The Empty Fortress

There are books that, more than others, carry legible traces of their history. This one is well-worn and shows signs of successive repairs. During one of these repairs, the original paperback cover was cut out and glued onto a sturdier one made of cardboard, preserving what was left of the image of a child pensively holding a doll. There is no tenderness on either of those faces – the child or the doll’s – just a sort of detached seriousness.

Because it comes from a library, this book includes a borrowing record, although there is no way to tell if it is exhaustive. The earliest stamp on the sheet affixed inside the back cover is dated 27 May 1982.

Who are you, who read the book at that time? Was it you who underlined in pencil the description of disabled children as ‘self-contained, narcissistic and empty’, or the part about the causes of their ‘emotional and intellectual death’? And what was your connection to those statements – did you evaluate them dispassionately or did they touch you personally? Were you sceptical or did you believe?

It’s thirty years later and the book still sits on its shelf at the library of my university, on the same floor as the letters of Antonio Gramsci, whose history it mirrors in reverse: one is an unlikely collection of writings by a man whom the state had tried, but failed, to extinguish intellectually; the other a work of bad scholarship whose popular appeal and institutional backing made invulnerable to criticism. One symbolises struggle and testimony; the other the persistence of an error. At Victoria’s library, both books have survived a number of cullings and cutbacks. Both books, as objects, carry material yet elusive signs of their successive encounters with their readers.

The Empty Fortress was first published in 1967 and is generally regarded as the text that either originated or popularised the ‘refrigerator mother’ theory of autism (neither charge is entirely correct, as I’m going to discuss later). The book’s author, Bruno Bettelheim, was an Austrian concentration camp survivor most famous in the field of psychology for a 1943 paper on the effects on the psyche of what he called ‘extreme situations’. The paper – one of the earliest attempts to document life in the Nazi camps – was largely based on Bettelheim’s own observations.

By 1967, Bettelheim had been for over twenty years professor of psychology at the University of Chicago and director of the Sonia Shankman Orthogenic School for emotionally disturbed children. It was in this setting that he engaged in the study of autism, a developmental disorder first (and independently) described by Hans Asperger and Leo Kanner in the early 1940s. One especially strong factor that drew Bettelheim to this condition was the resemblance of some autistic behaviours to what he had observed at Birkenau amongst the prisoners who exhibited extreme emotional withdrawal and a seemingly passive resignation to their fate – those that in the argot of the camp were called “moslems”. The book’s intent is therefore twofold: firstly, to argue that these two kinds of autism are manifestations of the same condition, in children and adults respectively; and secondly, that their common psychogenic origin points to psychoanalysis as the only effective form of treatment.

The latter hypothesis shaped the understanding of autism not only amongst laypeople but also in large parts of the scientific community for decades to come, and still enjoys considerable favour in some countries in spite of having been comprehensively discredited. However the larger issues is how it could possibly have been credited in the first place. Bettelheim’s book is so riddled with fallacies and circular reasoning, so devoid of scientifically falsifiable evidence – save for the results that he claimed to have achieved with the children treated by the school (it was decades before the data from his small and likely misrepresented sample was found to be unreliable) – as to beggar belief that it was ever taken seriously. In their study of the fortune of Bettelheim's ideas, Katherine DeMaria Severson, James Arnt Anne and Denise Jodlowski attribute this success to a conjunction of factors, namely

Post-war fascination with Freud and the Holocaust; his careful management of his public image; his careful attention to writing for a popular audience; and a generalized anxiety about the family in 1950s and 1960s America (Autism and Representation, 68).

The first and the last of these factors seem particularly relevant in the context of the persistent tendency to consider autism as a cultural condition, irrespective of its psychological, neurological and social specificity. When contemporary reviewer Eliot Fremont-Smith wrote in The New York Times that ‘The Empty Fortress is […] as much a philosophical and political book as it is a scientific one’ – a statement which he intended as praise and was duly recycled on the jacket of later editions of the book – he unwittingly emphasised one of the most damning and at the most time seductive aspects of Bettelheim’s rhetorical strategy, which consists in making statements of cultural as opposed to logical or scientific appeal.

To this we must add, in Bettelheim’s partial defence, that for contemporary audiences the contention – advanced by both Kanner and Asperger – that autism likely had a primarily organic (we would say genetic) origin was tantamount to denying that the life of the people who suffered from it could be improved. Belief in psychoanalytic treatment could therefore be perceived as the humane option, and in a sense it was: the prescription of conventional psychiatry and medicine at this time was long-term institutionalisation, frequently coupled with electroconvulsive therapy. Psychotherapy wasn’t the only alternative, but it was an alternative nonetheless. However in order for this option to be open at all, in adherence to Freudian theory it first of all had to be postulated (rather than argued or demonstrated) that autism had psychological causes that could be reversed by means of psychotherapy. The origin of autism had therefore to reside in severe psychological trauma analogous, in Bettelheim’s view, to the kind that made some concentration camp prisoners revert to a similar state as adults.

Enter the theory of the refrigerator mother, which wasn’t of Bettelheim’s own coinage but originated in comments that Kanner himself made at the time of his initial studies, when he observed that ‘emotional refrigeration’ was a common feature of the families of autistic children. Later he suggested that the condition might be related to ‘genuine lack of maternal warmth’. Bettelheim took this idea and put it at the centre of his theory of autism. However in so doing he also made it considerably more sophisticated and cruel.

Firstly we must note that the book is highly contradictory on this subject, in spite of its being absolutely central to Bettelheim’s entire thesis. Thus the author protests that ‘it serves no good purpose to make the parents of autistic children feel guilty as having caused the disturbance’ (403), and warns against the ‘myth of the perfect, all-giving mother we all wish we had had’ (28), or even goes as far as suggest that the mother’s role wouldn’t be so central (therefore so catastrophic when she proves to be inadequate) if the raising of infants wasn’t almost solely devolved to her for socio-historical reasons. However these few scattered comments do nothing to blunt the sustained attack mounted in the book against the parents – and most especially the mothers – of autistic children, an attack which admits no possible defence.

In light of the analogy with the concentration camp explicitly drawn by Bettelheim, we might expect him to trace the cause of autism to some comparably severe form of abuse or neglect, but of course that would clash against the observable fact that this is very seldom the case. The author’s only recourse therefore is to posit that there are critical times in the life of an infant when she must develop a deep bonding with her mother based on the carefully balanced and timely exchange of emotional signals. A mother that nurses her baby at fixed times, or who responds either too promptly or not promptly enough to the baby’s various demands, or allows the child to experience the world as frustrating at a particularly vulnerable time, might trigger the autistic Anlage and the inexorable onset of the condition. So it is not simply a case of a mother being ‘cold’ or ‘distant’ – for this too might not chime with real-world experience and call the hypothesis into question. The only way to lock in the Freudian approach is therefore to propose that even a warm or at any rate ostensibly warm mother could fail to read the signals correctly or read them but fail to respond correctly. In one of the book’s most chilling passages, Bettelheim observes:

A mother may be experienced as rejecting by the infant for a multitude of different reasons, connected with either her conscious or unconscious attitudes, her bodily or mental defects, her physical presence or absence, her unavoidable libidinal preoccupations, her aggressions, her anxieties, etc. (69)

In other words: there is no recipe to being a good mother, but if your child develops autism then it means you’ve been a bad mother. And in case the above passage might suggest to you – as the author himself seems inclined to provisionally conclude – that the ‘perception of rejection’ may derive from the child’s own neurosis and not constitute evidence of actual rejection, in the closing chapters Bettelheim shuts that particular door when he writes that ‘the background of all autistic children’ is that they are ‘utterly unacceptable to their parents for one reason or another.’ (355)

That is the indictment, finally: that while the behaviour of parents (by which again we mean mostly mothers) may vary widely, and even appear to be affectionate, the psychological injury stems from genuine rejection. And if a mother protests that she loves her child and seem convinced and convincing in doing so, it will be a simple matter of informing her that she harbours unconscious feelings of rejection.

There are many more examples in the book of such circular thinking, although few of them are quite as callous or as central to its argument. Bettelheim duly reinforced this core premise by selecting for his case studies three children from highly dysfunctional if not downright abusive families, and indeed it’s been speculated that he might have favoured children from that particular kind of circumstances for the institution’s very limited intake, which was of no more than seven or eight patients at a time.

Image via the Center for Social Media

By refining and expanding the refrigerator mother theory with the armoury of Freudian psychology, Bettelheim ensured it would speak to an audience that was culturally predisposed to hear such arguments, thus ensuring the popular legitimisation of his work, the vast increase of his personal prestige and a steady flow of benefactors and funds. Of course nowadays the appeal of pseudo-scientific hypotheses on the grounds that they fit in with dominant beliefs about causation is hardly diminished: so not only are Bettelheim’s ideas still in vogue in some countries, notably France and South Korea, but the equally disproved vaccination theory continues to enjoy significant support, causing unnecessary anguish and diverting resources from the investigation and treatment of autism.

In both instances we are dealing – and in a way that resonates with the autistic experience itself – with the problem of how we respond when faced with the limits of our knowledge. There is so much about the proximal causes and most importantly the dynamics of autism that remains mysterious to us, at the same time as we continue to seek better treatments and more effective ways to improve the lives of people on the spectrum, an effort that turns us increasingly towards treating society and questioning the notion of normalcy. But to acknowledge the mystery means to acknowledge the person. So this strikes me now as the single most enduringly offensive aspect of Bettelheim’s work: that cover, that full title, with its implication – that the fortress is empty, that the self with whom we are unable to communicate has no feeling and does not exist. This is the idea that more than any other we are still called upon to reject.

Bruno Bettelheim. The Empty Fortress: Infantile Autism and the Birth of the Self. New York: The Free Press, 1972.

This seems like the right place to recommend Hilary Stace’s doctoral thesis on autism policy in New Zealand, which you can read about and access from here. 

For a critique of the psychoanalitic approach to curing autism still prevalent in France, see – and consider supporting – Sophie Robert's corageous documentary Le Mur ou la psychanalyse à l’épreuve de l’autisme, available here with English subtitles.