Saturday, November 08, 2025

The origins of the concept of autism

The concept of childhood autism originated independently in the work of Leo Kanner and Hans Asperger. The original paper by Kanner (1943) ‘Autistic disturbances of affective content’ has been re-examined by Sausalito & Kocha (2025) in History of Psychiatry. Kanner described 11 cases in children with what he called "extreme autistic aloneness". Eugene Bleuler (1911) had identified the symptom of an active turning away from external reality to a preponderance of inner life as ‘autism’ and saw it as fundamental to ‘schizophrenia’, a term which he originated (see eg. previous post). Kanner emphasised that his cases were unlike schizophrenia because such aloneness had always been present rather than being a withdrawal from participation in the outside world. As Sausalito & Kocha point out, this relationship between Kanner’s autism and childhood schizophrenia tends to be overlooked today. 

Michael Rutter (1978) clarified the definition of autism based on Kanner's paper by identifying the key symptoms of general failure to develop social relationships; language retardation with impaired comprehension, echolalia and pronominal reversal; and ritualistic or compulsive phenomena. He also emphasised that autism and learning disability frequently coexist, although differentiating autism from learning disability on the basis of particular cognitive deficits that involve language and central coding processes. When I trained Rutter's perspective tended to influence the diagnosis of autism, which was rare, certainly in adults without a childhood diagnosis, and, if made, tended to be related to learning disability.

Another paper in History of Psychiatry focused on the history of Asperger's autism (see previous post). Lorna Wing (1981) resurrected Asperger’s original 1944 German paper and popularised the term Aperger’s syndrome, introducing the autistic triad of difficulties in social interaction, communication and imagination. She saw autistic spectrum disorder (ASD) as a group of disorders of development with life-long effects, having this triad of impairments in common. The spectrum included but was wider than the original syndromes of Kanner and Asperger. Rise in the incidence and prevalence of autistic diagnoses followed the change of criteria and the spectrum has become increasingly recognised. ASD over recent years is seen as part of neurodiversity, reframing various conditions such as ADHD, dyslexia and autism as differences rather than deficits (see eg. previous post). These historical perspectives on autism help to make sense of the epidemic increase in neurodevelopmental diagnoses and highlight the subjective aspect of psychiatric diagnosis. 

Psychiatric diagnosis is merely a hypothetical construct. We too easily assume such concepts are entities of some kind. They are justified by their clinical utility, but the ever expansion of the notion of neurodiversity questions its value. Diagnoses are relative rather than absolute with inevitable fuzzy boundaries between syndromes. More fundamentally, speculative notions about the causal biological basis of psychiatric diagnoses, like autism, mean than people tend to be reduced to their brains when they are given a psychiatric diagnosis. People are more than their brains and the reasons for their mental health difficulties, such as autism, may not be explained by brain abnormalities at all (see eg. last post). The kind of person they are, for example, may be more to do with their family origins and personal development than their brain. 

When I trained, schizoid personality disorder (SCD) was a diagnosis that overlapped with what is now seen as autism. It was originally a non-psychotic diagnosis of people that shared the feature of social aversiveness in schizophrenia. The key symptoms of SCD are social isolation and emotional detachment. It tends to be a diagnosis made in adults whereas autism was a diagnosis originally made in children. SCD in adults has now been replaced, at least to some extent, by higher functioning autistic spectrum in people without an original diagnosis in childhood, although signs of autistic traits in childhood are recognised retrospectively. Correspondingly, ASD diagnosis in children has also increased. 

The subjective aspects of psychiatric diagnoses are to be expected because diagnoses are merely idealised descriptions. There will be inevitable unreliability in their application. Of more concern is whether the diagnostic construct is valid. What worries me is when concepts like autism are seen as natural kinds, in the sense of being seen as reflecting the objective structure of the world. People are inevitably different and how well they relate to others and express themselves will of course vary. Blaming the brain for such difficulties may be more to do with wishful thinking than reality (see eg. previous post). Autism, like ADHD (see eg. previous post), is not a neurological condition as such. Of course it is mediated by the brain, but it’s not necessarily caused by it.