‘I have been fortunate in my professional life’

I set out wanting to study the mind and to help people in mental distress. I have been able to follow this path (although it is moot how much I've helped people) on the basis of three sometimes complementary, sometimes potentially conflictual, trainings: as a physician and then psychiatrist, as a developmental psychologist (with a PhD in experimental psychology from the University of Cambridge), and as a psychoanalyst trained in the British Psychoanalytical Society.

For a less than practical, bookish young man, my induction into medicine and surgery was an ordeal by fire. It was an experience I would not repeat, but I am grateful that I became a physician in the NHS when it was a happier institution than now, and had the privilege to serve ill and sometimes dying people.

Science and a 'quasi-religious aberration'

When I took up psychiatry at the Maudsley Hospital, which was much more 'me', I picked up the threads of a philosophically inclined research interest from my undergraduate days that concerned the early growth of interpersonal understanding. The philosophical part came from forays into philosophy of mind. I had been impressed by arguments (from Strawson) that the concept of 'persons', and not 'body' or 'mind', is foundational for our understanding of mental life, and (from Wittgenstein) that our shared form of life involving agreement in judgements provides a background for individuals' capacities to think and use language.

These thoughts were in the back of my mind when, about mid-way through my psychiatric training, it was suggested I visit a school for children with autism. I left the school that day convinced that the study of autism might reveal something profound about the bases for personal relations, shared experience, and our knowledge of persons-with-minds.  

At first I was supported in my research by an MRC Training Fellowship and two Senior Fellowships. Then there seemed no way to sustain a commitment to both psychotherapy and developmental research. A very famous and influential child psychiatrist told me I would need to decide between my interest in serious science and what he deemed to be a quasi-religious aberration, namely my involvement with psychoanalysis. In the event, I moved on to a specially created position at University College London that was linked with a clinical appointment as psychoanalytic psychotherapist in the Adult Department of the Tavistock Clinic.

At the time I embarked on research in autism, theoretical accounts of the disorder marginalised what Leo Kanner had described as autistic children's biologically based impairments in affective contact with other people. There appeared to be two reasons for this: firstly, affective contact was not so easily measured as cognitive and linguistic abilities, and what could not be measured had questionable status; and secondly, many psychologists could not countenance the possibility that disorders of interpersonal relations involving feeling could explain specific profiles of cognitive disability and ability.

I felt I might challenge these prejudices by designing experiments. Experiments can create a bridge between abstract developmental theorising and the nitty-gritty of children's behaviour and experience. One way to capture aspects of social-affective events, for example, is to demonstrate that independent raters can agree in their subjective (but not idiosyncratic) judgements of interpersonal goings-on recorded on videotape. Simplistically but with some success, colleagues and I have used this approach to measure limitations in affective contact and other aspects of communication between children with autism and other people. Or again, if we could predict what might seem tiny, even trivial, details of a child's behaviour to emerge within the structure of a controlled experiment, we had a means to explore social-developmental truths about autism and typical development.

Let me give one example. Imagine you have a coloured sticker placed on your forefinger. A typically developing child of, say, three or four years sits opposite on a mat. You hold up your finger, and say: 'Where shall I put my sticker?' The child points to her own shirt. You put the sticker on your own shirt, and the child smiles and nods. The child has identified with your propensity to identify with her own actions. She has anticipated you will understand her point to her shirt as communicating that you should put your sticker on your shirt. How do children with autism respond? By failing to communicate where to place the sticker? No, our participants with autism conveyed the message perfectly well. But whereas the majority of matched control children without autism behaved like the typically developing child I have described, most of those with autism either pointed directly towards the tester's body, or they verbalised their instruction. We had predicted this subtle but telling group difference on the basis of a hypothesis about autism, namely that affected individuals have a limited propensity to identify with the stance of other people – where 'identifying with' has cognitive, conative and affective aspects. A programme of studies on imitation, non-verbal and verbal communication, self-consciousness and empathy has reinforced our view of the value of this hypothesis.

A convergence

'Identifying with' is a psychoanalytic concept. How could psychoanalysis, according to some psychologists the very antithesis of science, yield such a powerful explanatory notion for accounts of mental development and developmental psychopathology?

Psychoanalysis has its own form of coherent investigative procedure. True, psychoanalytic thinking is prone to over-extend its domain of application, just as more respectable scientific branches of psychology are prone to tunnel vision and reductionism. Yet psychoanalysis represents a disciplined approach to the study of intersubjective experience, and in my view, intersubjective (that is, interpersonal) experience is an essential ingredient of our knowledge of what a mind is, and how minds work.  

But surely, psychoanalytic research is not worthy of the name, because it is irreducibly subjective? Not necessarily: in published research involving participants with borderline personality disorder and others with chronic depression, colleagues and I demonstrated the objectivity of clinicians' judgements about the presence of psychoanalytically defined paranoid-schizoid and depressive position functioning in videotaped excerpts of assessment consultations. Psychoanalytic procedures, and specifically the analysis of the transference relation between patient and analyst, yield insights into how we tick, and with this, a basis for conceptualising how development takes place. Perhaps it is no wonder that there is convergence between the psychoanalytic notion of 'identifying with' and ideas from elsewhere in developmental and social psychology, notably those of Vygotsky and G.H. Mead. And perhaps it is not so surprising that this concept from psychoanalysis proves especially apt, and perhaps necessary, to explain results from controlled experiments involving children with autism.

One final methodological thought on developmental psychopathology. Freud was among the first to recognise the value of studying typical and atypical development alongside one another, to achieve new depths of vision. So, too, there are insights to be gained from studying 'atypical atypicality'. Let me return to autism. Suppose there are diverse developmental pathways to autism, which is, after all, a constellation of clinical features (a syndrome) that happen to co-occur. With this idea in mind, we have conducted studies of congenitally blind children, among whom the syndrome is prevalent and sometimes unusual in presentation. Our hypothesis is that these children are at risk for developing this particular cluster of clinical features for a period early in their lives, because they are compromised in their engagement with other people's attitudes towards a shared, visually specified world – and therefore, at a disadvantage in coming to coordinate, understand and move among different mental perspectives, a critical problem for sighted children with autism.  The question arises: if autism is a heterogeneous developmental disorder, and not, as often supposed, a behavioural print-out of constitutional brain dysfunction (which is not to deny that neurological disorder is often causative), then how come it is relatively coherent as a syndrome? To seek an answer to this question, we would do well to trace the developmental implications of specific forms of limitation in very young children's experience of self-other-world relations.

I hope I have conveyed how much I respect, indeed relish, the methods of science. Having said this, it has been reflection on the nature of interpersonal experience, enriched by psychoanalysis and philosophy, that has inspired my own attempts to apply such methods to address vital but elusive issues in developmental psychopathology. It's a journey that has led me to explore the very origins of thinking, for example in my book The Cradle of Thought. And if this piece is more intellectual/academic than most of the 'Careers' profiles on these pages… well, that's just me!