Chapter 9

Updated 2026-07-02

"The law, which is an ass, knows no in-betweenness. It dichotomizes inflexibly, imposing a clumsy disjunction upon the subtly variegated flux of reality." — Graham Watson, Passing for White (1970), quoted in Bowker & Star, Sorting Things Out

On the night of the 1972 annual meeting of the American Psychiatric Association, a psychiatrist named John Fryer walked into a session wearing a rubber Richard Nixon mask, a wig, a tuxedo three sizes too large, and a microphone wired to distort his voice past recognition. He was introduced to the room as "Dr. H. Anonymous." Homosexuality was, at that moment, an official mental disorder, listed in the Diagnostic and Statistical Manual the same association published — and Fryer, a working psychiatrist, knew that if his colleagues in that room learned who he actually was, his career was over. From behind the mask, in the distorted voice, he said the sentence the costume existed to make possible: "I am a homosexual. I am a psychiatrist."¹

Fourteen months later, in December 1973, the APA's board of trustees voted to remove homosexuality from the manual entirely. The membership ratified the decision the following spring.² Millions of Americans woke up the next morning having been, the day before, clinically disordered by definition, and were not, by definition, any longer. Nothing about any of those people had changed overnight — not their feelings, not their relationships, not a single fact of their lives. What changed was a vote, taken in a room most of them would never see, by people most of them would never meet, about which box a professional manual would put them in. A diagnosis could be reversed by the same mechanism that had created it in the first place: not new evidence about the people being diagnosed, but a different decision about the category.

It is tempting to tell this story as a happy one, and in the long run it mostly was. But the APA vote proves something uncomfortable on the way to the happy ending: if a category can be removed by board decision and membership referendum, it was never describing a fact of nature — it was describing a decision. The same machinery that can lift a label off millions of people in an afternoon can put one on just as easily, and for exactly as little reason — and a few hundred miles and a few years away, a government was doing just that, with a thoroughness the APA never approached.

---

A South African legal scholar named Arthur Suzman, writing about his own country's race law a few years before the worst of it had even been built, put the whole coming project in a single sentence: "the legislature is attempting to define the indefinable."³ South Africa's Population Registration Act, passed in 1950, assigned every person in the country to one of a small number of official racial categories — broadly, white, "coloured," Asian, or "native" (the law's own term; "Bantu," dividing the African population into eight further subgroups, was the regime's preferred replacement, chosen partly because it let the government claim its policies served the people who were "really" African).⁴ By 1985, the body of law built on top of this single founding distinction ran past three thousand pages. Roughly a million people fell into the "coloured" category specifically — itself an internal patchwork of further subdivisions — and around a hundred thousand of them, over the decades, formally applied to be reclassified into a different one, with most applications denied. In one ordinary year, 1984, the government recorded 795 successful reclassifications: 518 people went from coloured to white, two white people became Chinese, one became Indian, eighty-nine Black Africans became coloured, five coloured people became African.⁵ Each of those numbers is a person whose legal race, and with it where they could live, whom they could marry, what schools their children could attend, changed by bureaucratic ruling, with no biological event of any kind having occurred.

The government did not rely on ancestry to sort people, which sounds, on its surface, like restraint — the alternative, after all, would be the American one-drop rule, descent traced backward through documentation. South Africa's classification boards explicitly rejected lineage as the test and wrote something far looser into the statute instead: "appearance and general acceptance and repute." In practice this meant a comb run through someone's hair to judge its curl; a study of complexion, eye shape, bone structure against an official skin-color comparison chart; folk theories about earlobe softness.

And, most infamous of all, the pencil test. A pencil was pushed into a person's hair, and the person was asked to bend forward. If the pencil held in place without falling, the verdict came down: "too woolly."

One man, stopped on the street and asked which soccer club he supported, had his answer used as race-classification evidence, on the official logic that "only natives play soccer." Hearings were held in private, lasting on average about twelve minutes, deciding the entire legal shape of a person's life in less time than a coffee break.⁶

The human cost sat where you'd expect, in families split down a line nobody in the family had drawn. A jazz musician named Vic Wilkinson was reclassified five separate times before he turned fifty, crossing the legal race line back and forth as different boards reached different verdicts on the same body. The scholars Geoffrey Bowker and Susan Leigh Star, who documented this system in exhaustive detail, have a word for what it does to a person: torque — the slow twisting of a life against a category it doesn't fit. Two preschool-age children were once held in a detention facility for three years while officials worked out a ruling on what race they were. One man's appeal succeeded when a judge declared him, in the judge's own words, "a white of the Mediterranean type" — a category invented, apparently, on the spot, to solve a single case. Families with members on both sides of a ruling developed their own quiet survival choreography: lighter relatives who had been classified white would pass on the street, while darker relatives who loved them learned to stare studiously into shop windows rather than risk a greeting that would expose the whole arrangement to a stranger — a practice with its own grim nickname, Venster-Kykers, window-lookers.⁷ Bowker and Star called the whole apparatus "an open degradation ceremony," borrowing the sociologist Erving Goffman's term for a ritual designed to publicly strip someone of standing. A South African legal scholar at the time put the underlying problem more simply than any of the bureaucracy's defenders ever managed to: "the law, which is an ass, knows no in-betweenness. It dichotomizes inflexibly, imposing a clumsy disjunction upon the subtly variegated flux of reality."⁸

The apartheid system's mechanics are familiar. A pencil in the hair and a soccer-club question are a prototype judgment — fuzzy, similarity-based, the kind a four-year-old uses on a pile of dinosaurs — wearing the uniform of a formal legal system and backed by the full violence of the state. The same two ways of sorting the world that sat side by side, harmlessly, in a child's mind — crisp rule and graded resemblance — sat side by side here too, and the combination was not harmless at all. A government got to claim the authority of a precise legal code while actually wielding the loose, on-the-spot judgment of a stranger's glance, and the conflation gave it the power of both at once: the appearance of objectivity, and the freedom of arbitrariness, available interchangeably, whichever served the moment.

And beneath both ran something older still. The boards' instruments were essence-detectors. A comb, a skin chart, a pencil pushed into hair — each one a tool for reading the underneath off the surface, built on the childhood conviction that what a person is lives somewhere deeper than anything that can be seen or changed. Every child who ever insisted the raccoon was still a raccoon under the dye holds that belief for free, about animals. The classification boards held it about people, with state power behind it — and their verdict about the underneath became, the moment it was stamped, legally true.

---

A category doesn't need a pencil test to have teeth. Sometimes all it needs is a form with too few lines on it.

The International Classification of Diseases is the system the world's death certificates run on — the list of approved, codeable causes of death that every country's vital-records office is expected to draw from. A death certificate generally records one underlying cause, chosen from the list, and whatever doesn't fit cleanly into a code is, for statistical purposes, simply not part of the official record of why a population died. Bowker and Star, the same scholars who documented apartheid's classification boards, point to a vivid case of what this does to data even now: in Japan, a death from heart attack carries a measure of social stigma a death from stroke does not, and they argue this shapes, in aggregate, which cause Japanese physicians are inclined to write down when a death could plausibly be coded either way — meaning international heart-disease statistics are, to some real degree, measuring not just bodies but reputations.⁹ The category is supposed to be neutral medical fact. What gets recorded inside it is also a negotiation with what a family can bear to see printed on the certificate.

This isn't a new problem invented by modern statistics. The seventeenth-century London Bills of Mortality, an early ancestor of today's vital-records systems, listed causes of death in categories a modern coroner would find bewildering: "frighted." "The itch." "Fainted in a bath."¹⁰ Behind every quaint entry sits the same constraint the ICD runs on today — a finite list, built by someone, that has to absorb every way a human being can stop being alive, and inevitably absorbs some of those ways more honestly than others. What can't be coded doesn't vanish from the world. It just stops being visible to anyone reading the aggregate numbers afterward, which, for purposes of public health funding, historical memory, and which causes of death get taken seriously a hundred years later, amounts to nearly the same thing.

---

Not every category that flexes its power does it gravely. Some do it for money, in broad daylight.

In 1893, the Supreme Court of the United States was asked to settle, with full judicial seriousness, whether a tomato is a fruit. The Tariff Act of 1883 taxed imported vegetables and exempted imported fruit, and an importer named John Nix, who had paid duties on tomatoes shipped up from the West Indies, sued to get the money back, arguing — correctly, botanically — that a tomato, growing as it does from the flowering part of a vine and carrying its seeds inside it, is a fruit by every definition science actually uses. The Court, in a unanimous opinion written by Justice Horace Gray, agreed with the botany and ruled against it anyway. Tariff law, Gray wrote, uses words "in the popular sense," not the scientific one — and in the popular sense, in the kitchen and at the dinner table, a tomato is served with the meal rather than after it, the way an apple or an orange is, and gets treated, ordinarily, as a vegetable regardless of what a botanist would call it.¹¹ The tomato lost its case. Nix paid the tariff. A perfectly correct scientific classification was overruled, with the full authority of the nation's highest court, by where people actually expected to find a tomato on the dinner table.

It is the same mechanism at its lightest: a category, when something real — status, money, a place to live, a clean death certificate, a clinical diagnosis — is riding on which side of the line a thing falls, will bend to serve whoever is asking, regardless of what the thing actually is underneath the label. The tomato's case cost an importer some money. The others cost people their families, their freedom, their place in the official record of how the world is shaped. Same mechanism, wildly different stakes. The question underneath is always the same: who decided, and what did they need the answer to be.

---