Verified by 
A recent Guardian article by Tom Clements both suggested that the autism classification is now meaningless, and criticised the neurodiversity movement for making autism into a mere "fashion label." Whilst Clements' article was a rehash of old arguments, many of which have long ago been refuted, it seems a good time to add some further clarity to these issues.
For context, let's begin with the medical framing of autism and its problems. In the medical framing, autism is a neuro-developmental spectrum disorder, with the severity of the disability being taken to range from "mild" to "severe." But this characterisation has come under attack from a number of angles.
First, conceptually, the "spectrum" seems overly broad, and conflates various different disabilities in a misleading way (e.g. often, "severe" autism is in fact just autism combined with other cognitive, developmental, or other disabilities). Second, the construct lacks pragmatic force, since what seems to be good for one autistic person may be a matter of indifference or even harmful for another. And third, many autistic people find it dehumanising to be described as “mild/high-functioning” or “severe/low-functioning.”
In reaction to the dehumanising aspects of the medical framing, neurodiversity paradigm proponents argue that autism should be seen as a neurominority. Instead of using a medical model to frame autistic disablement, this framing (typically) draws on a social model of disability in order to recast autistic disablement as stemming primarily from a society that fails to accommodate minority modes of neurocognitive functioning. On this view, functioning labels are generally avoided, and the term "autism" is used in a broad way to make it inclusive.
Whilst many autistic people prefer this framing, critics such as Clements (who is also autistic) argue that the rejection of functioning labels may lead to overlooking what he takes to be "severe" cases. By saying that a non-verbal autistic person with multiple learning disabilities has the same condition as someone able to speak, work, and live independently, the differences may be overlooked. Moreover, for such critics, more able autistic people may end up speaking over those who are multiply-disabled, leading to the latter's needs remaining unmet.
It is worth mentioning that such critics tend to systematically misrepresent the neurodiversity movement, as others have pointed out. Still, it is easy to understand why Clements and others have these worries, in so far as applying the more general term “autistic” without any sub-typing can give the impression of overlooking the complexity of autism. With this in mind, my aim here is not to revisit old debates, but to offer a constructive suggestion.
An Intersectional Social Model of Autism
My own suggestion is that we can better make sense of autistic disablement by drawing on the notion of intersectionality developed by feminist theorist Kimberlé Crenshaw. I argue that this can supplement the neurodiversity view and social model to avoid the criticisms noted above.
According to intersectional theory, when one person is simultaneously part of two or more marginalised groups, these intersect to form a third, more specific group that is different to just the combination of the former two (just as two roads can intersect to become a third road). This is important because this third group has group-specific needs, and will encounter group-specific problems, which may be different to or even conflict with the needs of the first two—meaning that they also need to be recognised as having a distinct voice.
Consider the case of black women (which I borrow from Crenshaw). Both being black and being female have long been recognised as marginalised identities. Crenshaw points out, however, that black women are often marginalised in a way that is different from either black men or white women.
For example, Crenshaw noted that in order to avoid anti-discrimination laws, many companies had to promote black people and women. In theory, this should have been good for all black people and all women. But in practice, Crenshaw spotted a worrying pattern: companies tended to only promote white women and black men, leaving black women in the most junior positions. Hence, recognising the intersectional identity "black woman" in its own right has important ethical, political, and legal implications (you can watch Crenshaw talk about this here).
I think we can avoid the criticisms of Clements by drawing on the core insight from intersectional feminism and applying it within our framing of cognitive disability. On what we might call the “intersectional-social model of disability,” when someone is, say, autistic and learning disabled, then they are part of a third intersectional category (i.e. that of the learning disabled autistics), and therefore will have different and sometimes conflicting needs to autistics who are not learning disabled.
This framing does two things. First, it allows us to acknowledge the complexity of autism without dehumanising or conflating different disabilities under the term "severe" (making my proposal in line with the neurodiversity paradigm). And second, it does this in a way that is nonetheless booby-trapped against autistic people (such as myself) talking over those autistic people with intersecting disabilities. For on this model, the only people who should be taken as the voice of any given intersectional identity are those that fall within their intersection.
It is worth ending by saying that, for my own part, from everything I have seen over many years, most neurodiversity proponents do already make their advocacy inclusive in just this way. But the intersectional version of the social model may still help make our acknowledgement of the complexity of autism, and how to conceptualise this, clearer. At the same time, I hope, it should help ease the worries of critics.
