The Structure of Scientific Revolutions

Thomas Kuhn's Revolution

Twenty years ago Thomas Kuhn's book, The Structure of Scientific Revolutions, was published. Few books create such a deep impression, especially in their authors' lifetimes. It is difficult now to think of an academic discipline, whether in the humanities or in the natural and social sciences, that has not been touched by it. Some of its concepts and insights - the notion of a 'paradigm' for example, or the contrast between 'normal' and "revolutionary" science - have passed into the language of academic discourse, and now colour and shape the way in which scholars think and talk about their specialisms.

Although many of Kuhn's ideas have become debased with overuse and thoughtless extension, few would dispute that he has wrought a revolution in the way we think about science as an intellectual and cultural phenomenon. Yet its very success has, paradoxically, led to a gradual obfuscation of Kuhn's achievement. His ideas become like the air we breathe - something we take for granted. This year, the twentieth anniversay of their original publication, offers the opportunity for a celebration - and a re-evaluation - of a remarkable essay.

To take the measure of Kuhn's achievement, we must first cast our minds back to the ways in which people thought and wrote about science in the decades before the publication of his book. In this connection, two intellectual traditions are particularly important.

The first is the monopolistic hold which philosophy of science, and especially the tradition known as 'logical reconstructionism', exercised over the explanation of the processes of scientific research and discovery. Successive generations of philosophers - from Ernst Mach and the logical positivists to Karl Popper and Imre Lakatos - had engaged in a massive research programme the main aim of which was to impose 'logical' or 'rational' reconstructions on the history of science. Their goal was to account for the development of science in purely logical terms.

This tradition, focusing as it did on purely internal factors in science, placed great emphasis on the intellectual processes of the individual scientist. In doing so, it reinforced a second, equally influential tradition, namely what might be called the Whig interpretation of the history of science. The was the tradition which tended to write history from the vantage point of the present. The Whig interpretation represented the history of science as a uniform progression from 'error' to more and more refined approximations to the 'truth' as embodied in modern scientific concepts and theories.

One of the inevitable consequences of this view was its tendency to undervalue the complexity and sophistication of the thinking of the great scientists of the past. Judged from the standpoint of present knowledge, the reasoning and experimentation of these men seemed riddled with error. To an observer familiar, for example, with the modern concept of oxygen, Joseph Priestley's obsession with 'phlogiston' seems bizarre - as does Sadi Carnot's apparent confusion of the concept of heat with the modern concept of entropy. In a very real sense, therefore, studies of the history of science before Kuhn seemed to boil down to an investigation of how gifted scientists of the past came to make such inexplicable mistakes.

Questions like these provided Kuhn with an entry point into the subject which was to become his life's work. In 1947 he began a study of the origins of 17th century mechanics. For that purpose, he needed to know what the predecessors of Galileo and Newton had known about the subject, and this led him to the discussions of motion in Aristotle's Physica and to later works descended from it. Like most historians, he approached these texts knowing what Newtonian physics and mechanics were, and this led him to ask: how much about mechanics was known within the Aristotelian tradition? Being posed in a retrospective, Newtonian framework, the answer was clear: even at a descriptive level, the Aristotelians had known very little of mechanics, and much of what they had to say about it was simply wrong.

This was a common view at the time, but Kuhn found it puzzling. When dealing with subjects other thn physics, Aristotle had been an acute and naturalistic observer. In fields like biology or political behaviour his observations had been penetrating and deep. How, then, could his talents have deserted him when it came to motion? How could he have held such apparently absurd views about it? And, above all, why had his views been taken so seriously for such a long time by so many of his successors?

Kuhn's great revelation came when he realised that Aristotle's concept of motion seems absurd only when viewed with Newtonian hindsight. But if this perspective is thrown away and an attempt made to view Aristotle's thought as an internally coherent system, then a radically different interpretation emereges. For Aristotle's concept of change was change-in-quality in general, a concept that lumped together both the notion of a falling stone and the growth of a child to adulthood. In his system, the subject that was to become Newtonian mechanics was an as-yet-unglimpsed special case. Aristotle's physics was based not on objects but on quality - of which position was one - and a body which changed its position therefore remained the same body only in the sense that the child is the same individual as the adult it becomes.

The inspiration Kuhn derived from this was that an insightful interpretation of Aristotle's intellectual framework can be obtained only from an attempt to understand it in its own terms. This seminal idea formed the basis for the thinking which led to The Structure of Scientific Revolutions.

The approach to the history (and, by interference, the philosphy) of science which is implicit in Kuhn's book is based on three basic propositions.

Firstly, when studying the work of past scientists, we should assume that their general mode of cognition was similar to ours. We should assume, in other words, that they were no less resourceful, intelligent, reasonable or logical than we are. We should approach their writings with the assumption that they are internally consistent and coherent. Their concepts should be taken seriously, and at the face value accorded them by their authors and by their contemporaries.

Secondly, we should assume that the conceptual usage of a scientist is that of the culture in which he worked. We should not reinterpret Carnot's concept of 'heat', for example, in the light of the modern concept of entropy.

Thirdly, when seeking explanations of why a particular scientist advocated or believed in certain concepts, we should check that any explanation that is offered is consistent with the specific historical context in which he worked. In the case of Carnot, for example, we should relate his concept of heat to the usage conventionally employed in the texts to which we know he had access.

Application of these principles to the history of science led Kuhn to an interpretation radically at odds with those current at the time of his writing. Instead of focusing almost exclusively on the rationality and perception of the isolated individual researcher and his experiments, he diverted attention to the complex interaction that took place between a research community, with its received culture, and its intellectual environment.

From this emerged the well-known Kuhnian picture of science as a cultural phenomenon characterised by two main phases. The first is what he called normal science, in which a community of researchers, clustered around a shared set of theoretical beliefs, accepted standards, procedures and exemplars (which collectively make up a paradigm), engage in routine puzzle solving. The 'puzzles' at the focus of normal scientific activity are usually those generated by perceived mismatches between the paradigm and the phenomena of the natural world, and the solving of puzzles in this way is, Kuhn maintained, the dominant mode of scientific activity.

He went to great pains to point out that this 'normal science' is in no way a radically innovative activity. The scientists engaged in it are not seeking to test their paradigm: and failure to resolve an anomaly which appears to threaten a paradigm is not taken as evidence that the paradigm is deficient. If - to take a trivial example - one carefully draws ten triangles, measures their internal angles and comes up with the result 179.91o +/-0.01, that does not shake one's conviction that the internal angles of a triangle sum to 180o. The good normal scientist, in other words, does not blame his intellectual tools.

No paradigm, however, is perfect, and gradually the perceived anomalies begin to accumulate, accompanied by a growing collective perception that the paradigm is inadequate. If, as usually happens in Kuhn's account, this develops into a state of intellectual crisis, then a radical reorientation of the scientific enterprise may occur. Alternative paradigms arise, each with their vociferous proponents, and a period of debate ensues in which work becomes more speculative and loosely structured. Eventually practice rearranges itself around new theoretical beliefs, procedures, standards and exemplars. A paradigm-shift - Kuhn's "scientific revolution" - has occurred.

Kuhn took a functionalist view of this process of paradigm change, seeing it as an essential phase in scientific development, just as conventional revolutions play a vital role in political development. However, he saw one feature of the revolutionary process as particularly important, namely the fact that rival paradigms are 'incommensurable'. By this Kuhn meant that there exists no metalanguage in terms of which we can rationally discuss and assess the merits of rival paradigms. The rivals may, for example, employ totally different theoretical concepts and terms - 'oxygen' and 'phlogiston', for example - or they may assign different meanings to the same terms. Thus the terms 'mass' and 'force' have different meanings in relativity theory from those they have in Newtonian dynamics. This implies that we cannot conduct a debate about the relative merits of rival paradigms in purely 'logical' terms, and therefore that the process of paradigm shift cannot be a purely 'rational' one. This element in Kuhn's picture of science generated most controversy when it was first revealed to the academic world.

From the outset, Kuhn's book was recognised as a pathbreaking work. His scholarly cast of mind, historical imagination and insight and the sheer novelty of his ideas won immediate attention. Many practising scientists, recognising in his account images which accorded with their own intuitive experiences, welcomed the book. And sociologists and scholars in other, related disciplines, recognising the openings offered by his emphasis on the cultural specificity of scientific activity, saw in it pointers to a new way of looking at science.

But some scholars also perceived the book as controversial because of the bold challenge it presented to established views of science. Its emphasis on normal science as the dominant mode of scientific activity, for example, challenged the Popperian image of science as a hypothetico-deductive process. The idea that a scientific community proceeds mainly by trying to make its paradigm 'fit', rather than by consciously seeking possible refutations of it, was an intolerable affront to this view. And Kuhn's insistence that crucial conceptual transitions in scientific thinking took place in non-logical ways was an open attack on the tradition devoted to imposing 'logical reconstructions' on the history of science. One critic went so far as to accuse Kuhn of portraying critical periods in the history of science as outbreaks of "mob psychology".

Furthermore, Kuhn's picture, by emphasising the incommensurability of rival paradigms, implied that Whiggish notions of scientific progress - with their implicit linear, cummulative representation of history - were untenable. And, of course, the fact that Kuhn was an historian rather than a philosopher added insult to injury. Overall, even his critics had to admit that Kuhn's challenge was indeed a thoroughgoing one.

Criticisms of Kuhn's book focused primarily on two areas: the nature and implications of normal science as he described it; and his claims about the incommensurability of rival paradigms and the consequences of those claims for the notion of scientific progress. The most serious criticism of his treatment of normal science turned on the vagueness of the concept of a paradigm. One commentator, Margaret Masterman, pointed out that Kuhn used the term in no fewer than 21 distinct senses, not all of which were compatible with one another. This highlighted Kuhn's slackness in allowing the terms to become bloated "to the point of being a philosophical analogue to philogiston", as the philospher Frederick Suppe put it. Another philosopher, Dudley Shapere, argued that Kuhn's view "is made to appear convincing only by inflating the definition of 'paradigm' until that term beccomes so vague and ambiguous that it cannot easily be withheld, so general that it cannot easily be applied, and so misleading tht it is a positive hindrance to the understanding of some central aspects of science".

Kuhn responded to these criticisms by admitting that his use of paradigm was counter-productively vague and that it tended to confuse two quite distinct and crucial notions: exemplars - concrete problem solutions which are accepted by the scientific community as being "paradigmatic" in the everyday sense; and disciplinary matrixes - the mix of accepted abstract generalisations, shared commitment to beliefs in particular models, shared values and shared examples that make it possible for scientific communities to communicate and research in a relatively productive and uncontentious way.

When this refined notion of paradigm is incorporated, Kuhn's version of normal science becomes that of an activity carried out by a research community bound together by a common disciplinary matrix. This is acquired through an apprenticeship characterised by the study of examples accepted by the scientific community as being archetypal of good science. (The main teaching components of such an apprenticeship are, of course, textbooks.)

Although some critics found Kuhn's account of normal science vague in certain important aspects, others found the whole concept distasteful on quite other grounds. For them, the representation of vast preponderance of scientific activity as mere routinised puzzle-solving was little short of an outrage. Predictably, some of the most forceful objections along these lines came from Karl Popper. Nothing could be further from his vision of scientists as fearless falsifiers, forever dispatching their most cherished theories on perilous voyages of prediction, observation and experiement than Kuhn's picture of cautious paradigm-slaves anxious not to endanger or blame their theoretical tools. Popper's attack on Kuhn, expressed at a famous symposium in 1965, could be boiled down to two points. In the first place, he maintained that "Kuhn is mistaken when he suggests that what he calls 'normal' science is normal". Secondly, Popper seemed effectively to say that insofar as Kuhnian normal science did exist he (Popper) regarded it as a highly disreputable, shoddy, second-rate activity.

These objections, though vigorous, were comparatively mild compared with the storm raised by Kuhn's account of the process of scientific revolution. In The Structure of Scientific Revolutions, remember, he maintained not only that rival paradigms are 'incommensurable' but also that, logically, they must be. This implies that the intellectual processes which lead scientists to switch allegiance from one paradigm to another cannot be exclusively rational. Kuhn himself emphasised this point by using terms like "faith", "conversion" and "Gestalt switch" in connection with the transition process.

For many critics, this raised the spectre of non-rational or even irrational forces in science. By likening the process of paradigm switch to religious conversion, Kuhn was explicitly challenging the whole of the 'logical reconstruction' thesis, and indeed seemed to some to be opening the door for the forces of 'mob psychology' in science. For many, this was too much to bear.

Not surprisingly, therefore, immediate rescue attempts were energetically launched. The main thrust of these is the argument that, if Kuhn viewed the process of paradigm-shift as non-rational (or even irrational), then there must be something wrong with this concept of rationality. Working on the diffidence with which Kuhn put forward the idea, philosophers like Wolfgang Stegmuller dismissed it as a "bit of musing" by a philosophical incompetent, and suggested that "experts" would have little difficulty in closing the "rationality gap" that Kuhn had opened. But it is noteworthy that no one, not even the ingenious Stegmuller, has actually succeeded in doing so.

However one measures it, there can be no disputing the extent to which Kuhn's ideas have pervaded the way we think about science and the nature of academic disciplines generally. The number and variety of citations of The Structure of Scientific Revolutions, for example, or the ubiquitous use of Kuhnian terms like 'paradigm', 'puzzle' and 'revolution' all point to this conclusion. Yet at the same time, there are reasons for thinking that the very pervasiveness of Kuhn's ideas has undermined their significance - that they become, as it were, diluted by the very process of assimilation.

There has, for example, been a noticeable tendency to try and reconcile the Kuhnian account with the myth of 'logical reconstructionism'. Many people have interpreted Kuhn's picture of science as one of long periods of rationalist 'business as usual', punctuated by occasional sporadic outbreaks of revolutionary madness.

The temptation to see science in these terms is understandable, but any such interpretation is bogus because it is based on a fundamental misapprehension of Kuhn's thinking. In particular, it disguises the extent to which Kuhn saw external social and cultural factors as a dominant influence in all phases of scientific activity. Although these factors are spectacularly visible at times of revolutionary change, they are present all the time, and in that sense Kuhn's concept of normal science is actually more subversive of rationalist myths than his account of scientific revolutions.

The significance of Kuhn's book is also obscured by the fact that the world has moved on since 1962. There is now, for example, a vigorous debate in progress between proponents of rationalist and non-rationalist models of scientific change. And opinions vary about whether Kuhn started the debate, or merely contributed to it. Some would argue, for example, that the Received View of scientific theories would have collapsed anyway under the accumulated weight of its own contradictions, and that Kuhn merely hammered the final nail into its coffin.

There is something to be said for this view, but it tends to undervalue Kuhn's achievement. For the fact that he approached his task as an historian gave a powerful impetus to the philosophical assault on the Received View. To those who argued that the Received View was untenable on epistemological grounds, Kuhn provided evidence that the actual history of science supported their position. His contribution may indeed have been to add just that one single argument, but for many it was the crucial piece of evidence that clinched the matter.

Whether Kuhn was the prime mover in all this or not, the fact remains that the collapse of the logical reconstructionist programme has had far-reaching implications, not all of which have yet permeated popular images of science. For although science as construed by Kuhn and others who share his views is still a very special human activity, it is not quite as special as its romantic image would have us believe. Scientists are fallible human beings, slaves of cultural norms as well as of purely intellectual considerations.

Scientists are rather like the rest of us, in fact. And one doesn't have to go as far as the philosophical joker, Paul Feyerabend - who once argued that organised crime in the United States could be described in exactly the same terms as Kuhn used to describe normal science - to see the implications of this reorientation. For if science is not so special after all, should we not rethink the status and legitimacy which our society accords it? That seminal thought is the real legacy of The Structure of Scientific Revolutions.