This book is directed against three types of relativism. First, there is the claim that history must be written without benefit of hindsight. This claim, which dates back to Butterfield’s book The Whig Interpretation of History (1931), had no discernible influence in the history of science until the 1960s. It can’t be right: it is only hindsight, for example, which enables one to identify Columbus’s discovery of America as a key moment in the development of modern science (see, in general, MacIntyre, ‘Epistemological Crises’ (1977)). Second, there is the claim that the concept of rationality is always culturally relative. This claim derives from Wittgenstein but began to have a major impact on history and philosophy of science after the publication of Peter Winch’s The Idea of a Social Science (1958). It is, I maintain, incompatible with any grasp of the achievements of modern science. And, third, there is the claim that, in science, successful claims and failed claims should be understood and explained in exactly the same way, an argument which originates with David Bloor’s Knowledge and Social Imagery (1976) and which Bloor named ‘the strong programme’. This argument involves denying that scientific claims are ever adopted because they fit the evidence better than the alternatives. Its consequences for the history of science have been, it seems to me, pernicious. Each of these arguments, of course, has become part of a larger intellectual movement which may loosely be labelled ‘postmodernism’. Postmodernism has, I believe, a great deal to teach naïve realists, but as naïve realism hardly gets a look in among historians of science these days I have concentrated here on its defects, not its merits.
1. See Shapin and Schaffer on truth as an actor’s judgement (i.e. truth is what you think it is): Shapin & Schaffer, Leviathan and the Air-pump (1985), 14 (compare Bloor, Knowledge and Social Imagery (1991), 37–45, and Shapin, A Social History of Truth (1994), 4: ‘For historians, cultural anthropologists, and sociologists of knowledge, the treatment of truth as accepted belief counts as a maxim of method, and rightly so’). Truth is only an actor’s judgement for statements which are necessarily subjective: e.g.‘That is the funniest joke I have ever heard’ is true if and only if I think it so. It hardly helps to make rationality an actor’s judgement either (Garber, ‘On the Frontlines of the Scientific Revolution’ (2004), 158), since the whole point of the concept is that it can be (and was) used to show that actors can be and often are mistaken. There is a difference between checkmate and death: changing the rules of chess may alter who wins and who loses, but we cannot bring ourselves back to life by changing our concepts (and to believe we can is a form of madness). If anything and everything is to be treated as an actor’s judgement then the ideas of truth, rationality and objective reality become meaningless, and we can all be immortal if we choose. But at least those who make this move avoid truly puzzling formulations, such as Newman’s and Principe’s claim that Starkey’s belief in the philosopher’s stone was ‘not unwarrantable’ (Newman & Principe, Alchemy Tried in the Fire (2005), 176) – thus they avoid claiming it was sensible, and also avoid acknowledging it was foolish.
2. Barnes & Bloor, ‘Relativism, Rationalism’ (1982), 23, formulate the core doctrine of the strong programme as the ‘equivalence postulate’: ‘Our equivalence postulate is that all beliefs are on a par with one another with respect to the causes of their credibility. It is not that all beliefs are equally true or equally false, but that regardless of truth and falsity the fact of their credibility is to be seen as equally problematic.’ And so Simon Schaffer insists that it would be mistaken to ‘account for the establishment of one version of natural philosophy [rather than an opposing one] through the superiority of its grasp over nature’ (Schaffer, ‘Godly Men and Mechanical Philosophers’ (1987), 57). But it should be obvious that not all beliefs are on a par with one another, and that the causes of their credibility vary greatly. The Galilean belief that ice is lighter than water is not on a par with the Aristotelian belief that ice is heavier than water; the modern belief that magnets are indifferent to garlic is not on a par with the classical belief that garlic disempowers magnets. In these cases the first belief has the facts on its side and the second doesn’t; one version of natural philosophy established itself over its opponent through the superiority of its grasp over nature. To insist that the issue of validity must be separated from the issue of credibility is to insist that well-founded beliefs be treated as if they are unfounded beliefs. Enquiries based on this premise are bound to conclude that the claims made on behalf of well-founded beliefs are excessive because that conclusion is built into the methodology.
Of course, the question of how best to interpret the strong-programme approach is much disputed: see the marvellous exchange of fire between Bloor, ‘Anti-Latour’ (1999) and Latour, ‘For David Bloor’ (1999): I find Latour’s reading of Bloor entirely convincing. For an effective critique, see Laudan, ‘The Pseudo-science of Science?’ (1981).
3. Secord, ‘Knowledge in Transit’ (2004), 657. The intellectual context within which Leviathan and the Air-pump was written is conveniently established by Shapin, ‘History of Science and Its Sociological Reconstructions’ (1982). For the strong programme, see Bloor, Knowledge and Social Imagery (1991); for other works by Barnes and Bloor: Bloor, Wittgenstein (1983); Barnes, T. S. Kuhn and Social Science (1982). The strong programme explicitly advocates ‘methodological relativism’, a term of art which means ‘All beliefs are to be explained in the same general way regardless of how they are evaluated’ (Bloor, Knowledge and Social Imagery (1991), 158: i.e. it is identical with the symmetry principle, on which see above, pp. 43–4 and below, No. 7, and a restatement of the equivalence postulate, on which see above, No. 2).
Harry Collins, founder of the Bath School, whose work is closely related to that of the Edinburgh School, is happy, at least on occasion, to employ the word ‘relativism’ unequivocally and identify those he regards as fellow relativists and aids to relativism: Collins, ‘Introduction’ (1981). But ‘relativism’ is rather like the word ‘atheism’ in the seventeenth century: lots of people attack it but few own up to it; and when they do they insist in defining the word in their own peculiar way (Bloor, ‘Anti-Latour’ (1999), 101–3). The result is a certain amount of confusion over who can fairly be called a relativist and who cannot. I have been repeatedly told, for instance, by people who should know better that Shapin is not a relativist, and certainly he rarely uses the word: nevertheless, he has recently explicitly identified himself as a ‘methodological relativist’, in other words as a supporter of the strong programme (which is not, from a sociological point of view, surprising, as he was a member of the Edinburgh Science Studies Unit from 1973 to 1989). Shapin practises what he preaches by giving an explanation for his own belief in the reliability of scientific knowledge which could equally be applied (in a different culture) to a belief in witchcraft: ‘My confidence in science is very great: that is just to say that I am a typical member of the overall overeducated culture, a culture in which confidence in science is a mark of normalcy and which produces that confidence as we become and continue to be members of it.’ (Shapin, ‘How to be Antiscientific’ (2010), 42 = Labinger & Collins (eds.), The One Culture? (2001), 111; compare Collins’s claim that those who believe in astrology are making a social mistake (in Labinger & Collins (eds.), The One Culture? (2001), 258–9); see also Shapin’s exposition of the ‘equivalence postulate’: Shapin, ‘Cordelia’s Love’ (1995), and the description of the ‘relativist genre’ in Ophir & Shapin, ‘The Place of Knowledge’ (1991), 5, which is evidently, on Shapin’s part, a self-description). I address Shapinesque relativism in Chapter 15.
I agree with Bricmont and Sokal in their contributions to Labinger & Collins (eds.), The One Culture? (2001), that ‘methodological relativism cannot be justified unless one also adopts philosophical relativism or radical skepticism’ (244). It is important to distinguish between methodological relativism (which is the adoption of relativism as a method) and a very different position with which it can easily be confused, methodological agnosticism, the claim that one cannot know a priori which method will work and which will not – a position I would defend – which is perfectly compatible with the claim that ex post facto one can see that one method is more successful than another (a claim that methodological relativists are committed to denying): see Kuhn, The Structure of Scientific Revolutions (1996), 173.
4. Shapin, A Social History of Truth (1994). Shapin advocates a ‘liberal’ rather than a ‘restrictive’ view of truth (4). Such an approach involves holding that garlic does disempower magnets, or did (for Pliny, Abertus Magnus, van Helmont, etc.). The claim that garlic does not disempower magnets becomes simply an alternative truth, not a discovery; the experimental method becomes one way of making truths, not a reliable way; and Boyle’s policy of disciplined suspicion becomes a new way of trusting others.
Shapin also advocates ‘a methodological disposition towards charity’ (4). In Leviathan and the Air-pump he and Schaffer wrote, ‘Following Gellner, we shall be offering a “charitable interpretation”’ of Hobbes, and cited an article by Gellner which had first appeared in 1962, and the use made of it by Harry Collins (on whom, see below, No. 9). In fact, Collins was quite clear that he was not following, but rather going against, Gellner (Collins, ‘Son of Seven Sexes’ (1981), n. 15), for Gellner’s article was, in his own words, a ‘plea against charity’ (Gellner, ‘Concepts and Society’ (1970), 48); he held that ‘Excessive indulgence in contextual charity blinds us to what is best and what is worst in the life of societies. It blinds us to the possibility that social change may occur through the replacement of an inconsistent doctrine or ethic by a better one . . . It equally blinds us to . . . the employment of absurd, ambiguous, inconsistent or unintelligible doctrines’ (42–3). It is my book, not Leviathan and the Air-pump, which advocates following Gellner. Indeed, Gellner stated my overall argument precisely: ‘In recent centuries, there has been an important shift from the use of merely social to genuinely cognitive concepts: this is normally known as the Scientific Revolution. Wittgensteinianism makes it impossible to ask any questions about this event, for on its terms nothing of the kind could ever occur, could make any sense’ (Gellner, Relativism and the Social Sciences (1985), 185). No wonder Shapin insists that ‘[t]here was no such thing as the Scientific Revolution’! (Shapin, The Scientific Revolution (1996), 1).
It should be stressed that those who were trying to build the new science were acutely aware of how one might claim to provide a sociology of knowledge but wanted to escape from a world in which knowledge was entirely socially determined: cf. Bacon on the Idols (Bacon, Instauratio magna(1620), 53–80 (Book 1, §§23–68) = Bacon, Works (1857), Vol. 4, 51–69), and Glanvill, The Vanity of Dogmatizing (1661), esp. 125–35, 194–5. For reviews of Shapin’s A Social History of Truth, see Feingold, ‘When Facts Matter’ (1996) and Schuster & Taylor, ‘Blind Trust’ (1997).
5. Thomas Kuhn is often credited with introducing the word ‘paradigm’ into English-language philosophy of science in Kuhn, Structure (1962) (e.g. Lehoux, What Did the Romans Know? (2012), 227, and Hacking, ‘Introductory Essay’ (2012), xvii–xxi), but in fact the word is used repeatedly in Hanson, Patterns of Discovery (1958): 16, 30, 91, 150, 161; some, though not all, of these seem distinctly proto-Kuhnian. Kuhn’s first use of the term ‘paradigm’ was in a conference paper delivered in 1959, after the appearance of Hanson’s book (‘The Essential Tension’, reprinted in Kuhn, The Essential Tension (1977), 225–39). Hanson also preceded Kuhn in stressing the importance of gestalt psychology and in laying emphasis on the philosophy of Wittgenstein. He is cited four times in Structure, and Kuhn later stressed the extent to which he was influenced by him (Kuhn, The Road since Structure (2000), 311; Nye, Michael Polanyi and His Generation (2011), 242).
This raises a larger issue in the interpretation of Kuhn. Joel Isaac has claimed that the apparent similarities between Kuhn’s work and a number of nearly contemporary works are a retrospective construction (Isaac, Working Knowledge (2012), 232), but he does not consider the influence some of these works had on Kuhn. Thus he says that Kuhn ‘hit upon’ the concept of a paradigm in 1958–9 (234), ignoring the possibility that Hanson had an influence on Kuhn. (Feyerabend, when he read Structure in draft, found it altogether too reminiscent of Hanson: Hoyningen-Huene, ‘Two Letters’ (1995).) Isaac also thinks that the apparent similarities between Kuhn’s Structure and Polanyi’s Personal Knowledge (1958) are misleading, despite the fact that in Structure Kuhn refers to Polanyi’s book as ‘brilliant’ (44; it is sometimes said that Kuhn plagiarized many of his ideas from Polanyi. Thus MacIntyre wrote that Kuhn’s view of natural science ‘seems largely indebted to the writings of Michael Polanyi (Kuhn nowhere acknowledged any such debt)’ (MacIntyre, ‘Epistemological Crises’ (1977), 465). The bracketed statement is simply false: the acknowledgement is there from the first edition, though it is overlooked in the index to the third and later editions.) So, too, with the similarities between Kuhn and Feyerabend, despite the fact that they were in close communication during 1960 and 1961 (Hoyningen-Huene, ‘Three Biographies’ (2005)). Isaac claims that reading Kuhn, alongside these other authors, as an opponent of positivism ‘conflates the reception of Kuhn’s book with the historical context of its composition’ (4; the classic reception text is Shapere, ‘The Structure of Scientific Revolutions’ (1964)). But Kuhn himself approved the interpretation Isaac seeks to overturn (Kuhn, The Road since Structure (2000), 90–1).
Thus Isaac understates the significance of Structure’s direct attack on positivism, which, according to Kuhn, underpinned ‘the most prevalent contemporary interpretation of the nature and function of scientific theory’ (Kuhn, Structure (1996), 98–103; Isaac, Working Knowledge(2012), 231–2; for a summary account of this contemporary interpretation, see Hesse, ‘Comment’ (1982), 704), and misrepresents the context of Structure’s composition. Isaac’s ‘local’ reading of Kuhn in a Harvard context is valuable, but Kuhn left Harvard in 1956; the key text for a Harvard reading should therefore be not Structure but The Copernican Revolution, published in 1957, a text which Isaac largely ignores – and Structure is (as Isaac occasionally seems to acknowledge) correctly read as an engagement in a much wider, international, anti-positivist debate.
6. Not everyone would agree that the truths of mathematics are necessary. Wittgenstein held that we ‘make’ or ‘invent’ mathematical truths, we do not ‘discover’ them (http://plato.stanford.edu/entries/wittgenstein-mathematics/revised 21/02/2011), and the strong programme seeks to extend this principle from mathematics to science (Bloor, ‘Wittgenstein and Mannheim’ (1973)). The question I ask is not ‘Were Regiomontanus and Hobbes right about mathematics?’ but ‘How did their understanding of mathematics help lay the groundwork for reliable scientific knowledge?’ Even Wittgenstein held that there is a reality which corresponds to mathematical truths, but ‘the reality which corresponds to them is that we have a use for them’ (Conant, ‘On Wittgenstein’s Philosophy of Mathematics’ (1997), 220). Science is one of the uses we have for our mathematics, and our mathematics and our science mutually support each other. Bloor, when discussing the utility of mathematics, tacitly assumes that it is useful in making possible certain types of social relationship – thus he thinks it might be right to call mathematics an ideology, like monarchism (189); but mathematics also involves what Wittgenstein calls ‘our practical requirements’ (188), and if 2 + 2 = 4 is a norm, it is not like the divine right of kings but rather like ‘when making a mayonnaise you should add the oil drop by drop.’
7. One way of escaping from the standard relativist argument that good science and bad science cannot be distinguished while avoiding an appeal to an independent reality is to argue that reality itself changes, so that one can then treat nature and society ‘symmetrically’ as part of the same history. This is the approach of Actor Network Theory (ANT); for an impressive example, see Law, ‘Technology and Heterogeneous Engineering’ (1987), and, for the thinking behind this approach, Latour, ‘The Force and the Reason of Experiment’ (1990), and Latour, ‘One More Turn after the Social Turn’ (1992). This approach is admirable in that it rejects the methodological relativism of the Edinburgh and Bath schools, but it leads to a radical historicism (‘My solution . . . is to historicize more not less’: Latour, Pandora’s Hope (1999), 169), according to which Tasmania did not exist before Tasman ‘discovered’ it in 1642 and tuberculosis did not exist before Koch ‘discovered’ it in 1882. All facts are thus artefacts (see above, pp. 256n and 540), which is not true. Nature and reality are also held to be artefacts, which brings us back to relativism by a different route: according to Latour, the laws of nature hold only where there are scientists and scientific instruments, just as frozen fishfingers can be found only where there are freezers and freezer trucks (Latour, We Have Never Been Modern (1993), 91–129).
8. Bloor, Knowledge and Social Imagery (1991): for a critique, see Slezak, ‘A Second Look’ (1994). A striking example of Bloor’s inability to acknowledge that nature constrains science is to be found on p. 39 (though the concession in the last sentence–‘Doubtless we are fully justified in preferring our theory [to Priestley’s] because its internal coherence can be maintained over a wider range of theoretically interpreted experiments and experiences’ – would seem devastating because incompatible with the equivalence postulate). It is important to distinguish between the principle of symmetry (that good and bad science should be explained in the same way) and the principle of impartiality (that failed science should be studied as carefully as successful science – a principle stated by Alexandre Koyré as early as 1933: Zambelli, ‘Introduzione’ (1967), 14). Thus Bertoloni Meli, Equivalence and Priority (1993), 14, appeals to a symmetry principle, but his argument requires only an impartiality principle. Indeed, his account of Leibniz’s conflict with Newton is not symmetrical, since Leibniz was a plagiarist and Newton was not.
9. My view is similar to that of Pickering, The Mangle of Practice (1995), although Pickering avoids the word ‘constraint’ because he thinks it implies social constraint (65–7), preferring ‘resistance’. Contrast Harry Collins’s defence of his assumption that ‘the natural world in no way constrains what is believed to be’ (Collins, ‘Son of Seven Sexes’ (1981), 54; Collins says his position became less extreme in 1980 (Labinger & Collins (eds.), The One Culture? (2001), 184n), so it is worth remarking that I am quoting here from statements of his mature or moderated position). If this were true, Columbus would have reached China, garlic would disempower magnets, and pigs could fly. It is important to grasp that Collins’s relativism (like that of the strong programme) is not the outcome of an empirical programme of enquiry (even though he calls it ‘the Empirical Programme of Relativism’ (Collins, ‘Introduction’ (1981)) but its premise: his whole enterprise ‘rests on the prescription, “treat descriptive language as though it were about imaginary objects”’ (Collins, Changing Order (1985), 16). Clearly, if this is your premise your only conclusion must be that science involves some sort of ‘artful trick’ (6), the trick of persuading people that imaginary objects actually exist. Even Collins, of course, succumbs to this trick (see Collins, ‘Son of Seven Sexes’ (1981), 34, 54), while insisting that it is wrong to do so: thus the empirical enterprise exists only to illustrate and not to test Collins’s relativist premises, and is profoundly implausible in that it requires one to think that things that can’t be true (are ‘literally incredible’) might be true. Some readers may think that Collins can’t be real and that I have made him up (after the Sokal hoax – Sokal, Beyond the Hoax (2008)–such a thought would not be unreasonable). I assure them he does exist and is not a crank: cranks do not get elected as Fellows of the British Academy.
For a more cautiously formulated rejection of ‘constraint’ talk, see Shapin, ‘History of Science and Its Sociological Reconstructions’ (1982), 196–7. What Shapin offers is essentially a circular argument: talk of constraint is incompatible with relativism, but historians are committed to relativism, consequently they must not talk of constraint. Second, he relies on the Duhem – Quine thesis to claim that what constrains scientists is not reality but a particular description of reality; but it is wrong to assume, as this claim does, that the outcome of scientific disputes is always open-ended. When Galileo saw the phases of Venus there was no alternative way of describing what he had seen; nor could there be, unless one was prepared to question assumptions that everyone, with good reason, held in common (that light travels in straight lines, for example).
10. It is easy to add to these examples: Mornet, Les Origines intellectuelles de la Révolution française (1933); Lefebvre, The Coming of the French Revolution (1947); Bailyn, The Ideological Origins of the American Revolution (1967); Trevor-Roper, ‘The Religious Origins of the Enlightenment’ (1967); Stone, The Causes of the English Revolution (1972); Weber, Peasants into Frenchmen (1976); Baker, Inventing the French Revolution (1990); Chartier, The Cultural Origins of the French Revolution (1991); Skinner, ‘Classical Liberty and the Coming of the English Civil War’ (2002); Bayly, The Birth of the Modern World (2003). Equally retrospective in character are books on decline, such as Thomas, Religion and the Decline of Magic (1997), or failure, such as MacIntyre, After Virtue (1981).
One of the reasons, of course, for abandoning the old retrospective stories is that they were profoundly unsatisfactory, as Elton and a whole series of scholars after him showed for the English Civil War (Elton, ‘A High Road to Civil War?’ (1974)), and as Cobban and a whole series of scholars after him showed for the French Revolution (Cobban, The Social Interpretation of the French Revolution (1964)). But the fact that a job has been done badly does not mean it cannot be done better, and it is hard to imagine how a situation in which we have no explanation for the English Civil War, other than that it was an unfortunate accident (which simply begs the question of why it was impossible to put Humpty together again) can be regarded as satisfactory. Nor can I see why historians should cede many of the most interesting questions to other disciplines – politics, philosophy, sociology – simply because they require a consideration of beginnings and endings.
The simple truth is that the definition of Whig history has got tighter and tighter year by year. But in history of science the issue of so-called Whig history is particularly vexed because it is used to censor any acknowledgement that there is progress in science, and so to entrench the equivalence postulate as a principle of historical method. Here, too, attitudes have become more restrictive with each passing year. In 1996 Roy Porter, an historian as opposed to Whig history as anyone, published a work (evidently written earlier, perhaps in 1989) in which he referred to the Scientific Revolution as resulting in ‘substantial and permanent achievements, full of future promise’ and wrote of ‘the advancement of science’ (Porter, ‘The Scientific Revolution and Universities’ (1996), 538, 560; compare Porter, ‘The Scientific Revolution’ (1986), 302). It is time to release the ratchet.
11. A fine analysis of the state of play when I began work on this book is provided by Daston, ‘Science Studies and the History of Science’ (2009); the difference between us is one of emphasis, for in my judgement Daston underestimates the extent to which fear of anachronism has debilitated history of science and overestimates the extent to which history of science has distanced itself from the symmetry principle. (For an earlier acknowledgement that history of science had lost its sense of direction, Secord, ‘Knowledge in Transit’ (2004), 671.) Golinski, ‘New Preface’ (2005), xi, summed up the situation immediately after the Science Wars: ‘Constructivism may have lost some of the bloom of its early promise . . . but it still informs much historical scholarship at the level of tacit assumptions.’
Golinksi is perhaps also typical in his muddle-headed view that the relativism of the strong programme can and should be used ‘as a tool rather than as the expression of a totalizing skepticism’ and in suggesting that constructivism can be regarded as ‘complementary to a range of other approaches’ (x–xi). It is true that sophisticated supporters of the strong programme insist that they are not relativists when going about their daily lives, but they do not suggest that you can be a part-time relativist when studying science as an historian or sociologist; their relativism cannot be picked up and put down like a tool because it is a methodological postulate which rules non-relativist questions out of order – in that respect, they are relativists through and through.
Golinski is also wrong to suggest that it is only with the outbreak of the Science Wars that the constructivist enterprise lost its sense of direction. In fact, by the time of the Sokal hoax (1996), the enterprise was already in deep trouble. From outside, it had been subjected to a devastating critique: Laudan, ‘Demystifying Underdetermination’ (1990). The gathering sense of crisis was marked by Bruno Latour’s statement, as an insider: ‘After years of swift progress, social studies of science are at a standstill’ (Latour, ‘One More turn After the Social Turn’ (1992), 272). They were, and (despite Pickering, The Mangle of Practice (1995), which represents an important attempt to get back on the right road) they still are.
It is now fifteen years since Victoria E. Bonnell and Lynn Hunt published a collection of essays entitled Beyond the Cultural Turn in which they sought – but did not find – a way out of what they called ‘our current predicament’ (Bonnell & Hunt, ‘Introduction’ (1999), 6). Alas, there are still plenty of people who think, with Nick Wilding, that ‘social constructivism does not go nearly far enough.’ Wilding suspects that in the seventeenth century ‘scientific practice was so localized and nontransferable that the idea of a norm belongs to an Enlightenment, rather than early modern, epistemological landscape’ (Wilding, Galileo’s Idol (2014), 136–7). Such an approach inevitably renders the Scientific Revolution totally invisible. It implies that Galileo’s claim that it is impossible, in defiance of nature, to turn falsehood into truth, was entirely misconceived; that Hobbes was wrong to admire Galileo as the founder of a new type of knowledge; and that Diderot’s dream represents the beginning, not the end, of the story of the birth of the scientific enterprise. And, of course, it is wrong: the transferability of Galileo’s new science is straightforwardly demonstrated by a list of the cities in which his work was published in the fifty years after his condemnation in 1633: Strasbourg (1634, 1635, 1636), Leiden (1638), Paris (1639, 1681), Padua (1640, 1649), Lyon (1641), Ravenna (1649), London (1653, 1661, 1663, 1665, 1667, 1682, 1683), Bologna (1655–6, 1664), Amsterdam (1682), to which may be added the popularizing works of Mersenne, Danese, Wilkins, and others. If this is localism, what would its opposite look like?