what’s wrong with attachment theory?

I’ve encountered enough examples of inadequate, chaotic, manipulative and abusive parenting to understand why people working with troubled children might view parents as prime suspects. However, there are many possible causes for children’s unusual behaviour and a focus only on parental behaviour means that other causal factors are likely to be overlooked.

In the next few posts, I want to explore three concepts – Attachment Theory, Fabricated or Induced Illness (FII) and the ‘cycle’ theories of deprivation and abuse – that attribute the cause of children’s abnormal behaviour primarily to parental behaviour. Although these concepts aren’t directly related to autism, parents of children with developmental issues and learning difficulties have reported these ideas being proposed as causes for their children’s behaviour despite evidence to the contrary.

Attachment theory

Attachment theory was an idea developed by John Bowlby, born in 1907, one of the six children of Sir Anthony Bowlby, surgeon to the King’s household, and his wife Maria. Like many children in well-to-do families of the period, Bowlby was brought up by a nanny and sent to boarding school at the age of seven. He later became a psychologist, psychiatrist and psychoanalyst who worked with maladjusted and delinquent children, studied the effects of maternal deprivation and after WWII became a mental health consultant to the World Health Organization. Bowlby’s work led him to the conclusion that ‘to thrive emotionally, children need a close and continuous caregiving relationship’ (Bretherton, 1992). He drew on new research in cybernetics, brain function and on Lorenz’s work on imprinting to develop his theory, and realized that a complete revision of Freudian ideas about child development was required. This he attempted to do in three books published under the title Attachment and Loss: Attachment (1969), Separation (1973) and Loss, Sadness and Depression (1980).

In 1950, a University of Toronto graduate called Mary Ainsworth had joined Bowlby’s research team. Three years later, having moved with her husband to Uganda, she carried out the first of several studies on infant-mother attachment. She then developed an assessment procedure for determining the nature of an infant’s attachment – the Strange Situation – and concluded that infants showed one of four distinct patterns of attachment to their mothers; secure, anxious-resistant insecure, anxious-avoidant insecure and disorganized/disoriented. The patterns of attachment developed with primary caregivers were seen as foundational for social interactions and mental health throughout later life – an insecure or disorganized attachment could lead to later problems.

It’s easy to see how attachment theory could be used to lay the blame for a child’s behavioural issues at the feet of parents, particularly mothers. Bowlby, however, doesn’t appear to have had any intention of blaming parents; his theory is firmly grounded in the idea of behaviour involving an interaction between genes and environment. But that doesn’t seem to be the way attachment theory is presented in texts that inform public policy. Before moving on to these texts I want to examine the ideas behind attachment theory in more detail because, like the concept of autism, it’s a theory constrained by the state of knowledge at the time of its inception.

Background to attachment theory

In the opening chapters of his first volume of the Attachment and Loss series – Attachment – Bowlby explains, systematically and in detail, the theoretical framework for his model. Bowlby brought together concepts from a number of different fields. Here’s a summary of his reasoning:

• Observations have shown that separation from the mother-figure can be extremely traumatic for children. There is evidence that this separation can result in problems with behaviour, personality and mental health in later life – e.g. stealing, depression and schizophrenia.

• Attachment theory is grounded in psychoanalytic theory because ‘despite limitations, psychoanalysis remains the most serviceable and most used of any present-day theory of psychopathology’ (p.xv).

• Data can be obtained by observing behaviour, as well as from introspective reports from participants.

• Freud drew attention to the importance of feedback in homeostatic biological systems; in behaviour, motivation is regulated by homeostasis in the same way.

• Instinctive behaviours can be complex and are the outcome of an interaction between ‘genetic endowment’ and environment – partly innate and partly acquired.

• Behavioural systems are goal-directed (ie they evolve in order to fulfil a specific purpose) – analogous to engineering control systems.

As I understand it, at inception, the Freudian psychodynamic model of behaviour was a novel idea; it was developed from first principles derived from contemporary understanding of biological mechanisms. Bowlby’s model wasn’t novel; it was an extension of the psychodynamic model based on new knowledge about those mechanisms. This means that Freud and Bowlby based their theories on the same assumptions:

1. Species-specific patterns of behaviour are biologically ‘provided’ – although they can be affected by the environment.

2. Biologically provided behaviour patterns are normative. That means that in the ‘right’ environment they will unfold naturally, but could be disturbed if something goes ‘wrong’ with either the genetic endowment or the environment.

3. Biologically provided behaviour patterns are goal-directed – they have evolved to fulfil a specific purpose.

4. Behaviour is driven by the need to maintain emotional (feelings) homeostasis.

5. Relationships between parents and children are central to child development.

Attachment theory is clearly a theory of its time, since research in all these areas has since moved on. That doesn’t necessarily invalidate attachment theory, but does suggest there are some problems with it.

What’s wrong with attachment theory?

1. Freud saw social and sexual development as central to human behaviour because of their importance in sexual reproduction – the means by which inherited characteristics (including behavioural drives) are passed on to offspring. This is why a child’s relationship with his or her parents was seen as so important; it could disturb the natural unfolding of social and sexual drives. Subsequent research, by contrast, shows that social and sexual behaviour is influenced by complex array of factors that change over time. Despite revising Freud’s framework, Bowlby still viewed social and emotional factors as central to the development of human personality rather than being two factors amongst many.

2. Much of Bowlby’s research was carried out prior to the structure of DNA being discovered and the consequent development of molecular biology. Subsequent research suggests that rather than behaviour patterns being biologically provided, they emerge from interactions between genetic expression and environment. Similarities between individuals in both genetic material and environment result in species-specific behaviours but differences mean that species-specific behaviour patterns vary between individuals. Oppenheim et al. (2009) noted, for example, secure attachment patterns in autistic children that were different to the one that Ainsworth described.

3. Biological behaviour patterns can only be goal-directed if the genome and the environment remain stable – but they don’t. Genetic make-up is unique to an individual and the environment changes constantly; behaviour patterns emerge from a dynamic interaction between the two. Although there’s no doubt that children do exhibit patterns of behaviour towards their primary caregivers, and it’s likely that those patterns can be seen across different individuals, the patterns are descriptive, rather than normative. Even if the patterns provide a useful way of identifying problems in infant-parent relationships, they show how children interact, not how they should interact.

4. Some biological systems – those that regulate body temperature or the levels of oxygen and water in the body, for example – are maintained via homeostasis because the biochemical reactions necessary for survival can occur only within certain narrow limits. Emotions and behaviour aren’t constrained in that way and tend to be cyclical rather than stable.

5. There’s no doubt that a child’s relationship with his or her primary caregiver is important. But in focussing on a single relationship, attachment theory by definition marginalises the role of genetic, biological and other environmental factors – including other relationships – in a child’s social and emotional development.

This brings us back to the concept of reification that cropped up in a previous post about Kanner’s model of autism – although I didn’t call it that at the time. Reification literally means ‘making a thing’ – the implication being that a ‘thing’ is made that doesn’t necessarily exist in the real world. There’s no question that ‘attachment’ can be used as a descriptive label for certain kinds of behaviour, just as ‘autism’ can. It doesn’t follow that attachment must be a clear-cut psychological function, nor that autism must be a distinct medical disorder.

When I studied psychology as an undergraduate in the 1970s, attachment theory was already being viewed with some skepticism for the above reasons. I haven’t kept in touch with child development research so I was surprised to find that attachment theory is still alive, well and influencing social policy in the 21st century. That’s the subject of the next post.


Bowlby, J (1969). Attachment and Loss vol 1: Attachment. Revised 2nd edition, 1997, Pimlico.
Bretherton, I (1992). The origins of attachment theory:John Bowlby and Mary Ainsworth, Developmental Psychology, 28, 759-775.
Oppenheim, D., Koren-Karie, N., Dolev, S. and Yirmiya, N. (2009) Maternal insightfulness and resolution of the diagnosis are associated with secure attachment in preschoolers with autism spectrum disorders. Child Development, 80: 519–527


I want to thank everyone who sent me links relevant to this and related posts. You know who you are!


6 thoughts on “what’s wrong with attachment theory?

  1. Hi, Bruno from France, I’ve just stumbled upon your blog which I find very interesting. I’m a social worker trained in psychology (systems and social). Attachment theory is all over the place in child protection over here, and it is often used to keep parents under legal supervision, even without clear signs of neglect or abuse. I had a feeling the reification bit was quite present in this theory, but I was having a hard time trying to criticise it, so you article is a bit of fresh air to me. Still, I don’t understand your points 2. and 3. : it sounds like a secure attachment pattern is supposed to be “biologically”, normatively and precisely determined in AT ; and then in point 5 you state attachment theory is all about one relationship, and not enough about biology and genetics. This confuses me, can you shed some light on this? Do forgive my clumsy English, best regards

  2. I’m concerned as to why it’s being rolled out in schools at the same time as the state assumes more control over children can yiu recommend further research. “professionals ” in school are wielding the scientific research like a mantra to make decisions which have huge impacts on children’s ‘ lives. I’m also curious about the political agenda informing social policy. Thank you. Any information I’d be grateful.

    • Some good points.

      Personally I think the current political agenda is often a knee-jerk reaction to specific issues. In the post-WW2 period there was a great deal of interest in the underlying causes of social problems, including how social and political systems might be involved in causing them. That systemic approach was to a large extent abandoned in the shift to neo-conservativism during the 1980s that frames social problems in terms of individual behaviour. That viewpoint seeps into the way we all think whether we like it or not. By now, many professionals in the UK will have grown up in a world that attributes social problems to individual behaviour and will have been trained to see attachment issues as being of primary importance in child development. Unless they’ve encountered examples of that theory not working or they’ve read widely, they might not be aware of any alternative explanations.

      If professionals come up with speculative ideas about what might be causing particular behaviour, it’s worth asking what evidence they are using. That question does make people think.

      I’ll have a think about the research.

      • What are your thoughts on Majorie Boxall and Nurture Groups? I am coming at this from the angle that it is pseudoscientific but that is mainly because it seems the Boxall Profile appears to have some very strange ideas about what attachment is. Indeed it is heavily based on Bowlby but only superficially in so far as it talks about the need for attachment. There are tons of schools with nurture groups and it concerns me that this is the case given its poor theoretical foundations and the lack of any kind of longitudinal study.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s