There seems to be an important historical connexion between changes in the concept of evidence and that of the person capable of giving evidence. Michel Foucault urged that during the classical age the relationship between evidence and the person was reversed: scholasticism derived statements’ authority from that of their authors, while scientists now hold that matters of fact are the most impersonal of statements.1 In a similar vein, Ian Hacking defines a kind of evidence which ‘consists in one thing pointing (...) beyond itself’, and claims that until the early modern period ‘testimony and authority were primary, and things could count as evidence only insofar as they resembled the witness of observers and the authority of books’.2 This captures a rather familiar theme of the ideology of early modern natural philosophy. Nullius in verba was the Royal Society of London’s motto. Robert Boyle, doyen of the Society’s experimental philosophers, tried to build up the credit of laboratory objects at the expense of untrustworthy humans. He reckoned that ‘inanimate bodies … are not capable of prepossessions, or giving us partial informations’, while vulgar men may be influenced by predispositions, and so many other circumstances, that they may easily give occasion to mistakes’. So an inanimate body’s deeds could function as signs of some other state of affairs in a way that the stories of vulgar humans could not.3 1. See Michel Foucault, L’Ordre du discours: Leçon inaugurale au Collêge de France prononcée le 2 décembre 1970 .2. Ian Hacking, The Emergence of Probability: A Philosophical Study of Early Ideas about Probability, Induction and Statistical Inference , pp. 34, 33.3. Quoted in Steven Shapin and Simon Schaffer, Leviathan and the Air Pump , p. 218. See also Peter Dear, ‘Totius in verba:’ Rhetoric and Authority in the Early Royal Society’, Isis 76 : 145-61. Simon Schaffer lectures in history and philosophy at the University of Cambridge. He is the coauthor of Leviathan and the Air Pump: Hobbes, Boyle, and the Experimental Life and coauthors of The Uses of Experiment: Studies in the Natural Sciences. (shrink)