Public infidelity and private belief?
The discourse of spirits in Enlightenment Bristol
Recent work on the history of witchcraft and magic has identified three themes or approaches as of particular importance in our understanding of a subject which, although it has been centre stage since the publication of Religion and the Decline of Magic in 1971, has continued to trouble historians. The first problem, acknowledged as 'the most baffling aspect of this difficult subject' by Thomas himself, is that of 'decline': by rendering early modern witchcraft beliefs intelligible, historians have highlighted the issue of why and how far they ceased to have meaning (or function), and many have regarded this as the least satisfactory feature of Thomas's account.1 The great interest now being displayed in the culture of the long eighteenth century, including its occult aspects, has rendered this theme of pressing concern.2
The second issue concerns the need for case studies. It is broadly agreed that witchcraft must be studied as a conjunctural phenomenon, operating at a whole series of levels and affected by the interplay of a variety of institutions, interests and languages. While these can all be analysed separately, it is equally crucial to study their interrelationship and this, for the moment at least, is best done in specific settings where the evidence survives to allow a full reconstruction of the development and resolution of a witchcraft episode.3 Third, historians have become particularly interested in witchcraft as a linguistic phenomenon and one imbedded in narrative. This involves a concern to reconstruct the role played by conflicts over the use of contested terms (such as 'witch'), and the provision of alternative stories of what was happening in a specific case. But it also involves an interest in the inter textuality of witchcraft cases, that is their shaping by reference to previous stories, authorities or cultural models of how, for example, a possession might develop and be resolved.4
The case to be discussed here fulfils all three of these criteria. It dates from 1761–62, some twenty-five years after the Witchcraft Act of 1736, and some ten years after the last English witchcraft episode that has been studied in any scholarly depth.5 Unlike that case, when a man was executed for