What Lynn Margulis, curved geometry, and the history of science can teach us about the medical and social models of disability
Some scientific arguments persist because one side is wrong.
Others persist because both sides are using maps drawn for different terrain.
Imagine planning a short walk across a city block. For that purpose, the ground can be treated as flat. Parallel streets remain parallel. A straight line is the shortest distance between two points. Ordinary Euclidean geometry works beautifully.
Now imagine using the same assumptions to navigate halfway around the Earth.
The Earth has not suddenly violated geometry. The problem is that a model that worked locally has been mistaken for a complete description of the larger surface. On a sphere, routes that look parallel on a flat map can converge, and the shortest path between distant points follows the curvature of the surface. Euclidean geometry remains useful; it simply operates within boundary conditions that must be acknowledged. The larger geometry does not make the local model foolish. It reveals the structure in which the local model is embedded.
Science repeatedly advances this way. A successful model explains a particular range of observations. Because it is successful, its assumptions gradually become difficult to see. The model stops being treated as one useful representation and begins to be treated as reality itself.
Then someone encounters a phenomenon the model cannot adequately describe.
The initial reaction is often to defend the model more forcefully. An anomaly is dismissed as noise, a methodological failure, an exceptional case, or an idea that is “not even wrong.” Sometimes that dismissal is justified. Scientific communities need guardrails. A field that cannot distinguish a productive hypothesis from an unfalsifiable story cannot accumulate reliable knowledge.
But guardrails can also become walls.
Lynn Margulis met such a wall in twentieth-century evolutionary biology. In a 1967 paper, published under the name Lynn Sagan, she proposed that crucial components of complex cells—including the mitochondria that power them and the plastids that let plants photosynthesize—had descended from formerly free-living organisms, incorporated through ancient symbioses. The paper drew together evidence from cytology, biochemistry, paleontology, and evolutionary theory, and it identified observations that could test the proposal.
The lasting lesson is not that the scientific establishment was foolish and Margulis was an isolated genius who replaced Darwin. That story makes for good mythology and poor philosophy of science.
Natural selection did not become false.
What changed was the explanatory architecture. Selection and symbiosis appear diametrically opposed only when each is forced to function as a complete explanation at the same level. Selection can explain differential persistence and reproduction among variants without, by itself, specifying every mechanism by which new levels of biological organization arise. Symbiosis supplied one such mechanism. Once those new organizations existed, selection could act upon them. Each mechanism kept its validity within its domain; what dissolved was the demand that one of them explain everything.
Margulis’s proposal was also not one indivisible claim to be accepted or rejected wholesale. Some components became foundational to modern cell biology. Other proposed symbiotic origins did not achieve the same empirical support. That is what a productive theoretical proposal should permit: separation into claims that survive testing, claims that require revision, and claims that fail.
The value of a scientific framework is not that every sentence bearing its name must be correct. Its value is that it creates deliverables—observations to make, variables to distinguish, predictions to test, and conditions under which parts of the framework must be abandoned.
This history matters to the current conflict over disability.
The medical and social models are often presented as irreconcilable.
The medical model directs attention toward the person’s body or mind. It asks about physiology, development, symptoms, impairment, diagnosis, treatment, and functional capacity.
The social model redirects attention toward the world surrounding the person. It asks how institutions, environments, expectations, attitudes, and distributions of power convert human variation into exclusion and restricted participation. The modern social-model tradition developed partly to contest the assumption that disability was simply a personal tragedy or an individual defect.
These models do not merely disagree about evidence. They have acquired different moral and political functions.
To many disabled and neurodivergent people, the medical model carries the history of normalization, institutionalization, coercion, and the treatment of difference as defect.
To many clinicians, patients, and families, an overly absolute social model can appear to deny pain, physiological instability, communication barriers, loss of function, or the desire for effective treatment.
Each side therefore hears the other not merely as an alternative explanation, but as a threat.
The resulting debate asks: Is disability located within the person? Or is disability created by the environment and society?
More relational frameworks already recognize that functioning occurs in context; the World Health Organization’s classification of functioning describes disability in relation to health conditions and environmental factors. Yet the practical debate repeatedly collapses back into a demand for one location and one responsible party.
That is the hidden problem.
We may have stopped believing that disability has one simple cause. We have not stopped believing that it must have one final location.
Inside the person. Between the person and the environment. Or in society.
Within a one-level framework, these answers are irreconcilable. If disability is entirely inside the person, then removing social barriers may improve opportunity but cannot alter the disability itself. If disability is entirely imposed by society, then intrinsic pain, fatigue, motor constraints, or physiological instability become difficult to discuss without appearing to retreat into pathology.
But a person is not separate from the environment in which that person must regulate, act, communicate, predict, recover, and remain viable.
The person and environment form a coupled system.
Quiet a noisy workplace, permit written communication, make expectations predictable, and something interesting happens. The intervention is social—a barrier has been removed. But its effects are biological: the nervous system has less to filter, less to predict, more capacity free for everything else. The accommodation is social; its effects are physiological.
Conversely, treat pain, stabilize physiology, or provide communication technology, and the person’s ability to work, sustain relationships, and participate politically may change. The intervention is medical; its effects are social.
The models were never describing separate worlds. Each measures a different projection of the same person–environment system.
Here, then, is the proposition this essay has been building toward:
Models can contradict one another when each is treated as a complete description. They can become compatible when their scale and boundary conditions are specified.
Selection and symbiosis. Flat maps and curved surfaces. The medical model and the social model. In each case, the contradiction lives at one level of description and dissolves at another.
But specifying the scale exposes a question that neither model was built to answer.
Two people can meet the same demand—hold the same job, earn the same grade, pass the same assessment—while what it takes to do so differs profoundly. One returns home with capacity to spare. The other has held the performance together through continuous effort that no observer records and neither model measures, and has nothing left. The outputs are identical. What producing them required is not.
Neither the medical model’s primary question—what is happening within this person?—nor the social model’s primary question—what barriers arise between this person and their world?—fully captures that difference. It lives in the relationship between them: in what maintaining function under these conditions demands, and what that demand displaces.
That missing dimension is the subject of the next essay.
For now, the conclusion is simpler.
Science sometimes advances by proving an earlier model wrong. But it also advances by discovering the scale at which an apparent contradiction disappears.
The medical and social models do not need a truce.
They need a larger geometry.
Want more? Read the 2nd article in the series here
And the 3rd Article here
Leave a Reply