The Salem witch trials ran from February 1692 to May 1693, with the first formal complaints sworn in Salem Village on February 29, 1692, and the final executions occurring on September 22, 1692. Nineteen people were hanged on Gallows Hill, one man was pressed to death with stones, and at least five more died in custody. The colony formally dissolved the court in October 1692, though related trials continued into the following spring.
The legal and social mechanisms that built the crisis
Understanding the Salem outbreak requires tracing how a fragmented legal system, a frontier war, and local factional disputes converged. Massachusetts operated under a revoked charter from 1684 to 1692, meaning the colony technically lacked lawful governing authority. When Sir William Phips arrived as the new royal governor in May 1692, he inherited a backlog of unresolved legal matters and a population primed for supernatural explanation.
The key mechanism was the "spectral evidence" doctrine. Puritan theology held that the Devil could assume the shape of an innocent person to torment others. A accuser's claim that a ghostly apparition of the defendant had attacked them was therefore admissible in court, even though no physical evidence existed. This inverted the normal evidentiary burden: defendants had to prove they had not sent a specter, which is logically impossible. The result was a system where accusation itself created probable cause, and denial compounded suspicion.
Local social dynamics amplified this. Salem Village had spent years feuding over its minister, land boundaries, and whether to separate from Salem Town. The Putnam family faction had lost influence to the Porter family faction. Most accusers came from the Putnam side; most accused came from the Porter side or from neighboring towns without local defenders. Historian Paul Boyer and Stephen Nissenbaum argued in Salem Possessed (1974) that these economic and factional lines tracked the outbreak closely, though later research has complicated the neatness of that thesis.
War also mattered. King William's War (1688–1697) had pushed refugees from the Maine frontier into Essex County. The Wabanaki were French-allied Indigenous peoples whom many colonists demonized; Tituba, the enslaved woman whose confession launched the trials, was herself an outsider in the household of Samuel Parris. Historian Mary Beth Norton argued in In the Devil's Snare (2002) that accusers often named people connected to the frontier conflict, suggesting the war shaped the imagery of spectral attack.
The colony's system of trying felonies required a grand jury, a petit jury, and judges sitting under a legal framework that had lapsed. The Court of Oyer and Terminer, established May 27, 1692, was a workaround—a special court with Chief Justice William Stoughton presiding and judges including Samuel Sewall. It lacked the ordinary procedural safeguards and accepted spectral evidence as standard. Plainly: the trials happened because the legal system was temporarily broken, and the people running it believed the accusations.
Quantified context: scope, death toll, and timeline
The raw numbers are sobering and worth stating precisely, because popular accounts often blur the figures.
Approximately 200 people were accused across Essex County and beyond. 19 were hanged: 14 women and 5 men, on Gallows Hill between June 10 and September 22, 1692. Giles Corey was pressed to death on September 19, 1692, for refusing to plead; pressing was a legal procedure to force a plea, not a punishment for witchcraft. At least 5 died in jail, including Sarah Osborne and the infant of another prisoner. The youngest accused was four-year-old Dorcas Good, daughter of Sarah Good, who was jailed and later released. The court sat from June 2 to September 22, 1692; the last hangings occurred September 22, 1692.
| Session or date | Event | Deaths from that session |
|---|---|---|
| February 29, 1692 | First complaints against Tituba, Sarah Good, Sarah Osborne | None (jail only) |
| June 10, 1692 | Hanging of Bridget Bishop | 1 |
| July 19, 1692 | Hanging of five defendants | 5 |
| August 19, 1692 | Hanging of five defendants | 5 |
| September 19, 1692 | Giles Corey pressed to death | 1 |
| September 22, 1692 | Hanging of eight defendants | 8 |
| October 29, 1692 | Governor Phips dissolves the court | 0 |
| May 1693 | Remaining prisoners released; Superior Court ends proceedings | 0 |
The chronology matters because the trials are often described as a single event. They were a rolling crisis: accusations spread outward from Salem Village to Andover, Topsfield, Ipswich, and beyond. The typical pattern was that one confession named others, who were then arrested and pressed to confess, generating a chain that expanded faster than the court could process it.
Edge cases and the mechanisms of escalation
Why did the accusations spread so rapidly? Several mechanisms deserve scrutiny because they explain how a local dispute became a regional panic.
First, confession was incentivized. Defendants who confessed and named accomplices were not executed; those who maintained innocence were hanged. This created a rational incentive to accuse others, even if the accusation was false. Confessions were sometimes extracted under torture-like conditions: prisoners were tied in contorted positions and deprived of sleep, a procedure called "watching." The 1692 confession of Tituba described the Devil's book and unnamed conspirators, and it set the template that later confessions followed.
A counterintuitive point: some historians argue the crisis was not irrational at all in its own terms. The Puritan worldview assigned real power to the Devil and to witches, and the legal system was designed to detect offenses that could not be directly observed. The escalation was a failure of institutional design, not simply mass hysteria. This distinction matters because it reframes the trials as a case study in how bad evidentiary rules produce catastrophic outcomes when the stakes are capital.
Second, the accused had limited ability to mount a defense. They could not call character witnesses against spectral evidence in the early sessions. Lawyers were allowed but few defendants had counsel; the judges themselves often questioned the accused aggressively. The accusers, often teenage girls, were treated as credible victims whose afflictions (convulsions, contortions, visions) were medical symptoms of witch attack.
Third, the geography of accusation mattered. People who lived on the periphery of Salem Village or in towns without kin networks were vulnerable. The few accused who escaped execution were either already confessed, pregnant, or had connections that generated public sympathy. The court eventually faltered not because of moral awakening but because governing authorities began to doubt the evidence, particularly after Governor Phips's own wife was accused.
The comparison table below positions the Salem trials against other major early modern witch panics, showing scale and institutional response.
| Event | Location | Peak years | Approximate executions | Legal response |
|---|---|---|---|---|
| Salem trials | Massachusetts Bay Colony | 1692–1693 | 20 (19 hanged, 1 pressed) | Spectral evidence accepted; court dissolved 1692 |
| Würzburg persecutions | Holy Roman Empire | 1626–1631 | Several hundred | Bishop-led; ended as war and politics shifted |
| Bamberg persecutions | Holy Roman Empire | 1623–1632 | Several hundred | Prince-bishop-led; ended after political backlash |
| Trier persecutions | Holy Roman Empire | 1581–1593 | Several hundred | Archbishop-led; ended after protests |
| Pendle witch trials | Lancashire, England | 1612 | 10 hanged | King James I's own skepticism on evidence helped end panic |
| Salem (England) trials | No major panic recorded | — | — | Separate from Massachusetts events |
The table's lesson: Salem was small by continental European standards, but it occurred later, in a colony with a functioning print culture that documented and criticized it in real time.
Misconceptions that stubbornly persist
The most common misconception is that the victims were burned at the stake. They were hanged, with one exception. In England and on the continent, burning was common, but Massachusetts used hanging for felony witchcraft.
Another common error: that the trials ran for years. The formal court sat for about four months, and the entire crisis from first complaint to final release was roughly fifteen months. The aftermath—compensation, apologies, and legal rehabilitation—took decades.
A third: that the accusers were all young girls. Accusers included children, teenagers, adult women, and men. The "afflicted" witnesses who convulsed in court were a core group of adolescents, but the broader accuser pool was socially mixed.
A fourth: that the accused were all poor outcasts. Some accused were wealthy and well-connected, including Rebecca Nurse, a respected matriarch of 71. Her conviction shocked the colony and prompted petitions for mercy.
A fifth: that the crisis ended because people became enlightened. The court dissolved because Governor Phips needed to stabilize the colony, and because judges like Sewall and Increase Mather publicly reconsidered the evidence. The political pressure and the specter of the governor's wife under accusation mattered more than any sudden moral conversion.
Practical actions: how to trace the actual record
For researchers, teachers, and readers who want to verify claims about the trials, the following steps work:
Start with the primary source collections: the Salem Witch Trials Documentary Archive and Transcription Project (University of Virginia) provides transcriptions of the legal papers, including warrants, depositions, and trial records. Check the date of a specific execution against the court's own records rather than using secondary summaries; the dates above can be confirmed from the legal transcripts. Distinguish "accused," "indicted," "convicted," and "executed," because these categories differ sharply; about 200 were accused, roughly 30 were convicted, 20 died as a direct result of proceedings, and 5 or more died in jail. Note that "Salem" is geographically imprecise: the trials involved Salem Village (now Danvers) and Salem Town (now Salem), plus Andover, Topsfield, Ipswich, and other Essex County towns. For later restitution, consult the 1706 reversal of attainders, the 1711 Massachusetts act compensating survivors and families, and the 1712 formal reversal of Sarah Good's and others' attainders; the historical timeline spans more than two decades. Treat modern secondary sources that cite no page numbers or archival references with skepticism; the primary archive is well-organized and searchable.
Red-flag warnings:
A source that says the victims were burned in Massachusetts is wrong for all but possibly one later case; the hangings were the method in 1692. A source that gives a precise total of "25" or "30" without breaking down deaths by category is probably conflating accused, convicted, and deceased. A source that treats Tituba as fully documented in motivation is overreaching; her 1692 confession is a legal document, not a reliable autobiography, and her status as an enslaved Indigenous woman complicates any account of her choices. A source that says the trials "led directly" to the American legal principle of separation of church and state is anachronistic; the connection is thematic and was argued much later.
Limitations and what the record cannot settle
Several questions about Salem cannot be answered with confidence from surviving evidence. We cannot reliably reconstruct the motives of individual accusers; the depositions record what they said, not why. We cannot fully account for the medical symptoms (convulsions, paralysis, visions) that began the crisis; historians have proposed ergot poisoning, encephalitis, and social performance, but none is confirmed, and the diagnosis is not the mechanism of the trials anyway.
We also cannot treat trial testimony as literally transparent. The transcripts were produced under adversarial conditions, in a legal system that we now consider unjust. Many deposition statements were shaped by the questions asked. The records show what was said in court, not what happened in the accused's life.
The best-supported conclusion is institutional: Salem happened because a colonial legal system under stress accepted an evidentiary standard that made accusation nearly sufficient for conviction. The fix arrived not through philosophy but through political intervention, judicial reversal, and later compensation. The number to remember is twenty: nineteen hanged, one pressed, plus at least five jail deaths. The dates to remember are February 29, 1692, and May 1693, with the trial court active from June to October 1692.