When Henri Boguet, a judge in Franche-Comté, wanted to know whether witches could have any children by having sex with the devil he simply asked two of them. One replied that she was too old, the other that God would not permit it.Footnote ¹ In this short yet powerful and also extraordinarily well-written book Virginia Krause points to the sheer oddness of such situations: an elite male judge seeking knowledge from unlettered village women. While historians of witch trials have spent considerable time parsing the confessions of witches, the role that these played in the demonological literature has not yet been systematically studied. Krause is certainly right to insist on the importance of confession to early modern demonology. Beyond the borders of France (and hence not mentioned by Krause), Petrus Binsfeld devoted an entire treatise to the confessions of male and female witches (Tractatus de confessionibus maleficorum et sagarum, 1591). Perhaps not surprisingly given her field (French literature) and the subject (confession) Krause draws considerably on Foucault's insight that knowledge is produced, and that its creation is tied up with the methods of its production: ‘without confession, there would have been no demonology’ (p. 20). Krause's sustained study of the main French demonologists (as well as the sceptical Michel de Montaigne) yields a number of important insights. She ties demonology to a declining epistemology which preferred hearing over vision (confession, of course, is auricular), paying particular attention to Jean Bodin who himself possessed a spirit that communicated through sound. With considerable insight she identifies the ways in which demonologists edited these confessions and drew on humanist techniques and arguments to heighten their plausibility. She also stresses the considerable unease with which these men, ‘by necessity, Satan's editor and exegete’ (p. 65), at times fulfilled their duty. In her final chapter, Krause engages with but also critiques Foucault's work on the advent of subjectivity (that is, our emergence as ‘self-narrated beings’) and assigns demonology in this larger narrative. A case study of a 1582–3 trial in Marlou, in central France, shows how suspects were called on and forced to play the role of witch, just as in Foucault's work the mandate to annual confession by the Fourth Lateran Council, ‘compell[ed] us to speak our inner truth’ (p. 103). Historians may feel that this last part of Krause's argument is too bold, whereas the earlier part is perhaps unduly narrow. Witchcraft belief, as Stuart Clark has taught us, operated within wider systems of belief. Demonologists believed in witchcraft for other reasons (for instance, because it offered proof of the spirit world) and possessed other sources which confessions were meant to confirm. In his chapter on demonic offspring Boguet also discussed classical mythology and the alleged coupling of Luther's mother with the devil. Demonology in its sources and motivations could be (to paraphrase one of Krause's demonologists, Pierre de Lancre) as inconstant as the devils themselves. There is no denying, however, that Krause has thrown considerable light in some very dark corners.