Thomas Kuhn

Alexander Bird

• Publisher: Routledge
• Year: 2014
• eBook ISBN: 9781317490128
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

Thomas Kuhn (1922-96) transformed the philosophy of science. His seminal 1962 work "The Structure of Scientific Revolutions" introduced the term 'paradigm shift' into the vernacular and remains a fundamental text in the study of the history and philosophy of science. This introduction to Kuhn's ideas covers the breadth of his philosophical work, situating "The Structure of Scientific Revolutions" within Kuhn's wider thought and drawing attention to the development of his ideas over time. Kuhn's work is assessed within the context of other philosophies of science notably logical empiricism and recent developments in naturalized epistemology. The author argues that Kuhn's thinking betrays a residual commitment to many theses characteristic of the empiricists he set out to challenge. Kuhn's influence on the history and philosophy of science is assessed and where the field may be heading in the wake of Kuhn's ideas is explored.

Chapter 1
Kuhn's context

Revolution in the history and philosophy of science

Thanks in large part to the work of Thomas Kuhn it has become something of a commonplace among historians and sociologists of science, and even among philosophers, that in order to appreciate fully the significance of a scientific idea as much as any other historical act, we must properly understand the historical context in which it arose. Kuhn’s own work concerned only ideas in science, but soon after the publication of his book The Structure of Scientific Revolutions in 1962 his model of theory-change in science was being applied to the development of thought in a wide variety of areas of academic, intellectual, and social activity. While not all extensions of Kuhn’s thought are equally sound, it would be odd if the injunction to consider the historical context were applicable to science but not to the history and philosophy of science. Accordingly, in this chapter I shall sketch an outline of some of the more important developments in the philosophical, social, and historical study of science in the years preceding Kuhn’s work; a brief résumé of the central claims and concepts of Kuhn’s thinking will conclude the chapter. An historical introduction is particularly apt and important in Kuhn’s case. One reason for this lies in the very fact of Kuhn’s basic ideas having been so hugely influential. As Kuhn himself argued, a new and revolutionary idea may come to form the basis of subsequent thinking in such a way that it becomes difficult to imagine not thinking in terms of that idea, and difficult therefore to appreciate the thinking that occurred before that idea came along. Consequently the requirement of sensitivity to historical circumstance is both less easy to achieve but also all the more important if we are to see why the new idea was so new and so influential. Since, necessarily, the originator of an idea started his thinking before the revolution his idea causes occurs, his perspective on its content and significance may well be different from ours – and there is reason to think this is true of Kuhn too. Some of his more fervent admirers and would-be disciples expressed thoughts or even found them in his writings that he had to repudiate as different from, or even repugnant to, his own thinking. And so, in order to be fair to Kuhn, we must be careful to read him as addressing a set of problems and concerns that he believed existed in the history and philosophy of science at the time when he was writing. These are unlikely to be the same as ours now. For that matter those concerns and problems did not remain constant throughout his life, and so we will need to give some thought to the changing nature and context of Kuhn’s own work.

Because revolutionary thinkers are born and bred before the revolution they create, inevitably much of what such thinkers believe or, more especially, take for granted, will be shared with the thinking of those whose ideas they help to overthrow. In some regards revolutionaries will seem more advanced to their followers than they actually are. We are apt, for example, to think of Copernicus as the first modern cosmological thinker, the revolutionary who overthrew the Aristotelian-Ptolemaic picture of the Earth-centred universe. Yet, as Kuhn is at pains to illustrate in his first book The Copernican Revolution (1957), Copernicus may equally be regarded as one of the last Aristotelian cosmologists. While the eventual outcome of the revolution he ignited was the rejection of Aristotle’s physics, there is no hint of this from Copernicus himself. On the contrary, Copernicus adhered to the Aristotelian premise that natural motion in the heavens is uniform and circular, and even claimed it as an advantage of his system over Ptolemy’s in that it followed this precept more closely.¹ In many other respects, such as his belief in the finiteness of the universe, the existence of crystalline spheres, and the distinction between “natural” and “violent” motions, Copernicus was closer to Aristotle and Ptolemy than to us.

This provides the second reason why attention to Kuhn’s historical circumstances is important. For, like Copernicus, Kuhn did not represent a total break with his predecessors. There are the seeds of Kuhn’s own revolution in such historians and sociologists as Ludwik Fleck, Karl Mannheim and Robert Merton, as well as philosophers such as Toulmin and Hanson. More importantly, there are several significant respects in which Kuhn accepted the assumptions of those against whom he was reacting. Indeed, I shall argue that part of what makes Kuhn appear so radical is, odd though this may sound, precisely that he failed to reject much of the philosophy of science of his predecessors and contemporaries. The idea that revolutions exhibit some continuity with the past is perhaps well established; it is one to which we shall return in the scientific context. The analogy between scientific and political revolutions predates Kuhn.² The Russian Revolution was certainly a revolution, and a dramatic one at that. But both Tsarist and Bolshevik regimes used repressive techniques – secret police, spy networks, internal exile or labour camps – to maintain and exercise authority. Governments on neither side of the Revolution trusted the Russian people to control their governments democratically. Those few who advocated a constitutional monarchy or democratic republic along western European lines were seeking a revolution that would have been less dramatic but in reality more radical than that achieved by the Bolsheviks. Thus it was the similarities with Tsarism that made the Bolshevik revolution possible, and the dissimilarities that made liberal democratic institutions improbable. With no more analogy intended than this, I shall be arguing that likewise the dramatic aspects of Kuhn’s own revolution are consequences not simply of what was new in his thought but in equal part attributable to what he retained. In this book I shall not only describe Kuhn’s thought but also attempt a critical analysis. The outcome is that had Kuhn rejected more of his predecessors’ thinking he might have come to conclusions less apparently dramatic than those he actually maintained – and also closer to the truth.

The Old Rationalism and the New Paradigm

It is now time to look at the world pre- and post-Kuhn – the world of the philosophy of science, that is. The name I shall give to the dominant network of ideas in pre-1960s philosophy of science is the “Old Rationalism”, while its post-Kuhnian rival will be the “New Paradigm” in honour of Kuhn’s most famous notion. The main players representing the Old Rationalism were Rudolph Carnap, Carl Hempel, Karl Popper and Imre Lakatos. These philosophers did not agree on all their opinions – quite the contrary. But they did share some views, both explicit and implicit. In particular, I use the term “Rationalism” to highlight the opinion common to all those mentioned that a central task of the philosophy of science is to say what scientific rationality consists in, to describe how scientists ought to make inferences from evidence or to choose between competing hypotheses. This is normative philosophy of science. The rationality of scientific procedures, especially with regard to theory assessment and choice, may, as adherents of the Old Rationalism understood it, be encapsulated in the existence of the scientific method. Those adherents did not always call their accounts of scientific rationality by this name. But we may use it to capture a set of ideas to which they by and large subscribed. All Old Rationalists regarded the task of the philosophy of science as the articulation, in as much detail as possible, of the content of the scientific method.

One attempt to describe the scientific method, as here defined, was Rudolph Carnap’s programme of inductive logic. Carnap hoped that philosophers could do for inductive reasoning something akin to what they had already done for deductive reasoning. As regards the latter, Frege and Russell laid down rules whereby the validity of supposedly deductive inferences could be assessed. Since inductive inferences are never valid in this sense, inductive logic would have to aim for something slightly different. Rather than establishing the truth of a conclusion given the truth of premises, as deductive logic does, Carnap’s inductive logic aimed to establish the logical probability of truth of an inductive conclusion, given the evidence.

A quite different Old Rationalist proposal for the content of scientific method was Karl Popper’s falsificationism. Popper fully accepted Hume’s argument that inductive knowledge is impossible. The reliability of induction cannot be known a priori by reason alone. But any attempt to establish the reliability of induction on the basis of experience would beg the question of the reliability of induction, since it would inevitably employ induction. Thus Popper held that no rationally acceptable pattern of reasoning could establish the truth or even probability of an inductive hypothesis. Since he held that science is indeed rational, it follows that in Popper’s view science does not in fact proceed by inductive inference. While a general proposition cannot be verified by corroborating evidence (results of experiment, observations etc.), such a proposition’s falsity can be established deductively from refuting evidence. A single black swan (as may be observed in Australia) refutes the hypothesis that all swans are white. Popper built his picture of scientific method on the basis of the deductive validity of refutation. Science proceeds by a cycle of bold conjecture followed by the attempt to falsify that conjecture. Some theories will be refuted while others may survive (for a time). Following the falsification of a favoured hypothesis, scientists will try to come up with new conjectures and then set about attempting to falsify these. This process of conjecture and refutation Popper likened to Darwinian natural selection.³

Ultimately neither Carnap’s logical probability nor Popper’s falsificationism worked. Carnap’s approach was limited in scope and in any case depended on the choice of language in which to express the hypotheses and evidence. It is clearly unsatisfactory to have the probability one attaches to a hypothesis dependent on the language one chooses to express it – unless there are rational grounds for preferring one language to another. As we shall see, Carnap thought this choice arbitrary. But if one thinks differently, that science should make our choice of language for us, then one faces a problem of circularity. For some language must be adopted in which to carry out science, but that will determine which scientific conclusions we may draw, including a decision about the language that should be in use.

On the other hand, Popper’s avoidance of induction fared even worse. At best, according to Popper, we can only know that a hypothesis is false – we can never know that it is true. Despite Popper’s attempts to suggest that we might get to know that one theory is nearer to the truth than another, even this cannot be done if, like Popper, we eschew induction altogether. This is because claims about the relative verisimilitude (truth-likeness) of general hypotheses always concern infinitely many consequences of whose truth or falsity we are ignorant. Consequently there is no deducing of facts about verisimilitude from our evidence; and so, according to a consistent anti-inductivist, there is no knowledge of verisimilitude. In fact things are worse than this, because we cannot even claim negative knowledge of the falsity of hypotheses. Popper accepts that observations are theory-dependent. An observation is theory-dependent (in one sense) if when making that observation we depend upon the same theory. For example, one might use a radar gun in order to measure the speed of a car or some other object. In so doing one will be depending on the theory of the Doppler effect in order to come to one’s conclusions. A number of philosophers, Kuhn among them, argued that this theory-dependence is ubiquitous and Popper agreed. But if we do not know the theories in question to be true, as Popper holds, then he is committed to agreeing that we do not know the truth of any observation claims either. But observations are what falsify hypotheses. So, if we don’t know any observations, we don’t know that any hypotheses are falsified. And so Popper is committed to a very radical form of scepticism indeed.

It was partly this failure of the Old Rationalism to provide a satisfactory normative philosophy of science that gave rise to the “New Paradigm”. One of the key representatives of the New Paradigm, Paul Feyerabend, challenged the notion of a method that would be applicable in all circumstances – the only injunction that could have such generality would be the empty rule “Anything goes”. But if the Old Rationalists were right in thinking that a philosophically explicable scientific method or a priori account of scientific inference is an essential part of the picture of science as a rational enterprise, then the lack of such a method would seem to suggest that science is not so rational after all.

The Old Rationalists believed not only that they could give an account of scientific rationality but also that such an account could explain the actual development of science. They were impressed by the actual achievements of science and regarded scientific knowledge as the model for all knowledge. It is because scientists by and large do what a normative philosophy of science tells them to, or would tell them, that science converges on the truth and scientific knowledge accumulates. The Old Rationalism endorsed what might seem to be quite a natural picture of the history of science. We are inclined to think that we now know many scientific facts that were not known 1,000 years ago, and indeed many that were not known 100 years or even a decade ago. Our daily newspapers have columns devoted to recent scientific discoveries; the burgeoning of high-tech industries and the ubiquity of their products are everyday, if indirect, testimony to the more basic science that underlies them. It is only natural that we should think of modern scientific knowledge as cumulative. As time passes, new discoveries are made and new knowledge added to the common stock.

Reflection and a little knowledge of the history of science may refine this picture. For it is clear that not everything scientists have believed is true. Sophisticated observers before Copernicus believed that the Earth stood at the centre of the universe, while now we know it to be rotating (more or less) about the Sun, which is itself moving through space. And more recently too, once seemingly well-established theories have come to be held in doubt. It was thought that dinosaurs were cold-blooded creatures whose demise explained simply by the forces of natural selection favouring warmblooded mammals; today it is widely believed, if not yet conclusively known, that they were warm-blooded creatures whose end was precipitated by the impact of a huge meteor. The simple cumulative picture of progress will have to be amended to allow for the existence of false theories. But it is no great modification to allow that along with the large bulk of true theories, some false theories will be allowed to prosper temporarily. Furthermore, the amended picture will add that part of the growth of knowledge will involve the replacement of false theories by true ones – or at least by ones that are nearer to the truth.

If this traditional, cumulative picture were correct, the Old Rationalism had a ready explanation of that fact. According to the Old Rationalism, the reason why the history of science is one of increasing knowledge and of the replacement of falsehood by truth is that scientific beliefs are formed and theories accepted for purely rational reasons. As time passes our stock of evidence grows. Since the likelihood of a theory being true depends on the strength of the supporting evidence, and since rational individuals proportion their beliefs to the evidence, it follows that the rationality of science will explain its accumulation of truth. And if such an explanation is not available to Popper, he was at least able to make use of the Darwinian analogy. If poor theories are weeded out by falsification, those that remain will be more sophisticated and less liable to refutation than their predecessors. Explaining events in the history of science, according to this view, is much more straightforward than explaining events in human history outside science. To explain a scientific discovery it is usually sufficient to detail previous discoveries and the relevant available evidence, along perhaps with a description of the train of reasoning leading from these to the rational scientific conclusion, the discovery in question. Not invoked are factors external to the rational gathering and assessment of evidence. Whereas in political history we must refer to the characters, ambitions, prejudices, alliances, organizations and powers of men and women, in science we can ignore such matters. Or at least these need only be referred to on those rare occasions when it is necessary to explain a wrong turning in the history of science.

The fact that the Old Rationalism was so readily linked with this picture of science as a rationally ordered exercise in the accumulation of knowledge was a second source of dissatisfaction. Despite its attractions, it is a view that has come under considerable attack. Feyerabend argued that scientific heroes, such as Galileo, advanced precisely by going against the proposed canons of normative philosophy of science. Historical studies have suggested that the influence of what I called “external” factors is far more pervasive than the traditional picture allows, and that they are to be found at work in successful, “correct” discoveries as well as erroneous science. Sociologists of science have argued that social relations between scientists are among the factors that determine which theories get accepted. Indeed, they claim that the very kinds of thought a scientist can have are culturally or socially determined.

The failure of the Old Rationalist normative philosophy of science, along with the rejection of the traditional history of science that accompanied it, marked the advent of the New Paradigm. According to the New Paradigm there is no single, universal, all-encompassing scientific method that...

Table of contents

• Cover
• Half Title
• Title
• Copyright
• Contents
• Dedication
• Preface
• 1 Kuhn's context
• 2 Normal and revolutionary science
• 3 Paradigms
• 4 Perception and world change
• 5 Incommensurability and meaning
• 6 Progress and relativism
• 7 Kuhn's legacy
• Notes
• Bibliography
• Index