Is neurodivergence a disability or a difference?

What each model claims

The medical model treats a condition as located in the individual: something to be identified, described against criteria, and where possible treated. It’s the model that clinics run on, and it’s the one that produced the diagnostic categories the rest of this site discusses. Its strength is that it’s actionable—it gets people assessed and supported. Its weakness is that it can make a person the site of the problem.

The social model separates impairment from disability. In this framing, impairment is a feature of a body or mind, and disability is what happens when a society is arranged so that the feature becomes a barrier. Stairs disable a wheelchair user; a lift doesn’t. Applied to neurodivergence: open-plan noise, unstructured instructions, and interviews that score social performance are the stairs.

Mike Oliver, who did more than anyone to establish the social model in Britain, wrote a short reflection on it three decades in. His framing there is worth keeping: he described the model as a practical tool for improving lives rather than as a complete social theory of disability. It was built to shift where the intervention goes, not to explain everything.

The neurodiversity paradigm shifts the ground again, treating neurological variation as natural human diversity. Under it, the question isn’t only what accommodations remove barriers, but whether “typical” was ever the correct benchmark.

A worked example helps. Autistic and non-autistic people often struggle to read each other. The medical model reads that as a social communication difficulty in the autistic person. Damian Milton’s “double empathy problem” reads the same breakdown as mutual—two parties with different styles each failing to interpret the other, which relocates half the difficulty to the non-autistic side. Same observed event, two accounts, and they imply different responses: teach the autistic person, or change what both parties expect.

Where the neurodiversity idea came from

This gets miscredited often enough to be worth a section.

The concept is regularly attributed to one person, the Australian sociologist Judy Singer, writing in the late 1990s. A 2024 correction in the journal Autism, co-authored by Monique Botha, Robert Chapman, Morénike Giwa Onaiwu, Steven Kapp, Abs Stannard Ashley and Nick Walker, sets out the case against that. Drawing on archival findings, they argue the term “neurological diversity” was in use earlier than previously thought, and that the concept and the body of theory around it were developed collectively by neurodivergent people rather than originating with a single author.

That’s a correction about credit, and it also matters for how the idea is read. A paradigm built collectively by the people it describes is a different sort of object from one proposed by an individual academic about a group.

What the community endorses

Here’s where the argument gets conducted with assumptions instead of evidence, so it’s worth naming what has been measured.

Dwyer and colleagues surveyed 504 autistic and autism community members—278 autistic and 226 non-autistic, including 100 researchers, 122 professionals and 162 parents or caregivers—asking them to rate the neurodiversity movement, the social model, the medical model, and a range of intervention goals.

Several findings cut against the caricatures.

Neurodiversity support didn’t mean rejecting help. Support for the movement went with endorsing societal reform and more supportive environments, and with lower support for normalisation and adaptive skills interventions. But teaching adaptive skills was widely supported overall.

The “strong” social model wasn’t universal. Participants often described the social model as attributing disability solely to society. That view existed but wasn’t held by everyone, and even the people who did hold it still endorsed some interventions aimed at individual characteristics.

The disagreement is narrower than the rhetoric. The authors read their results as showing shared goals alongside real tensions, and suggest oversimplified rhetoric on either side causes confusion about what is in dispute.

If you’ve felt out of place in this argument because you want the barriers removed and you want help with the parts that are hard regardless, that combination is well populated.

Why people hold more than one

The models answer to different situations, and most lives contain several.

A framing is doing a job when you use it. “This is a difference” is the right frame for describing how you think, and the wrong frame for a conversation with an employer about why you can’t take minutes in a loud room. “This is disabling” is the right frame for that conversation, and the wrong frame for explaining to a friend why you’re delighted by train timetables.

That isn’t inconsistency. It’s using the tool that fits the task, which is what the models were built for.

There is a version of the question underneath this one that the models can’t settle: whether you personally want the word. That’s a separate matter, and the practical side of it—what a definition requires versus what you call yourself—is covered in am I disabled if I mask well?.

What changes depending on the frame

The choice isn’t only philosophical. It routes attention and money.

Read down the last column and the stakes get clearer. A school that adopts the first frame sends a child for intervention. A school that adopts the second changes the lighting, the instructions and the timetable. A school that adopts the third asks why the day was designed that way to begin with. All three can be done well and all three can be done badly.

The framing also reaches the people being described, which is the part that’s easiest to miss. Pearson and Rose’s conceptual analysis of autistic masking argues that masking is an unsurprising response to the deficit narrative and the stigma built around it—that if the prevailing account says your ordinary responses are symptoms, concealing them is a rational move rather than a personal quirk. On that reading, a framework isn’t only a lens on a person. It’s part of what the person is responding to.

Which is why the answer to the question in the title isn’t a tiebreak. It’s that you’ve been handed a genuine dispute among people who mostly want the same outcomes, and you’re allowed to take what’s useful from each without joining a side.