What’s in the bottle?

Donald MacKenzie

  • The One Culture?: A Conversation about Science edited by Jay Labinger and Harry Collins
    Chicago, 329 pp, £41.00, August 2001, ISBN 0 226 46722 8

For nearly a decade, heated debates about science have split academia and sometimes spilled onto the pages of newspapers. Although the ‘science wars’ were well underway by 1996, they came to wider attention in that year when Alan Sokal succeeded in publishing his brilliant pastiche in Social Text. Sokal’s hoax implicitly condemned – and a fair number of further books and articles raged against, often, alas, without Sokal’s wit – views of science akin to the following: modern science ‘resembles much more a stock-market speculation than a search for the truth about nature’; scientists ‘do not find order in nature, they put it there’; ‘the picture of the scientist as a man with an open mind, someone who weighs the evidence for and against, is a lot of baloney’; ‘modern physics is based on some intrinsic acts of faith’; ‘at any historical moment, what pass as acceptable scientific explanations have both social determinants and social functions.’

To assert things like this about science, Sokal and his fellow ‘science warriors’ argued, traduced its nature as objective knowledge, the result of careful experiment and painstaking analysis, and the source of techniques of great power and predictions of extraordinary accuracy. Even – perhaps especially – when put forward by those who considered themselves on the political Left, such assertions opened the door to creationism, astrology, the occult and all the other forms of present-day anti-rationalism.

The statements I have just quoted, however, do not come from sociologists of science, post-structuralist literary critics, feminist theorists, French philosophers or any of the other varied targets of ‘science wars’ criticism. They actually come from scientists: Erwin Chargaff, Jacob Bronowski, Gunther Stent, Brian Petley, and the trio of Richard Lewontin, Steven Rose and Leon Kamin. In a modest ‘anti-Sokal’ hoax, one of the contributors to The One Culture?, Steven Shapin, leads the reader initially to assume that the quotations come from critics of science in the arts and humanities wing of C.P. Snow’s Two Cultures.

That scientists themselves, often eminent ones, sometimes say such things shows that the ‘science wars’ are more complicated than a simple clash between anti-scientific irrationalists and science’s rationalist proponents. Nor, of course, are they really ‘wars’. As Ian Hacking has pointed out, to call them that is to trivialise the horrors of real war. The ‘science wars’ attracted that name because in many ways they are a specialised version of the wider, but equally misnamed ‘culture wars’ of academia (particularly American academia). Splits within the humanities have been as evident as those between science and arts: many of science’s defenders have been philosophers of traditional (decidedly non-French) sympathies rather than natural scientists. Just as feminism has been a central pole of the ‘culture wars’, so feminist writing about science seems to have attracted particular opprobrium.

Michael Lynch, another contributor to The One Culture?, suggests that the debates about science should be seen not as a ‘war’ but rather as an adversarial legal dispute. It is a closer analogy, but not an entirely reassuring one. The rhetorical resources of the unscrupulous courtroom attorney have been deployed all too frequently: ad hominem attack, selective quotation, scorn, sarcasm, ridicule, accusations of incompetence, charlatanry and intellectual dishonesty, denial of the right of reply. Dismayingly, these have been the tools predominantly of reason’s defenders rather than of its supposed opponents. Minor and reluctant participants in the science wars, such as myself, have not been able to rid ourselves of the feeling that academics and intellectuals just shouldn’t behave like that, even if we often do.

Against this background, The One Culture? is a brave initiative. Its editors are a leading chemist and science administrator (Labinger), and a well-known, controversial sociologist of science (Collins). The volume brings together the science warriors – Sokal, the Nobel laureates Steven Weinberg and Kenneth Wilson, and their fellow physicists Jean Bricmont and David Mermin – with a selective group of the targets of their criticism: sociologists and social historians of science rather than, for example, literary or cultural studies theorists. The debate is joined also by a number of natural scientists and others who have not played a prominent role in earlier disputes. The format is dialectical: twelve initial position statements are followed by eleven commentaries (by the same authors) and a further round of ten responses. Refreshingly, the tone is almost always that appropriate to the seminar room rather than courtroom. Disagreement is expressed, sometimes sharply, but courtesy prevails, and divergent positions are examined and argued against rather than stereotyped and dismissed out of hand. No consensus emerges, but the two ‘sides’ end the book closer than they began it.

Much of the passion of the science wars rests on mutual misunderstandings, and the successive rounds of debate, along with face-to-face discussion at the two conferences on which the book is based, clarify a great deal of what is at issue. Inevitably, the format means that coherence is sacrificed in favour of free debate, and a reader will sometimes feel that he or she is eavesdropping on the slow resolution of multifaceted disagreement among the members of a large and fractious family. Specialists may occasionally even conclude that, in Lynch’s words, they are watching ‘sand-lot philosophy with pick-up teams’. But philosophy’s pros are hardly any closer to agreement on the issues at stake than they were a couple of millennia ago, and those issues are in any case too wide-ranging to be the property of any one academic discipline.

Perhaps the deepest misunderstanding fuelling the science wars – one not focused on squarely in the book – concerns what is involved in analysing or explaining a scientific development in terms of a social process or social ‘factor’. Scientists themselves do this all the time, as anyone who talks to them in any depth about the content of their work will discover, but the way in which they do it tends to be quite different from the way in which it is done by a sociologist of science.

Consider, for example, Weinberg’s assertion that much current work in the quantum theory of gravity is shaped by an ‘ill-placed loyalty to general relativity in its original form’, which ‘persists because of the enormous prestige that the theory earned from its historic successes’. Weinberg is citing a factor that is in a broad sense social, but he does it in order to criticise work that he regards as misguided. Bricmont and Sokal generalise this form of argument when they advocate, against more ‘radical’ sociology of scientific knowledge, the following: ‘First one shows, using conventional scientific arguments, why the research in question is flawed according to the ordinary canons of good science; then, and only then, one attempts to explain how the researchers’ social prejudices (which may well have been unconscious) led them to violate these canons’ (emphases in original). Weinberg again: ‘Although scientists recognise that their theories often bear the stamp of the social environment in which they are formulated, we like to think of this as an impurity, some slag left amid the metal, which we hope eventually to eliminate.’

But do social processes produce only slag to be refined away? Should we turn to social explanations only when the science involved can be shown to be inadequate? Surely not. Consider what science free of social processes and social influence would be like. Scientists would not be educated in their disciplines, for education is a social influence, the authoritative passing on of what a community of practitioners knows. They would work as individuals, not communicating with each other, not assessing or criticising each other’s work, not building on it. If they even read each other, they would have to do so indiscriminately and thus inefficiently, for they couldn’t draw on the social knowledge that tells them which authors are worth reading and which are cranks, whose results can be relied on and whose should be viewed with suspicion. They could not invoke the essentially social trust – a major focus of Shapin’s work – that allows them to be confident that what is in the bottle labelled ‘copper sulphate’ (or ‘sheep’s brains’) is indeed that substance, or that a table of logarithms (or its computerised equivalent) or the electron microscope does not lie systematically. True, such things are sometimes doubted and checked, but if everything had to be doubted and checked all the time, science would come close to paralysis. In sum, a science that was not social could possess almost none of science’s actual power, beauty and accuracy. The social influence of the scientific community on its members – which is profound – is in no sense in tension, in general, with the empirical adequacy, theoretical rigour or practical reliability of scientific knowledge.

Hence the core divide. Natural scientists tend to cite social influences or social processes when they wish to criticise. Sociologists (or modern historians of science) invoke them even-handedly, seeing them at work in the production of well-founded as well as of inadequate knowledge. This is the case even when the social influence in question arises from the wider society (the situation which Bricmont and Sokal have in mind). Take the example of an old piece of sociological analysis of my own that has survived stringent science wars criticism. (I would think that, wouldn’t I?) The development of the mathematical theory of statistics in Britain was shaped substantially by the commitment of statisticians such as Sir Francis Galton and Karl Pearson to eugenics. They would not have done what they did without that commitment, and it affected what they did at the most technical, mathematical level. But the resulting knowledge was not in consequence necessarily flawed: much of it has become part of the corpus of modern statistical theory, and I teach it to my students happily. This is not because there is anything praiseworthy about eugenics – there isn’t – but simply because the history of science is not a morality play. Social influences that have their roots in the wider society can’t be assumed a priori to be detrimental in their effects on knowledge, and this, alas, is true even of those influences of which one disapproves.

The sociologist’s even-handedness is what prompts the frequent accusation of relativism (a term used, fortunately, only in discriminating ways in The One Culture?). It can seem as if sociologists of science don’t know the difference between truth and falsehood, between rationality or irrationality, or that they don’t care about the distinction. In fact, though, what they are saying is that all knowledge (true or false, rational or irrational) tends to arise from a combination of inputs from the natural world and biological, psychological and sociological processes.

This claim underpins the controversial ‘symmetry’ principle of my colleague David Bloor’s ‘Strong Programme’ of the sociology of knowledge (famed, at least, to science wars aficionados). Note carefully what is, and isn’t, implied: even the contributors to The One Culture? tend to translate Bloor’s claim into the related, but not equivalent, precept that the sociologist or historian should set aside, in her or his analysis, the truth or falsity of the science in question. An analogy suggested by Bloor may help. When psychologists analyse visual illusions, they don’t claim that psychological processes are at work in illusions but not in cases of accurate perception. Input from the natural world and biological and psychological processes are held to be at work in both. Indeed, if I understand it correctly, the psychologists’ engagement with the issue, the interest to them of illusions, is at least in part to do with what illusions reveal about the psychological processes of correct vision. Psychologists of perception are thus relativistic and symmetrical, in Bloor’s sense: they believe that the same machinery (so to speak) is involved in both accurate and erroneous perception. But that doesn’t imply that they don’t know, or don’t care, whether one line is longer than another, or even that they have to set that matter aside.

The sociologists’ stance is similar, in that they look for social processes influencing both those collective understandings that are correct and those that are incorrect (including collective understandings in the sphere of science). Rather than social processes leading to error, and truth resulting solely from input from the natural world, both sorts of input are likely to be involved in the production of both truth and error. As in the case of the psychology of perception, that is a methodological precept and ultimately an empirical claim. It does not imply any wild ‘Postmodern’ assertions about truth.

Of course, more is at issue in the science wars than such things as explanatory symmetry. The most compelling passages in The One Culture? come when contributors drop the philosophy and speak of the feelings generated. Most interesting in this respect is one of the newcomers to the debate, Peter Saulson. ‘It’s been just my luck to live my entire scientific career inside a case study,’ he writes. For twenty years he has worked on gravitational wave physics, the topic of a series of sociological investigations by Collins. He talks of his growing respect for Collins and his work: ‘If his is supposed to be the face of the Devil, we should all lighten up.’ But Saulson also notes his and his colleagues’ ‘vicarious thrill’ at the success of Sokal’s hoax. ‘Science studies’ – the generic name for the diverse social-science and humanities disciplines studying science – seemed to have missed something of the essence of science. ‘Deep in my heart,’ Saulson says, ‘I confess that I still hold the faith that something about it’ – science – ‘is both unique and uniquely good.’

While mulling over Saulson’s confession, it struck me how many historians and sociologists of science (myself included) share his faith. We have technical quibbles over the nature of science’s uniqueness, and we have learned that what Hacking nicely calls ‘elevator words’ – Truth, Rationality, Objectivity, Method – fail to capture science’s core and, paradoxically, diminish it. There is, for example, no Scientific Method – no rulebook which, if followed faithfully, guarantees scientific progress. But science is the more admirable for its absence.

Scientists know, Saulson writes, that ‘individually, we are no better human beings than anyone else.’ That imperfect human beings can create an extraordinary product rests, it seems to me, on science’s unique characteristics as a social institution: something which was a central focus of an earlier tradition of the sociology of science, initiated by Robert Merton, but on which recent sociology has been too silent. Science is humanity’s finest cognitive achievement. It is also a social achievement, through and through, and can and should be analysed as such: to do so diminishes it not a whit. And here, surely, is the central tragedy of the science wars: that, quite falsely, and quite unnecessarily, those who see themselves as defending science counterpose science and social processes, and thus reinforce a disabling cultural divide. If The One Culture? helps heal this rift, it will have served its purpose well.