The transition from a paradigm in crisis to a new paradigm from which a new tradition of normal science can emerge is far from a cumulative process achieved by clarifying or extending the old paradigm. Rather, it is a reconstruction of the field from new foundations, a reconstruction that changes many of the field’s paradigmatic methods and applications, as well as even its most basic theoretical generalizations. During the period of transition, there will be considerable overlap between the problems that can be solved by the old paradigm and by the new, but the overlap will never be complete. There will also be decisive differences in the modes of solution. When such a transition is complete, the profession will have changed its view of the field, its methods, and its aims.
One perceptive historian of science, recently considering a classic example of the reorganization of science through paradigm change, described such a transition as “picking up the other end of the stick,” a process that involves “handling the same bundle of data as before, but placing them within a different framework so that they stand in a new system of relations to one another.”8) Others who have noted this aspect of scientific progress have emphasized that such a transition resembles a change in visual gestalt. Marks on paper that at first appeared to be a bird now appear to be an antelope, or vice versa.9) Such analogies are easily misunderstood. Scientists do not see something as something else. Rather, they simply see it. We have already examined some of the problems caused by saying that we have seen something. Nor does the scientist enjoy the freedom of the gestalt subject, who can shift back and forth in his way of seeing. Nevertheless, the gestalt switch, especially because it is so familiar today, provides a useful elementary prototype of what happens in a full-scale paradigm shift.
The preceding anticipations help us recognize crisis as an appropriate prelude to the emergence of new theories, particularly because, in our earlier discussion of the emergence of discoveries, we already examined small-scale modifications of that very same process. Because a new emergence breaks the relation with one tradition of scientific activity and introduces a new tradition conducted under entirely different rules and within an entirely different universe of discourse, crisis can occur only when the first tradition is felt to have gone badly astray. But such signs are no more than a prelude to the examination of the crisis state, and unfortunately the questions they lead to require the talents of a psychologist rather than the abilities of a historian of science. What, after all, is extraordinary research? How does an anomaly become lawlike? How do scientists continue their research when they have realized only that something has gone fundamentally wrong at a level their existing training gives them no means to handle? These questions require deeper consideration than this, and not all of it need be historical. What follows must necessarily be more tentative and less complete than what has come before.
Before a crisis has advanced very far, or before it has been clearly recognized, a new paradigm often appears, at least in a less developed state. Lavoisier’s work provides one example of this. His sealed memorandum was deposited with the French Academy of Sciences less than a year after the first thorough consideration of weight relations within phlogiston theory, and before the crisis in pneumatic chemistry had been fully revealed by the publication of Priestley’s papers. As another example, Thomas Young’s first account of the wave theory of light appeared at the very earliest stage of the crisis then developing in optics, a crisis that, without Young’s assistance, would scarcely have been noticed except that within ten years of his first publication it had become an international scientific event.
In cases such as these, we can say only that a minor breakdown of the paradigm and the first weakening of the paradigmatic rules of normal science were enough to make the scientist look at the field in a new way. What intervened between the first perception of trouble and the recognition of a possible alternative must have been largely unconscious.
In other cases, however—such as those of Copernicus, Einstein, and modern atomic theory—a considerable interval opens between the first recognition of paradigmatic breakdown and the emergence of a new paradigm. When this happens, the historian of science can obtain at least a few hints about what extraordinary science is like. Confronted with a clearly fundamental anomaly in theory, the scientist often first attempts to isolate it more precisely and to give it structure. Though he now knows that they are not necessarily correct, he will employ the rules of normal science more forcefully than before in order to see where, and to what extent, they can still be made to apply in the troubled area. At the same time, he will seek ways to enlarge the breakdown, to turn it into a crisis more dramatic and more suggestive than what has appeared in experiments whose results could be predicted in advance. And in these attempts, more than at any other stage in the post-paradigm development of science, the one who seeks such a path will appear in the image of the scientist at his most scientific. Above all, he will often appear to be someone pursuing anything at random, conducting experiments merely to see what will happen, and searching for results whose essence he cannot properly infer. At the same time, since no experiment can be understood without some sort of theory, the scientist in crisis will constantly strive to put forward conjectural hypotheses, which, if successful, will open the way to a new paradigm, and, if unsuccessful, can be abandoned without much consequence.
Kepler’s account of his years-long struggle with the motion of Mars, and Priestley’s description of his reactions as new gases appeared one after another, provide classic examples of the more random form of research provoked by the recognition of anomaly.10) But perhaps the most fitting descriptions of all can be found in modern research on field theory and fundamental particles. Without a crisis that required knowing how far the rules of normal science could be extended, could the immense effort required to detect the neutrino truly have been justified? Or, if the rules of normal science had not definitely broken down at some undisclosed point, would the extreme hypothesis of parity non-conservation ever have been proposed or tested?
As with much other work in physics over the past decade, these experiments were, in one respect, still an attempt to find and define the source of a scattered series of anomalies.
Extraordinary research of this kind is usually, though by no means universally, accompanied by another form of research. I have in mind, in particular, the turn scientists make toward philosophical analysis as a device for solving the riddles of their field during especially conspicuous periods of crisis. Scientists generally need not be philosophers, nor do they wish to become philosophers. Indeed, normal science tends to keep original philosophy at a distance, and it seems to have good reason to do so. To the extent that normal scientific research can be conducted by taking a paradigm as its model, rules and assumptions need not be made explicit. We have already noted this in Section V through philosophical analysis. But that does not mean that the search for assumptions—even for nonexistent ones—cannot be an effective way of weakening the hold of tradition on the mind and of suggesting the basis for a new tradition. It was no accident that the emergence of Newtonian physics in the seventeenth century and the birth of relativity theory and quantum mechanics in the twentieth were both preceded and accompanied by fundamental philosophical analyses of the research traditions of their times.11) Nor was it any accident that in both of these periods the so-called thought experiment played so decisive a role in the progress of research. As I have indicated earlier, the analytical thought experimentation that occupies so much of the writings of Galileo, Einstein, Bohr, and others is perfectly designed to expose the old paradigm to existing knowledge in a way that isolates the root of crisis with a clarity unattainable in the laboratory.12)
Along with the development of these extraordinary procedures, singly or in combination, one more thing occurs.
By concentrating scientific attention on the narrow area where trouble has arisen, and by preparing the scientific mind to recognize experimental anomalies as such, crisis often produces a profusion of new discoveries. We have already seen how the recognition of crisis distinguished Lavoisier’s work on oxygen from Priestley’s. And oxygen was not the only new gas that chemists’ recognition of anomalies could have found in Priestley’s work. New optical discoveries also poured forth before and during the emergence of the wave theory of light. Some phenomena, such as polarization by reflection, were the result of accidents brought about by intensive research in the troubled area—the discoverer had just begun work on the Academy’s prize essay on double refraction. Other phenomena helped to transform the new hypothesis into a paradigm for the study of the bright spot appearing at the center of the shadow of a rotating disk. Still others, such as the colors of scratches and the colors of thick plate glass, were effects often seen and sometimes mentioned before, but those phenomena had been assimilated in the old way to already well-known effects.13) The various discoveries that occurred around 1895 and with the emergence of quantum mechanics can be explained in much the same way.
Extraordinary research must have many other forms and effects besides these, but in this area we have not yet taken even the first step toward discovering the questions that need to be raised. Perhaps, however, nothing more is needed at this point. The preceding remarks are sufficient to show how crisis loosens stereotyped frameworks while at the same time providing the incremental data necessary for a fundamental paradigm shift. Sometimes the form of the new paradigm is foreshadowed in the structure that extraordinary research has given to anomalies. Einstein wrote that, before he had anything to replace classical mechanics, he could see the interrelations among the already known anomalies of black-body radiation, the photoelectric effect, and specific heats.14) More often, such a structure is not consciously anticipated in advance. Rather, the new paradigm, or the sufficient hint that permits its later clarification, comes all at once and reveals itself, sometimes in the middle of the night, in the mind of a person deeply immersed in crisis.
What the nature of that final stage is__how an individual devises a new way of giving order to the data now all at hand (or discovers that he has devised it)__must here remain an inscrutable matter, and may remain so forever. Let us now note only one thing about it. Almost without exception, those who have achieved such fundamental creations of a new paradigm have been either very young or very new to the field of the paradigm they transform.15) And perhaps that point need not be made explicit, for certainly these are the people who, because of their previous activities, are least bound by the traditional rules of normal science, and who are particularly likely to see that the former rules no longer define a game worth playing and to conceive of other rules
to replace them.
The resulting transition to a new paradigm is a scientific revolution, the topic we are at last prepared to approach directly. But first note one final, apparently elusive characteristic for which the preceding three sections have prepared the way. Until Section VI, where the concept of anomaly was first introduced, the terms “revolution” and “extraordinary science” may have seemed coextensive. More important, there was a circularity that may have seemed suspect to at least some readers, for neither term seemed to mean anything more than “non-normal science.” In fact, it need not have.
We may now seem to face a similar circularity, in which neither term means more than “non-normal science.” In fact, it need not have. We have now reached the point of discovering that a similar circularity is characteristic of scientific theories. Whether troubling or not, however, such circularity is no longer improper. This section and the two preceding sections of this book have drawn out a number of limiting criteria in the breakdown of normal scientific activity, criteria that are wholly independent of whether a revolution occurs after the breakdown. When faced with anomaly or crisis, scientists take a different attitude toward the existing paradigm, and the nature of their research changes accordingly. The proliferation of competing articulations, the willingness to try anything, the expression of explicit discontent, the recourse to philosophy and to debate over fundamentals—all these are symptoms of the transition from normal to extraordinary research. The concept of normal science depends not on the existence of revolutions but on the existence of these symptoms.
“Notes”
1) See especially N. R. Hanson, Patterns of Discovery (Cambridge, 1958), pp. 99-105.
2) T. S. Kuhn, “The Essential Tension: Tradition and Innovation in Scientific Research,” The Third (1959) University of Utah Research Conference on the Identification of Creative Scientific Talent, ed. Calvin W. Taylor (Salt Lake City, 1959), pp. 162-77. For a comparable phenomenon among artists, see Frank Barron, “The Psychology of Imagination,” Scientific American, CXCIX (September, 1958), 151-66, especially p. 160.
3) W. Whewell, History of the Inductive Sciences (rev. ed.; London, 1847), II, 220-21.
4) On the speed of sound, see T. S. Kuhn, “The Caloric Theory of Adiabatic Compression,” Isis, XLIV (1958), 136-137. On the variation of Mercury’s perihelion, see E. T.
Whittaker, A History of the Theories of Aether and Electricity, II (London, 1953), 151, 179.
5) Quoted from T. S. Kuhn, The Copernican Revolution (Cambridge, Mass., 1957), p. 138.
6) Albert Einstein, “Autobiographical Note,” in Albert Einstein: Philosopher-Scientist, ed. P. A. Schilpp (Evanston, Ill., 1949), p. 45.
7) Ralph Kronig, “The Turning Point,” in Theoretical Physics in the Twentieth Century: A Memorial Volume to Wolfgang Pauli, ed. M. Fierz and V. F. Weisskopf (New York, 1960), pp. 22, 25-26. Much of this paper describes the crisis in quantum mechanics during the years immediately preceding 1925.
8) Herbert Butterfield, The Origins of Modern Sciencem, 1300-1800 (London, 1949), pp. 1-7.
9) Hanson, op. cit., chap. i.
10) For an account of Kepler’s work on Mars, see J. L. E. Dreyer, A History of Astronomy from Thales to Kepler (2d ed.; New York, 1953), pp. 380-93.
Dreyer’s summary is occasionally inaccurate, but it is quite sufficient for the materials needed here.
On Priestley, see his own works, especially Experiments and Observations on Different Kinds of Air (London, 1774-75).
11) On the philosophical correspondence that accompanied seventeenth-century mechanics, see Rene Dugas, La mecanique au XVII siecle (Neuchatel, 1054), especially chap. xi. For a similar nineteenth-century episode, see the same author’s earlier study, Histoire de la mecnaique (Neuchatel, 1950), pp. 419-43.
12) T. S. Kuhn, “A Function for Thought Experiments,” in Melanges Alexandre Koyre, ed. R. Taton and I. B. Cohen, 1963, Hermann (Paris).
13) For new optical discoveries in general, see V. Ronchi, Historie de la lumiere (Paris, 1956), chap. vii. For an early account of one of these effects, see J. Priestley, The History and Present State of Discoveries Relating to Vision, Light and Colours (London, 1772), pp. 498-520.
14) Einstein, loc. cit.
15) This generalization about the role of young scientists in fundamental scientific research is common enough to be called a cliché. Moreover, even a glance through any list recording fundamental contributions to scientific theory confirms it impressively. Nevertheless, the generalization surely requires systematic examination. Harvey C. Lehman (Age and Achievement [Princeton, 1953]) provides a great deal of useful data. In his study, however, there is no apparent effort to single out contributions involving fundamental reconceptualization. Nor does his study consider the special circumstances—if there are any—that may accompany relatively late productivity in science.
IX. The Nature and Necessity of Scientific Revolutions
The Nature and Necessity of Scientific Revolutions
With this examination, we can at last consider the problems that have given this essay its title. What are scientific revolutions? And what function do they perform in scientific development?
The answer to these questions has been largely anticipated in the preceding sections. In particular, according to the discussion in the immediately preceding section, scientific revolutions are here taken to be those non-cumulative developmental episodes in which an older paradigm is replaced in whole or in part by an incompatible new one. But there is more to be said, and its essential element can be grasped by posing a question. Why should a change of paradigm be called a revolution? Despite the vast and essential differences between political development and scientific development, what analogy justifies the metaphor that finds revolutions in both?
One aspect of that analogy must already have become clear in the course of the preceding discussion. Political revolutions begin with a growing sense, often widespread among groups within a political community, that existing institutions can no longer adequately solve the problems posed by the surrounding situation. In a quite similar way, scientific revolutions begin with a growing sense, confined to a narrow sector of the scientific community, that an existing paradigm, which had previously guided the many-sided exploration of natural phenomena, has ceased to function adequately. In both political and scientific development, the recognition of a functional defect that can lead to crisis is a prerequisite for revolution. Moreover, though it obviously strains the metaphor, that analogy applies not only to major paradigm changes such as those associated with Copernicus and Lavoisier, but also to the more local paradigm changes involved in the assimilation of new phenomena such as oxygen or X-rays. Scientific revolutions, as we saw at the end of Section V, need seem revolutionary only to those whose paradigms are affected by them. To outsiders, they may appear, like the Balkan revolutions of the early twentieth century, to be a normal part of the developmental process. Astronomers, for example, accepted X-rays as merely one more addition to the stock of knowledge, because their paradigms were not altered by the existence of the new radiation. But for such scholars as Kelvin, Crookes, and Roentgen, who in the course of their research dealt with radiation theory or cathode-ray tubes, the appearance of X-rays necessarily violated one paradigm while creating another, new one.
This is why those rays could be discovered only after something had first gone wrong in normal research.
This aspect of the fundamental similarity between political and scientific development is no longer open to doubt. But that fundamental similarity has a second, more profound aspect. In general, political revolutions aim to reform existing political institutions in ways that those very institutions prohibit.
Therefore, the success of a political revolution necessarily entails the partial destruction of existing institutions in favor of others, and during that process society is not fully controlled by the old institutions. Just as we saw earlier that only a crisis weakens the role of a paradigm, so too only a crisis can weaken the role of political institutions in the first place. As their numbers grow, more and more people become gradually alienated from political life and act within it in ways increasingly outside the normal course. Then, as the crisis deepens, many of these people begin to articulate some concrete alternative for reorganizing society within the framework of new institutions. At that point, society becomes divided into several competing camps or parties: one side takes a position defending the old order, while the other seeks the establishment of new institutions. And once this polarization of camps occurs, reliance on politics breaks down.
Because those camps differ over the institutional model within which political revolution is to be carried out and evaluated, and because they know of no supra-institutional framework for reconciling differences in a revolution, the parties engaged in revolutionary struggle ultimately often have to appeal to techniques of mass persuasion, including force. Revolutions have played a decisive role in the progress of political institutions, but that role depends on the fact that revolutions are, in part, extra-political or extra-institutional events. The remainder of this essay aims to demonstrate that a scientific examination of paradigm change reveals very similar characteristics in the evolution of science. It becomes clear that the choice between competing paradigms is like the choice between competing political institutions. Because it has such a character, the choice is not, and cannot be, determined simply by the methods of evaluation characteristic of normal science. The reason is that the choice depends in part on a particular paradigm, and that very paradigm is what is under discussion. When a paradigm, as it necessarily must, enters into the debate over paradigm choice, its role inevitably becomes circular. For each group uses its own paradigm in the arguments defending that paradigm.
Of course, the resulting circularity does not make the arguments false or even powerless. Rather, those who take the paradigm as a premise in arguments defending it can present clear evidence of what scientific activity would look like to those who accept a new view of nature. Such evidence may be enormously persuasive, and it is often pressed in that way.
Nevertheless, whatever its force, the status of a circular argument is only that of persuasion. It cannot logically, or even probabilistically, compel those who refuse to enter the circle to do so. The premises and values shared by the two sides in a debate over paradigms lack the comprehensiveness required for that. As in political revolution, so it is in paradigm choice—there is no criterion higher than the assent of the relevant community. Therefore, to discover how scientific revolutions are achieved, we must examine not only the impact of nature and logic, but also the techniques of effective persuasive argument within the rather distinctive group that constitutes the scientific community.
In order to find why this issue of paradigm choice cannot be settled firmly by logic and experiment alone, we must soon examine the nature of the differences that distinguish supporters of the traditional paradigm from their revolutionary successors. Such an examination is the principal aim of this section and the next. But we have already seen several examples of such differences, and no one will doubt that history can provide many others. What is indeed more doubtful than the existence of such examples—and what must be considered first—is whether such cases provide decisive information about the nature of nature. If we accept that the rejection of paradigms has been a historical fact, does it reveal anything more than humanity’s tendency to believe too readily and the existence of confusion? Is there an intrinsic reason why the assimilation of a new kind of phenomenon or a new scientific theory must compel the rejection of the older paradigm?