The Swimming Test
Lacking direct physical evidence to prove witchcraft, contemporaries had to rely on other methods for the discovery of witches The three main remaining evidential possibilities were supernatural signs, circumstantial evidence and testimonies.
When Judge Hale presented the dilemma before the jury at Bury St Edmunds Assizes, he ‘desired them, strictly to observe their Evidence; and desired the great God of Heaven to direct their Hearts in this weighty thing they had in hand’.1 His statement, of course, should not be interpreted literally as a call for the use of supernatural means, but as an acknowledgment of human inferiority in discovering the truth. In addition to strict observance of the evidence, this experienced judge called on the help of God. Before the ordeals were abolished in 1215, guilt or innocence was proved, so the contemporaries believed, by the hand of God. Although determination of criminal proceedings was transferred to human hands, in cases of witchcraft, unlike other crimes, supernatural signs of guilt were still widely sought, even centuries after the abandonment of the ordeals.
The most famous method was the swimming test of witches, but other supernatural methods or signs were the inability of the witch to shed tears or the causing of the witch to appear by burning an object. The early modern cultural toolbox included magical means and supernatural signs that both cured the bewitched and discovered witchcraft. Scratching of the suspect by the victim, for example, was a technique aimed at seeking relief of the symptoms but also at ascertaining the identity of a witch. Similarly, healing of the victims subsequent to the suspect’s arrest or conviction was a reassuring sign. The burning of an allegedly bewitched cow annulled the bewitchment. Since the witch was allegedly compelled to appear if the disease had been caused by bewitchment, it also assisted in finding out whether the cattle were bewitched or naturally sick.
Supernatural signs enjoyed popularity, but they were not part of the official procedure. In some instances, those ways of proof were specifically denounced as illegal, and there was much controversy around their use. The chief supernatural method for the discovery of witches was the swimming test. The ordeal by swimming, judicium aquae frigidae, was mentioned in English laws and decrees as early as the tenth century.2 According to these texts, the swimming test was used in cases of theft, adultery, homicide and witchcraft. The suspects were cast into the water. If they sank, they were innocent, and if they floated, they were guilty. There were formal rules regarding the size of the pond and the manner of tying the suspects before throwing them into the water.3 In the thirteenth century the ordeals became illegal in England and were abandoned.4 Despite this fact, the seventeenth century witnessed a resurgence of the swimming test only in witchcraft cases. The revival of the swimming test cannot be accounted for by the enactment of anti-witchcraft legislation, as the re-emergence was subsequent to the statutes of 1542, 1563 and 1604. Despite its illegality and a lack of ecclesiastical backing, the practice gained popularity. It took place at the fringes of the official proceedings at the pre-trial stage and was usually conducted by fellow villagers trying to bolster the case against the suspect.
The test was at the centre of a debate about its validity and its moral and religious adequateness. The approach toward the test was both class- and profession-related. Members of the elite were generally opposed to it, though it was popular with the lower classes, while the middling sort was in a somewhat ambivalent position. The middling sort was active in the reconstruction of the swimming test as an experiment, a strategy that made it easier to cope with the dissonance between the two extreme positions. Professional affiliation shaped the types of arguments the elite used against the test. Physicians emphasized natural explanations, clergymen focused on its religious inadequacy and judges deplored its illegality.
Three early and significant treatises that discussed the test and the theory behind it were Reginald Scot’s Discouerie of Witchcraft, James I’s Daemonologie and William Perkins’ A Discourse of the Damned Art of Witchcraft.5 The first accounts of actual cases of swimming of witches in early modern England were two pamphlets from 1612 and 1613.6 It seems that the test was most widely used at the time of the Hopkins witch hunt of the 1640s.7 The practice persisted, although some of its performers were eventually convicted for assault and even murder (in case the ordeal had ended with the drowning of the suspect). There is evidence that swimming continued even after the repeal of the anti-witchcraft act in 1736 until almost the end of the nineteenth century.8
The resurrection of the test in witchcraft cases in seventeenth-century England, about four centuries after the abolition of the ordeal system, has puzzled historians, but no satisfactory explanation has yet been offered. Holmes pointed to the functional aspect of this culturally embedded practice as a tension-relieving device,9 but it is not clear why tensions were relieved in this particular manner. Continental influence was not ruled out.10 Rosen and Gaskill suggested that in the latter part of the seventeenth century, as the judicial system became reluctant to prosecute witchcraft cases, private persons found informal violence or the swimming test to be an alternative solution.11 That may be true, but it does not explain the flourishing of the test in the earlier part of the seventeenth century. Pihlajamäki dedicated a whole paper to the mystery of the ordeal’s revival and concluded that it was mostly used to cross the evidential threshold needed for the application of torture. He assumed that swimming was seldom used in England because its legal system avoided torture and let jurors evaluate the evidence.12 However, although there might have been more swimming tests in Germany, the practice was definitely not rare in England.13 Clark linked the revival to the nature of witchcraft as an opaque crime and to its theological context. Evidence ‘of a supernatural or quasi-miraculous character’ matched the nature of the crime.14 James I’s approval of the swimming test was also mentioned as a factor that encouraged actual implementation.15 In his Daemonologie, James valued the test as ‘good helpe’:16
so it appeares that God hath appoynted (for a super-naturall signe of the monstruous impietie of the Witches) that the water shal refuse to receiue them in her bosom, that haue shaken off them the sacred Water of Baptisme, and wilfullie refused the benefite thereof.17
James wrote Daemonologie as James VI of Scotland, before ascending to the English throne in 1603. The book expressed his theological views and clearly did not reflect English law. As the king of England, James I took no steps to encourage the implementation of the test. On the contrary, he was not the witch hunter one might have expected as the author of Daemonologie, but a cautious fact-finder and the exposer of a few fabricated bewitchments.
Still, James’ words echoed in subsequent pamphlets. The pamphlet of 1612 that gave the first account of swimming described how Arthur Bill, a man with a bad reputation as a witch and who was suspected of killing, was thrown into the pond with his parents:
The Iustices and other officers (thereby purposing to trie the said Arthur by an experiment that (many thinke) neuer failes) caused them all to bee bound, and their Thumbes and great Toes to bée tied acrosse, and so threw the father, mother and sonne, and none of them sunke, but all floated vpon the water.18
The author then repeated the arguments of James (without due credit) and declared the swimming test, together with the devil’s mark, ‘more certaine’ signs of witchcraft, although his statements were softened by expressions such as ‘mee thinke’. Witches were rejected by the water that signified the holiness of baptism and therefore were discovered with the help of God.
Yet, the logic of the author seemed to be flawed, considering the following events he described:
These thrée, the Father, Mother and Sonne, beeing thus séene floating vpon the water, the suspition that was before not well grounded, was now confirmed: Whereupon the said Arthur Bill beeing the principall or (I thinke) the onely Actor in this Tragedy, was apprehended and sent to Northampton gaole.19
The questions remain that, if all three members of the family floated, why the case was pursued against Arthur Bill alone, and if the parents floated as well, why the author saw Arthur Bill as the principal, and even only, perpetrator. James VI & I’s rationale was presented by the author as a hypothesis that could be empirically tested, as ‘an experiment that (many thinke) neuer failes’. A sense of dissonance is created because the identical result of the experiment for all three subjects (all three floated) did not lead the author to an identical conclusion (that all three were guilty).
Despite the inconsistency, this case is notable because the supernatural ordeal-type rationale (reliance on divine inscrutable determination) was moulded into a template of an empiricist experiment. It is, in my view, this reframing of the rationale behind the swimming test that contributed to its revival in the seventeenth century. The idea that supernatural presuppositions could be tested empirically was a bridge between old and new concepts of knowledge about the world and justice.
It should be noted that in Bill Arthur’s case, it was the justices ‘and other officers’ who ordered the swimming test. The typical scenario, however, was for neighbours to perform the test before taking the suspect to the JP. Such was the case recorded in Witches Apprehended. The full title included the promise ‘With a Strange and Most True Triall How to Know Whether a Woman Be a Witch or Not’, and an illustration of the swimming test appeared on the title page.20 The detailed description of the test made the pamphlet a practical manual for swimming witches.21 The pamphlet tells the story of Mr Enger, a local gentleman who suspected Mary Sutton and her mother of killing his cattle, bewitching his servant and killing his seven-year-old son. His friend advised him to swim the suspect, saying ‘I have seene it often tried in the North countrey’.22 The friend suggested a three-step procedure. First, swim the suspects, stripped to their undergarments with their arms tied and a rope under their waist so they could be pulled out of the water in case of drowning. Second, if the suspect floats, have her searched by women for the devil’s mark. Third, if such marks are found, swim her again, this time with her hands and feet tied diagonally.23 The friend’s advice was followed, and next morning Sutton was forcibly taken by Mr Enger and his men to be swum. She was cast into the water, but ‘shee sunke some two foote into the water with a fall, but rose againe, and floated upon the water like a planke’. Afterward she was searched by women that Mr Enger had ready for the task. The women found suspicious signs on her body, and Mary Sutton was thrown to the water again, this time with her thumbs tied to her toes. Unfortunately, she ‘sunke not at all, but sitting upon the water, turned round about like a wheele, or as that which commonly we call a whrilpoole’. The men, who were holding the rope on both sides, persisted in their efforts, ‘tossing her up and downe to make her sinke, but could not’.24 The swimming test was followed by questioning. Only on the suspect’s insistence on her innocence was she carried to a justice.25 It is notable that the swimming test in this leading case was not a spontaneous eruption of mob justice, but seems to be a carefully planned and prearranged experiment composed of orderly and predetermined stages and orchestrated by gentlemen.
The swimming test persisted despite fierce attacks on it by the judiciary and clergy. William Perkins listed the test with other insufficient proofs of witchcraft. He was very clear in his denial of James VI & I’s rationale.
And yet to iustifie the casting of a Witch into the water, it is alledged, that hauing made a couenant with the deuill, shee hath renounced her Baptisme, and hereupon there growes an Antipathie between her, and water.
Ans. This allegation serues to no purpose: for all water is not the water of Baptisme, but that onely which is vsed in the very act of Baptisme, and not before nor after. The element out of the vse of the Scrament, is no Sacrament, but returnes again to his common vse.26
The treatise was published posthumously, but it definitely took courage on the part of the publisher to publish it while James still sat on the English throne. The fact that the publication was not censored can perhaps demonstrate that, as a king, James did not try to enforce the opinions he himself had expressed years before in his Daemonologie.
Filmer, writing a treatise aimed to refute Perkins and demonstrate the futility of his signs for the discovery of witches, regarded Perkins’ discussion of the swimming test as a direct refutation of James. Filmer stressed that Perkins ‘condemnes point blanke King James’s judgment as favouring of Witch-craft in allowing the triall of a Witch by swimming as a principall proofe’.27 As Filmer’s treatise was an attack against Perkins, it was rhetorically convenient for him to portray Perkins as critical of the king, although he himself also opposed the swimming test. Filmer quoted Perkins, who considered the swimming test to be among the ‘lesse sufficient’ proofs: ‘so far from being sufficient, that some of them, if not all are after a sort practices of witch-craft, having no power by Gods Ordinance.’28 This argument, Filmer contended, should have led Perkins to reject the devil’s mark, rather than consider it a certain proof, as it was not ordained by God as well.
Cotta, like Perkins, rejected James VI & I’s rationale for the test, but for other reasons. In his view, the practice was vulgar: