Why social workers shouldn’t use ‘attachment’ in their records and reports

After a recent court case criticised a social worker's use of attachment in evidence, David Shemmings looks at the implications

Photo: aquarius83men/Fotolia

by Professor David Shemmings OBE PhD

There are a number of problems that emerge from the recent judgment that criticised the use of attachment.

I will restrict my thoughts and observations to those concerning the use of attachment theory, except to say that, for me, some of the most important points to note from Mr Justice Mostyn’s summary are his reminders that decisions made under Sections 39 and 31 of the Children Act 1989 must not be based upon ‘less than perfect parenting’ and that, furthermore, the principles of ‘necessity’ and ‘proportionality’ always need to be uppermost in local authority assessments.

I think this is where the social worker’s use of attachment theory begins to falter.

As an explanation of the three attachment patterns – secure, avoidant and ambivalent – the first section of the first social worker’s report was, I thought, reasonably clear. It was the conclusions drawn that caused problems. Simply put, because around 40% of any representative population is insecurely attached (adults and children) this is nowhere near the threshold levels to keep children away from their parents.

David Shemmings has worked with Community Care Inform on a wide range of resources to help social workers understand and apply contemporary attachment theory in their work. Subscribers can listen to David talk about using concepts from attachment theory to help develop relationships with parents or read a review of the key messages from research into disorganised attachment behaviour. Find these and much more in the attachment knowledge and practice hub.

The social worker explicitly acknowledges this when using the term ‘less optimal parenting’. In other words the existence of insecurity is given far too much weight.

Attachment behaviour

Additionally, from recent research into separated identical twins it appears that attachment organisation is not as stable through the lifespan as it was once thought to be: we can change our attachment patterns later on, depending on who we meet through the life course. Similarly, if carers change their parenting style, a child’s attachment behaviour can also change.

Surprisingly, perhaps, recent research also indicates that change is more likely if the child shows ‘disorganised attachment behaviour’. But, unwittingly, when referring to this phenomenon in her report, the social worker makes a number of errors, as the construct has undergone a major revision over the past few years.

There are other circumstances when a child’s behaviour – disorganised attachment is not so much a ‘pattern’, as with the other three, than a set of behaviours – can indicate attachment disorganisation, but which may be less to do with parenting and more about the circumstances in which they are living e.g. adversity, which can leave their carers emotionally drained and unavailable (but through no ‘fault’ of theirs). For more information see the work of Professor Pehr Granqvist et al, 2017 and Dr Robbie Duschinsky.

We also need to bear in mind that to ‘assess a child’s attachment’ is only possible after very extensive training. For example, the training needed to administer the Strange Situation Procedure with a 12-18 month-old toddler is 10 days. But that only enables the practitioner to administer the procedure.


To become fully accredited to ‘assess’ attachment requires the practitioner to rate a number of tapes to a very high level of concordance with acknowledged experts in the field. Other approaches are required as children grow older, including Representational Doll-based techniques, the Child Attachment Interview, the Adolescent Attachment Interview and the Adult Attachment Interview (and they also require extensive training and expert-assessed accuracy in rating to gain accreditation).

For these and other reasons, I have come to the view that an understanding of the three main patterns of attachment i.e. secure, ambivalent and avoidant should be far less concerned with trying to categorise children’s behaviour and more about helping practitioners learn about their own attachment history, especially how they process powerful emotions.

Doing this enables them to become more aware of how, when their own attachment system is activated – as it is most of the time when working with families – they may, inadvertently, end up emotionally unavailable.

My advice to practitioners is actually not to use word ‘attachment’ in their records and reports (I encourage them to substitute ‘relationship’, as it usually does the trick).

It’s important also to remember that any assessment of a child or young person’s ‘attachment’ can only be done with any confidence if their attachment system has been activated. Serendipitous observations of parent-child interactions in the home or elsewhere are not reliable.

‘Say what you see’

Additionally, I also advise practitioners to take heed of Professor Sue White’s tip ‘to say what you see’. Of course, that may assume that we all see the same thing but if we then expose fully what assumptions we are making, including the use of theory, then we remain accountable, both to the court and to family members.

Being clear about how one interprets parent-child interactions and relationships is essential in social work. Unfortunately this judgment is peppered with examples of problems with the specificity – and arguably at some points the relevance – of the observations. Similarly, the conclusions drawn are sometimes flimsy. Judge Mostyn commented as follows:

“This witness statement was very long on rhetoric and generalised criticism but very short indeed on any concrete examples of where and how the mother’s parenting had been deficient.”


Mr Justice Mostyn asked her to name her ‘top four criticisms’ of the mother’s parenting of her son, referred to as M and L respectively, which were: “ i) The mother’s questionable ability to meet L’s emotional needs; ii) The dynamics of M’s relationship with L; iii) The dynamics of the relationship between L and his siblings; iv) The mother’s questionable ability or willingness to work with the local authority.” But when pressed to give examples, she struggled.

“(The second social worker) was asked to identify her best example of the mother failing to meet L’s emotional needs. Her response was that until prompted by the local authority mother had not spent sufficient one-to-one time with L and had failed on one occasion to take him out for an ice cream. This struck me as utterly insubstantial criticism … A further criticism in this vein was that the mother had failed to arrange for L’s hair to be cut in the way that he liked …”

So do we conclude that attachment theory and research are irrelevant? Not in my view, in fact quite the reverse: used properly, contemporary attachment theory and research can offer powerful insights into parent-child interactions, especially when they involved challenging situations.

They are used best when aimed at helping family members understand some of the complexities of what’s going on; but too often, it seems, attachment theory is used as a stick with which to beat them. 

David Shemmings is a professor of Child Protection Research, at the University of Kent and is a visiting professor at Royal Holloway, University of London 

More from Community Care

6 Responses to Why social workers shouldn’t use ‘attachment’ in their records and reports

  1. londonboy June 28, 2018 at 12:07 pm #

    And what of the many families separated in the past because a judge accepted ‘evidence’ from an ‘attachment expert’? Do all social workers, psychologists and lawyers just walk away from the possible/probable inter-generational harm caused in many cases?

    Where was the challenge within the system into how misapplied attachment theory became a cornerstone in child removal practice – from social workers, psychologists and lawyers – and what does that say about entire systems failures and dangerous and unjust power imbalances?

    Do professionals and policy makers simply not care about entire systems failures because it did not affect them personally?. It certainly seems that way from here. The long term view is that likely to be that this is a shameful period in the history of social workers as a result.

    • Zoe June 29, 2018 at 12:23 pm #

      Psychologists would have substantial understanding of attachment theory and be trained to administer the strange situation in order to accurately measure attachment status at that particular time. I have been a practitioner family outreach worker and I have a psychology degree and I have observed less qualified practitioners making sweeping judgments on this exact area that is misinformed and inaccurate.
      So I agree with the judge in this instance and think it is right to use the correct language when diagnosing or when providing an evaluation of a family you may be working with.

    • Dianne Kirkham June 29, 2018 at 3:25 pm #

      So very well said
      I agree completelim now interested in how long and what further appropiate training will be given before we see chsnges

  2. lisa June 29, 2018 at 11:47 am #

    I’d like to comment on social workers. Every child deserves the right to survive thrive and forfil their potential. Be protected if there is a problem in the home.
    Well I tried to help a child I stuck up for a child I got named as reportee so I have to put up with anything or any bad manners thrown my way and social workers are helping parents not the children. So could someone tell me where the protection is for the children? Do these social workers take up resident with these bad parents protecting these children? I think you will say no. So who’s helping these neglected abused children?

  3. John Ramsey July 4, 2018 at 12:46 pm #

    I am a Fostering Social Worker and a large part of my role is to assess prospective carers. I spend a great deal of time talking to them about their childhoods, the patterns of attachment they developed and how this might have changed over the course of their lives. I would love to use the discourse marking technique to which Dr. Shemmings refers, but I have never persuaded my employers to send me on such a major piece of training!
    However, I do look at carers in terms of the ecological-transactional model of development and the Secure Base model; the latter examining the extent to which the applicants parenting gave them a secure base and how they could offer this to foster children.
    I believe that Attachment is absolutely central to fostering and adoption. My aim is therefore to write strictly evidence based assessments of how it works for foster carers. Panels to whom I have presented my assessments seem to appreciate my efforts. If I followed David Shemmings’ advice, I am not sure how I would fill the theoretical hole it left.

  4. Julie Wilkes July 22, 2018 at 9:24 am #

    The advice to only describe what is seen sounds sensible. Until you remember the constraints of trying to assess a relationship in the highly contrived circumstances of a foster placement. A child has been separated for over two years. The parent visits for brief periods that are observed. How would you ever gain direct evidence to the original standard of significant harm? What would the parent have to be observed doing? It’s fairly obvious that an authority is only going to commission a report on attachment when it is anxious and looking for backup evidence. So, as has already been concluded, elsewhere, the 89 Act operates as a parents’ charter. Without some way of referring to history and the bigger picture of the child’s experiences and psychological vulnerability, we can’t protect children. This may be expedient given budget constraints?