Classification: Documented

Salem: how accusations became trials, sentences and memory

Between 1692 and 1693, the Massachusetts witchcraft crisis spread from a minister's home through several communities and left 25 people dead. Examination records, petitions and later reversals show how untestable evidence gained force, and how survivors fought to undo the judgments.

Salem, Massachusetts, Estados Unidos · See stories from this place and nearby

One name for several places

These events are usually called the “Salem witch trials,” but they began in Salem Village, a rural community now part of Danvers, Massachusetts. Hearings and trials also took place in the port town of Salem, while accusations reached Andover and other settlements. Nocturna's marker indicates the Salem area approximately; it does not compress every event into one address.

In early 1692 the colony faced religious and political tensions, local disputes and fear generated by war on the frontier of what is now Maine. The new colonial charter issued in 1691 altered relations with the English Crown. Such conditions did not automatically “cause” the accusations. They formed a setting in which supernatural explanations for illness and conflict found institutions ready to hear them. No single explanation, such as spoiled bread or a simple land dispute, has been demonstrated as the cause.

The minister's household and the cake that backfired

In January, nine-year-old Betty Parris, the daughter of Reverend Samuel Parris, and his eleven-year-old niece Abigail Williams developed alarming symptoms and behavior. Fasting, prayer and home remedies did not end the problem; talk of witchcraft took hold. As others reported similar afflictions, neighbors began to be named.

On February 25, neighbor Mary Sibley proposed a “witch cake,” a counter-magic practice involving rye meal and the children's urine. It was fed to a dog in the hope of revealing a supposed tormentor. Tituba and John Indian, both enslaved in Parris's household, prepared it following her instructions. This easily missed detail defeats the simple story that Tituba alone introduced some exotic ritual. Sibley received a religious reprimand and resumed her life; Tituba was imprisoned for more than a year.

Warrants went out on February 29 for Tituba, Sarah Good and Sarah Osborne. Good lived in poverty; Osborne was ill and had stayed away from church meetings; Tituba lacked the freedom and protection available to her accusers. Their circumstances help explain why these three were exposed first, without implying that everyone later accused shared the same status.

Questions that supplied their own answers

The three were examined on March 1. In Sarah Osborne's surviving examination, a magistrate asks why she hurts the children; she answers that she does not. The question presumes precisely what needs proving. Tituba initially denied harming the children, then described demonic figures and other women who supposedly compelled her to act. The record preserves the shift in her testimony; it does not verify the apparitions.

Her enslavement, the pressure of questioning and the possibility of abuse make it wrong to read the “confession” as a freely given admission or evidence of supernatural acts. By naming others and describing a wider group, her account opened paths for further accusations. Tituba was not executed. Tried only in 1693, she was found innocent; Parris refused to pay her jail fees, and she was sold to another owner to secure her release. She then disappears from the surviving record.

From examining witnesses to accepting specters

The first hearings were preliminary: magistrates decided whether a person would remain jailed for trial. In May, Governor William Phips created the Court of Oyer and Terminer, meaning “to hear and determine.” Presided over by William Stoughton, it began trying cases in June. Bridget Bishop was convicted at its first sitting and hanged on June 10, the first of nineteen people hanged that year.

“Spectral evidence” gave weight to a claim that an accused person's invisible shape had appeared and tormented a witness. Someone unable to see that figure could scarcely disprove the identification. Fits in court, older rumors, objects attributed to the accused and behavior judged improper could accompany these accounts. That a witness reported suffering is a historical fact; the conclusion that another person caused it through witchcraft is not a verifiable observation.

Even people who believed witchcraft existed criticized this procedure. Ministers advised the court in June not to rely on spectral visions alone. The warning did not end convictions. These were not simply trials “without rules”: a real judicial system adopted a standard of proof that could turn untestable suspicions into death sentences.

Rebecca Nurse: when the first verdict was not enough

Rebecca Nurse, an elderly and respected church member, did not fit the stock image of an isolated defendant with no allies. Neighbors signed a petition affirming her conduct. She requested a chance to contest a supposed bodily mark interpreted against her. Those papers survive in the University of Virginia archive. They are a defense mounted at the time, not an innocence invented centuries later.

The jury initially returned a not-guilty verdict. The judges asked it to reconsider, and the verdict became guilty. Her later explanation of a misunderstood remark and a question she had not heard did not change the outcome. Rebecca was hanged on July 19 with four others. The episode shows a process that could absorb a favorable petition and even an initial acquittal without abandoning the court's preferred conclusion.

Rebecca's family provides another break from simple retellings: her sisters Mary Easty and Sarah Cloyce were also accused. Mary was released after an initial detention, accused again and sentenced. Before her death she wrote to the judges, seeking above all to prevent more innocent bloodshed rather than merely to save herself; she asked them to examine accusers separately and test the confessions. Her petition is a critique of the evidence-making process from inside the unfolding tragedy.

The crisis moves beyond Salem

In September, Andover witnessed a scene that exposes the mechanism of expansion. In a “touch test,” accused people were blindfolded and led to those who said they were afflicted. If distress stopped after contact, the effect was taken as confirmation of witchcraft. At least seventeen supposed “witches,” including children, were identified in one gathering. An effect observed in an uncontrolled test cannot establish a supernatural cause; it shows how the test's rule generated suspects.

Nor were only women accused. Five men were hanged. Giles Corey, the husband of Martha Corey, refused to enter a plea before the court. On September 19 he died under stones used to force an answer. He was neither burned nor convicted by a jury and then executed. The last eight hangings took place on September 22. In all, nineteen people were hanged, one died by pressing, and five died in jail or awaiting release: 25 deaths linked to the crisis in the Salem Witch Museum's count. Estimates of how many were “accused” vary because a complaint, warrant, arrest and trial are different categories.

An end to executions did not free everyone at once

As criticism of spectral evidence grew, Phips dissolved the special court in October 1692. The court that heard remaining cases in 1693 no longer admitted it. Three people were still convicted, but the governor prevented further executions. That did not make release automatic: prisoners owed jail fees. Lydia Dustin, though acquitted, died in jail in March 1693 because she could not pay the amount needed to leave.

A public fast and day of atonement followed in 1697. Jurors acknowledged error, and Judge Samuel Sewall apologized for his part. Civil repair came slowly. In a petition of 1703, survivors and relatives argued that the weakness of the evidence and the harm to their families' names were plain, and asked for the convictions' stigma to be removed. In 1711 the colony reversed some judgments and approved compensation, but did not include everyone. Corrections continued for centuries: Elizabeth Johnson Jr., convicted in 1693 and spared execution, was exonerated through legislation in 2022.

The execution site and the record

For many years the summit of Gallows Hill was said to be the site of the hangings. A fresh study of maps, landscape descriptions and accounts by seventeenth-century neighbors led researchers in 2016 to confirm a lower site, Proctor's Ledge. A later memorial bears the names of the nineteen people hanged. The shift looks small on a map but matters to historical method: even a place that tradition had seemingly settled needed to be checked against records.

Salem was not a secret burning of accused women. It was a documented succession of complaints, warrants, examinations, testimony, verdicts, deaths, petitions and incomplete restitution. The papers tell us what people said and what officials decided; they do not prove witchcraft or reduce every accuser to a single motive. The precise weight of each individual's pressures remains open. What is clear is the collective failure of a system that treated an unanswerable allegation as sufficient evidence to destroy lives.

Open in Google Maps