Defining Abnormality

In brief

Before a manual can list disorders, someone must decide what a disorder is. Defining abnormality is that prior question: what turns a mental state into a condition, rather than an eccentricity, a vice, a talent, or simply a life the culture around it dislikes. Four criteria have been offered, and each, taken alone, fails in a way worth studying. A state can be statistically rare without being disordered, and common without being healthy. It can violate a social norm that later looks like prejudice. It can distress the person, or fail to. It can impair functioning, or leave it intact. The most influential modern attempt, Jerome Wakefield's harmful dysfunction, tries to fuse a factual claim (something the mind was built to do has broken) with a value claim (the breakage harms the person by the standards of their culture). Whether that fusion succeeds is the live debate. What is not in dispute is that the line has moved, and that some of its historical positions are now indefensible.

The full treatment

The prior question

Diagnostic systems presuppose their subject matter. To classify disorders you must already know which states count, and that decision is not made by the manual; it is made by the concept of abnormality the manual inherits. The concept is surprisingly hard to state. Physical medicine has an easier time: a broken bone can be pointed to, and most cultures agree it is bad. Mental states resist this. Grief, fear, obsession, and despair are all part of ordinary life at some intensity, and there is no thermometer that reads disorder. So psychology has reached instead for criteria, general marks by which an abnormal state might be told from a normal one.

The four criteria

The oldest is statistical: abnormal means rare, a deviation from the average. It has the appeal of objectivity, and it fails immediately. Genius, unusual altruism, and perfect pitch are statistically rare and none is a disorder, while conditions like mild anxiety are common enough to be near universal. Rarity tracks distance from the mean, not pathology, and cannot say which direction of deviation is bad.

The second is the social norm, or deviance criterion: abnormal means violating the rules a society lives by. This captures something real, since much of what gets called madness is conduct others cannot make sense of. But it makes the clinician an enforcer of local custom. What one culture treats as a symptom, another treats as a gift or a duty, a point pressed by the broader argument of cultural relativism. Worse, it licenses the labeling of dissent as illness, which history has done repeatedly.

The third is distress: abnormal means suffering. This puts the person's own experience at the center, which is humane, but it under-counts and over-counts at once. Some serious conditions, mania at its height or certain personality disorders, feel fine to the person and painful only to those around them, while ordinary bereavement is agonizing and is not a disorder. Suffering is neither necessary nor sufficient.

The fourth is dysfunction: abnormal means a breakdown in the ability to carry on the tasks of a life, at work, in relationships, in self-care. This is closest to the medical intuition and does the most work in real diagnosis. Its problem is the word itself. Functioning is measured against an expectation, expectations are set by roles, and roles are cultural, so a state that impairs one life may suit another. Teaching heuristics bundle these into four Ds, deviance, distress, dysfunction, and danger, precisely because no single D holds alone.

Wakefield's harmful dysfunction

Jerome Wakefield's 1992 analysis, published in American Psychologist, is the most cited attempt to do better. A condition is a disorder, he argued, only when two things are true at once. First, a factual condition: some internal mechanism is failing to perform the function it was shaped by evolution to perform, a genuine dysfunction in the biological sense, not merely a difference. Second, a value condition: that failure causes harm to the person as judged by the standards of their culture. Neither half suffices. A part of the mind can work exactly as designed and still be disliked (rebelliousness, a taste for risk), and a state can be unwelcome without any mechanism having failed. Only where a real breakdown meets real harm, on the culture's own terms, does Wakefield grant the word disorder. The account is deliberately hybrid: it concedes that harm is evaluative and culture-bound while insisting that dysfunction is factual, anchoring the concept to something outside opinion.

Distinctions that matter

Three lines are easy to blur. Abnormal is not the same as bad, since much abnormality is admirable or neutral. Disorder is not the same as disease, since a disorder need not have a known lesion or cause. And the question here, what makes a state a disorder at all, is prior to and separate from how a manual sorts and names disorders. That downstream work of criteria, codes, and revisions belongs to psychiatric-classification and the DSM and ICD systems it describes. The concept examined here is the ground those systems stand on.

Lineage

The problem is old. Émile Durkheim (1858 to 1917) argued that even crime is statistically normal in every society, sharpening the point that frequency cannot define pathology. The modern debate opens with the antipsychiatry challenge of the 1960s. Thomas Szasz (1920 to 2012), in The Myth of Mental Illness (1961), argued that mental illness is a metaphor and a category mistake: illness belongs to the body, and what we call mental disorder is really a name for problems in living and for conduct we disapprove of. Michel Foucault (1926 to 1984), in Madness and Civilization (1961), traced how the mad were redefined across centuries by the institutions that confined them, treating abnormality as a product of social power rather than a fact of nature. Against this dissolving pressure, Wakefield's harmful dysfunction (1992) was an attempt to rescue the concept as real without denying its evaluative core, and it descends visibly from the effort to keep psychiatry honest that its critics forced.

The strongest case for it

The reason to keep a worked-out concept of abnormality, rather than abandon the idea as Szasz urged, is that the alternatives are worse. Real suffering and real breakdowns of mind exist, and people need help with them; a psychology that refused to name any state disordered could not offer that help or study it. Wakefield's hybrid does something valuable: it holds the factual and the evaluative apart instead of letting them collapse. It gives a principled reason why drapetomania was never a disorder (no mechanism had failed; the enslaved were responding rationally to bondage) and why homosexuality is not one (nothing is broken, and the harm was imposed by prejudice, not by the state itself). By requiring a genuine dysfunction it blocks the medicalization of mere difference; by requiring harm judged on the culture's own terms it refuses to pretend the concept is value-free. It explains our clearest historical mistakes as mistakes, which neither a purely statistical nor a purely normative criterion can do.

The strongest case against it

The account has drawn hard, specific fire. The deepest objection targets its factual half. Wakefield leans on the function a mechanism was designed by evolution to perform, but evolutionary function is often unknown, contested, or plural, and evolution optimizes for reproduction, not well-being, so what counts as a dysfunction can be as arguable as the value judgment it was meant to anchor. Critics including Rachel Cooper and Dominic Murphy have pressed that the factual leg is not the firm ground Wakefield needs. A second line, from the social constructionists, holds that no such repair can succeed because the categories are shaped by social forces all the way down, and the history is their evidence. David Rosenhan's study, On Being Sane in Insane Places (Science, 1973), in which healthy volunteers admitted to hospitals reporting a single fabricated symptom were kept for weeks and discharged as schizophrenic in remission, was read for decades as showing that clinicians cannot reliably tell the abnormal from the normal. (The journalist Susannah Cahalan has since questioned the study's honesty, which complicates but does not dissolve the worry it raised.) A third line accepts a concept of disorder but attacks how loosely it is applied. Allen Frances, who chaired the DSM-IV task force, argued in Saving Normal (2013) that expanding diagnostic categories medicalizes ordinary distress, turning grief, shyness, and childhood temper into treatable disease and widening the net far past anything harmful dysfunction would license. Here the concept may be sound while its use is not.

Where it stands now

No single criterion has won, and the field has largely stopped expecting one to. The working consensus is that abnormality is a hybrid, part fact and part value, best triangulated from several marks at once rather than defined by any one. Wakefield's harmful dysfunction remains the reference point both sides argue against, as it has for three decades, without settling the matter. Two indefensible historical cases stay in every textbook as fixed stars: drapetomania, coined by Samuel Cartwright in 1851 to cast the flight of enslaved people from bondage as a disease, and homosexuality, listed as a disorder until the American Psychiatric Association's 1973 vote removed it (formalized in DSM-III in 1980), after Evelyn Hooker's 1957 research found no difference in adjustment between homosexual and heterosexual men. They mark the low-water line the concept must never return to, and they are the reason the prior question, what makes a state abnormal at all, cannot be handed off to a manual and forgotten.

Test yourself

Think of a trait in yourself or someone close that has, at some point, been called a problem: a temper, a shyness, an intensity, a way of grieving. Run it through the four criteria. Is it rare, does it break a social rule, does it cause distress, does it impair a real function, and by whose standard? Notice how the verdict shifts as you change the criterion and the culture you measure against. That shifting is the whole difficulty, and it is why the line was drawn wrongly before and can be drawn wrongly again.