Salem Witch Trials: 3 of 52 Convicted in 1693, Phips Wrote

What was spectral evidence at Salem?

Bridget Bishop was examined on April 19, 1692, before the magistrate John Hathorne. Samuel Parris and Ezekiel Cheever both took minutes, and the historian Charles Upham printed their substance in 1867 (Exhibit H).

The “afflicted children,” Upham wrote, charged her with trying to persuade them to sign the Devil's book. Bishop answered, “I am innocent. I know nothing of it. I am no witch. I know not what a witch is.”

Bridget Bishop's death warrant, signed by William Stoughton on June 8, 1692, with Sheriff George Corwin's return of June 10 written at the foot (Exhibit A). Photograph of the manuscript by Frank Cousins.

The same minutes show how a specter could leave a physical trace, at least in the examiners' eyes. Asked “Is not your coat cut?” Bishop said no, and the examiners found it “cut or torn two ways.” Jonathan Walcot then said that “the sword that he struck at Goody Bishop was not naked, but was within the scabbard.” Cheever wrote that the rent “may very probably be the very same that Mary Walcot did tell” of after Jonathan struck at Bishop's “appearance.”

Giles Corey's examination before Hathorne and Jonathan Corwin the same day, printed in the appendix to Samuel G. Drake's 1866 edition of Robert Calef (Exhibit B), shows the charge in its plainest form. “Giles Cory, they accuse you, or your Appearance, of hurting them,” the questioner said. Corey said, “I never did hurt them.” The reply: “It is your Appearance hurts them, they charge you.”

A suspect could deny what her hands had done, but the charge was about her likeness, which the accusers said they could see. Upham printed the theory in one exchange from George Jacobs's examination: “The Devil can take any likeness,” Jacobs said, and the answer came back, “Not without their consent.” According to Upham, opinion split on that point: “Some maintained that the Devil could employ only the spectres of persons in league with him,” while others held that he could send the specters of innocent people “without their consent or knowledge.” “The chief-justice held the former opinion,” Upham wrote, so on that view a specter pointed back to its owner.

Phips later described the courtroom version (Exhibit E). Accusers fell “to the ground in strange agonies” when a suspect looked at them and revived when touched. Then they swore they “saw their shape or spectre come from their bodies.”

Did Salem convict people on spectral evidence alone?

For some defendants it may have come close. For others the files hold more, and for most of the 1692 cases no one can fully check. In 1869 Upham described Bishop's trial and execution as “resting as they did, entirely upon spectral evidence” (Exhibit D). That is the charge this section tests.

Bishop's file says more than that. Cotton Mather's account of her trial at the Court of Oyer and Terminer on June 2, 1692, printed in 1693 (Exhibit C), sums up the sworn testimony of John Bly and William Bly. Hired to take down the cellar wall of her old house, they found “several Poppets, made up of Rags and Hogs-bristles, with headless Pins in them, the Points being outward.”

Cotton Mather, whose 1693 account of Bridget Bishop's trial is quoted in this section. Mezzotint by Peter Pelham, 1727.

Mather's account adds that “a Jury of Women found a preternatural Teat upon her Body: But upon a second search, within 3 or 4 hours, there was no such thing to be seen.”

The judges also told Phips that they began with the afflicted persons' evidence “but had humane testimony against such as were condemned” (Exhibit E). Other cases look thinner. Robert Calef, a Boston critic of the trials, wrote that against Samuel Wardwell “his former Confession and Spectre Testimony was all that appeared against him” (Exhibit B). Phips himself wrote that when he stopped the court, at least fifty people sat in prison, “most of them having only spectre evidence against them.”

The weak point in any answer is the archive. Benjamin Ray built the University of Virginia's digital archive of the trials. He writes that the trial records of those convicted in 1692 are missing (Exhibit I).

The surviving papers hold examinations, sworn statements and indictments, but not what each trial jury heard and weighed. The record puts spectral evidence at the center of the 1692 cases with other testimony beside it; it does not say which one carried each conviction.

What did the ministers tell the judges?

The warnings came early, and in writing. On May 31, 1692, the Boston minister Cotton Mather wrote to John Richards, one of the judges, in a letter Upham printed in full in 1869 (Exhibit D). He asked Richards not to “lay more stress upon pure Spectre testimony than it will bear.”

According to Upham, the same letter endorsed evidence from “witch-marks.” It also called wounds given to specters and later found on suspects “intimations, broad enough, in concurrence with other things, to bring out the guilty.”

Bishop was hanged on June 10. The governor and council then consulted several ministers. Their answer is dated June 15 and survives as the “Return of the Ministers consulted by his Excellency and the honorable Council” (Exhibit D).

The Return's Section VI said that grounds for conviction “ought certainly to be more considerable than barely the accused persons being represented by a spectre unto the afflicted.” The reason it gave: “a Demon may, by God's permission, appear, even to ill purposes, in the shape of an innocent, yea, and a virtuous man.”

Increase Mather's own answer to the same question, Cases of Conscience, was dated October 3, 1692, and printed at Boston in 1693. It is quoted below, in the section on why the trials stopped.

Section VIII pulled the other way. It recommended “the speedy and vigorous prosecutions of such as have rendered themselves obnoxious.” Upham, who attributed the Return to Cotton Mather, read it as caution “in the use” of spectral evidence rather than “against the use.”

Look at the dates in the first chart. Of the 19 hangings, 18 came after June 15, and so did Giles Corey's death in September. The ministers' paper did not stop the court; what it changed inside the jury room, the surviving records do not show.

How did Rebecca Nurse's “not guilty” become “guilty”?

One file shows what the 1692 court did with a jury that disagreed with it. On July 4, 1692, the juror Thomas Fisk put in writing why the jury “brought her in Guilty, after her Verdict not Guilty” (Exhibit B). “When the Verdict not Guilty was, the honored Court was pleased to object against it,” he wrote. The judges pointed to words Nurse had spoken about two confessed witnesses who testified against her: “What do these persons give in Evidence against me now, they used to come among us.”

Illustration

An illustration of the kind of moment Thomas Fisk described: the jury back in court after the bench objected to its verdict. Not a record of the room or of the people in it.

Several jurors asked to go out again. Back in court, Fisk raised the words. Nurse, at the bar, “made no reply, nor interpretation of them.” So, Fisk wrote, “these words were to me a principal Evidence against her.”

In her own declaration to the court, Nurse said she was “something hard of hearing, and full of grief,” and had meant only that the two women were fellow prisoners. She was hanged on July 19.

By Fisk's account, the turn came from the bench's objection and from her silence. His statement names no specter. Calef later claimed the court had “cleared none upon Tryal” (Exhibit B). Nurse's file shows that one jury did try, briefly.

Why did the Salem witch trials stop?

The end came from the top of the colony. On October 3, 1692, Increase Mather, Cotton's father and president of Harvard College, dated his Cases of Conscience. It said: “It were better that ten suspected Witches should escape, than that one innocent Person should be Condemned” (Exhibit C).

On October 12, Phips wrote to London from Boston (Exhibit E). Back from fighting in the east, he had found “that the Devill had taken upon him the name and shape of severall persons who were doubtless inocent.” He had forbidden jailing anyone else who was accused “without unavoydable necessity” and had “put a stop to the proceedings of the Court.” In late October, the judge Samuel Sewall wrote in his diary that the Court of Oyer and Terminer “count themselves thereby dismissed” (Exhibit F).

From Cotton Mather's letter of May 31 to Sewall's diary entry of October 26, 1692: dated documents and Bridget Bishop's trial above the line, deaths by court action below. Drawn from the tables published with this page.

The documents, with sources (CSV) (where the data comes from) Data: Charles W. Upham, Salem Witchcraft and Cotton Mather: A Reply (1869); Cotton Mather and Increase Mather, The Wonders of the Invisible World (1862 reprint); George Lincoln Burr, ed., Narratives of the Witchcraft Cases (1914); Diary of Samuel Sewall, vol. 1 (1878)

Two cautions apply to Phips as a witness. He wrote partly in self-defense; the same letter says “my enemies are seeking to turn it all upon me.” He also did not doubt that the Devil was at work. His objection was that the Devil could wear an innocent person's shape, so “the evidence of the afflicted persons only” should not decide who was jailed or tried.

What happened at the 1693 trials?

A new Superior Court of Judicature opened at Salem on January 3, 1693, a date written 1692/3 at the time (Exhibit E). Its chief judge, Phips wrote, was again the lieutenant governor, William Stoughton. Stoughton had led the 1692 court and signed Bishop's death warrant (Exhibit A). Phips wrote that Increase Mather and other ministers had judged “that the look and the touch of the suspected persons was not sufficient proofe,” and so “these things had not the same stress layd upon them as before.”

Phips's count of the result is short: “all that were brought to tryall to the number of fifety two, were cleared saving three.” By his figures, 49 of the 52 were cleared. His letter does not say how he counted, which sessions he included, or what “cleared” covered.

Trial of George Jacobs, August 5, 1692, painted by T. H. Matteson in 1855 — an imagined scene, not a record. Shown from a Detroit Publishing Co. postcard reproduction, cropped to the painted picture.

Two other published accounts cover the Salem session alone. Calef wrote that bills against thirty people were thrown out, that bills against twenty-six more were found, and that “of all these Three only were found Guilty by the Jewry upon Trial” (Exhibit B). Upham named twenty people tried at Salem that January and wrote that the last three, Sarah Wardwell, Elizabeth Johnson Jr. and Mary Post, “were condemned, but not executed: all the rest were acquitted” (Exhibit H).

The court's own record of the 1693 sessions is transcribed, verdict by verdict, in the University of Virginia's Salem archive (Exhibit K). Every account read here ends with the same three convictions.

Phips's letter explains why those three did not hang. Stoughton signed a warrant for their “speedy execucion,” along with five others condemned in 1692. Phips “sent a reprieve.” Stoughton, Phips wrote, was “inraged and filled with passionate anger” and refused to sit when the court met at Charlestown.

How were the Salem convictions reversed?

One judge went first. Samuel Sewall handed the Rev. Samuel Willard a bill to read on a public fast day, and stood while it was read; in it he desired “to take the Blame and shame of it” (Exhibit F). The province's proclamation, printed with the diary, set “Thursday the Fourteenth of January next” as the fast: January 14, 1696/7 in the dating of the time, which the diary's 1878 editors give as January 14, 1697.

From the hanging of June 10, 1692, to the act of 1711, on an axis broken between spring 1693 and 1696. The dates in early 1693 and in January 1697 were written 1692/3 and 1696/7 at the time.

From 1692 to 1711, dated events with sources (CSV) (where the data comes from) Data: George Lincoln Burr, ed., Narratives of the Witchcraft Cases (1914); Robert Calef, More Wonders of the Invisible World (1700), ed. Samuel G. Drake (1866), vol. 3; Diary of Samuel Sewall, vol. 1 (1878); W. Elliot Woodward, Records of Salem Witchcraft, vol. 2 (1864); Charles W. Upham, Salem Witchcraft (1867)

On October 17, 1711, the General Court passed “An Act to reverse the attainders of George Burroughs and others for Witchcraft” (Exhibit G). It declared the convictions of 22 named people “null and void.” It also kept the old frame of the events. The act blamed the “Influence and Energy of the Evil Spirits” acting “in and upon those who were the principal accusers and Witnesses,” and said some accusers had since shown themselves “persons of profligate and vicious conversation.”

The act did not name everyone. Comparing its list with the deaths in this page's table, 14 of the 20 people who died by court action are named. Six are not: Bridget Bishop, Susanna Martin, Alice Parker, Ann Pudeator, Wilmot Reed and Margaret Scott.