In the autumn of 1671, Elizabeth Knapp of Groton, Massachusetts, experienced fits, periods without speech and episodes which neighbours and ministers interpreted as possession. Samuel Willard, the town’s minister, recorded an unusually detailed sequence of observations, conversations and changing explanations. His account is valuable evidence of a seventeenth-century crisis. It also preserves contradictions and uncertainty that prevent its religious interpretation from being treated as a demonstrated supernatural cause.
The account and its transmission
The principal source is Willard’s narrative, reproduced independently from the manuscript by Samuel A. Green in Groton in the Witchcraft Times, published in 1883. Green identifies Knapp as the daughter of James and Elizabeth Knapp, born in 1655. She was therefore sixteen during the main episode. The manuscript’s title uses the spelling Knap; Knapp is retained here as the familiar name for the same person.
Green explains that the manuscript belonged to the Mather Papers at the Boston Public Library and had also appeared in the Massachusetts Historical Society’s collections. His edition varies slightly from that earlier printing. The article follows the report in Green’s edition, with key passages checked against its printed page images. It does not claim a new examination of the original handwritten document.
The report itself distinguishes some events Willard witnessed from others he learned about. He says he was away from home on 30 October, when a severe episode began. The opening description of that evening consequently relies on information received from others. This distinction is easy to lose when the entire narrative is presented as one minister’s uninterrupted eyewitness testimony.
From unusual behaviour to a household crisis
Willard begins with roughly a fortnight of sudden cries, laughter and changes in expression. On 30 October, Knapp reportedly complained of pain and difficulty breathing. The following evening she said that two people were in the cellar. Others searched without finding anyone. Later that night came violent movements, shouting and an episode in which those present restrained her to prevent injury.
The narrative then follows alternating periods of distress, relative calm and conversation. Some fits involved speech; others prevented it. The physical struggles could require several people to hold her. These are descriptions of a distressed young person supplied by participants in her care. They do not establish an independent external force, even when Willard describes the strength involved as beyond ordinary simulation.
A physician initially considered a natural disturbance involving the stomach and bodily processes understood through contemporary medicine. Willard records a period of treatment and some improvement. After a more violent episode in late November, the physician changed his view, judged the condition diabolical and advised religious measures. That change documents the interpretation of a seventeenth-century practitioner, rather than a medical test proving possession.
Prayer and pastoral visits were central to the response. Ministers named in the narrative include Rowlandson, Bulkley and Estabrook. Willard repeatedly questioned Knapp about temptation, wrongdoing and possible dealings with the Devil. Her answers emerged in a setting where spiritual danger was already an organising explanation, rather than through an interview designed to compare modern clinical alternatives.
Accusations, admissions and retractions
At one stage Knapp implicated a woman whom Willard regarded as a trustworthy neighbour. He resisted accepting the charge on appearance alone. Knapp later withdrew it, explaining that an evil spirit had appeared in that person’s likeness. Another accusation was also reconsidered. The report therefore contains a warning against treating a claimed apparition as evidence that a real neighbour had caused the illness.
Her accounts of dealings with the Devil changed substantially. She spoke of promises involving money, clothes and relief from work, and gave inconsistent durations for these experiences. At different points she denied or affirmed having entered a covenant. A particularly elaborate claim involved signing with blood despite her inability to write. Later she withdrew important parts of that story.
The retractions matter, but they do not amount to a single confession that every fit was deliberately staged. Willard reports that she denied the alleged blood covenant and much of the apparition material, while continuing to describe distress and temptation. Some experiences were likened to dreamlike images. Her statements about what was happening, and whether an evil presence was within her, remained unstable.
Willard acknowledges this problem at the end. He considers the covenant question unanswerable because the declarations contradict one another. His confidence that her suffering was real sat alongside uncertainty about what some statements meant. A careful account preserves both positions, rather than selecting only the most dramatic confession or only its later denial.
The reported voice
One of the best-known episodes concerns a voice which insulted Willard and answered people in the room. He described a hollow sound and thought Knapp’s ordinary organs of speech were not visibly moving. He called for a light to inspect her and later listed the apparent absence of lip movement as an argument for an unusual cause.
That was a visual judgement made in a charged household setting, without a recording or a controlled examination. Other people addressed the voice, which Willard discouraged. His narrative records their participation and his concern that conversation might encourage the disturbance. It does not provide a way to establish the voice’s source independently of the observers’ expectations and Knapp’s condition.
The episode is sometimes reduced to a claim of an external entity speaking through a motionless mouth. Willard’s actual argument depended on what he thought he could see, the unfamiliar language of the insults and the religious significance he attached to them. Those considerations explain his conclusion; they do not independently verify the existence or identity of a spirit.
Knapp also requested a journey to ministers nearer Boston. Willard opposed an immediate trip because of deep snow, extreme cold and the danger of a fit during travel. His response shows that practical care and physical safety remained part of the household’s decisions, even while its members interpreted the illness through a theological framework.
Where the contemporary narrative stops
Towards the end, Willard says that the fits had become less violent: one person could then restrain her where several had previously been needed. He also noticed that visitors could precipitate episodes and that larger numbers were associated with greater violence. That observation is part of the original report; it should not be converted into a complete modern diagnosis or proof of conscious performance.
The narrative continues into January 1672. On printed page 20, Willard says that she remained without speech and subject to fits at the time of writing, 15 January. Although a preceding marginal entry marks 12 January, that does not replace the closing date in the text. The report ends with continued illness, not a fully documented cure or a conclusive demonstration of how it had arisen.
Green’s book also reproduces later commentary. Thomas Brattle’s 1692 discussion used the Groton episode to question the reliability of accusations during the Salem prosecutions. Later writers offered ventriloquism, fraud or hysteria as explanations. These comments belong to different periods and purposes. They cannot be folded into Willard’s original observations as though all were contemporary findings from one investigation.
The outcome is Insufficient evidence. There is a substantial narrative of distress, communal interpretation and contradictory statements, but no evidence capable of establishing a supernatural cause or securely diagnosing the whole episode at this distance. Describing the uncertainty also protects the neighbours whose alleged appearances were withdrawn and avoids treating religious accusation as proven wrongdoing.
Location and sources
The map uses a rounded reference for Groton town in Massachusetts, not Groton in Connecticut. It does not identify a current private house or imply that a surviving residence is open to visitors. The lead image reproduces the opening printed page of Willard’s account and identifies its nineteenth-century publication context.
The main source is Samuel Willard’s narrative in Samuel A. Green, Groton in the Witchcraft Times (1883), pages 7–21, with the later commentary on pages 21–23 kept distinct. The Hanover Historical Texts transcription provides another reading route but is marked unproofread. Coordinates use the 2025 Census Gazetteer. Sources and image rights checked 13 November 2026.