Practice makes... we are not sure

Thomas Kuhn's message to science in the 1960s was 'cling to your paradigm' and the discipline has never been the same since, says John Turney

It was the word "revolution" that caused most of the confusion. If Thomas Kuhn had called his pithy little book The Pre-eminence of Paradigms or Commitment and Conversion in Science , or even Punctuated Equilibria in Scientific Theory , he might have had fewer sales. But The Structure of Scientific Revolutions it was, and it was easy to assume that it would be a deeply radical text. Kuhn, it appeared, not only described multiple scientific revolutions but told you how to stage one.

Through the 1960s and 1970s, the Kuhnian account of scientific practice attracted hundreds of thousands of readers. Normal science solved puzzles whose rules were set by a paradigm - Newtonian dynamics, say, or Ptolemaic astronomy. Over time, it accumulated observational anomalies, which eventually triggered a crisis of explanation and interpretation. Such crises were eventually resolved by the advent of a new paradigm, incommensurable with the old. A paradigm was a two-edged sword. It furnished taken-for-granted assumptions that bound together a community of investigators, and thus permitted truly productive investigation. It obviated continual dispute over the presuppositions of inquiry. But it could be displaced only through a clean break. There could be no floating voters in science. Adherents of successive paradigms, Kuhn maintained, literally lived in different worlds.

This miniature reprise is in deference to the fact that, not many years ago, a THES sub-editor insisted to me that "paradigm" must be explained for the general academic reader, because he did not know the term. So Kuhn is not quite the icon some maintain, and Steve Fuller exaggerates when he suggests that Structure has had more influence on views of science than any other book. Stephen Hawking or James Watson, for example, have a much stronger claim to have influenced the public image of science.

But Kuhn is undoubtedly important for intellectuals, and his influence on them is largely to be deplored. This is the main message in Fuller's painstaking reappraisal of Structure , its author, his milieu, and the reception, and later widespread appropriation, of his most famous book.

This negative verdict does not apply to such an extent to Kuhn's inspiration of various studies of how scientific practice works, which contradict the formerly received view of objective, rational investigators working cooperatively towards unquestioned truths about the universe. These, on the whole, Fuller rather approves of, although he has strong reservations about some features of contemporary science studies. What he abhors are the conservative aspects of the Kuhnian scheme, the disciplining of disciplines. The way to establish a proper research programme, after Kuhn, was to get yourself a paradigm and hold on to it. Forget debate about fundamentals. The path to progress lies in puzzle solving, generating solutions that satisfy internally validated standards. Knowledge production becomes the province of autonomous experts, apprenticed to a working paradigm, and incapable of thinking effectively outside their own, ever more fragmented disciplines.

The focus on normal science, rather than revolution, means that Fuller's critique carries strong echoes of Karl Popper's response to Kuhn. Popper, always prescriptive, argued famously that even if Kuhn's depiction of scientists working in an established paradigm was accurate, it ought not to be. In order to conform to Popper's own scheme for the growth of knowledge, researchers must be perpetually alert to the possibility of refutation of their most cherished assumptions. Good science meant permanent revolution, at least in principle. Fuller wants to extend this to fashion a "social epistemology", in which the entire community, not just a section of the scientific community, gets to debate the means, ends and findings of any important field of inquiry. Authority is continually renegotiated, not taken as given. Science becomes a series of overlapping social movements, not a set of exclusive paradigms.

This is so far from our present experience of science that it may be hard to envisage. But the rest of the book helps. For there is a great deal more to Fuller's study than this rehabilitation of the Popperian old guard as more radical than their Kuhnian successors. His revisionist treatment rests on a detailed account of Structure's origins in postwar Harvard and on a deeper excavation of classic debates about the social relations of science, epitomised by a 100-year-old dispute between Max Planck and Ernst Mach.

The key figure at Harvard was not Kuhn but his mentor, James Conant, president of the university, defence adviser and cold warrior, who instituted the General Education in Science Curriculum, in which Kuhn, as teacher, first developed his ideas. The role of history in that curriculum was to convey the essentials of scientific thinking to the future managerial and political elite, and to elide the differences between older scientific practice and the industrialised "big science" that dominated the physical sciences after the second world war - crudely, to pretend that the Manhattan Project changed nothing.

Kuhn, a disenchanted physicist seeking refuge in history, readily followed Conant's lead. "From this standpoint," writes Fuller, " The Structure of Scientific Revolutions is an exercise in wish fulfilment, as it proposes a general model of scientific change based on examples drawn almost entirely from the 300 years of European physical science prior to the first world war." At the same time, he presents Kuhn as a historian both strikingly unaware of his own position in history and who presented a picture of scientific change that legitimates scientists' own erasure of their past. Real history of science, ironically, becomes the sole concern of historians.

If Kuhn failed to recognise his own historical position, Fuller makes good that omission many times over. Writing with what at times seems near-compulsive erudition, he details the philosophical and political forebears and descendants of just about every significant thinker whose work has any relevance to science or policy.

Most important for his overall narrative, he relates the clash between Mach's view that science's overarching goal should be to find the best ways to manipulate the world for human ends and Planck's conviction that science is the quest for the most complete picture of reality. This instrumentalist versus realist dispute is also, Fuller argues, about whether science is for society, or society for science. The Planckian view dominates, according to Fuller, and is writ large in Kuhn. But Mach must be revisited if we aspire to a more democratic science policy, and to less divisive ways of dealing with expertise.

This episode, and many others that Fuller recounts, offer intriguing perspectives on debates about the place of science. As with any book by Fuller, there are provocations aplenty. Those in science and technology studies will quarrel with his characterisation of the banality of many of the case studies that adorn the field, although Fuller's critical eye, here as elsewhere, is acute. Again, his fundamental point is political. The narratives of science and technology studies involve "more people and more things" than the ones typically produced by scientists or philosophers. This complicates our picture of science, but also "makes it difficult to hold anyone accountable for anything".

For a wider readership, there is more to take issue with in Fuller's view of Kuhn's malign influence. Structure is, after all, little more than a sketch, and a little thought soon suggests that Kuhn's revolutions are pretty unrepresentative of science in general, and that the incommensurability of different paradigms is never as complete as Kuhn originally made out - a point he himself conceded in his 1969 postscript.

More serious, Fuller's book is at heart an extended lament for the costs of ever more refined specialisation. These costs are real, and we face them daily. But Kuhn's case for the benefits of a division of intellectual labour is also persuasive, and I think Fuller's take on the politics of expertise makes him underestimate them. Nor does he quite give full due to Kuhn for opening up for scrutiny a host of questions about how communities of scientific practice actually operate. That, for my money, is the main reason why for many readers science after Structure never looked quite the same again.

It is fair to say, though, that Structure will never look quite the same again after Fuller. In that sense, he has achieved one of the main aims of his ambitious and impressively executed project. Whether he will succeed in cancelling Kuhn's less desirable influence I doubt. Rather, he will induce many to re-read Structure , perhaps with deeper reservations but with no less pleasure than before. For all its flaws, it remains a book impossible to ignore.

Jon Turney is senior lecturer in science and technology studies, University College London.