The Myth of Autism is an ambitious work that seeks to encompass three broad areas of enquiry.

1. The social and cultural factors that influence our beliefs about childhood: Timimi et al.  concentrate on what they call “the cult of child development” and how it informs our child rearing practices and shapes our expectations for adult outcomes.
2. A critical review of the scientific literature on autism.
3. Recent political and cultural developments in Western society which form the context for our current understanding of the autistic spectrum.

The lead author, Sami Timimi tells us in the Preface that the science of autism is so unsound that it will be judged as a dangerous pseudoscientific fad, comparable to phrenology and eugenics.  Timimi acknowledges that this view is controversial and likely to be unpopular both with parents of autistic children and with autistic adults. He concludes that, “there is no such thing as autism and the label should be abolished.”

But this is followed by a statement a few pages later in the Introduction that, “If we believed that our position would have only negative effects on those with the label we would not have written this book.”

I find this incredible. If the “autism industry” they describe had no basis in science and was no better than phrenology or eugenics how could Timimi et al. not publish? Are they seriously arguing that we should judge science primarily by its social consequences even if that means giving the pass to pseudoscience that serves a useful purpose? That is the logic of their words but they seem oblivious to this contradiction. I think their intention is to soften criticism from the autism community by showing sympathy for our situation. But they do so at the expense of clear thinking.

From the outset the authors advise us that “Readers should be aware that we are not about to ask you to digest the umpteenth rehashing of the autistic spectrum theme, but are inviting you to dissect it, apply critical appraisal to the current literature and think through the implications of whatever model we choose as an ‘explanation’ for the diverse set of behavioural presentations currently being categorised as ‘autism’..”

I checked the References and found over 300 citations taking up 22 pages. Judging from the titles less than half these citations referred directly to autism. That in itself is not an issue. The authors are open about their intention to place their discussion of autism within a broader historical, economic and sociological context and it is right that this should be fully referenced. But the reference to current literature is an issue. According to Timimi, “We met in York in November 2006 and from here work started on the book in earnest.” The literature may have been current then but I counted only twenty five references dated 2007 or later. Six of these were books or articles by one of the authors. Only three of the twenty five, all from 2007, mentioned autism and one of those was a literature review.

This is a real weakness. Since 2007 some important books have been published that speak directly to the authors’ concerns. Grinker (2007) published an anthropological account of autism that examined its social and cultural context across the globe and was not just restricted to western societies. Murray (2008) wrote an account of the cultural discourse on autism in novels, plays and films. Barnbaum (2008) produced an original, if flawed account of the ethical dimension to debates about causation and cure. Feinstein (2010) has produced an important history of autism that draws strength from its face to face interviews with many of the pioneers of autism research. It also reveals the fierce divisions within the research community that give the lie to arguments about a hegemonic approach by an autism establishment of the type suggested by Timimi et al.

In the field of research Soulières, Mottron, Dawson and Gernsbacher are among those currently challenging the idea that all autistic features are necessarily defects. Tony Attwood has openly criticized DSM-IV criteria for Asperger Syndrome at the NAS International Conference in 2005. Eric Schopler, who developed TEACCH as a model for service delivery to autistic people in North Carolina and was one of the first to challenge the refrigerator mother ideology, disputed the claims of Ivor Lovaas that Applied Behaviour Analysis was the only scientifically proven treatment for autism. See Cohen 1998 for a summary of the dispute.

The primary educational goal of TEACCH is to increase the student’s level of skill. Recovery is not a term used in this system. While the Lovaas program is based on the premise that the child must overcome his autistic characteristics so as to adapt to the world around him, in TEACCH the child is provided with an environment designed to accommodate the characteristics of autistic children.

Although both men are now dead the argument still resonates, especially in the campaigns of parent advocates to make ABA mandatory for the treatment of autism and to oblige health insurance policies to pay for this treatment.  Timimi et al ignore these debates.

The history of Autism

The book not only fails to keep up to date with current thinking about autism, it also reveals a limited understanding of history. The authors go back to the Middle Ages when discussing the concept of childhood. But they delve no further than Bleuler in 1911 when considering the existence of autism. This ignores Frith’s exposition on the Holy Fools of Russia in “Autism; the enigma.” Then there is Itard and the Wild Boy of Aveyron from 18^th century France. Closer to home, Downs published an account of children who were arguably autistic in 19^th century London (see Treffert) and Shattock and Waltz found similar cases when trawling Victorian health records.

There is a constant message that case histories are privileged over data. Data cannot be trusted because it is collected in order to justify the pre-existing bias of the scientists. This bias derives from the priorities of the society that pays their wages. What is not clear is how Timimi et al have escaped from this straitjacket. Have they found the truth or are we just being invited to select a more meaningful narrative? Timimi talks of his own clinical practice in which therapy is based on “more meaningful narratives … than those attached to the superficial focus of ‘symptoms’ and ‘diagnosis’. By what measure are these narratives to be judged. Timimi et al. argue that the claims of science have to stand up to Popperian tests of falsification.   Do the same standards apply to personal narratives? Having read the three personal narratives with which the book opens, the answer is obviously, “No.” There is a serious mismatch here. We are invited to dismiss the science of autism on scientific grounds and replace it with personal narratives that cannot meet the same tests. This is a catergorical mismatch that places us firmly in the post modernist mire.

Autism: a disability or a disadvantage?

Timimi et al dismiss the autism rights movement in six pages, in part because in their view it accepts the monolithic spectrum handed down by the autism establishment. They offer no citations from the work of anybody on the spectrum. In fact autism rights campaigners are far from monolithic in their views. Some autistic self advocates are influenced by the gay rights movement. There is a reason for this. When the psychoanalytic dominance of autism theory was successfully challenged by those who argued that it was an organic condition and not the result of bad parenting Skinner’s theory of operant conditioning was proposed as an alternative. Ivor Lovaas held that early, intensive behavioural therapy could reverse autism. Timimi et al. discusses Lovaas but, in a book where gender looms large it is surprising that they do not comment on a previous study by Lovaas to reverse homosexuality in boys at a time when it was still regarded as a psychiatric disorder. This fact was not lost on the early pioneers of the autism self advocacy movement who looked to the emerging gay rights movement as a model for their own efforts to have autism regarded as a difference rather than a disease.

Others look to campaigns by disability rights activists.  The social model of disability has attracted support from campaigners who understand that that if society were to make accommodations to meet their needs, creating autism friendly services for example, we could do much to mitigate the drawbacks that come with autism. Then there are those like Ros Blackburn who are open about the disabling impact of autism on their lives and would welcome medical interventions if they were available.  This should not be taken to imply that the autistic community is divided into separate camps. The arguments are more nuanced than that.

Rita Jordan attempted to tackle this in an article in Communication, the members’ magazine of the National Autistic Society. It is the one issue that is missing from my files (Winter 2007) but if memory serves she proposed the term Autistic Spectrum Condition in order to accommodate those who see autism as a way of being and not a disorder. Some people’s autistic condition might be so disabling as to necessitate a lifetime of care. Others might manage most of the time and only need support at key moments of crisis or transition in their lives. While others need not be considered disabled at all if only other people were aware of and respected their difference. This drew criticism in the next issue from autistic people who were leading successful lives and would not count as medically disabled under such a definition. Nevertheless they had to work very hard to fit in and cope with bullying, exploitation and the constant stress. Another autistic person referred to his problems with communication and short term memory along with anxiety, sensory overload and repetitive behaviour, all of which led him see his autism as a deficit. I believe Rita Jordan was right to tackle the question. I do not think we can properly judge the success of her endeavour on the basis of a thousand word magazine article. I do find it significant that a leading figure in the alleged “Autism Establishment” as characterized by Timimi et al. was sensitive to the aspirations within the autistic community and felt it right and necessary to respond in a magazine read mainly by parents who make up the majority of the membership of the NAS. Similarly, the concepts of Neurodiversity and Neurotypicality which were developed by autistic people are increasingly used in professional discourse about autism.

The Science of Autism

Timimi et al. are at their most interesting when they challenge the nature of autism science. They place great emphasis on the importance of case histories and personal narratives. Indeed we are treated to 25 pages of the authors’ own personal narratives (pages 15 – 41) before the book properly begins. Their critique of the science of autism is really a critique of social science in general. It is ideologically driven and heavily influenced by the physical sciences in its attempt to “measure physical and mental competence” (page 91). Timimi et al accept that science is governed by Popper’s principles of hypothesis making and falsifiability. But this narrow view of science is not universally accepted either by working scientists or philosophers of science.  It is not even accepted by Popper himself, who, according to Rose (2005) had to modify his position when it became apparent that, by his criteria, Darwinian evolutionary theory was unscientific.

But it suits Timimi et al to define science so narrowly in order to contrast “true” verifiable hypotheses of real science with the pseudoscience of autism that fails the Popperian test. They attack the data because of its alleged weakness. But they also attack the idea that any data about autism can be reliable. This is because of the ideological construct they place upon autism science. If they can show historical, economic and sociological influence on scientists and clinicians this is enough to invalidate the science and leave the way clear for their own personal narratives.

For people who display their left wing credentials so prominently throughout the book this shows incredible ignorance and misunderstanding of socialist theory of science. When Hessen delivered his paper on The Social and Economic Roots of Newton’s Principia in 1931 he was not suggesting that Newtonian physics should be dismissed because it was influenced by the socio-economic demands of the time. Rather it should be understood within that sociological framework. Timimi et al pose a false dichotomy between true science that is evidence based and false science that is ideologically led. In fact all science is influenced to some extent by social and historical factors and has to be understood in that context.

Their specific claims about the data are not always rigorously argued. They are at pains to demolish the myth that autism is genetic while ignoring the fact that in 1993 Simon Baron-Cohen and Patrick Bolton published a book, Autism the Facts, which clearly stated that on one model, “The final common pathway,” genetics was only one contributory factor alongside viral infections, birth and pregnancy complications and other causes as yet unknown. Timimi et al seek to discredit twin studies as the major pillar of the hypothesis that autism has a genetic content. But they only consider five papers published between 1977 and 1996. Yet, the first decade of this century produced over thirty twin studies.  Some of them are products of the Twins Early Development Study (TEDS)   TEDS researchers recruited 15000 families after contacting all the parents of twins born between 1994 and 1996 in the UK. Fifteen years later 13000 families are still participating and 5000 have provided DNA samples. This study has generated a number of papers and goes a long way towards meeting the objections of Timimi et al.

They question the reliability of parental reports of zygosity in the twin studies they review. As early as June, 2000 TEDS found that parental reports were reliable.  They also object to selection bias. Twins are recruited because one of them already has a referral. Researchers know that one is autistic. They assume a genetic basis and are predisposed to “see” autism in the second twin. But TEDS recruited thousands of twin pairs without reference to diagnosis. They came across their autistic twin pairs by accident as it were.

The issue of shared environmental experience or the equal environmental assumption (EEA) is another matter. Timimi et al suggest that we cannot assume an EEA between MZ and DZ twins, still less between twins and singleton siblings. And they are right. TEDS has confirmed this in a series of papers that are not directly concerned with autistic traits. With hindsight it is only common sense that twins have a different environment to their siblings. The shared environment of the womb is denied to singletons. After birth identical twins experience a different upbringing to their non-identical and singleton siblings.

Common sense is one thing. But now we have some data.  A recent American study claimed that, “Susceptibility to ASD has moderate genetic heritability and a substantial shared twin environmental component.” This paper is not without its critics but I include it because it lends weight to the arguments of Timimi et al. Why didn’t they include it?

I have some sympathy with their criticism of the science around autism and mental health issues in general, especially the perennial claims of genetic researchers that they are close to finding a cause and hence a cure for autism.But the autism rights movement that they deride has made some of the most trenchant criticisms in this area without denying that autism exists, that it can be disabling and that people with autism are entitled to support and services.

I do object to the muddled thinking and poor research in Timimi et al. They do not do their argument justice and if taken seriously are likely to set our movement back. But the authors sacrifice any claims to be taken seriously when they indulge in personal attacks on autism researchers. Simon Baron-Cohen is accused of McCarthyism over the questions in the Childhood Asperger Syndrome Test. They accept the criticism of Professor Gillberg who risked his professional reputation and career rather than breach confidentiality agreements with reseach subjects and hand over his data to anti-psychiatry campaigners influenced by the Church of Scientology. All this is documented in the British Medical Journal. The authors provide no sources for their uncritical acceptance of the attack on Gillberg.

Worst of all is when the authors commit egregious errors like stating that John Wing was Lorna Wing’s father when he was in fact her husband. They demean themselves with the suggestion that Lorna Wing embraced autism as a result of her having a severely handicapped daughter and discovered the milder form, Asperger Syndrome when she gave birth to a less handicapped child in 1981. This is an even more serious error. I do not know which is worse, mistakenly claiming that Dr Wing had a second child in 1981 (she did not) or suggesting that she made up the science to suit her personal circumstances.

In conclusion the authors raise some important questions but provide disappointing answers. The Myth of Autism is marred by dated research, confused arguments and is potentially libellous in places. Avoid this book.