Summary
There is a long tradition of looking back at historical figures and recognising autistic or ADHD traits in their biographies. Einstein, Newton, Mozart, Turing, Darwin, Cavendish, Dickinson, Warhol: the list grows with each popular book on neurodivergence. The practice is sometimes called retrospective diagnosis, sometimes posthumous diagnosis, and sometimes (by its critics) “armchair diagnosis at a distance of centuries.”
It is not one thing. At its best, retrospective identification offers late-diagnosed people a sense of lineage, challenges the assumption that autism is a modern invention, and reframes historical eccentricity as recognisable neurodivergent experience. At its worst, it projects modern diagnostic categories onto people who cannot consent, reduces complex lives to a checklist of traits, and reinforces a narrow image of autism as white, male, and genius-adjacent.
This page examines the practice honestly: acknowledging what is illuminating, naming what is problematic, and arguing that the most valuable version of looking backward is not diagnosis at all but recognition: finding kindred spirits in the historical record without importing clinical labels they never carried.
Why people do it
Retrospective diagnosis serves several distinct purposes, not all of which its practitioners acknowledge.
Representation and belonging. For people diagnosed late in life, discovering that historical figures they admire may have shared their neurology is powerful. It provides a lineage. The message is: people like you have always existed, have always contributed, have always been here. For a community whose existence was unnamed until the mid-twentieth century and pathologised for most of the time since, that matters.
Historical reframing. Autism and ADHD are sometimes treated as modern phenomena, products of changed diagnostic criteria or shifting social expectations. Identifying plausible cases in the historical record pushes back against this. If Newton’s contemporaries described behaviour that maps recognisably onto an autistic profile, that tells us something about autism’s presence across time and culture, independent of when the label was invented.
Public engagement. Books and articles about famous autistic people sell. This is not inherently cynical: Temple Grandin’s account of her own experience, Simon Baron-Cohen’s analysis of Einstein and Newton, and Steve Silberman’s NeuroTribes all reached audiences that no academic paper could. The famous-autistic-person genre is one of the most effective vectors for changing public understanding of autism. It also shapes that understanding in particular ways, which is where the problems begin.
What makes it problematic
You cannot diagnose someone you have never met
Autism diagnosis requires observation of behaviour across contexts, developmental history, and the person’s own account of their inner experience. For historical figures, none of this is available in clinical form. What exists are letters, biographies, anecdotes recorded by contemporaries with their own biases, and retrospective interpretation filtered through centuries of cultural distance.
Muramoto (2014) examined the epistemological foundations of retrospective diagnosis in Philosophy, Ethics, and Humanities in Medicine. He argued that the practice is not inherently invalid but requires careful separation of nosology (the classification of diseases) from the act of clinical diagnosis. A historical figure can be described as displaying traits consistent with a modern category without claiming they “had” the condition in any clinically meaningful sense. The distinction is important and routinely ignored.
The evidence is thinner than it looks
Professor Fred Volkmar of Yale has described the practice as “a sort of cottage industry.” Oliver Sacks called the evidence for many retrospective diagnoses “thin at best.” The Thomas Jefferson Foundation, responding to claims about Jefferson, noted that such analyses rely heavily on secondary sources rather than primary evidence, and that childhood behaviour, which is central to autism diagnosis, is precisely what the historical record most often lacks.
The problem is not that historical descriptions never resemble autism. Some of them do, strikingly. The problem is that resemblance is not diagnosis. Biographical anecdotes are selected, curated, and retold for narrative purposes. The quiet, focused, socially awkward aspects of a person’s life make good evidence for autism; the warm friendships, collaborative work, and social competence that the same person also displayed are quietly set aside.
Consent and posthumous harm
The person cannot agree to be diagnosed. This seems like a minor point when the subject is Einstein, but it is not. A diagnosis reshapes how a life is understood. It reframes achievements, relationships, and difficulties through a particular lens. Muramoto’s analysis notes that while public figures are sometimes assumed to have forfeited privacy, “most philosophers agree that posthumous harm does occur,” and medical information extracted from historical evidence may deserve more protection than is typically given.
For historical figures whose reputations are secure, retrospective diagnosis is unlikely to cause harm. For those whose lives were more complicated, or whose descendants are alive, the stakes are higher. The practice also sets a precedent: if we are comfortable diagnosing the famous dead, what prevents the same logic being applied to the less famous living?
The representation problem
The most commonly cited candidates for retrospective autism diagnosis are overwhelmingly white, male, and associated with scientific or mathematical genius: Einstein, Newton, Turing, Darwin, Cavendish, Dirac, Tesla. This reflects something real about whose biographies have been preserved in enough detail to permit retrospective analysis. It also reinforces the most damaging stereotype in autism discourse: that autism is a condition of brilliant white men.
Autistic women, autistic people of colour, autistic people with intellectual disabilities, and autistic people whose contributions were not in STEM fields are almost entirely absent from the retrospective-diagnosis canon. The practice, whatever its intentions, builds a historical mythology that looks nothing like the actual autistic population.
The candidates
What follows is not an endorsement of any retrospective diagnosis. It is an honest look at the evidence that has been cited, and what it does and does not show.
Albert Einstein (1879–1955)
Baron-Cohen (2003) assessed Einstein against autism criteria and identified delayed speech development (reportedly not fluent until age nine, though accounts vary), intense and narrow interests pursued with extraordinary persistence, social difficulties including a well-documented discomfort in formal social settings, and a preference for solitary thinking.
The case has obvious appeal but also obvious limits. Einstein was also warm, funny, politically engaged, and capable of deep personal relationships, including complicated ones. His letters show emotional range and social awareness that sit uncomfortably with a straightforward autism narrative. As several commentators have noted, the honest answer is that nobody knows.
Isaac Newton (1642–1727)
Baron-Cohen and Fitzgerald have both identified Newton as a strong candidate. The evidence includes extreme absorption in work (he reportedly forgot to eat and could lecture to an empty room without noticing), few close relationships, social awkwardness that was remarked upon even by the standards of his era, rigid routines, and what his contemporaries described as a fierce temper when his work was interrupted or challenged.
Newton’s case is complicated by period. Seventeenth-century English social norms differed substantially from those used to define autism in the twenty-first. Solitary scholarly life was an established social role, not necessarily evidence of neurological difference. Newton’s behaviour was unusual even for a reclusive scholar, but translating “unusual for a seventeenth-century Cambridge don” into a modern diagnostic category requires leaps that honest analysis should acknowledge.
Alan Turing (1912–1954)
O’Connell and Fitzgerald (2003) assessed Turing against the Gillberg criteria for Asperger syndrome and concluded he met all six. The evidence includes social difficulties from childhood, intense and narrow interests (particularly in mathematical logic and long-distance running), idiosyncratic behaviour (chaining his mug to a radiator, counting bicycle pedal rotations), literal communication style, and motor clumsiness noted by colleagues.
Turing’s case raises a particular ethical concern. His life was marked by persecution for his homosexuality, culminating in chemical castration and death at 41, probably by suicide. Retrospectively diagnosing him with autism risks adding another label to a man who was already defined, constrained, and destroyed by the labels others imposed. This is not a reason to avoid the discussion, but it is a reason to conduct it carefully.
Wolfgang Amadeus Mozart (1756–1791)
Mozart’s letters are remarkable documents. They are scattered, associative, full of sudden topic changes, wordplay, and scatological humour. Some researchers have identified ADHD traits in the correspondence: rapid shifting of attention, impulsivity, difficulty with the administrative aspects of his career despite extraordinary musical productivity. Others have pointed to possible autistic traits: sensitivity to sound that went beyond professional interest, strict compositional routines, and social behaviour that his contemporaries found erratic.
A separate line of analysis has proposed Tourette syndrome, based on accounts of involuntary movements and vocalisations. The Tourette hypothesis and the autism/ADHD hypothesis are not mutually exclusive, but they illustrate the problem: the same biographical material can support multiple, incompatible retrospective diagnoses.
What Mozart’s letters show most clearly is a mind working differently. Whether that difference maps onto any modern diagnostic category is a question the evidence cannot answer with certainty.
Socrates (c. 470–399 BCE)
This is the wiki’s own speculative contribution, so it deserves to be framed as such.
The descriptions of Socrates that survive through Plato, Xenophon, and Aristophanes paint a portrait that is, to a reader familiar with autistic experience, immediately recognisable. Not as a checklist of diagnostic criteria, but as a way of being in the world.
Start with the sensory profile. Socrates walked barefoot through Athens year-round, including on military campaign. Alcibiades’s account in the Symposium describes him marching barefoot across ice, faster than booted soldiers, apparently unbothered by cold that incapacitated others. He wore the same rough cloak in every season. He bathed rarely. His contemporaries interpreted this as philosophical asceticism, a deliberate indifference to bodily comfort. It could equally be read as a sensory profile: reduced sensitivity to cold and discomfort, strong aversion to certain textures or sensations (shoes, varied clothing), and a relationship to the physical world that his peers found baffling.
Then the absorption. In the Symposium, Alcibiades describes an incident during the siege of Potidaea. Socrates stopped in the open one morning, apparently seized by a problem. He stood motionless through the entire day, through the night, and into the following dawn. Other soldiers, bewildered, brought their bedding outside to watch. At sunrise he said a prayer and walked on. This is twenty-four hours of unbroken, apparently involuntary cognitive absorption. It is not meditation. It is not performance. It is a mind that locked onto a problem and could not disengage. To anyone who has experienced monotropic flow states, or has watched an autistic person enter deep focus, the description is arresting.
The daimonion is perhaps the most intriguing element. Socrates described an inner voice that spoke to him, always negatively (telling him what not to do, never what to do), always reliably. He treated it as divine. His contemporaries were uncertain. At his trial, it was cited as evidence of impiety. Modern interpretations range from auditory hallucination to metaphor for intuition. But there is another reading: a strong interoceptive or proprioceptive signal that Socrates experienced as a felt sense of “wrongness” about a course of action, which he interpreted through the only framework available to him as a divine sign. Autistic people frequently describe gut-level aversion signals that precede conscious reasoning, a felt “no” that arrives before the analysis. The daimonion could be interoception, described in fifth-century BCE Athenian terms.
And then the social pattern. Socrates spent his days in public, talking to anyone who would engage. He was not reclusive. But his mode of engagement was distinctive: relentless, questioning, apparently unable to let a logical inconsistency pass unremarked, indifferent to social hierarchy, and frequently infuriating to the people he spoke with. The Socratic method, as described by Plato, looks less like pedagogical technique and more like a mind that processed social interaction through logic rather than convention, and could not stop itself from pointing out where the logic broke down. His interlocutors often left angry. He seemed genuinely surprised by this.
None of this constitutes a diagnosis. It cannot. The distance is too great, the sources too mediated, the cultural context too different. But if retrospective identification has any legitimate purpose beyond the parlour game, it is this: recognising patterns of being that persist across cultures and centuries, patterns that modern categories can sometimes illuminate even when they cannot definitively classify. Socrates was almost certainly something. The word “autistic” is as good a guess as any, and better than most.
Diagnosis or recognition?
There is an important distinction the retrospective-diagnosis literature rarely makes explicit, and it changes the entire exercise.
Retrospective diagnosis takes a modern clinical category, a label developed in a specific institutional context, refined through successive editions of diagnostic manuals, defined against statistical norms of a particular population at a particular time, and projects it backward onto someone who lived in a different world. This is the version that draws legitimate criticism. It treats the diagnostic label as a timeless natural kind rather than a historically situated construct. It imports twenty-first-century clinical infrastructure into contexts where it does not belong. And it cannot meet its own epistemic standards: you cannot diagnose someone you have never met, cannot observe, and cannot ask.
Retrospective recognition is something different. It is a neurodivergent person reading about Socrates standing motionless for twenty-four hours, seized by a problem he could not let go, and thinking: I know that. It is reading Newton’s contemporaries describe his fury when interrupted mid-thought and feeling the recognition in your own body. It is reading Mozart’s letters and hearing a mind that moves the way yours moves. This is not clinical. It is not trying to be. It is the experience of finding a kindred spirit in the historical record, someone whose way of being in the world rhymes with your own across centuries of distance.
The second is far more valuable than the first, and it does not need the diagnostic label to work. Why bring the clinical apparatus with us when we travel back in time? The DSM did not exist in fifth-century Athens or seventeenth-century Cambridge. What did exist were people whose sensory experiences, cognitive patterns, and social styles would be recognisable to neurodivergent readers today. Seeing those people clearly, in their own context, and recognising the areas of shared lived experience: that is worth doing. Slapping a modern label on them adds nothing and risks reducing a complex life to a retrospective case study.
The best retrospective work sits closer to recognition than diagnosis. It asks: what can we learn about the persistence of certain ways of being human by noticing that the same patterns appear across cultures and centuries? Not “Socrates had autism” but “what does it tell us about the constancy of certain cognitive styles that a fifth-century Athenian philosopher displayed patterns we now associate with a twenty-first-century diagnostic category?”
What the practice reveals about us
Which figures get retrospectively identified, and which do not, reflects current ideas about what autism looks like: brilliant, male, white, solitary, STEM-focused. The candidates who make the lists are people whose neurodivergent traits are already legible to a modern audience primed by Rain Man and The Big Bang Theory.
This is why the practice needs the corrective of the late-diagnosis literature (see Late diagnosis). The people whose diagnoses came late, after decades of being missed, are disproportionately women, people of colour, and people whose autistic traits were masked by compensatory strategies that the historical record would never capture. If we only look backward at the people whose difference was visible and documented, we reproduce the same selection bias that delayed diagnosis for everyone else.
Recognition, unlike diagnosis, has room for this corrective. You do not need a famous name or a well-preserved biography to recognise a kindred spirit. The anonymous weaver who could not bear the texture of certain threads, the medieval anchoress who chose a sensory environment she could control, the unnamed clerk whose colleagues found him baffling. These people left fewer traces, but they were there. A recognition-based approach honours that presence without demanding the evidence base that formal diagnosis requires.
Key sources
- Baron-Cohen, S. (2003). Autism in historical figures. Discussed in The Essential Difference. London: Penguin.
- Fitzgerald, M. (2005). The Genesis of Artistic Creativity: Asperger’s Syndrome and the Arts. London: Jessica Kingsley.
- Muramoto, O. (2014). Retrospective diagnosis of a famous historical figure: ontological, epistemic, and ethical considerations. Philosophy, Ethics, and Humanities in Medicine, 9, 10. doi: 10.1186/1747-5341-9-10
- O’Connell, H. and Fitzgerald, M. (2003). Did Alan Turing have Asperger’s syndrome? Irish Journal of Psychological Medicine, 20, 28–31.
- Plato. Symposium, 174d–175b (the delayed arrival), 220a–d (Potidaea, barefoot on ice, twenty-four-hour trance).
- Plato. Apology, 31c–d (the daimonion).