A label mistaken for the mechanism it was meant to stand in for
Carlos Léger Sherman-Palmer, a lecturer at the Université de Bordeaux’s management school, has published a position paper — dedicated, touchingly, to his sister — titled Deconstructing the Ontology of ADHD. Before anything else, the frame he sets deserves respect, because it is the frame this piece holds too: the paper’s stated purpose is not to deny suffering, impairment, or the practical role the diagnosis plays. Its target is one specific claim — that ADHD, as currently codified, corresponds to a unitary, discretely bounded biological entity. The suffering is evident. The question is whether the label names a single thing.
Earlier this week I reported evidence that this label’s consequences include a mortality gap of seven to nine years. These two pieces are not in tension, and it matters to say why: a heterogeneous construct can still track real, lethal outcomes, because the difficulties it compresses are real even where the ontology is borrowed. The earlier piece showed what the label protects — recognition and treatment reduce the risk. This one asks what the label is. Both questions deserve honest answers, and the honest answers point in different directions.
The paper’s foundation is Steven Hyman’s account of diagnostic reification: operational categories, built from surface criteria for practical purposes, get mistaken over time for naturally given entities. The tell is when the label starts functioning as pseudo-explanation. “He’s inattentive because he has ADHD” sounds like an explanation, but nothing has been explained — a descriptive shorthand has been circled back on itself. The sentence does not say whether the process involved is unstable sustained attention, weak interference control, altered reward valuation, delay aversion, arousal dysregulation, or a working-memory limit — genuinely different mechanisms, differently distributed across the people who carry the same label. Hyman’s phrase for what reification does is exact: it creates epistemic blinders that impede progress toward valid diagnoses.
Sherman-Palmer’s most useful contribution is a three-level distinction that should be tattooed onto every argument about whether ADHD is “real.” Clinical-pragmatic: does the label organise care and communication? Often yes. Descriptive: does it summarise recurring patterns of difficulty? Yes, heterogeneously. Ontological: does it denote a discrete biological entity? The current evidence does not support this. Most public argument about ADHD is people shouting across these levels at each other — one side defending the first two, the other attacking the third, both calling the fight “does ADHD exist.” Keep the levels separate and the fight mostly dissolves; the interesting questions start.
This is also where the paper walls itself off from the slogan it will inevitably be conscripted into, and I’ll reinforce the wall: none of this is “ADHD isn’t real.” The difficulties are real, the impairment is real, the deaths Tuesday counted are real. What is under examination is the claim that one bounded thing underlies them all.
The field's own institutions conceding the category is not a natural kind
The striking feature of the evidence Sherman-Palmer assembles is that almost none of it comes from critics of psychiatry, but psychiatry itself.
The NIMH’s Research Domain Criteria framework was created explicitly because symptom-based categories were hindering research — hindered, in the NIMH’s own words, by heterogeneity, comorbidity, and criteria that are “somewhat arbitrary.” The sentence the paper rightly leans on is the NIMH’s own: two people can, in some cases, be diagnosed with the same disorder despite having few or sometimes no symptoms in common. The HiTOP consortium — mainstream quantitative psychopathologists — describes traditional nosology as drawing arbitrary boundaries between psychopathology and normality. And in January 2026 the American Psychiatric Association released its roadmap for the future of the DSM, foregrounding dimensions, biomarkers, environmental determinants and functioning — an institutional admission that the symptom-category model is insufficient as a scientific account. The genetics runs the same way: Plomin’s formulation that common disorders are quantitative traits — thresholded extremes of continuously distributed variation, not discrete kinds — and ADHD’s own genetics literature showing high heritability delivered through thousands of tiny polygenic effects, which makes the phenotype biologically serious while dissolving any single causal core. Executive functions, per Diamond and the Kofler review, are graded human capacities on which everyone sits somewhere. And after decades of well-funded search, the consensus statement stands: no reliable ADHD biomarker has been described to date.
A cut-point for clinical action, as the paper puts it, is not a boundary in nature. Sherman-Palmer’s own positive proposal, borrowed from Kendler, is the sensible middle: psychiatric conditions as mechanistic property clusters — fuzzy sets, causally grounded, with paradigmatic centres and marginal members, real without being essentialist. Neither natural kind nor social fiction.
Notice what this evidential pattern means for how the argument should be received. When the NIMH, the APA, the HiTOP consortium and behavioural genetics all converge on “the category is pragmatically useful but not a natural kind,” the ontological critique of ADHD is not an anti-psychiatry position. It is approximately the institutional consensus, stated plainly rather than diplomatically. What the institutions have not done is follow the concession to its uncomfortable question — if the category is known to be a coarse compression, why does it remain so immovable? — and that is where the paper’s real contribution begins.
The corporatism turn — what the diagnosis does for the profession that holds it
The answer Sherman-Palmer assembles is sociological, and readers of my corpus will recognise its logic immediately: look at what the category does, and for whom. He draws on Abbott’s analysis of professional jurisdiction — professions maintain authority by controlling the classification of the problems assigned to them — Freidson’s professional dominance, Conrad’s finding that the engines of medicalisation are now commercial as much as professional, and Pilgrim’s blunt question: the issue is not only what is wrong with psychiatric diagnosis, but how and why it survives, through the “interest work” of the groups defending it.
Run ADHD through that lens. The category is load-bearing for far more than patients: professional training pathways, licensure, service commissioning, reimbursement codes, research funding structures, expert testimony, school resource allocation, and a pharmaceutical and coaching market of considerable size (to name only two of many) all rest on its continued stability. A category that much infrastructure depends on does not need to be a natural kind to persist; it needs only to keep the infrastructure running. The paper is careful — and I’ll match the care — that this is not conspiracy: no bad faith is required, only the normal self-preserving behaviour of organised institutions. Structural self-preservation, not scheming.
This is POSIWID applied to nosology: the purpose of a system is what it does, and what the diagnostic category demonstrably does — beyond its clinical function — is stabilise jurisdiction, secure reimbursement, and reproduce the profession that administers it. The paper’s sharpest line deserves quoting exactly: “When a category becomes safer for the profession than for the patient, science has already begun to retreat.”
The test that line implies is worth making explicit, because it converts critique into diagnostic. Ask of any defence of the current category: does this argument protect patients, or does it protect jurisdiction? “The diagnosis secures access to support” protects patients — and is true, and Tuesday’s mortality data shows how much it matters — but notice that it is an argument about what the label does, not what it is; it defends the clinical-pragmatic level while borrowing the authority of the ontological one. The institutions concede the levels are different. The public settlement pretends they are the same. That pretence is what the corporatism analysis explains: the conflation is not an intellectual error that nobody has noticed. It is a load-bearing ambiguity.
And the ethical stakes of the conflation are not abstract. A construct treated as an essence directs research money at validating the category rather than decomposing the mechanisms; it tells a person “you have this thing” when the honest statement is “you sit at this region of several dimensions, and here is which ones”; and it makes support conditional on crossing an arbitrary threshold that people one symptom short of it fall beneath, unsupported, while carrying most of the same difficulty.
An environment driving a bounded architecture past its range
The paper’s closing move is the one that connects it to everything this corpus argues, and it runs through Lewin and Herbert Simon. Lewin’s formula — behaviour is a function of person and environment — makes impairment relational: some part of what is diagnosed reflects the transaction between a cognitive profile and settings organised around prolonged stillness, delayed reward and monotony tolerance, rather than a defect wholly internal to the person. Simon supplies the modern half: in an information-rich world, what becomes scarce is attention itself. Human cognition is bounded — bounded attention, bounded working memory, costly selection — and contemporary informational ecologies place demands on that architecture unlike anything in the environment that shaped it.
Sherman-Palmer’s synthesis: modern environments do not merely reveal attentional limits; they actively amplify them, and their pressures are then reified as evidence of individual disorder. What the “epidemic of attentional complaint” may partly show is a species-typical nervous system driven beyond the range in which it was shaped to operate — “until everyone, or almost everyone,” in his phrase, qualifies for the same provisional diagnosis. A construct behaves exactly this way under environmental load: its prevalence tracks the mismatch. A natural kind would not.
Readers will recognise this as the cybernetic argument I have made from the other direction — digital environments producing ADHD-like presentations at population scale, attention strain read as technostress, the environment formatting the substrate and the diagnostic apparatus then locating the result inside the person. Sherman-Palmer arrives at the same junction from bounded-rationality economics and philosophy of psychiatry, apparently without contact with the neurodiversity literature — the paper never uses the word. Convergence from independent starting points is worth something.
The honesty notes, clearly: this is a single-author position paper from a management school, not a peer-reviewed review; it synthesises rather than adds evidence; and its author’s institutional distance from clinical psychiatry cuts both ways — freedom from the jurisdiction he analyses, but also distance from the clinic. The primary sources it rests on, though — Hyman, Kendler, Plomin, the NIMH and APA’s own statements, the biomarker consensus — are load-bearing regardless of who assembled them.
Where does this leave someone who carries the label — who perhaps, like me, has built an understanding of their own mind partly through it? Not dispossessed. The difficulties the label compresses are real; the support it unlocks is real and, per Tuesday, life-preserving; and nothing about “mechanistic property cluster” makes anyone’s struggle less legitimate. What changes is the direction of explanation. The label is the compression, not the cause — a threshold drawn across several continuous dimensions, held in place partly by the institutions that need it, naming a real region of human variation without carving nature at a joint. The mature position is to use the category for what it does while refusing to worship it for what it isn’t — and to keep asking the question that survives its deconstruction intact: not “do I have the thing?” but “which mechanisms, in which environments, produce my difficulty — and which of those environments could change?”
Citations
Sherman-Palmer, C. L. (2026) — Deconstructing the Ontology of ADHD: A Neuroscientific Review of Dimensionality, Behavioural Genetics, and Cognitive Architecture — position paper, École Universitaire de Management, Université de Bordeaux
Hyman, S. E. (2010) — The diagnosis of mental disorders: The problem of reification — Annual Review of Clinical Psychology
Kendler, K. S., Zachar, P. & Craver, C. (2011) — What kinds of things are psychiatric disorders? — Psychological Medicine
Plomin, R., Haworth, C. M. A. & Davis, O. S. P. (2009) — Common disorders are quantitative traits — Nature Reviews Genetics
Thome, J., Ehlis, A.-C., Fallgatter, A. J. et al. (2012) — Biomarkers for attention-deficit/hyperactivity disorder (ADHD): A consensus report — World Journal of Biological Psychiatry
Kotov, R., Krueger, R. F., Watson, D. et al. (2017) — The Hierarchical Taxonomy of Psychopathology (HiTOP): A dimensional alternative to traditional nosologies — Journal of Abnormal Psychology
Conrad, P. (2005) — The shifting engines of medicalization — Journal of Health and Social Behavior
Faraone, S. V., Bellgrove, M. A., Brikell, I. et al. (2024) — Attention-deficit/hyperactivity disorder — Nature Reviews Disease Primers
