curious incident

Gregory Hollin is a researcher at Nottingham University’s Institute for Science and Society whose work explores changing perceptions of autism, both scientific and cultural. He recently wrote a superb essay for the anthropology website Somatosphere which looked at the relationship between the rise of autism as a cultural phenomenon and changing scientific conceptions of human nature and of the ‘social’, and at how these changes have led to the view of ‘the social hole in autism’ as ‘a window to the soul’. My thanks to Greg, and to Somatosphere, for allowing me to republishing the essay on Pandaemonium.


Gregory Hollin
Autism, sociality and human nature

There are, I believe, a few reasons to suppose that autism is a particularly fascinating area to be studying at the moment. What are those reasons? Firstly, prevalence rates of autism have soared in recent decades, from 1:2,500 in 1978 to around 1:100 today: a staggering 25-fold increase. Secondly, and simultaneously, the nature of those receiving a diagnosis of autism has changed considerably. To give just one example, in the 1980s no more than twenty percent of individuals diagnosed with autism had an IQ above 80. Today, by contrast, it is widely argued that ‘intellectual disability is not part of the broader autism phenotype…[and] the association between extreme autistic traits and intellectual disability is only modest’ (Hoekstra et al 2009: 534). Whatever you make of IQ scores, this changing profile means that it is reasonable to assume that when you meet somebody with autism today they are quite unlikely to be similar to someone you would’ve met with the same diagnosis just thirty years ago. Thirdly, as the number of people diagnosed with autism has increased, and as the capabilities of those individuals has increased, a (self-)advocacy network of enormous importance and influence has arisen, perhaps on a scale hitherto unseen. When woven together, these dynamic elements have led Ian Hacking to claim that, in autism, ‘we are participating in a living experiment in concept formation of a sort that does not come more than once in a dozen lifetimes ‘(Hacking 2009: 506). This, I think, is quite exciting.

Finally, and perhaps differently, over the past thirty years autism has become an all-pervasive cultural experience. ‘Autistic fiction’, for example, has become a recognised genre. And when I talk of ‘autism fiction’, think not only of Rain Man and The Curious Incident of the Dog in the Night-Time but of all those times that autism is used as a ‘prop’ or ‘prosthetic device’ to explore humanity in toto (Murray 2008: 163). Just last week I found myself watching The Machine, a dystopian film in which badly brain-damaged war veterans have computer chips implanted into their brains with the aim of allowing them to return to ‘normal functioning’ (read: become super-soldiers). As you might imagine, this experiment does not end well. What I find particularly interesting, however, is the manner in which these scientists come to realise that these militarised cyborgs are less than human: they fail the Sally-Anne Test, one of the oldest psychological tests for autism. ‘Facts are just facts’ says Paul the cyborg, unable to grasp that the world could appear different to a second person. And so it is within The Machine: as with a great deal of fiction (and, as I’ll argue below, within particular academic disciplines) what is missing in autism is taken to reveal something fundamental about what needs to be present in order to be human.

How did this situation occur? How did autism which, until quite recently, was an unusual diagnosis of little broader concern, come to hold a central place in debates over human nature? That’s what I’d like to think about in this essay. My argument, in short, is that the thing which is ‘missing’ in autism, crudely put, is assumed to be social functioning and this is crucial when it comes to understanding why autism is taken to be so important for the human.


Re-constructing the social

Those working in disciplines such as science and technology studies and anthropology are used to debates over the nature of ‘the social’. It is perhaps more surprising, however, to find that the experimental human sciences have engaged in similar debates (see, Danziger (2000) for an overview). Consider someone praying, alone, in front of an altar. Is this a social behaviour? Most psychologists working before 1950, certainly 1920, would probably have answered ‘yes’; the activity is demonstrably being shaped by, and takes the form it does, because of that person’s previous social experiences and group membership. It seems exceedingly unlikely that someone who had never been immersed in the traditions of the church would find themselves praying at this alter, in this physical position. When the social is understood in this way, it is not that some behaviours (or cognitions, or attitudes, or emotions…) are always social and others are always non-social; rather, individual background and context is crucial to making a judgement. John Greenwood gives the example of a pro-life stance on abortion – this could be a social belief if it is held because of, for example, membership of the Catholic Church, and it could be a non-social belief if that belief has been arrived at ‘individually’ and ‘rationally’ (Greenwood 2004: 21).

the machine

If you were to ask experimental social psychologists and neuroscientists the same question today, we would find the opposite answer most frequently given: praying is not a social behaviour. The reason for this is that, within today’s experimental psychology and neuroscience, the social is characterised by two features. Firstly, within contemporary thinking, the social refers to objects of cognition (the things which our cognitions are directed towards) and not forms of cognition (the particular shape of those cognitions). Cognitions, or behaviours, which are present, or altered, by group membership (such as praying) are not social under this framework. Instead, a social cognition is simply one related to the understanding of other people in the immediate vicinity. Similarly, a social emotion is something like empathy or sympathy: an emotion whose object is another person. The social is, therefore, now primarily about interpersonal engagement. Kurt Danziger has called this construction ‘a social in the shape of a crowd’, a concept that captures the idea eloquently; immediate, interpersonal behaviour is always social, while nothing else ever is.

A second feature of this novel understanding of the social is that the cognitive processes responsible for social behaviours (like mimicry or empathy) are also responsible for various non-social behaviours. To give a famous example, cognitive psychologist Alan Leslie (1987) proposed that the ‘innate cognitive module’ which allows us to understand other people’s minds (i.e. to empathise with other people) also allows us to engage in pretend play (e.g. to pretend that a banana is a telephone). Thus, the social ceases to be a qualitatively distinct from the non-social, relying upon the same cognitive and neurological process.

Within the experimental human sciences this is a really significant shift in understandings of the social. Under this new regime the social is individualised, essentialised, and biologised, becoming a property of individual persons outside of context, individual or institutional history. I have an innate, biological capacity to feel empathy and this capacity lies at the heart of my social being. In other words, ‘the social’ is not something that shapes us throughout one’s lifetime, it is something that we are inherently and naturally.


Giving autism its form

An exception to the rule that we are inherently social creatures is believed to be found in autism. As described in the introduction, social abnormality is taken to be a, or even the, primary symptom in autistic spectrum conditions. At the most general level, I think we can easily show that the description of autism as social disorder is reliant upon the contemporary construction of the social, outlined above. In psychology’s first sense of the social, where praying is social, individuals with autism are demonstrably able: as noted earlier in this essay, many individuals with autism take part in one of the most significant self-advocacy movements of all time. People with autism are clearly able to join groups, have their behaviours shaped by membership of those groups, and so forth. It is only when the social is understood as being related to interpersonal conduct that autism becomes conceivable as social disorder residing within an individual who has difficulty with, for example, feeling empathy.

What I’ve argued elsewhere (Hollin 2014) is that as cognitive theories of autism began to dominate the field of research during the 1980s, autism became dependent upon this contemporary construction of the social in a more profound way, and this dependence marked a definite break from previously dominant psychoanalytic conceptions of the condition. In a sense, and attempting to follow Mary Douglas, I think that the nature of autism has been made to conform to this idea of the social.

Rain Man

To take the most obvious example of autism cohering to this new construction of the social: autism symptomology has come to increasingly place cognitive capacity alongside overtly interpersonal behaviour. This is exactly what one would expect given that, as outlined above, the modules which allow us to be social and interact with other people are also taken to govern other cognitive capacities. Take the theory of Weak Central Coherence (WCC), for example, first articulated in the late 1980s (Frith 1989) and still popular today. At least in its initial guise, the theory of WCC suggests that one cognitive mechanism, which integrates disparate pieces of incoming information (think here of the skill to ‘read for context’ – you know that ‘the minute speck of dust’ and ‘the minute past the hour’ should be pronounced differently because you’ve integrated the words’ surrounding context) is responsible for both interpersonal impairment and exceptional or savant abilities on various jig-saw like tasks. This linkage, I suggest, has been made possible by the belief that the cognitive processes responsible for social behaviours are also responsible for various non-social behaviours, something basically unimaginable in other frameworks. Within this framework some behaviours (like puzzle solving skills) are drawn closer to the centre of autism as order and coherence is seen where previously there was none. Meanwhile, the threads linking autism and other behaviours are cut; Bonnie Evans has written, for example, on disappearance of ‘autistic fantasy’ from within the contemporary scientific literature (Evans 2013).


Autism and human nature

For those of us interested in autism I think these changes are both interesting and important. However, they do not explain, to return to a point made in the introduction, how what is missing in autism is now taken to reveal something fundamental about what needs to be present in order to be human. This change, too, can also be related to the shifting conception of what it means to be social.

From around the middle of the twentieth century, psychology, neuroscience, and evolutionary biology have begun to coalesce around the idea that being social, and in particular being able to emphasise with other people, lies at the core of human nature. There is not space here to investigate this trend within the bio- and human sciences more fully (Vincent Duclos’ interview with Allan Young in Somatosphere is a good place to start; Duclos 2013), but the point can be made rather simply. It is increasingly believed that human evolution has not only been shaped by ancient physical environments but also ancient social environments – and by social environments we very much mean social in the contemporary sense, the presence of other people. Our ‘social brain’ has evolved and now functions specifically to understand those around us. Given that autism is so often characterised in terms of a lack of empathy, an inability to comprehend those around us, it is unsurprising that the condition has begun to take on importance within narratives of the social brain. Continuing a long-standing tradition within the psy-disciplines of examining ‘normal’ cognitive functioning in those cases where that function is perceived to be lacking, autism offers a pure case of human-minus-social. The individual with autism has thus become an example of what Will Viney has called inherently useful humans, useful ‘not for what they do, simply for what they are’ (Viney 2013). Thus, the social hole in autism is actually a window to the soul. This, I believe, is how autism has come to stand at the centre of what it means to be human.

curious cover

From what I can see, there is no reason to suppose that these relations between the social, the human, and autism have stabilised. Indeed, this recent belief that autism is key to understanding human nature hints at an avenue of potential change. For, while I’ve argued above that autism has been made to ‘conform to an idea of the social’, I think one could legitimately claim that, increasingly, the social is made to conform to the idea of autism. If your logic stresses that autism is a form of social dysfunction and, therefore, anything functioning in autism cannot be social, then the concept of autism has become very powerful indeed. I would (very tentatively) suggest that autism research which found mirror neurons to be functioning in individuals with autism contributed significantly to the declining status of those neurons when it comes to understanding ‘normal’ social functioning. For these reasons, and many more besides, I think an understanding of ongoing autism research within the bio- and human sciences will be one of the key tasks of the medical humanities in the coming decade.



Danziger, K., 2000, Making social psychology experimental: A conceptual history, 1920-1970. Journal of the History of the Behavioral Sciences, 36(4), pp.329–347.

Duclos, V., 2013, When anthropology meets science: An interview with Allan Young, Somatosphere. Available at: [Accessed June 18, 2014].

Evans, B., 2013, How autism became autism: The radical transformation of a central concept of child development in Britain, History of the Human Sciences, 26(3), pp.3–31.

Frith, U., 1989, Autism: Explaining the Enigma 1st ed., Cambridge, MA: Blackwell.

Greenwood, J.D., 2004. The Disappearance of the Social in American Social Psychology, Cambridge, UK: Cambridge University Press.

Hacking, I., 2009, How we have been learning to talk about autism: A role for stories, Metaphilosophy, 40(3-4), pp.499–516.

Hoekstra, R.A. et al., 2009. Association between extreme autistic traits and intellectual disability: Insights from a general population twin study, The British Journal of Psychiatry, 195(6), pp.531–536.

Hollin, G.J., 2014, Constructing a social subject: Autism and human sociality in the 1980s, History of the Human Sciences.

Leslie, A.M., 1987, Pretense and representation: The origins of ‘Theory of Mind’, Psychological Review, 94(4), pp.412–426.

Murray, S., 2008, Representing Autism: Culture, Narrative, Fascination, Liverpool: Liverpool University Press.

Viney, W., 2013, Useful humans: The Wonder of Twins. Available at: [Accessed May 14, 2013].


Gregory Hollin is based at the Institute for Science and Society, University of Nottingham, and is supported by a Mildred Blaxter Post-Doctoral Fellowship from the Foundation for the Sociology of Health and Illness. Entitled ‘Locating autism: Diagnosing a social disorder’, this post-doctoral project continues work commenced under an ESRC funded PhD and attempts to understand both how autism emerged as a disorder of social functioning and how that diagnosis is maintained in practice. Greg can be contacted at gregory.hollin [at] and @GregHollin.


The top image is from the National Theatre production of The Curious Incident of the Dog in the Night-time.


  1. Thanks for this, I’lll have to check out The Machine

    Speaking as a diagnosed autistic (Asperger’s), Rain Man can fuck off. The person he was based on wasn’t even autistic.

    I’m currently putting the finishing touches to on an essay about Blade Runner and autism and the best aspie character in cinema history is Brion James’s as the replicant Leon Kowalski. In the opening sequence he is subjected to an empathy test remarkably similar to the diagnostic tests for autism by a bounty hunter/psychologist. Not only does this sequence demonstrate his difficulty with empathic responses, it also shows his difficulty maintaining eye contact, his lack of fluency in conversational turn taking (talking over and interrupting the tester), his difficulty transitioning between different modes of discourse (‘Is this the test now?’) and pedantic fixation on the details of the questions rather than the gist of them (‘What desert? How come I be there?’).

    At the heart of the movie is the concept of a spectrum; the replicants are becoming more human – at least one believes she is human – while the humans are becoming more like the replicants. It is based on Philip K Dick’s novel Do Androids Dream of Electric Sheep which was published just a few years after his novel Martian Time-Slip which featured an autistic boy. At one point the replicant Racheal convinces the bounty hunter Rick Deckard that she is human but ‘schizoid’, a term Dick uses synonymously with ‘autistic’ throughout his work; in fact many children with Asperger’s would have been diagnosed as schizoid before Hans Asperger’s work was translated into English and Uta Frith defined Asperger’s as an autistic disorder.

    The movie directly addresses the theme of prejudice against those defined as less than human simply because they are supposed to lack empathy despite having normal or high intelligence; and unlike Rain Man it doesn’t subordinate the autistic characters to the neurotypical protagonist’s need to feel good about himself.

  2. Incidentally, there’s a growing body of science fiction that addresses autism directly, including Elizabeth Hand’s novel Speed of Dark (which feature’s an autistic protagonist), Greg Egan’s Distress (featuring a subplot about a Voluntary Autistic Society) and Kim Stanley Robinson’s 2313 (which features a protagonist who takes oxycytin to manage his condition).

  3. I should add that autism isn’t just a social disorder, there are common sensory issues too: hyper- or hypo-sensitivity to pain or noise, repetitive or stereotypical movement, sensory self-stimulation (‘stimming’), etc.

    My inability to function with a hole in my sock is not the result of social deficits.

    Another thing that Philip K Dick got right was the portrayal of Mercerism, a religion based entirely on empathy. You only need to look at the self-help section of a bookshop or listen to commentators to see that ’empathy’ is cited as the panacea for all social ills despite little agreement on what ’empathy’ is. Whatever that is, it’s a quality autistics are supposed to lack and what leads to us being dehumanised by society.

  4. Shatterface (@Shatterface)

    Just watched The Machine and can confirm it is indeed the Sally Anne test. The scientist who gives the test is working on artificial intelligence in the hope of discovering a cure for his daughter’s Rett syndrome, a neurodevelopment disorder rather more severe than most forms of autism.

    There’s an episode of The Sarah Connor Chronicles in which the cyborg Cameron (a terminator in the form of a teenage girl) poses as a student in order to meet a psychologist who previously worked in artificial intelligence. The psychologist fails to spot she’s a cyborg and diagnoses her with Asperger’s…

Comments are closed.

%d bloggers like this: