Revolutions and Relativism: From Kuhn to the ‘Strong Programme'
Meanwhile, the historical approach to philosophy of science which had been pioneered by the French tradition found its way into the English-language debate through the work of the American historian and philosopher of science Thomas Kuhn.
Kuhn’s study of the Copernican revolution in astronomy (Kuhn 1959) exemplified this approach, but it was not until his later, thoroughgoing conceptualization of his understanding of the history of science in The Structure of Scientific Revolutions (Kuhn 1970) that his ideas began to have their profound effect. In this work, Kuhn set out to use historical evidence to undermine the prevailing view of science as involving a progressive accumulation of knowledge towards a greater approximation to the truth, and of the scientific community rationally and disinterestedly fitting fact with theory, rejecting the latter when the facts did not fit and so on. But in the process Kuhn also developed his own alternative account of the nature of scientific activity, with a focus on processes of historical change in scientific understanding.Kuhn uses examples taken from the history of several natural sciences to construct a typical pattern of change. The period prior to the establishment of a scientific approach to any topic is characterized by a plurality of competing approaches. One of these eventually acquires the adherence of the emergent scientific community, and is from then onwards the basis for further research. Once consensus has been reached, the approach which governs subsequent research is called by Kuhn the ‘paradigm’. Research carried out according to a paradigm is generally regarded as ‘routine puzzlesolving’, and Kuhn calls it ‘normal science’. It is taken for granted within the scientific community that problems can be solved in the terms provided by the currently accepted paradigm, so that failure is seen as a failure on the part of researchers, rather than as grounds for rejecting the paradigm (the opposite of what should be happening, according to both the empiricist and the Popperian accounts of scientific rationality).
In Kuhn’s view, paradigms are necessary for scientists to define problems and select methods in their research. To reject the prevailing paradigm just because of a little local difficulty in fitting together facts and theory would leave the scientists without any guidelines for further research (note the parallel here with Kant’s criticism of the empiricist view of experience - see Chapter 3). In fact, there are always problems which have so far defied solution within the prevailing paradigm - this is what keeps scientists busy.However, there are some problems which, either because they come to be seen as particularly persistent and challenging for the paradigm, or because they are posed by external demands on science, do eventually lead to a loss of confidence in the paradigm itself. Kuhn terms these problems ‘anomalies’, and an accumulation of them can lead to a period of ‘crisis’ in the science concerned. In the context of such a crisis, a plurality of alternative and conflicting approaches again emerges. Crises are only resolved when a consensus forms around one of these competing approaches, and normal science can resume, but now under the guidance of a new and different paradigm. Kuhn likens the whole process to a social revolution.
Kuhn's concept of a paradigm is central to his model of science, and it has generated a huge critical literature. For him, the key point is that it is a source of guidance for conducting and evaluating research which is consensual within a particular scientific discipline. It may take the form of a set of shared theoretical assumptions, a commonly accepted view of the subject-matter (ontology), an accepted set of standards for evaluating explanations, a metaphysical view of the world or merely a generally acknowledged past scientific achievement (an ‘exemplar') which provides rules for subsequent practice. Of course, it can be, and often is, all of these things.
The shift of paradigms which occurs through scientific revolutions involves far more than a mere accumulation of new factual knowledge.
The appropriate analogy, in Kuhn's account, would be the shift of perspective one experiences when viewing an ambiguous figure, such as the famous ‘duck-rabbit' (see Figure 3.1, p. 31). There is a sense (just what this sense is has, of course, been very controversial!) in which scientists after a scientific revolution inhabit a different world: everything in their field is redescribed, and interpreted through different theoretical concepts. The new paradigm does not in any straightforward sense build on the past achievements of the science, but, rather, involves rejection of what had previously been accepted. The history of science is, thus, a ‘discontinuous' process, periods of cumulative and consensual ‘normal' science alternating with periods of crisis, ‘revolution' and ‘paradigm shift’.The peculiar character of paradigm shifts in Kuhn's account is what makes his model of science so challenging to earlier views of scientific rationality. The all-embracing character of paradigms is such that they carry with them their own standards for evaluating explanations and judging between rival theories, as well as their own ways of interpreting new evidence. There are, in short, no standards for judging rival candidates to become the new paradigm which are neutral as between the contenders. It is as if there are two political parties, one offering a way of keeping inflation down, the other a way of reducing unemployment, but with no way of deciding which achievement would be the more desirable. The problem may go even deeper than this. Even where, as in the paradigm shift from Newtonian mechanics to the relativistic physics developed by Einstein, there are terms that the rival theories have in common, such as ‘mass' and ‘time', they are defined in different ways. The consequence of this ‘meaning variance' is that dialogue between advocates of the rival theories is at cross-purposes. This is the strongest sense in which rival theories can be said to be ‘incommensurable': that they are mutually unintelligible.
This very radical thesis of incommensurability has the paradoxical consequence that rival theories do not contradict, but just talk past one another. If this were really so, then it is hard to see how they could be in conflict. Kuhn subsequently retreated into a less radical version of incommensurability, but still held that there were no objectively justifiable decision-procedures for basic theory choice. This leaves the way open for a sociological approach to understanding such changes in science. In the absence of objectively rational, paradigm-neutral criteria for theory choice, scientific revolutions are accomplished by way of power struggles in the scientific community, in which editorial control over key journals, capture of particular university departments, the use of rhetoric and propaganda may all have a place. In fact, however, Kuhn himself remained committed to the view that science does progress through revolutions, and that the successful new paradigm does constitute an advance over its defeated rivals.
Kuhn's great achievement was to have brought to the centre of the debate the recognition of historical transformations in scientific beliefs (including transformations in concepts of what science itself was) and the role of social processes within the scientific community in bringing about such transformations. Although Kuhn himself was not a relativist, and believed that science does progress through paradigm shifts, many of his arguments pointed in a relativist direction. This sparked off a major debate in the philosophy of science, and made possible a revolution in the sociology of scientific knowledge.
The philosophical debate turned on whether it was possible to accept the social and historical processes of scientific change which Kuhn had drawn attention to, but still believe in the rationality of science as providing ever-more adequate knowledge of an independently existing reality. However, for other philosophers of science, Kuhn's earlier thesis of radical incommensurability between different paradigms, and the absence of theory-neutral tests to choose between them, led to more radically relativist conclusions.
P. K. Feyerabend was the boldest and best known of these. He was a self-styled ‘anarchist' (Feyerabend 1975) in the theory of knowledge, but, in case this implied that he took himself seriously, he was more inclined to call himself a ‘Dadaist' (referring to the interwar avant garde art movement), implying playful but subversive irreverence. Like Kuhn, Feyerabend took his philosophical views from his interpretation of episodes in the history of science. He argued that what came to be recognized as key advances in scientific thought were in fact achieved by deliberate breaches of accepted scientific method. If sciences progress by breaking the rules, then the appropriate motto should be ‘Anything goes'. For Feyerabend, there are no methodological principles which distinguish science from non-science, and so no reason for thinking science is superior to other forms of understanding of the world. This position supports a tolerant pluralism within science, but also, more broadly, a dethroning of science itself from its privileged social position:And yet science has no greater authority than any other form of life. Its aims are certainly not more important than are the aims that guide the lives in a religious community or in a tribe that is united by a myth. At any rate they have no business restricting the lives, the thoughts, the education of the members of a free society where everyone should have a chance to make up his own mind and to live in accordance with the social beliefs he finds most acceptable. The separation between church and state must therefore be complemented by the separation between state and science.
(Feyerabend 1978: 299)
However, others, most influentially Imre Lakatos, a follower of Karl Popper, attempted to use historical analysis to defend the rationality of science (Lakatos 1970). Lakatos's term ‘research programme' is close to Kuhn's concept of ‘paradigm'. Like the paradigm, a scientific research programme provides rules (‘heuristics') and topics for empirical research - it is able to recruit and sustain the activities of researchers in the field concerned (what Kuhn called ‘normal science').
Unlike Popper, Lakatos accepts that early on in the development of a research programme scientists may be justified in holding on to their basic propositions, or hypotheses (the ‘core' of the programme) in the face of apparently adverse evidence. They may justifiably protect this core from falsification by surrounding it with a ‘periphery' of ad hoc hypotheses. This is necessary to give the supporters of an embryonic research programme a chance to develop it to a point where its explanatory potential can really be shown. In the longer term, the programme can be shown to be succeeding to the extent that it generates unlikely predictions which turn out to be confirmed. Where research programmes fail to do this over a prolonged period of time, they are said to be ‘degenerating, and a shift to a successful programme is ‘progressive'. Since Lakatos's work, considerable ingenuity has been devoted by philosophers of science to the task of offering historically and sociologically sensitive defences of the rationality of science. Examples include Newton-Smith (1981), Hacking (1983), Brown (1994), Longino (1990) and others. The work of the critical realists (discussed at greater length in Chapter 8) can also be understood in this way.So far as sociology of scientific knowledge was concerned, Kuhn's work was seen as liberating the discipline from the requirement to treat the content of natural scientific beliefs as a special case: as outside the scope of sociological explanation because it is determined exclusively by evidence and logic. Influential texts by Barry Barnes (1974) and David Bloor (1976, 1991) inaugurated what presented itself as a ‘strong programme’ in the sociology of knowledge. In his classic statement of this approach, Bloor defined it in terms of four commitments: to seek causal explanations; to treat ‘true' and ‘false' beliefs impartially; to explain ‘true' and ‘false' beliefs, as far as possible, in the same sociological terms (the ‘symmetry' principle); and to accept these commitments as also applying to the explanations provided by the sociology of knowledge itself (the principle of reflexivity).
This new approach to the sociology of science satisfied at least one of Lakatos's criteria for a progressive research programme: it inspired a large number of sophisticated and insightful empirical studies of science in the making, by MacKenzie (1990), Pickering (1984), Collins (1985), Shapin and Schaffer (1985), Pinch (1986) and many more.
These studies showed the extent to which social processes of negotiation and consensus-formation were involved in the construction and authorization of scientific knowledge-claims. Despite some disclaimers (for example, Bloor 1976, 1991: 7), the tendency was to give the impression that sociological explanation could account wholly for the content of science. So scientific knowledge-claims about nature were presented as ‘constructs' of the social processes whose outcome they were. Further, given the symmetry principle, no special status could justifiably be assigned to scientific as against other sorts of belief. There was no direct access to nature-in-itself by which to compare alternative representations of it. The drift towards radical relativism about knowledge, and a social constructionist view of nature itself, seemed both necessary to sustaining the empirical enquiries, and a conclusion supported by the results of the studies themselves.
This led to sharp divisions both within the sociology of science and between its more relativist wing and philosophers and natural scientists who were determined to defend the objectivity and rationality of scientific work; see, for example, the entertaining report on the confrontation between Wolpert and Collins at a British Association meeting by Irwin (1994) and the debate between Collins and Murphy in the UK journal Sociology (Murphy 1994; Collins 1996).