Sam Fellowes: Historical and Philosophical Problems with Neurotribes

Guest post by historian and philosopher Sam Fellowes

I am a diagnosed autistic individual who studies philosophy of psychiatry and history of autism. I am a Wellcome Trust funded academic researcher on philosophy of psychiatry, exploring the conceptual foundations of psychiatric diagnoses. I think this is important because unquestioned assumptions about psychiatric diagnoses can unhelpfully constrain how people think about psychiatric diagnoses, potentially harming scientific research and negatively impacting how people view diagnosed individuals.

I find the popular success of Steve Silberman’s Neurotribes concerning because I feel that it uncritically promotes one approach to autism whilst not giving alternative possible approaches fair consideration. It does this through Silberman’s flawed historical picture of Leo Kanner’s notion of autism and through Silberman’s own notion of autism resting on unargued for and questionable philosophical assumptions. I have outlined my claims in publications (Fellowes 2017; Fellowes 2020a); I shall provide an overview of my concerns here.

Silberman argues that Kanner knew of the autistic spectrum through Asperger’s work but suppressed it in favour of his own narrow notion of autism. Silberman thinks that, until Kanner’s approach was overturned in the 1980s, many people who could have received a helpful diagnosis went undiagnosed because they did not fit Kanner’s narrow diagnostic criteria. However, Silberman provides a very inaccurate account of Kanner’s approach.

Firstly, Silberman claims Kanner’s notion of autism was “monolithic by definition” (2015, p.44), meaning autism, in contrast to modern notions of a spectrum, could only manifest in a few narrow ways. However, Kanner explicitly talks of how autism can manifest with different degrees of intensity in the 1950s. Also, Kanner regularly contrasts the divergent development of individuals he diagnosed with autism, describing how some would remain non-verbal whilst others became active members of the social life of their community (Kanner & Eisenberg, 1956, p559).

Secondly, Silberman claims that Kanner only believes that autism was “limited to childhood” (2015, p.44). However, Kanner followed the development of individuals he diagnosed as autistic, describing them as autistic when they were adolescents and adults. Also, in 1951 he described a 34 year old adult as autistic. She was 26 in 1943 (when Kanner first published on his notion of autism) yet he considered her autistic (Kanner in Darr & Worden, 1951, p.570). Kanner clearly considered it legitimate to diagnose adults as being autistic even if they did not receive the diagnosis in childhood.

Thirdly, Silberman describes Kanner’s autism as “vanishingly rare” (2015, p.44). It certainly was a rare diagnosis but Silberman needs to contextualise this claim. He does not mention how in the 1950s Kanner explicitly considered autism to be a more severe form of a more common psychiatric diagnosis named childhood schizophrenia (Kanner & Lesser, 1958: 728). Kanner was quite restrictive over who could receive the diagnosis of autism but he also thought individuals who did not meet the restrictive criteria of autism could instead be diagnosed with a different diagnosis which covered most the same symptoms but in a less severe form.

I have also criticised Neurotribes on philosophical grounds. Silberman seems to treat the autistic spectrum as a fact waiting to be discovered. He seems to think autism exists as a spectrum and this fact was discovered by Asperger and later Lorna Wing. This faces two problems. Firstly, he offers no proof that autism exists as a spectrum, and instead seems to assume the DSM-5 account of autism is the truth. However, modern scientists could be mistaken and autism might exist in a form other than a spectrum. Secondly, he seems to assume autism exists in a form which is independent of our views. Whilst this is not necessarily wrong, modern philosophers of psychiatry typically believe that psychiatric diagnoses must involve classification. They think that there are facts about people and which symptoms they exhibit but to form psychiatric diagnoses we must make decisions about how to classify those facts, such as deciding whether to group together symptoms into discrete entities or group them into a spectrum. Silberman implicitly employs one way to think about what psychiatric diagnoses are but does not discuss alternative possibilities.

Silberman also favours only one notion of a spectrum and discounts alternative notions of a spectrum. Silberman favours a non-subtyped spectrum, whereby there is only an autistic spectrum and it has no subtypes. However, it is possible to think there is an autistic spectrum and think that the spectrum also has subtypes. So if Silberman is correct that we should see autism as a spectrum this does not then mean that we should not consider subtyping the spectrum. Silberman promotes one notion of the autistic spectrum without providing a fair comparison to other alternative approaches to an autistic spectrum. Finally, Silberman’s position is often considered to support a neurodiversity position but I have argued that neurodiversity is better supported by a subtyped spectrum rather than a non-subtyped spectrum (2020b).

Pre-prints of my articles on Neurotribes can be found on my website ( These contain more detail on these criticisms and contain some additional criticisms.


Darr, G, C. & Worden, F, G. (1951). Case Report Twenty Eight Years After an Infantile Autistic Disorder. American Journal of Orthopsychiatry, 21/3: 559-570.

Fellowes, S. (2017). Putting the present in the history of autism. Studies in the History and Philosophy of the Biological and Biomedical Sciences, 61: 54-58.

Fellowes, S. (2020a). The Importance of Getting Kanner’s Account Right in Debates over First Descriptions of Autism. Journal of Autism and Developmental Disorders, 50/12: 4329-4330.

Fellowes, S (2020b). Additional Challenges to Fair Representation in Autistic Advocacy. American Journal of Bioethics, 20/4:44-45.

Kanner, L. & Eisenberg, L. (1956) Childhood Schizophrenia Symposium 1955: Early Infantile Autism. American Journal of Orthopsychiatr y, 26/3: 556-566.

Kanner, L. & Lesser, L, I. (1958). Early Infantile Autism. Pediatric Clinic North America, 5/3: 711-730.

2 responses to “Sam Fellowes: Historical and Philosophical Problems with Neurotribes

  1. Failure of language development, repetitive movements, and diminished level of consciousness (or environmental awareness) are neurological problems. Brain damage will at some point be found that leads to the language, movement, and conscious state impairments. Then Silberman’s and other non-neurologically based theories will vanish into thin air. Findings of brainstem auditory pathway impairments in autism (by Kulesza & Lukose et al.) seem most promising to me.

    Liked by 1 person

  2. Pingback: Guest post on Neurotribes | History of Autism Blog·

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.