SOURCE: "The Salem Witchcraft Prosecutions: The Invisible World at the Vanishing Point," in Witchcraft, Magic, and Religion in 17th-Century Massachusetts, Amherst: The University of Massachusetts Press, 1984, pp. 160-83.

… Even before the Salem trials, there are ample indications that the clergy regarded the discovery of witchcraft as problematic. In the pre-Salem litigations, adherence to theological strictures had rendered the translation of popular suspicions into convictive proofs inoperational. During the Salem trials, the ecclasiastical recommendations of June 15 had advised against the use of spectral evidence without offering any alternative criteria for the validation of imputations of witchcraft. Now, in the aftermath of the Salem trials, some members of the clergy were prepared, however reluctantly, to reexamine the epistemological assumptions in terms of which acts of witchcraft and the identity of witches were believed to be humanly ascertainable.

It is in this context that Increase Mather's Cases of Conscience may be appreciated as a final resolution to the long-standing theological uncertainties about witchcraft, for in spite of its conciliatory stance toward the magistrates, Mather's treatise was far more than an exercise in political diplomacy. The questions Mather raised about spectral evidence were questions that applied to other forms of evidence as well. Cases was a demonstration within the framework of orthodox doctrine that the workings of the invisible world were not available to human understanding and that, accordingly, errors in the discovery of witchcraft could be neither controlled nor eliminated.

As Mather formulated it, the testimony of the afflicted was unreliable not because it overestimated the powers of witchcraft but because it underestimated the powers of Satan. The use of spectral evidence was based upon an assumption that was crucial to the discovery of witchcraft in general, namely, that God imposed humanly recognizable limits upon Satan's capacity for evil. It followed from this assumption that God would make the activities of Satan and his confederates available to humanity, either by preventing a witch from reciting the Lord's Prayer or by refusing Satan the power to impersonate innocent men and women in spectral representations or by some other sign. It was this belief about the laws of supernatural causation that allowed courts of law to proceed with confidence in their determination of responsibility for acts of witchcraft.

Mather's work undermined this confidence by positing an inscrutable God who would permit Satan to act in ways that were too subtle for human comprehension. Thus, with regard to spectral evidence, Mather argued that it was entirely possible for Satan to represent the images of innocent and pious persons to the victims of bewitchment.⁶² Indeed, Satan was so cunning that, in order to deceive the unwary, he could arrange for concomitances between empirically observable events. Accordingly, the recovery of afflicted persons when touched by their alleged assailants was as likely to be a diabolical trick as a providential sign.⁶³ Even the question of how to distinguish between possession and bewitchment could not be decided with certainty. It was quite conceivable that afflicted persons might only appear to be in conflict with the devil when, in fact, they were instruments of Satan's will. The outcome of Mather's critique was to suggest that human perceptions could not be trusted at all in matters relating to afflictions.

For Mather, however, the exclusion of spectral evidence as a method for the detection of witchcraft entailed the exclusion of all other forms of evidence in which it was assumed that the activities of Satan could be comprehended by humanity. The testimony of confessed witches was dismissed as the devil's testimony, and popular allegations of maleficia were not even granted the status of presumptions. What remained as criteria for the validation of imputations of witchcraft were most unlikely to furnish a basis for successful convictions.

Mather allowed as sufficient proofs, first, the free and voluntary confessions of the defendant and, second, the sworn testimony of two credible witnesses that the accused party and not any spectral representation had performed supernatural feats. With respect to the second criterion, Mather urged extreme caution: "It were better that ten suspected Witches should escape, than that one innocent person should be condemned."⁶⁴ Henceforth, the criteria for the validation of imputations of witchcraft—confession and the two-person rule—would in no way differ from the criteria applied to the validation of other capital crimes.

In the course of formulating the theological grounds for the rejection of spectral evidence, Mather arrived at conclusions he himself was at pains to disclaim. Indeed, it is clear that he was no less ambivalent about the theological implications of his work than he had been about its political implications. The postscript to Cases included reassurances not only to the magistrates but to all believers in witchcraft. For those among his readers who might be troubled by possible deviations from orthodox doctrine, Mather promised to publish another discourse in which he would demonstrate the existence of witches and the need for decisive action against them.⁶⁵ Nevertheless, the promised sequel to Cases was not published, and Mather's drastic reformulation of the rules of evidence was left to stand as one of the final statements by a New England divine on the subject of witchcraft.

Ultimately, it was a later work by another minister that most directly confronted the epistemological problems relating to witchcraft. John Hale's Modest Inquiry into the Nature of Witchcraft made explicit the full theological implications of Mather's dissertation. Hale's work was published posthumously in 1702, some five years after it had been written. That the work evoked strong reactions from readers familiar with the manuscript may account for the delay. Despite his own public misgivings about the Salem trials, Sewall noted in his diary on November 19, 1697: "Mr. Hale and I lodg'd together; He discours'd me about writing a History of the Witchcraft; I fear lest he go into the other extream."⁶⁶

Hale's critique of the evidences used in the Salem trials added little to what had already been articulated in Cases of Conscience. The criteria he recommended for the discovery of witchcraft were identical to those proposed by Mather.⁶⁷ What was distinctive about Hale's work was not its recommendations for future prosecutions but rather its analysis of past prosecutions, including the Salem trials. Hale's accomplishment was to formulate the problem of error in the Salem trials in such a way as to call into question the entire history of witchcraft prosecution in New England. Of all the contemporary commentators on witchcraft, it was Hale alone who dared to suggest that what made the Salem trials problematic made the discovery of witchcraft problematic altogether.

Certainly, by 1697, there were few inhabitants of Massachusetts Bay who would have challenged the claim that errors had been committed during the Salem trials. Far more controversial was the question of what errors had been committed. For some members, it was enough merely to note the consequences of judicial action. As the bill of October 26, 1692, had stated, the trials had resulted in the indictment and arrest of an extraordinarily large number of persons, many of whom were individuals of exemplary piety. Accordingly, the number and quality of the defendants were proof in themselves that mistakes had been made. More often, members pointed to the judicial reliance on spectral evidence. For some critics, the magistrates had violated the theological recommendations on the use of such testimony; for others, the court had confounded afflictions with possessions. In general, however, the tendency of contemporary postmortems was to rescue witchcraft from the Salem trials by conceiving of the errors of the Court of Oyer and Terminer as specific to the events of 1692 rather than as applicable to all witchfinding activities.

For Hale, the problem of error was more complex. He himself had been one of the most active supporters of witchcraft prosecutions among the clergy, and he had not offered this support on the basis of spectral evidence alone. The problem posed by the Salem trials was not that the rules of evidence were treated lightly but rather that they had been taken so seriously. The Court of Oyer and Terminer had organized the most cautiously empirical and systematic investigation into witchcraft ever to occur in New England. In addition to collecting spectral evidences as well as allegations of malefic harm, the magistrates had conducted searches for witch's marks, puppets, and special healing potions. Moreover, they had gathered testimony from confessed witches who implicated other suspects. They had even conducted experiments in the courtroom in order that the devil might reveal his confederates before witnesses.

The cumulative effect of this testimony was to corroborate in detail the magistrates' version of conspiracy. Indeed, during the trials, one of the best educated of the defendants, George Burroughs, conceded that the testimony against him was convictive in view of the fact that eight confessed witches had identified him as their leader.⁶⁸ Since he nevertheless claimed he was innocent because his accusers were false witnesses, Hale proceeded to question one of the confessors who had named him. She stood by her account even after Hale had urged her to reconsider her statement while Burroughs was still alive.

As Hale understood it, the Salem trials constituted a faithful and valid application of approved theological recommendations for the discovery of witchcraft. The failure of the Salem trials was the failure of a paradigm:

If there were an Error in the proceedings in other places and in New England, it must be in the principles proceeded upon in the prosecuting of the suspected, or in the misapplication of the principles made use of. Now, as to the case at Salem, I conceive it proceeded from some mistaken principles made use of.⁶⁹

That the identifications of the Court of Oyer and Terminer were in error was an anomaly that could not be explained in terms of the principles embodied either in earlier legal actions against witchcraft or in authoritative texts for the discovery of witchcraft. If the Salem trials had been based upon mistaken assumptions, then so also were the prosecutions against Margaret Jones, Ann Hibbins, and others in which an earlier generation of New England divines had participated.⁷⁰ If the policies followed by the Court of Oyer and Terminer were ill advised, then so also were the recommendations of Richard Bernard, Richard Baxter, and other revered authorities on witchcraft. For Hale, an acknowledgment of the errors of the Salem trials entailed an acknowledgment of past errors as well: "May we not say in this matter … We have sinned with our fathers?"⁷¹

While Hale's specific criticisms of the rules of evidence were primarily extensions of Mather's thesis on the inadequacy of human perceptions in understanding Satan's intentions, he went beyond Cases in another respect. Where Mather's work had expressed doubt about the possibility of objective criteria for the determination of witchcraft, there are intimations that Hale had arrived at an even more extreme attitude of disbelief. Various passages in Modest Inquiry verge on the radical interpretation that witchcraft might have no ontological basis apart from the perceptions of believers.

Thus Hale explained that the efficacy of charms, curses, and puppets as instruments of malefic harm had little to do with invisible forces:

And probably the cause may be, that Satan, the Lord permitting him, may inflict his mischief on the person, the Spectators or Actors herein supposed to be concerned, suiting thereby his design to man's faith about it. And, if so, the reason why any suspected person is hereby concerned is not because they are guilty but because they are suspected.⁷²

Elsewhere he offered a similar interpretation of imputations of malefic witchcraft:

Some persons will put an evil construction upon an innocent person, and so raise an evil fame against a person; and then others believing it, are apt to look upon other actions with a squint eye, and through the multiplying glass of their own jealosies, make a Molehill seem a mountain.⁷³

Modest Inquiry was very nearly the work that Sewall had feared it might be. It was an indication that at least one member of the New England clergy had begun to reflect on witchcraft from outside the context of belief.

Cases of Conscience and Modest Inquiry articulate the theological closure around the problem of witchcraft. Together, the two works constitute an admission that witchcraft was no longer available as a demonstration of the existence of the invisible world. To be sure, Mather and Hale had met doctrinal requirements by acknowledging the reality of witchcraft, but such acknowledgments notwithstanding, both authors had served notice that, in the future, theological confirmation of imputations of witchcraft as real empirical events was most unlikely. The new rules of evidence virtually precluded the further production of witches. On the one hand, no provision had been made for the translation of popular suspicions into theologically acceptable proofs. On the other hand, the allowance for free and voluntary confessions as a sufficient proof was a condition that had been fulfilled by only a single earlier prosecution in Connecticut. In the aftermath of the Salem trials, witchcraft had become an impossible crime.

Symptomatic of the altered epistemological status of witchcraft was a proposal for the recording of special providences, signed by eight leading ministers on March 5, 1695.⁷⁴ The reasons for this undertaking were the same as those given in Increase Mather's earlier proposal of 1681: to help demonstrate the existence and agency of the invisible world. This time, however, there was a significant deletion in the list of unusual occurrences about which information was to be gathered. No mention was made of witchcraft. In this act of omission, the clergy tacitly acknowledged that witchcraft no longer had a place within orthodox doctrine. In the future, New England divines would look to other signs for the presence of God.

THE DECLINE OF WITCHCRAFT PROSECUTIONS

For all its apparent suddenness, the decline of witchcraft prosecutions after the Salem trials involved no marked discontinuities with the past. Apart from the events of 1692, popular allegations had formed the primary basis for legal action against witchcraft, and these testimonies had ceased to carry convictive weight after the last execution in 1656. Furthermore, though allegations of maleficia were included among the evidences gathered by the Court of Oyer and Terminer, there is no reason to believe that they were regarded with greater seriousness in 1692 than in the earlier prosecutions. Now, in the light of a further separation between popular and theological understandings of the problem of witchcraft and in view of the new rules of evidence, witchcraft soon disappeared entirely from the legal records of Massachusetts Bay.⁷⁵ Within two years after the Salem trials, witchcraft was no longer an actionable legal offense.

The final case to be recorded, that of Mary Watkins, well demonstrates that, by 1693, the legal possibilities for witchcraft as a valid form of deviant imputation had been exhausted. Mary, a Boston servant, confessed to witchcraft shortly after failing in an attempt at suicide.⁷⁶ She was brought before a magistrate, and a bill of indictment was presented to the grand jury of the Superior Court of Judicature on April 25, 1693. The jury refused the bill in spite of the confession. This time, when asked by the magistrates to reconsider their verdict, the jurors maintained their original finding of ignoramus. Although the grounds for the decision are not mentioned, it is likely that the defendant was judged to be distracted and therefore not competent to offer credible testimony. This possibility notwithstanding, the case of Mary Watkins was an indication that, after the Salem trials, not even voluntary confessions would be accepted as proof of witchcraft.

The decline of witchcraft prosecutions by no means coincided with the decline of witchcraft accusations. Despite official and ecclesiastical discouragement, villagers continued to believe in the efficacy of malefic witchcraft until well into the eighteenth century.⁷⁷ Indeed, fragmentary records indicate that, even without the benefit of legal authority, witchcraft allegations remained an effective weapon against suspected adversaries. An account in 1728 by a Massachusetts minister, the Reverend Mr. Turell of Medford, furnishes a detailed description of the local response in a small community in 1720 to three children believed to be afflicted. The report makes clear that the children and their supporters were able to mobilize local opinion against the woman whom they identified as responsible for their condition.⁷⁸ And, in the town of Colchester in neighboring Connecticut in 1724, one woman regarded the suspicions against her with sufficient seriousness to sue her accusers for 500 pounds in damages.⁷⁹

At the same time, the outcome of these events suggests that if, in certain communities, popular beliefs in mystical harm had not changed greatly in the thirty years since the Salem trials, the response of civil and ecclesiastical authorities had nevertheless changed appreciably. In his narrative, Turell questioned not only the validity of the accusations but also the motives of the accusers. According to Turell, the afflictions were not only not genuine, they were deceitful strategies employed by the children to attract attention to themselves.⁸⁰ Similarly, the decision rendered in the Connecticut case also reveals an inclination to view as problematic not the behavior of the accused but the perceptions of the accuser. On appeal, the complainant was awarded a nominal compensation of one shilling in damages. More significantly, the same ruling included a judgment that the defendants were not insane. In the decades following the Salem trials, public attention began to focus less on the guilt of the accused and more on the credibility of the accuser.

Ultimately, the withdrawal of legal recognition from imputations of witchcraft entailed far more than the decriminalization of a category of deviant behavior. The loss of witchcraft as an actionable offense divested contemporary theories of supernatural causation of their last remaining claim to legal authority. With the decline of witchcraft prosecutions, questions about the availability of the invisible world ceased to be a matter of practical concern for the state.

Notes

…⁶² Included with Cotton Mather's Wonders, pp. 255 ff.

⁶³ Ibid., p. 258 ff.

⁶⁴ Ibid., p. 283.

⁶⁵ Ibid., p. 285.

⁶⁶ Burr, Narratives, p. 389.

⁶⁷ Hale, Modest Inquiry, pp. 162-65.

⁶⁸ Burr, Narratives, p. 421.

⁶⁹ Hale, Modest Inquiry, p. 40.

⁷⁰ Hale may have overestimated the involvement of the clergy in these cases; see chapter 7, above.

⁷¹ Burr, Narratives, p. 426.

⁷² Hale, Modest Inquiry, p. 81.

⁷³ Ibid., p. 69.

⁷⁴ Calef, More Wonders, pp. 92-94.

⁷⁵ Interestingly, the only case in which the new rules of evidence were applied occurred in Connecticut in 1693. A young woman who had been convicted on the basis of spectral evidence, the water test, popular allegations, and positive findings in the search for witch's marks, was reprieved by the governor of the colony. Increase Mather's Cases of Conscience was cited as part of the justification for reversing the decision of the court. See Taylor, Witchcraft Delusion, pp. 62-78.

⁷⁶ Burr, Narratives, pp. 383-84; also W. Watkins, "Mary Watkins," pp. 168-70.

⁷⁷ For further discussion of survivals of folk belief in witchcraft after the decline of prosecutions, see Herbert Leventhal, In the Shadow of the Enlightenment, pp. 66-125.

⁷⁸ Turell, "Detection of Witchcraft," pp. 10-11.

⁷⁹ Taylor, Witchcraft Delusion, p. 155.

⁸⁰ With regard to one of the afflicted, Turell wrote: "This little girl had observed what sort of treatment her sisters had met with during their disorders, viz., they seemed to be more the object of their parents' care and love, as well as pity, than ever … and, accordingly, she feigned herself afflicted, said and acted as they did, to the very last, without being found out"; Turell, "Detection of Witchcraft," p. 15.