‘Myth-busting around attachment theory’ Professor Pasco Fearon

Matt Kempen
Marketing Manager for ACAMH

Posted on

Professor Pasco Fearon on ‘Myth-busting around attachment theory’.

Recorded on Friday 8 March 2019 at the Emanuel Miller Memorial Lecture & National Conference ‘Attachment & early intervention: Improving emotional wellbeing and relationships in the family, and at school’

Watch this lecture via our YouTube channel

ACAMH members can now receive a CPD certificate for watching this recorded lecture. Simply email membership@acamh.org with the day and time you watch it, so we can check the analytics, and we’ll email you your certificate.

Dr. Pasco Fearon

Dr. Pasco Fearon is Chair of Developmental Psychopathology in the Division of Psychology & Language Sciences at University College London (UCL). His research focuses on the mechanisms of socio-emotional development and the processes involved in risk for psychopathology. He is particularly interested in the intersection between social and biological processes in children’s development, and accordingly he uses a range of diverse methods, including observational techniques for measuring family processes (attachment, parenting), quantitative and molecular genetics for examining genes and gene-environment interactions, and biological measurements for capturing peripheral physiology (e.g. cortisol, heart rate) and brain function (EEG, fMRI). As his primary interest is in development, much of his work involves longitudinal cohort studies, and multivariate longitudinal data analysis. Dr. Fearon is the deputy Editor in Chief for JCPP.


Good afternoon. Yeah, so I’m going to talk about myth busting. Right, now, this was not my title, I have to tell you, I think ACAMH specifically said, would you talk about myth busting around attachment theory. And I thought to myself, well, this is a subject I’ve been studying for the last 20 years. And you’re asking me to have a pop at it. But so I’m going to do my best. But I’ve given it a subtitle, which is a rather more boring and academic and considered rephrasing of the title, which is, or rather some empirical, empirical challenges to commonly held interpretations of attachment theory. Let’s, although I’m going to sneak at the end of this talk, some proper myth busting, but it will just be the last minute or so of my talk. What I would like to do is tell you about some of the work that colleagues and I have been doing over the last few years, which is to actually try to, in a sense, establish where are we at with this really important subject? What does the data tell us about what’s more or less accurate or correct about attachment theory? What needs revising? What’s the status of this theory with respect to a lot of accumulated evidence over the last 30 or 40 years. Before I start to have a pop up my favourite theory, I want to say something about how incredibly significant attachment theory is, I think it’s quite easy, certainly, for me to be, you know, very kind of engaged in at the coalface of doing research on attachment theory, and, and not to step back and think about the influence that this theory has had or what you know what this theory really represents.

And it’s, I still believe, despite some of the empirical qualifications, to the theory that I’ll talk to you about in a moment, that it’s one of the most extraordinary pieces of theoretical enterprise that we’ve seen, in the field of psychology in the last 50 or so years. John Bowlby made the most incredible contribution to our social, our clinical and our intellectual lives when he developed attachment theory deeply, profoundly rooted in evolutionary thinking, and in, you know, interesting ideas around cybernetics, and on biology, but also integrating psychoanalysis and cognitive theory and so on. It’s the most amazing, eclectic, but rigorous and coherent developmental theory, I think that has actually probably ever been invented. And not only that, but it stands the test of time rather well, and some of the fundamentals that we think are encapsulated by that theory, particularly the basic idea that children have a biologically predisposed set of instincts to seek out the care of, of reliable others, as a way of affording them protection in the face of potential threats to their development, that I cannot believe is not basically true. So I’m not going to have a pop up that basic idea, it must be surely one of the things, one of the few things we can genuinely agree on.

And in fact, where the qualifications have to come in are around some of the margins of this theory, and particularly the way that the theory developed in contemporary research. The other thing I wanted to add, actually picked up a point that was just absolutely beautifully made by David Olds at the beginning of today, which is just the sheer reach of the impact of this theory, because there’s the theory itself and the contemporary framework developed by Mary Ainsworth and others over the last, and very influentially, also by Marina Van Eisendawn. But if you think about the, the sort of the flavour of what John Bowlby and Mary Ainsworth was trying to do back in the, in the late 50s, and 60s, you see the influence of that theory in almost every field of developmental and clinical psychology to this day. So, as David was pointing out, much of what we think of as contemporary developmental psychopathology really had its roots in some of the basic ideas that Bowlby developed. And it’s, it’s had an enormously widespread impact and that that, and almost all of that, I would say, has been extremely positive. That’s my little…now I move on to some of the more, let’s say, critical issues. I’m slightly put off by this screen, because it keeps flashing at me in a bright pink, looks a little bit, anyway, so maybe I’ll try not to look at it, because it’s actually really putting me off.


So the rest of my talk I’m going to touch on, and what I would say, are probably three of the most important propositions of attachment theory, as we currently understand it, and they are the following. So one of them, of course, is that there’s a proposition within attachment theory, they’re quite strong claims that attachment theory is making, that the that attachment insecurity is fundamentally something about the environment and the caregiving environment around the child. So of course, touching on what Gordon was just talking about the counter factual to that if you like, is that you know, it could would also be genetic. But attachment theory makes the very strong and very clear prediction that actually, when we look at variation in attachment, security and insecurity, what we’re doing is seeing variation that is all entirely due to variability and the quality of care provided to the child is fundamentally an environmental set of influences and not about the child’s genes. Very clear prediction, we’re going to find out whether that prediction is in fact correct.

The second really important proposition that Bowlby and others subsequently put forward about attachment is that patterns of attachment develop very early in life. They are laid down early, and they and they remain persistent influences on a child’s development throughout the lifespan, and indeed may be transmitted from one generation to the next. That’s a fundamental tenet, if you like of attachment theory. And again, I’m going to present to you some evidence about whether or not the data really holds up on this crucial question about attachment and its stability over time. And its stability to be as it were transmitted from one generation to the next. And then finally, I’m sure there are more, but these are my these are the ones that we’ve worked on particularly but they are very central. I think the final one is that attachment security is important for children’s mental health. So it is not a given that that might be that that is the case. But it’s certainly a proposition that people make that a secure attachment bond. I feel like I’m about to be, or that as if I am, quoting from pretty much every developmental psychology textbook you can think of attached attachment security is fundamentally important for children’s future, socio emotional development, their adaptation and their wellbeing right, those, that’s the entree to most papers on attachment. Although I’m trying to be a little bit careful about how I say that these days, because of the work that I’ve done on this subject recently. But these are fundamental questions and predictions of the theory.

And I suppose one reflection before we dive into this, I mean, one of the lovely things about doing research is that now and then you do accumulate enough fairly clear evidence, where you can step back and think, do you know what I think we almost know the answer to these questions. And that’s really, for those of you who aren’t researchers, it’s depressingly not the case, most of the time, most the time, we just don’t really know anything, and we’re trying our best, and everything’s a bit equivocal. But I feel like we’re in a position where I can say with a little bit of confidence, more or less where we’re at with those subjects. And I hope that will be helpful, both for asserting the continuing vigour and usefulness of attachment theory as a framework, but also for asserting some caution and care, and pointing out where we clearly need to either draw back some of the strong claims we might make about attachment, or that we just need to do much more careful research in order to better understand some of the some of the puzzles posed to us by the evidence. Okay, so that’s, that’s the plan.

So the first one is, genetics and the environment. Gordon already mentioned that a twin studies are probably the most powerful, and commonly used method for addressing these questions about heritability. The attachment field went for many, many years without ever doing a twin study, which is really interesting, because, you know, from very early days, people in the field had wondered, and certainly people outside the field had claimed that a lot of this attachment business was a lot of nonsense, really, because it’s probably almost certainly all to do with the child’s temperament. Remember, people who followed this subject, there was a very famous paper by Lamb and colleagues back in 1984, which was making that basic point. It took a long time before anybody did any actual behavioural genetic research, to test the very testable claim that actually, even though most things look quite genetic, apparently, variation in attachment security or insecurity is not like that at all. And it’s all about the environment. I often say that I never really set a question in behavioural genetics exams on how much of the variation in something or other to do with children is genetic, because the answer is always somewhere between about 40% and 60%, right. I.e. a lot. So attachment theory is making a very bold claim to say that 100% of the variation in attachment security and insecurity is actually to do with the environment and genetics is irrelevant. A bold claim that hadn’t been subjected to a proper test until relatively recently.

So here’s the answer to that question. This is more or less what I would say, I would summarise the evidence suggests, this is the results of a study that we did back in, in 2003, twin study looking at the proportion of variation in attachment security or insecurity that you could attribute to genetics, and to what’s called the shared environment, and what’s also called the non-shared environment. What you can see there is that the bit that’s to do with genetics is entirely absent, so there’s nothing there that’s representing genetics. The estimate of the genetic influence on attachment security, insecurity was indeed absolutely zero. That study has been replicated both by Tom O’Connor and Carla Croft, or replicated, in fact, they came to it first. And on a larger scale by Glenn Reisman and Chris Fraley in 2008, using a rather different measure of attachment, but the picture is remarkably consistent in that genetic contributions to attachment seem to be really very, very small. And that’s a very fascinating and interesting and pretty robust defence, as it were, of one of the most basic predictions of attachment theory. And of course, can you imagine, I remember at the time not really thinking about this when we did this study, but what would we have done if it turned out to be about 98% genetic? Oh, my God, I would have had to have left town.

So it turned out, it’s funny that, yes, I didn’t contemplate that at all, just thought, this is fascinating, let’s do it. But it would have been interesting. But as it turns out, the evidence from our study and several others is really quite consistent, and very consistent with attachment theory. Or is it? So we then decided, well, hang on a minute, that’s very interesting. These are all very young children and it is very consistent with the evidence that attachment security is very much about the environment, it is almost certainly telling us something about the quality of care. But we want to study attachment across the whole lifespan, don’t we? And we, and we know that it changes in some ways, certainly, we measure it in a very different way, in later stages of the lifespan. And so we went out to do the study again, some years later, collaborating with Robert Boy, who was mentioned earlier, using data from the twins early development study, we conducted with a fairly large sample of 15 year old twins, the child attachment interview, which is a very nice, interesting in depth, sort of semi clinical interview that is designed to reveal something about attachment status in in teenagers. And the question would be well, you know, are we going to see exactly the same thing. In fact, Robert, by the way, in case everyone’s slightly biased against, you know, assuming that Robert would be excited, would always think everything is genetic, he really wanted to see lots of shared environment, because he thought that was a really interesting counterexample to so much of what he’d studied in the past.

And this is what we found, in a nutshell. The shared environment, which we saw so much of when the children were very young, is absolutely zero, in adolescence, and we see a very substantial, about 40% of the variation, in the security of attachment, as measured by the child attachment interview, in adolescence is heritable. I find this absolutely fascinating, because it’s so different to what we saw in the young children. But it’s also very consistent with what we see in many other areas of development, which is a very strong signal, that there’s something heritable in the child that is impacting in some way, on the way that they, in this case, think about, reflect on and talk about in the child attachment interview, their attachment relationships with primary caregivers. So that’s the that’s the first take home message. Yes, in early development, it looks like attachment theory with basically right, it’s very strongly influenced by, attachment security is very strongly influenced by the environment. And really shows very little evidence of being influenced by genetics. But it looks and this is only one study, but it’s quite large and quite well designed, though I say so myself study. And in adolescence, the picture looks very, very different. So the take home message is, we can’t assume that what we see in young children will always be the case. And of course, development is interesting and complex. In a way, I don’t think we should be that surprised that genes in complicated ways, genes, the child’s personality, their traits, their behaviours, their ways of seeing things, their ways of reacting, might start to impact on attachment in later life. Okay, I shall skip that.

So that’s the first one genetics. What about continuity? So I said that attachment theory makes this claim that attachment patterns are laid down in early life, and they stay with us relatively unchanged over time. And they may even influence the patterns of attachment that a child forms in the next generation. Again, a really interesting claim about how cycles of relational disadvantage might be mediated over time within the same individual, but also cascading down the generations. So what does the evidence say on that? Well there have been quite a few studies. It takes so long to do this kind of research, we’ve got what one of the studies they’ve done that, because you have to wait 20 years or something to a strange situation at the age one and then follow them for 20 years. Think of something to do while you’re waiting. There are a very small number of studies that have done that, and a lot of them are very, are also quite small scale. Couple of them within the early days found that there was quite a lot of continuity. So you measured the attachment, the strain situation, then you measured it 20 years later in the adult attachment interview, and a couple of studies found quite a lot of continuity, and people were excited about that, but they were really small studies. Relatively recently 2014, Ashley Grow and Glen Reisman and colleagues followed up by far the largest sample that where it was possible to do this into late adolescence at age 18. And this is what they found. So they managed to follow up about nearly 819 children who were seen in infancy, and then they conducted the adult attachment interview with them in adolescence.

So in terms of sample size, blowing all of the studies out of the water. And this is what they found, SSP predicting AEI 18 years later, a correlation of .04, I don’t know whether you know what correlations usually look like. But they go from minus one to plus 1, .04 is absolutely tiny. And this wasn’t even statistically reliable. They had some other measures of attachment. So you could say, well, maybe it’s something funny about the strange situation. Well, there’s a hint that that might be the case, because there’s a little bit more evidence of continuity, when you use a slightly different measure attachment called detachment key set. And it’s also a little bit later, it’s at 24 months, two years, rather than 15 months, you see a little bit more of a correlation there, it’s still pretty small, but it’s possibly lurching into the not entirely trivial category, but it’s certainly small.

And if we look at some other assessment assessments, we also find that the association is very small. So the idea that continuity from infancy to adulthood is very strong, I think we can set that aside, it clearly is not the case. There may be some continuity, but it’s weak. But it’s certainly, all of this evidence suggests that change must surely be the norm. In case you think there’s something funny about this particular study, and maybe this is just, maybe they just did the measurements wrong or something methodologically weak about that study, just for you, by the way, you should feel very special, I did this last night, I collected together all of the studies, I could find that had ever done this. It’s not absolutely 100% systematic review there. So if I was going to publish it, I’d be a little bit more careful. But it’s, it’s a pretty good sample of all the studies that have actually measured attachment in early life, and then conducted the adult attachment interview, in adolescence or later or into adulthood. And these are all the effect sizes here, you can see there’s a really big one there .5, that was one of the first studies actually by Mary Main, all of the others pretty much, there’s another strange one there, that’s a very small sample, all of the others are rather consistently low. And the average is about  .09, which is actually very similar to what we saw in the large Nic HD study. If you take out the large Nic HD study, you could say what such a big sample, maybe it’s biassing everything? Well, if you do that, the correlation stays pretty much the same. If you take out these two rather large effect sizes, because you think maybe they’re a bit suspect. Again, it doesn’t change very much, it stays pretty consistently a correlation of about .08, very small, but it is significant. So there’s little signal of continuity, there’s an enormous amount of change.

What about intergenerational transmission, because that’s a slightly different thing, actually, this is not continuity of a large periods of time, necessarily, or long periods of time, I should say. It could be, it’s really associations between parental attachment status, and the attachment that the child forms to that caregiver, there been a lot of very interesting studies on that. One of the great studies actually was done in London by Miriam Steele, Howard Steele and Peter Fonaghy, looking at adult attachment status in pregnant women, and then assessing the attachment pattern that the child showed in the strange situation to to that parent, a little bit more than a year later and finding a very strong associations between those two things suggesting there’s something in the parental state of mind with respect to attachment that influences the evolution of their relationship with their child. Back in 1995, Marinas Van Eisendawn, we saw earlier today, did a meta-analysis of all studies that had looked at these two things, parental attachment, and whether that predicted the child’s attachment to that parent, and he found what is it actually, that’s a huge correlation, in the social sciences, with things that are measured, you know, without too much bias, that’s a really strong association. For many of you will also know that that study found that parental behaviour was able to account for some of that relationship, but not all of it, there’s quite a bit missing. And that gave birth to this idea of the transmission gap. But I’m not going to talk about that, particularly, what I want to focus on is how strong that effect size was.

Now, a group of us had been talking about this. There is a bit of a funny story about this because we are all members of the attachment human development journal and one meeting. We have a nice dinner afterwards and I think we did get slightly drunk. And we started saying, you know that effect size, have you got any studies that didn’t really find this association between adult attachment and child attachment? And we all started confessing that, yes, indeed we did. And we didn’t really know what to make of it. And we started to have doubts. And we worried about whether this was something you could even say, given that this was such an established finding. And then, and then we thought, well, wait a minute, this is a really important question. And actually, that, you know, we need to look at this. And we, we started to contact all of the colleagues that we knew in the field, anybody we could think of, that had ever done a study like this. And, and we did a systematic review of the of the literature and so on, and managed to create an entirely new meta-analysis with a much larger data set than was available to Marinas back in 1995. And, in fact, it’s extraordinary to see how much that field has changed.

So, in 95, Marinus had 19 studies that he could better analysed, in other words, pool the results and find out how strong that relationship was, in 2015, we were able to study with 95 different papers that are looked at this relationship, 95 independent studies, it’s enormous. Obviously, it’s quite possible for 95 studies to totally wipe out the earlier findings from just 19. And again, this was an exciting moment, because you know, what’s this going to tell us? And again, what would it mean for the field, if it turned out that that association had completely disappeared, or turned on his head or whatever, you know, but this is the fun of good science, right, is to get some data, you know, the more nervous you are about the result, the more useful the data is, because it’s telling you that there’s a there’s a very testable hypothesis here. And here’s the result. Well, the correlation was .7 in 1995. And it’s dropped down to point .31.

So what does that mean? Well, what it implies is that, first of all, the association is still very clearly there, that’s not a trivial association at all, is very actually consistently replicated in all of those studies. The other really interesting little side line here is that many of it, half of the studies we reviewed, were not published. And even amongst those sad file drawer papers that nobody knew what to do with, that, if you put all of those papers together, they still showed an association from one generation to the next, even though individually, a lot of them didn’t probably because they were just too small and lacks statistical power. So in some ways, you could say, well, that’s a really resounding confirmation of one of the basic principle, one of the basic proposition propositions of attachment theory. But on the other hand, is also a reason to be a little bit more sober, this is no way near a kind of fully deterministic relationship, it’s a much weaker association. It’s not small, but it is much more modest than we might have expected. There’s lots of possibility for that intergenerational transmission process to not leave faithfully, sort of from one classification to the other classification that you’d expect. So it’s a much more probabilistic process than that that kind of association might have led us to believe.

Okay. Oh, and one other little interesting thing is that we found that if you look closely at the data, what you find is that families where there’s a lot of social risk, for example, parental postnatal depression, psychiatric diagnosis or high stress environments, what you find is that the intergenerational transmission process is weaker and not stronger. And our inference there is that what’s happening is that circumstance and stress is very similar, I think, to what David was saying earlier, is really driving a lot of the struggles if you like, that are happening in the parent child relationship. And the parents, internal working models of attachment are being sort of overwhelmed by that. Okay, so last topic is the question of what’s the relationship between attachment and children’s mental health outcomes? We’ve done three meta analytic reviews, looking at this question, the first focusing on children’s behavioural problems, externalising problems like aggression, the second on anxiety and depression, and then the third on children’s social competence. And I’ll summarise those reviews for you now.

So here’s the first one on externalising problems. And what I’m going to show you is just a quick sort of summary of how strong the relationship was between a measure of attachment in relatively early life and some later measure of behavioural problems. And there’s your answer. This bar here is if you just pull all of the children who are classified as insecure in a strange situation. And the bar, the length of the bar tells you sort of how much more likely insecure children were to show some level of behavioural problems compared to the children who were secure. And that effect size is just over what’s a what’s called a Cohen’s D of .3, that’s very statistically reliable. So the first answer to this question is yes, insecure children clearly do show more behavioural problems than secure ones. For those of you who are a little bit statistically minded, you might notice that the D of .3 is a small effect size. You can’t translate that directly or you can, it’s not the same as a correlation. But this will be something like a correlation of about point .16, or .17. So it’s so it’s actually quite a weak effect. So it’s there, and it’s quite reliable. But we could nowhere near say that there’s a strong connection between attachment insecurity and children’s behavioural problems. It’s weak, it’s probabilistic. It’s not deterministic. The other interesting thing is that the different patterns of insecure attachment show different levels of kind of affects, a quite a weak association for avoidant children and resistant ones, and almost all the action is coming from the children who got a disorganised classification.

Another thing to notice is that we found a few factors that seem to make a difference, we found stronger associations in older children, we found stronger associations in any other measure of attachment apart from the strain situation, detachment keyset stronger, the Cassidy and Marvin attachment assessment, which all of these actually tend to also be done a little bit later showing stronger effects. We found that there were stronger associations in clinical groups, and we found stronger associations in males. Again, I think that’s all interesting food for thought. Is suggests to me that we might need to think quite developmentally about when and why does attachment affect behavioural problems, it may be that actually as children get older, their attachment relationships change. And they become more important in predicting later developmental difficulties with aggression, for example. It also suggests to me that there are measurement issues here, how you measure it makes quite a difference. And that probably needs to carefully looking at and the level of risk in the family comes up again, although actually, interestingly, in the other direction, to what we saw in the intergenerational transmission question. So a weak effect, not a strong one, if anyone tells you that attachment security and insecurity completely predicts, you know, how a child’s going to be doing and whether they have aggression later on, it’s just not true. At least so far, as far as we can tell, a sober reading of the data is there is a relationship. But there’s lots of other things that impact on how children develop. We shouldn’t be surprised by that, right? Human Nature is a complicated thing. Child Development is multi determined, we know that. And actually attachment theory just or at least the evidence suggests that that’s true in attachment, just like many other outcomes.

What about internalising problems, maybe all the actions there, maybe we’ll see whopping great effect sizes when we look at internalising problems. Well, these are the effect sizes for internalising problems from a similar review. And you can see the effect sizes are smaller, not bigger. I’ll put them side by side just for fun, that’s externalising again, that’s the one you’ve seen. That’s internalising and you can see it’s substantially stronger association between attachment and aggressive behavioural problems, let’s say, relative to attachment, and internalising problems like anxiety and depression. So again, a bit of a surprise there. And then what about social competence? Well, this is what we find for social competence. I’m going to do this in reverse order like some cheesy Game Show. So, coming in last is the internalising problems association, never even really pushing above a .2, next you see the externalising problems superimposed on that. And again, it just shows you how much more robust the association is for externalising problems. And then finally, when you add in social competence. That’s what you see. So social competence actually ends up being the most consistent and robust association.

It’s actually not dramatically different to the externalising one. That’s not a huge difference. Neither is that. That’s quite a big difference, interestingly enough, again, that’s fairly similar to what we saw in the case of externalising problems. So this is very interesting, I think, because it suggests to me that there’s something profoundly social about the mechanisms linking attachment to later outcomes, because partly we see the most consistent effects for social competence. And of course, we know that a lot of difficulties associated with externalising problems are really about social norms and about social relationships, aggression, peer relationships, and so on. So my suspicion is that that’s really why we’re seeing this more consistent association for your attachment and externalising problems and social competence in particular, and internalising problems so slightly, clearly, slightly different story. I should add here, just as a caveat that these are children up to about the age of 12 or 14, so basically pre adolescence, so the story could be different after that. So there you go. The final thing I wanted to highlight, I would really, for those of you who are interested in the role of attachment in social policy, and social care policy in particular, I’d really encourage you to read this paper because this paper is a position statement by very many people who’ve been working in the field for many years, in response to concerns around the way in which the concept of disorganised attachment is used in child protection context in particular,

We refer to various forms of evidence that suggests that it would be an extremely unwise idea to use the concept of disorganised attachment, certainly in isolation in any kind of child protection proceedings, and as forms of evidence. The assessment itself is extremely difficult and many people in in the social care well use the concept without even having been trained in how to assess it. And many people who have been trained struggled to do it very consistently, the measurement is actually quite susceptible to exactly how you run the procedure, even if you are going to run the procedure. And we already know from what I showed you before that disorganised attachment is not strongly implicated in very poor outcomes for children. It’s a weak and unreliable association. And we also know that there are lots of other causes of disorganised attachment, including social deprivation, parental psychopathology, and trauma and so on. So if we, if people are tempted to make a direct read across from disorganised attachment to maltreatment, which is happening in some contexts, this paper, please you’re not to on the basis of the empirical evidence.

Okay, so there’s the full story about where I think we are at more or less, in terms of the evidence supporting and clarifying the importance of attachment in children’s development, is it half empty or half full, I think I’ll leave it to you to decide. Thank you.

[round of applause]

I2: Thank you for an excellent talk, most stimulating, you were actually slightly ahead of time. So we’ve got plenty of time for questions, and then tea break. And then our final speaker, Jonathan green.

PASCO: In that case, can I show I did put on I said I would I would do some cheeky genuine myths. Can I just quickly, close if we have time, so I was going to I was going to skip it, but I called it, sorry. I think these things you can all agree with. Miss Bingo, anyone want to play? Insecurity of attachment in a fixed early period of development determines later outcomes? Yes, no?

I2: No.

PASCO: No, I don’t think so. This is a really weird kind of game show. Attachment must be with a mother?

[collective no]

PASCO: Thank you. Children only have one attachment or one main attachment.

[collective no]

I2: No.

PASCO: Very good. You’re such a good and well informed audience. Attachment doesn’t change?

[collective no]

PASCO: Absolutely not, clearly, it changes. Attachment is a social construct doesn’t exist in some societies?

[collective reply]


PASCO: I don’t know, I put a little hmm, there is a very interesting challenge from the anthropologists around attachment. I don’t think there’s enough really good research to be able to robustly say we know that that’s not true. My instincts are everything I’ve seen, and I have studied attachment in some very different, non UK contexts. And I am yet to see any society where I don’t think I can see attachment. But I would love to be proved wrong, because that will be so interesting, wouldn’t it? But, so I think the answer is almost certainly is very likely that it’s a universal construct, and is present in all societies. But wouldn’t it be great to do some really interesting and properly rigorous research that’s not just anthropologists pondering and going around with a notebook and having ideas but actually properly observing parent child relationships, in very kind of non-Western context, where the structure of care is extremely different. I think that would be very, very interesting. I’d like someone to do that, or maybe, maybe I’ll try and do it sometime. All problems experienced by maltreated children or who are in care are due to attachment insecurity?

I2: No.

PASCO: Yeah, you sure? Yeah. Okay, good. Steve has spoken up particularly loudly there, that’s right, attachment is obviously relevant to these children, but we can’t think that that’s everything. Oh I gave it away, oh damn.

I2: No.

PASCO: Okay, attachment can only be to an adult?

I2: No.

PASCO: You will you know what I think but what do you think?

I2: No.

PASCO: I think it’s a really interesting one. Some people take a very strong view on this, that, you know, Bolby’s original thinking was that the whole point is that you’re attaching to someone who is stronger, wiser, more, a more reliable, you know, provider of protection to you. Personally, I mean, I’ve seen communities where quite young children are often actually primary caregivers or very consistent caregivers. And I’m pretty sure I see attachments. But again, I think it’s really–

Q: The idea of a significant other.

PASCO: Yes, right exactly and who can be a significant other is a really interesting thing. And to me, it raises the whole question of what little we know about how that process happens. How does the child figure out? Do you know what I’m not going to attach to you’re a bit hopeless or whatever, I are going to go with this person or, you know, we know nothing about the way in which attachments form. For people are interested in the children who went through Romanian orphanages, for example, the question of how do attachment form, how does the attachment form? And what impact might there be when there are severe challenges to that process as it’s happening, in formation, I think it’s extremely clinically important, but we know nothing about it. We just wait until the 12, and we check when the strain situation has all happened by that stage. So that’s an interesting one. And I think the jury’s out, but I suspect that there was a beautiful presentation by Judy Mesmer at SRCD a few years ago, where she showed some videotapes of a little toddler playing, well a little infant playing with a four year old sibling, I think in the in the South American jungle somewhere. And they’re playing and having a nice time. And then something frightens the little baby. And the baby just goes straight to this four year old and the four year old, gets top marks on Ainsworth sensitivity scale, beautifully sensitive picks, and there is just this gorgeous little scene of the four year old patting the little one year old on the back, lifting him off the ground, which is quite impressive in itself. And so, you know, it really made me question, you know, what is the range and scope of the sorts of attachments that children might be able to form.

Okay, so that’s that one. Sorry. Disinhibited social engagement disorder, that’s me being disinhibited, is unrelated to attachment. This is controversial. And my view anyway, is that we don’t know yet. My strong suspicion is that Charlie Zener is right that disinhibited social engagement disorder is not as such a disorder of attachment, because some of these children show very normal looking attachment behaviour. But it may be the consequences of a severe disturbance to attachment processes happening early in life. And so there’s probably a halfway house there. The strain situation measures everything we need to know about attachment perfectly.

I2: No.

PASCO: Thank you. Disorganised attachment means the child is highly likely to been abused, neglected?

I2: No.

PASCO: Great. I think you got them all right. Thanks very much.

I2: Questions. We have 10 minutes for questions.

Q: Thank you. Would you be able to comment on attachment in cognitive development, not in terms of ability but in terms of thinking pattern?

PASCO: That’s interesting. It’s really difficult to say. I mean, I think it depends what you mean by thinking pattern. We do have some evidence that there is an association between attachment security and cognitive development measured somewhat traditionally. And that’s a weak association. It seems to be slightly stronger for language development than it is for sort of pure fluid intelligence skills, for example. But if you think much more broadly about sort of the idea of thinking patterns, then obviously, it’s much more difficult to answer but and also more interesting. So I mean, we, I reviewed some literature looking at the relationship between attachment and how children think in a sort of socio emotional realm. So were quite a lot of nice studies looking at how children will make sense of a social, a challenging social situation. Somebody is playing with your toy, and then breaks it, what do you think that means? How would you respond, and so on, and getting children to think about that. And they do see quite consistent relationships with secure children showing a sort of a more sophisticated and mature way of understanding, navigating and responding to social experience. So I think if I were if I had to put my money on it, I’d say that if that’s what you mean, to a certain extent, at least it partially overlaps with what you’re talking about, then I would say yes, I think there probably is some reasonable evidence that there’s a relationship there.

Q: I’m thinking more if the adult attachment interview, that measures other things?

PASCO: Oh, I see what you mean. Oh, I beg your pardon. Yes. No, well, I think yes. I mean, that’s….so I think the question is, is the adult attachment interview measuring attachment, is that what you mean? As opposed to thinking patterns or other processes? Well, if that’s what your question is, I’ll pretend that it is, because it’s a subject like that. I think it’s a really important and interesting question. So because with our twin study, one interpretation is that attachment was measured, you know, we measure it here, we then measure it here at 15, it’s the same construct, and it’s, the factors causing it have changed from one time to the next. Another interpretation is that we’ve measured attachments, I’m pretty confident the strain situation measures attachment, but then in adolescence, we measure something else, which is actually a psychologically more complex phenomenon that is, is informative about attachment, and may influence it, but may not be a measure of attachment. I’m actually fairly convinced that that’s probably true. And it certainly is not hard to understand why that kind of measurement, which is tapping into emotional regulation, how one sort of manages social interaction in an interview, how one process his early experiences reflects on mentalizing, and so on. That’s a really rich combination of psychological skills. To me, it seems that it’s not surprising that that shows some role for genetics, but also that it’s sort of obvious in a way that it’s not tapping into precisely the same construct of attachment that we think of when we talk about children. I’d love to know whether attachment, as in genuine, you know, proximity seeking, when frightened and distress, is still present in adolescence, and could be measured if we just had the right machinery to do it. For the moment, I don’t think we really do and that that will, you know, then you could actually tease apart properly whether this is a different construct, or it’s a different time that we’re picking up.

I2: Sorry, we’re just doing that microphone, we’ll get we’ll get to in a minute.

Q: It was just a quick related point, sort of the 18 seems a bit young to being an adult attachment interview, knowing what we know about adolescent brain development.

PASCO: Yes, I agree. Yeah, that’s absolutely right. And there’s there actually is a few, there have been a couple of studies that show that if you measure it at that age, things do change a bit. And it’s in fact, funnily enough, the patterns of response that you see, feed directly into all your stereotypes about teenagers. So you see a lot of dismissing responses, particularly amongst males, we found that in our child attachment interview study, so, so you’re right is picking up is something that’s probably a work in progress in terms of how attachment, or whatever we want to think of it as, is developing, having said that, when I listed all those papers, you know, the ones that I managed to find last night, some of those are well into the mid-20s, and some late 20s. So, you know, I’d be surprised if we found that by the time we get to a certain age, suddenly that continuity reappears. I mean, that that will be, it will be amazing, but it’s sort of unlikely, I think.

Q: I’m delivering some training next week. I’m a specialist teacher for social, emotional, mental health, to teachers and teaching assistants in our district. And I’m just wondering what you would like me to emphasise, what would be your key messages that perhaps they haven’t heard before? Because the only thing they’ve ever known about attachment is the strain situation and maybe a couple of bits about the key kind of types of attachment. So now we know a little bit more, what are your key messages to put out to the, to the education professionals who are working directly with these children in classrooms? And they’ll be primary school age?

PASCO: Absolutely brilliant question, oh it’s so hard. I think I suppose I would say, erm, erm, I would say, well, I think, I would, I think for teachers, it’s helpful to know that children do for the whole stream of research about whether or not children form attachments to their teachers. I think it’s quite hard to be definitive about these things. But I can say for sure that children need to feel that there are adults who care for them, and are responsive to their needs. And so the concept of nurturing, responsive caregiving, and feeling safe and understood, which are all kind of elements that we think are really important in attachment are really important for children wherever they go. And it’s especially important, I think, in school. Some children will need more of that from their teacher than other children, for all sorts of reasons to do with them and their families and so on. But that basic principle, I think, stands the test of time. What doesn’t stand so well is that there’s, you know, that their only source of kind of attachment related security is with a single figure and that’s their mum, it is nowhere near as straightforward as that. And I think it’s really important that they understand that not everything they see is about their attachments. There may be stuff going on with their peers, you might need to look at your, the way you manage the classroom as Tamsin was picking up earlier, so important. Is this child a child looked after, yes, it might be about attachment, but it might not be it might be that they’re experiencing trauma, or they have other psychiatric difficulties. So the lesson is, it’s important, but it’s not the whole story. So you have to actually put it all together in a sort of comprehensive view of what might be happening for this child. And remember that this child will probably value your relationship, the relationship they have with you, and you’re a very helpful source or feelings of comfort, and security.

I2: Next question.

Q: Thank you very much for the wonderful talk, John Richford, Oxford. Many years ago, somebody called Margaret Manning wonderful researcher, and sadly now dead, did some work on hostility in children. And she noted three types of hostility, and saw that some children were specialists in some sorts of hostility, although all children were capable of doing all three types. When Pat Crittenden talks about attachment, she talks about strategies that children use, the avoidance strategy, or the ambivalent strategy and so on. And I suppose one could say that people who get classified as avoidant are avoidance specialists, even though they’re capable of other…is there any mileage in looking at attachment in terms of strategies that where we’re capable of all these strategies, we tend to specialise?

PASCO: Thank you for that question. I absolutely agree. I think it’s a really great, great point. I think I had it buried in a slide. But I mean, I think there are profound measurement problems at the moment, with the way that we do attachments, in fact, I would go as far as to say that the attachment paradigm in terms of measurement tools we’ve got has, will run out of steam quite soon, because there are fundamental issues that have never been properly sorted out. And one of those is about the continuity, or sorry, of the continuum of attachment related behaviour. We’ve stuck with this kind of classifications, when we know I mean, for example, one really important take home message, by the way, for practitioners is that if you read the textbook, it says that children who are avoidant don’t tend to their attachment figures when they’re worried or upset or distressed. That’s just not true. It’s true in the very kind of refined circumstances of the strain situation. But if you do, home observations, Mary Ainsworth did lots of very rich and extensive home observations, where she found the children who were classified as avoiding in the strain situation certainly did approach their carers in the home. In fact, they were often the most fussy and difficult ones. So actually, it’s a much more dynamic process that we’re trying to capture here. And that it changes. And it’s probably as you’re saying, it’s about thresholds. So under certain circumstances, in the strain situation, what we’re getting at is something about the threshold that a child has, and for an avoidant child, their threshold to seek care when they’re a bit worried or anxious, you know, is a little bit higher, I think, for whatever reason, than a secure child. And that’s probably a better way of understanding it. But we don’t have tools that allow us to think about all of that nuance and flexibility and, and probably the continuum that exists between these kind of, at the moment, rather discrete categories that probably don’t capture nature at its joints.

The other thing is that I think, related to I think the other point you made, which is that the way in which children engage in social interactions, so going forwards a bit in terms of the example you gave to think about how they interact with their peers, I think is a really, really interesting question. We again, we know very little about that. There are clues, for example, that the children who are resistant, they show incompetent social behaviour, but are not aggressive and actually were more likely to get bullied or be anxious in social interactions, that the disorganised children might be a little bit more likely to become proactively aggressive, for example, but you know, there’s a whole interesting set of questions there that hasn’t been thoroughly studied.

I2: Right. Before we thank you again, I just want to remind you what’s happening now. We have 20 minutes of tea break, exactly. And then we’re back here in 20 minutes for Jonathan Greene’s talk. Then we have a panel discussion for 20 minutes where we’d like all the speakers to come back here. And then you can be thinking of further questions you’d like to ask the speakers for that last 20 minutes before we close. So thank you again to Pascal.

[round of applause]

Add a comment

Your email address will not be published. Required fields are marked *