the nature/nurture debate head-on and is the source of the brain photograph on the cover of Graham Allen’s report Early Intervention: the Next Steps. The Schore paper “Attachment and the regulation of the right brain” seeks to map Bowlby’s attachment theory onto recently acquired knowledge about how the brain develops, and a shortened version of it forms the foreword to a recent edition of Bowlby’s classic book Attachment. Attachment is the first of three volumes dealing with attachment; the others are entitled Separation and Loss.

The frequent mention of the same few names isn’t unusual in a specialist field, and what I expected to find when I read Perry’s and Schore’s work was the painstaking, step-by-step hypothesis testing typical of researchers working in a little-explored area. That wasn’t quite what I found. Before tackling either of the papers, I need to re-visit Bowlby’s attachment theory because Schore’s paper starts where Bowlby leaves off, and Perry’s paper opens with a discussion about the evolution of human behavioural characteristics – a topic central to Bowlby’s thesis.

The title of this post refers to a brief return to Bowlby’s theory, not to the length of the post. I’ve evaluated Bowlby’s ideas in some detail because later interpretations of his ideas are many and varied and sometimes haven’t taken into account Bowlby’s often meticulous reasoning.

Schore and Bowlby

Schore begins by summarising Bowlby’s view of the biological systems underpinning attachment, and then shows how recent findings about the development of regulator systems in the brain’s right hemisphere support Bowlby’s predictions. In the abstract to his paper Schore summarises Bowlby’s model of attachment like this:

“… attachment is instinctive behavior with a biological function, that emotional processes lie at the foundation of a model of instinctive behavior, and that a biological control system in the brain regulates affectively driven instinctive behavior”.

The first thing that struck me about Schore’s paper is that he doesn’t define terms such as attachment, instinct and emotion. Biologically speaking, none of these constructs is straightforward and Bowlby discusses their definitions at length. It’s a pity Schore doesn’t explore these terms, because re-reading Bowlby’s explanations half a century after publication (revisions in the second edition of the book in 1983 notwithstanding) suggests that there are some implicit assumptions in his thinking that are open to question. And anyone unfamiliar with either Bowlby or the animal behaviour literature might be unaware of the complexity of the issues involved.

Let’s start with some of the key concepts that underpin Bowlby’s theory of attachment – first, psychoanalysis.

Psychoanalysis

Bowlby opens Attachment with a chapter called “Point of View” in which he explains the basis for his model. He begins with a quote from Freud, and takes as his starting point an underlying principle of psychoanalysis; that personality, both healthy and pathological, is shaped by events that occur in childhood. But Bowlby differs from Freud on several key points;

Despite these differences, Bowlby assumes that personality (one construct that he doesn’t attempt to define) is shaped by childhood events and that pathologies of personality originate in traumatic experiences in childhood, observations suggesting that a significant trauma is separation from, or loss of, the mother. In the two later volumes of his Attachment trilogy Bowlby suggests that separation is an underlying cause of anxiety and that loss is a cause of sadness and depression. Bowlby points out that his model’s underlying mechanism involves biological processes rather than psychical energy. A central biological process is evolution.

Evolution

Bowlby’s theory is rooted firmly in the Darwinian model of evolution. Darwin’s theory is based on two main concepts, inherited characteristics and natural selection. Neither of these concepts was new when Darwin proposed his theory. People had known for millennia that offspring inherit physical and behavioural characteristics from their parents and that plants and animals thrive in a particular environment only if their characteristics are well adapted to it. Farmers had been using artificial selection for centuries to breed characteristics into or out of plants and animals to suit a specific purpose or environment. What was new about Darwin’s theory was the idea that environmental factors acting on inherited characteristics could result in changes to a species, or in new species arising. What was missing from his theory was the mechanism by which characteristics were passed on to subsequent generations. Although DNA had been discovered in the mid-19th century, its structure and function were unknown and it wasn’t until the turn of the 20th century and the re-discovery of Gregor Mendel’s work showing that characteristics were inherited via discrete units of heritability, that the concept of genes was developed.

Genes

Bowlby sees ‘instinctive’ behaviours like attachment as emerging from the interaction between genetic endowment and environment, but essentially limits his comments on genes to the following paragraph:

The basic concept of the genetical theory of natural selection is that the unit central to the whole process is the individual gene and that all evolutionary change is due to the fact that certain genes increase in number over time whereas alternative genes decrease or die out. What this means in practice is that, through the process of differential breeding success, individuals that are carrying certain genes increase in numbers whilst individuals that are carrying others diminish. A corollary is that the adaptedness of any particular organism comes to be defined in terms of its ability to contribute more than the average number of genes to future generations. Not only, therefore, does it have to be designed so that it is capable of individual survival but so that it is capable also of promoting the survival of the genes it is carrying. This is commonly done through reproducing and promoting the survival of offspring.” (pp.55-56)

Although what Bowlby says is true at one level, his comments don’t give any indication of the complexity of genetic variation in a population. This is despite de novo (spontaneous) mutations having been discovered in fruit flies by the 1920s, and by the time Bowlby published Attachment in 1969 population genetics was a well-established field. The second edition of Attachment came out in 1983, only five years before the commencement of the Human Genome Project.

Bowlby is clear that a debate about whether development is primarily a matter of nature or nurture is meaningless because individual development is a product of an interaction between genes and the environment (pp.38, 296). Nonetheless Bowlby plays down the likelihood of genetic causes of behavioural anomalies in favour of environmental causes;

There are many reasons why, in the course of development, one or another feature of an animal’s diverse biological equipment may fail to develop satisfactorily… Though occasionally one or more genes are responsible for the failure, more often some anomaly of the embryo’s environment is the cause – a virus, a chemical, a mechanical trauma and so on. It is probably the same with failures in development of behavioural systems. Whilst genes may well account for some forms and cases of failure, anomalies of a juvenile’s environment beyond those to which behavioural equipment is adapted are likely to be the cause of most of them.” (pp. 129, my emphasis).

Bowlby couldn’t have known about the frequency of de novo mutations in humans and he was studying a behavioural interaction that’s highly susceptible to environmental factors, but his speculative ruling out of genetic causes is rather surprising. His emphasis leads to a detailed discussion of environments.

Environment

A central concept in Bowlby’s model is what he calls the environment of adaptedness. Although human beings occupy a wide range of environments, all those environments fall within limits outside which human beings can’t function efficiently. Bowlby argues that each biological system of each species has its own environment of adaptedness; the one to which it is evolutionarily adapted and the one in which it functions best. He points out that cardio-vascular systems work only within certain limits of oxygen and carbon dioxide, and that those limits vary for different species. So, he concludes, the behavioural systems responsible for maternal behaviour will work within certain limits of the physical and social environment, but not outside them. In short, the environment of adaptedness produced instinctive behaviour so instinctive behaviour will work effectively only within the environment of adaptedness.

Instinctive behaviour

Bowlby presents the traditional definition of instinctive behaviour as;

• species-typical
• a sequence of behaviours running a predictable course
• having obvious survival value for the individual and/or the species
• arising in the absence of opportunities for learning the behaviour.

During the heyday of research into instinctive behaviours there was considerable debate about whether they were innate (inborn) or acquired (learned). Bowlby thinks this is a pointless distinction, because all the characteristics of biological organisms are products of an interaction between genetic endowment and environment (p.38). He follows zoologist Robert Hinde in suggesting that instinctive behaviour forms a continuum ranging from characteristics that remain stable regardless of environment such as nest-building behaviour, to those that are labile and environmentally dependent, such as show jumping or piano playing. Although he wouldn’t include reflexes as instinctive behaviour, Bowlby’s continuum could be seen as extending from reflexes and simple stimulus-response reactions at the stable end, to complex chains or hierarchies of behaviour that include both innate and acquired behaviours, at the labile end. Although ‘instinctive’ might be a more useful concept than ‘innate’ or ‘acquired’ for researchers studying behaviour, for those investigating the biological mechanisms underpinning behaviour, a distinction between innate and acquired behaviours might be more useful than the umbrella term ‘instinctive’.

Emotion

Schore’s summary of the concept of attachment says ‘that emotional processes lie at the foundation of a model of instinctive behavior’. Bowlby devotes an entire chapter to emotion, which is well worth reading if you’re interested in behavioural theory. In it, he opts to use the term ‘feeling’ rather than ‘emotion’ because ‘feeling’ can be applied to a wide range of… well, feelings, but ‘emotion’ tends to have a more restricted use. (And one that’s proved notoriously difficult to define.) Essentially Bowlby sees feeling, not as a stand-alone biological phenomenon, but as having an important role in an organism’s appraisal of its internal and external state. The appraisal process involves;

• assigning a value (e.g. nice/nasty) to current sensory input
• comparing the current situation with previous ones
• selecting an appropriate behavioural response
• evaluating the behavioural response in terms of the organism’s goals

Bowlby shares the view of philosopher Susanne Langer that ‘being felt is a phase of the process itself’ (p.108). In other words, feelings are a phase of the process of appraisal.

He explores the role of feelings in communication – via facial expressions, for example, and whether feelings cause behaviours. Bowlby’s analysis is very thorough. So although feelings clearly play an important role in the interaction between mother and infant, I find it difficult to understand how Schore can conclude that Bowlby sees emotional processes as at the root of attachment, especially as Bowlby warns repeatedly about the danger of reifying feelings and emotions as if they can exist apart from the process of appraisal of the internal and external environment.

Strengths and weaknesses of Bowlby’s model

This is my understanding of Bowlby’s model of attachment;

Over time, human beings have evolved instinctive, biologically regulated behaviours that increase the likelihood of the survival of the individual and the species. Mating, parenting and attachment behaviours are critically important (p.179). Attachment involves infants seeking proximity to a primary caregiver, usually the mother. Separation from the attachment figure causes anxiety, and loss of the attachment figure causes depression.

Elsewhere I’ve been quite critical of attachment theory so it might be worth highlighting in more detail where I think Bowlby is wrong – and where he’s right.

There are some aspects of Bowlby’s model I wouldn’t question. It’s clear that childhood experiences affect development and adult behaviour. Some behaviour fits Bowlby’s definition of instinctive, and it’s valid to describe the proximity-seeking behaviour of young children toward a caregiver as ‘attachment’. Attachment behaviours would generally improve an infant’s chances of survival. And there’s no question that there’s usually an emotional bond between infants and their mothers and that separation from the mother can have lasting effects.

What I am questioning is some of Bowlby’s underlying assumptions about nature and nurture. I think there are four questionable assumptions – about genetic endowment, the environment of adaptedness, typical development and what causes developmental differences. Bowlby sees the nature/nurture debate as pointless because development – of the species and the individual – is the outcome of interactions between genetic endowment and environment. Yet he sees individual development as being influenced predominantly by environmental factors. How does he come to what appears to be a contradictory conclusion?

genetic endowment

As far as I can tell, the contradiction originates in his first implicit assumption about genetic endowment; that the human genome has already evolved and is unlikely to evolve further. In Bowlby’s model, instinctive behaviour has evolved once and for all in its environment of adaptedness. It’s as if human beings, like all other species, have ended up in a kind of evolutionary cul-de-sac. That’s true in the sense that past genetic changes limit future ones, but it doesn’t preclude the minute de novo changes in genetic material between generations that are now believed to be involved in many developmental disorders.

The interaction between genes and environment can occur at a molecular level. A small genetic variation can result in changes to levels of a specific protein that can in turn trigger a cascade of developmental abnormalities, including abnormal social interaction, with variations between individuals resulting from interacting developmental and environmental factors interacting in different ways; this is what occurs in Williams syndrome, for example.

environment of adaptedness

The second assumption involves the environment of adaptedness – the environment that shaped the genetic endowment and the instinctive behaviour of a species. Bowlby says;

So long as the environment is kept within certain limits, it seems likely that much of the variation in the behaviour of different children is attributable to genetic differences. Once environmental variation is increased, however, the effects to which such variation gives rise are plain to see”. (p.296)

The problem with this view is that although the limits of the environment of adaptedness are clear for something like the cardiovascular system – if they are exceeded the system stops functioning and the organism dies – we don’t actually know for sure what those limits are for various behavioural systems. We can only speculate by observing the points at which behaviour begins to depart from typical patterns. And typical behaviour varies not only between individuals, but also within individuals – over time and in different environments. Although some behaviour patterns can be described as typical of a species, the typicality itself isn’t clear-cut; all we can say is that a species has a tendency to behave in particular ways in particular circumstances.

typical development

Although he doesn’t say so explicitly, Bowlby appears to be moving towards a normative view of child development. So a third assumption is that not only are some patterns of behaviour typical of human beings, those are also patterns that healthy human beings should be showing.

causes of developmental differences

The fourth assumption is that psychoanalysis is correct in locating the root of many pathologies in attachment patterns;

If the satisfactory development of attachment is as important for mental health as is claimed [by Freud], there is an urgent need to be able to distinguish favourable development from unfavourable and also to know what conditions promote one or the other.” (p.331)

Bowlby explains Freud’s view of separation anxiety as follows;

“… we try at times to withdraw or escape from a situation or object that we find alarming, and … we try to go towards or remain with some person or in some place that makes us feel secure. … So long as the required proximity to the attachment-figure can be maintained, no unpleasant feeling is experienced. When, however, proximity cannot be maintained … The consequent searching and striving are accompanied by a sense of disquiet … and the same is true when loss is threatened. In this disquiet at separation and at threat of separation Freud in his later work came to see ‘the key to an understanding of anxiety’.” (p.330)

Bowlby supports this thesis by referring to studies that show typical attachment patterns in infants. I think this is where his theory runs into problems because the studies also show a great deal of variation in behaviour in both babies and mothers. Despite this, Bowlby focuses on the correlation between infant and maternal behaviour and concludes;

Whatever the causes of a mother’s behaving in one way or another towards her infant, there is much evidence suggesting that whatever that way is plays a leading part in determining the pattern of attachment he ultimately develops.” (p.345)

At one level Bowlby understands the importance of the interaction between genetic endowment and environment in behaviour and that genetic endowment and environment can vary. But because of his underlying assumptions about the human genome, the environment in which it evolved, the nature of instinctive behaviours and a psychoanalytic model that locates the source of mental health in childhood experiences, it’s almost inevitable that Bowlby ends up placing a disproportionate emphasis on environmental factors – especially on the mother’s behaviour.

Does the nature/nurture debate really matter?

Despite child development theorists from Darwin onwards espousing the idea that behaviour develops from an interaction between genetic make up and environment, child development theory has throughout its history swung between an overemphasis on genes and an overemphasis on environment. An overemphasis one way or another isn’t just an obscure theoretical issue. Overemphasising the role of genes resulted in the eugenics movement that had catastrophic outcomes for minority groups worldwide. Overemphasising environmental factors (coupled with a misunderstanding of probability) led to the relatively recent imprisonment in the UK of several mothers wrongfully convicted of murdering their children – with tragic consequences.

Attachment theory has experienced something of a resurgence in recent years, but professionals using it don’t necessarily have sufficient biological knowledge to critique it. In the publications on child development referred to by the Munro review of child protection Bowlby’s ideas were generally presented as givens, with little discussion. In other words, there appears to be an assumption implicit in the child protection literature – and amongst politicians – that the most likely cause of abnormal behaviour in children is parental behaviour. The problem with an overemphasis on parental behaviour is that there’s a serious risk of genetic and medical disorders and alternative environmental factors being overlooked. I’ve lost count of the number of accounts I’ve read from parents who have been puzzled as to why their child refusing to go to school is attributed by teachers and social workers to a child having a problem with attachment (rather than a problem with school), or where problems with attachment have transformed into a diagnosis of autism or ADHD once practitioners who are specialists in those fields get involved.

My fundamental problem with attachment theory is not that I think attachment behaviours don’t exist – they clearly do. It’s that Bowlby’s theory consists of a number of phenomena over which there’s little disagreement, held together by a series of assumptions. Those assumptions implicitly rule out a host of other possible causes for mothers and children behaving in the ways they do. In the next post, I’ll look at what Allan Schore makes of Bowlby’s model.

Edited for clarity 11/7/15.

Image of Bowlby from http://tinyfootprints.wikispaces.com/John+Bowlby