Between February 1692 and May 1693, in and around the village of Salem in the Massachusetts Bay Colony, more than 200 people were accused of witchcraft. Thirty were convicted. Nineteen were hanged. One man, Giles Corey, was crushed to death under stones for refusing to enter a plea. At least five more died in jail.
The phrase "witch hunt" entered the language because of these months. But the story most people know — hysterical girls, a superstitious mob, a mad clergyman — flattens something more interesting and more uncomfortable. Salem was not a story about ignorance. It was a story about what happens when a legal system accepts a category of evidence that cannot be disproved.
The village before the trials
Salem Village, now the town of Danvers, was not prosperous Salem Town on the coast. It was an inland farming community of a few hundred people, chronically at odds with itself.
The tensions were real and old. The village had fought for years to have its own church rather than travelling to Salem Town. It had run through three ministers in sixteen years, each departure bitter. Families split into factions, broadly between the farming interior and households tied to the merchant economy of the port. Land disputes ran through generations.
Add to this a genuinely frightening moment in the colony's life. King William's War was being fought on the Maine frontier, and refugees from destroyed settlements had arrived in Essex County with stories of massacre. A smallpox epidemic had passed through. The colony's charter had been revoked in 1684, leaving the legal status of every land title uncertain for years.
None of this caused the trials in a simple mechanical way. But a community already convinced that it was under siege, spiritually and physically, was primed to read misfortune as attack.
How it started
In January 1692, nine-year-old Betty Parris and her eleven-year-old cousin Abigail Williams — the daughter and niece of the village minister, Samuel Parris — began to have fits. They screamed, contorted, complained of being pinched and pricked by invisible things.
A local doctor, finding no physical cause, suggested bewitchment. Under questioning, the girls named three women: Sarah Good, a homeless beggar; Sarah Osborne, an elderly woman who had scandalised the village by marrying her indentured servant; and Tituba, an enslaved woman in the Parris household, probably of Native American origin.
All three were easy targets: poor, marginal, already gossiped about.
Then Tituba confessed. She described a tall man in black, a book to sign, animal familiars, night flights. Historians still debate why — pressure, possibly beating, possibly a shrewd read that confessors were treated better than deniers. Whatever her reason, the confession was catastrophic. It converted the case from "are these girls ill?" to "how large is the conspiracy?"
Because Tituba said there were others. And now the court had to find them.
Spectral evidence
The single most important thing to understand about Salem is a legal one.
The Court of Oyer and Terminer, convened in June 1692 by the new governor William Phips, admitted what was called spectral evidence: testimony that the accused person's spirit or shape had appeared to the victim and tormented them, even though nobody else could see it.
Consider what this means as a rule of evidence. The accuser is the only witness. The event is by definition invisible to everyone else. The accused cannot produce an alibi, because being physically elsewhere is no defence — it was your spectre, not your body. There is no possible testimony that refutes the charge.
The theological objection was raised at the time. Increase Mather, one of the colony's most influential ministers, argued that the Devil could take the shape of an innocent person, which would make spectral evidence worthless. His son Cotton Mather expressed similar doubts while still supporting the prosecutions. But the court accepted spectral evidence anyway, and once it did, the outcome was essentially determined.
The second structural problem: confession saved you. Nobody who confessed to witchcraft was executed at Salem. Those who maintained their innocence were hanged. A system that rewards confession and punishes denial will generate confessions, and each confession named more names.
The accused
The pattern of accusations broke down quickly. It began with the marginal and the poor, as such things usually do. Within months it had reached Rebecca Nurse, a 71-year-old church member of unimpeachable reputation, whose neighbours signed a petition on her behalf. The jury first found her not guilty; the judges asked them to reconsider a remark she had made, and they reversed. She was hanged in July.
By autumn the accusations had reached a former minister of the village, George Burroughs, hanged in August after reciting the Lord's Prayer perfectly on the gallows — something a witch was supposedly unable to do. They had reached wealthy merchants, a Boston minister's mother-in-law, and eventually, reportedly, the wife of Governor Phips himself.
Giles Corey, an 81-year-old farmer, refused to plead either way, knowing that a defendant who did not plead could not be tried and so could not have his estate forfeited. Under the old English penalty of peine forte et dure, heavy stones were piled on him over two days to force a plea. He gave none, and died.
Why it stopped
Salem did not end because someone proved the accusations false. It ended because the accusations climbed high enough to threaten the people running the trials, and because the theological argument against spectral evidence finally won.
In October 1692, Increase Mather published a work arguing, in a line often quoted since, that it were better that ten suspected witches should escape than that one innocent person be condemned. That same month Phips dissolved the Court of Oyer and Terminer and forbade further arrests. A new Superior Court convened in January 1693 with spectral evidence excluded. Of 56 people tried, 53 were acquitted. Phips pardoned the rest.
Remove one rule of evidence, and the entire machinery stopped working.
The reckoning
What followed is unusual for its era: the community admitted it had been wrong.
In 1697 the colony held a day of fasting and repentance. Samuel Sewall, one of the judges, stood in his Boston church while his confession of guilt was read aloud, and observed the day privately for the rest of his life. Twelve jurors signed a statement saying they had been deluded and asking forgiveness. Ann Putnam Jr., one of the principal accusers, publicly apologised in 1706, saying she had been deceived by Satan.
The colony began reversing convictions and paying restitution to families in 1711. Legal exonerations continued in fits and starts for centuries; the last accused person was not formally cleared until 2022.
What actually caused it
Theories abound and most explain part of it.
Ergot poisoning — a fungus on rye that produces LSD-like compounds — was proposed in 1976 and remains popular, but it struggles to account for the pattern of who had fits and when.
Conversion disorder, in which psychological distress produces genuine physical symptoms, fits some of the early behaviour. Outright fraud and social opportunism plainly explain some later accusations; the map of accusers and accused correlates suspiciously well with the village's existing factional and property disputes.
Trauma from the frontier war may explain the intensity: several of the most prominent accusers were orphans of raids on Maine settlements.
But the mechanism that turned distress into twenty deaths was procedural. A court accepted evidence that could not be tested, in a system where denial was punished and confession rewarded, during a period when the community already believed it was under supernatural attack. Any of those elements alone is survivable. Together they were lethal.
That is why Salem stays useful as a warning. The villagers were not stupid. They had a bad rule, and they followed it carefully, all the way to the gallows.
Be the first to share a thought.