Science Learns Autism Is Not One Thing (But We Already Knew That)
Cracks Revealed in the Deficit Frame
A new wave of genetics research says autism is “not one condition but many” (studies here and here, explainer here).
It is an easy line to headline, and an easy one to misread. The word condition gives the game away — a deficit frame dressed in clinical neutrality. Still, there is something worth pausing over in the slip from one to many. Not because it rescues the study from its medical frame, but because it fractures the framing researchers have tried for decades to keep whole. Note that the word “condition” is a media gloss, not a term the study authors themselves actually used. But it is easy to understand why the journalist saw it that way.
The Monolith Was Always a Myth
For years, autism research has leaned on a unitary model. Autism was one condition, varying only in severity. Early diagnosis meant “severe,” later diagnosis meant “mild.” This simplicity kept funding tidy. It also kept autistic lives flat. A spectrum reduced to one axis: how broken you looked to someone else.
The new study out of Cambridge, published in Nature, unsettles that tired trope. It finds genetic and developmental differences between people diagnosed early and those diagnosed later. The groups followed different pathways, not just the same path at different speeds. Instead of presenting autism as “many conditions,” as some coverage has suggested, the study actually finds evidence for at least two partially distinct genetic and developmental pathways. That is a scientific way of saying: the monolith never really held.
When Pathology Splits, the Cracks Show
Here’s the irony: fragmenting autism into “many conditions” is still pathology talk. But it is a pathology that has lost its neatness. Instead of one bucket, there are several. Instead of one genetic profile, there are overlapping sets, weakly correlated. The cracks multiply, and in those cracks other truths can emerge.
For autistic people, plurality is not news. It is how we live. The fact that one child masks their differences until adolescence while another is identified at three years old is not proof of separate diseases. It is proof that human development refuses to collapse into a single trajectory. The study names this in clinical code. We name it in lived experience.
What “Many” Makes Possible
If research insists on calling autism a condition, then better many than one. Many resists a single cure project. Many complicates the fantasy of a biomarker that will sort the world into autistic and not. Many gives us entry points to ask different questions:
- If genetics only explains eleven percent of the variation in diagnosis age, what explains the rest?
- What role do bias, access and stigma play in who gets named autistic and when?
- If later-diagnosed profiles overlap with ADHD and depression, what does that say about psychiatry’s habit of siloing lives into discrete labels?
While healthcare access and diagnostic bias clearly influence when autism is recognized, the study also shows that developmental trajectories themselves play a role — genetics explain about eleven percent of the variance, with the rest shaped by environment and lived development.
The answers are not in the genome. But the genome’s failure to be simple is an invitation.
Headlines Are Not Harmless
Still, it matters how this research gets carried into public view. Technology Networks ran the headline “Autism Is Not One Condition but Many.” That phrasing is not neutral. It multiplies the deficit frame rather than reducing it. The authors themselves emphasize complexity. It is in translation to headlines and press releases that “many conditions” language risks flattening autism into pathology again. A better headline would have been “Autism Takes Many Pathways” or “Autism Emerges in Many Trajectories.” The words pathway and trajectory carry room for variation. The word condition closes that room, locking difference back into disease.
Autistic people feel the weight of these words. When our lives are cast as multiple conditions, it is not long before policymakers imagine multiple treatments, multiple surveillance tools, multiple ways to miss who we are.
The Plural Was Ours First
What researchers call subtypes, we call variance. Our plurality is not their discovery. It is our daily fact. Some of us speak, some of us do not. Some of us were identified early, some much later. Some of us need support every hour, some only in certain contexts. These are not different conditions. They are different lives, equally autistic, equally valid.
So yes — let science break its monolith. Let the word many do its work. But let us also remind them: autistic people are not waiting for a genome-wide study to tell us we are not one thing. The paper’s real contribution is evidence for plural pathways — a finding that should encourage more nuanced research, stratified supports and public messaging that avoids turning plurality into pathology. We already knew. We have always known.