100 Best Books for an Education

A Revision and Update of Will Durant's 100 Best Books for an Education

Note 4


The Structure of Scientific Revolutions


   How does science progress? The conventional answer involves using a logical algorithm entitled The Scientific Method. But is that how it really works? Of course the answer is “Yes,” up to a point. That point is reached when the scientific method cannot answer certain nagging questions that are built into the system or framework from which we ask questions. What happens then? In 1962, the physicist-turned-historian of science, Thomas S. Kuhn (1922–1996), published The Structure of Scientific Revolutions, a book that presented an account of the nature of scientific activity that diverged radically from those offered by mainstream philosophers of science. At the same time, Kuhn retained the conventional supposition that science differed from most other intellectual activities in its “progressive” character. Kuhn’s account, which has played a tremendous role in stimulating the entire social studies of science movement, centered on the role played by paradigms, or “universally recognized scientific achievements that for a time provide model problems and solutions to a community of practitioners.” Like most seminal concepts, a certain vagueness surrounds the term paradigm.

   According to Kuhn, a paradigm is generated when some particular work implicitly determines the legitimate problems and research methods of a research field for a succeeding generation of practitioners. Historical examples of paradigm-setting works include Ptolemy’s Almagest, Newton’s Principia and Optics, Lavoisier’s Principles of Chemistry, Lyell’s Principles of Geology, and Bohr’s On the Quantum Theory of Line-Spectra. Dominant paradigms emerge only when the work in question has two critical characteristics: first, its achievement must seem so substantial as to attract an enduring group of followers, and second, it must be sufficiently open-ended to suggest a large number of problems waiting to be solved. In essence, the scientific community shifts its focus even though looking at the same data and comes to a different perspective. Is it a vase, or two faces?   

   A paradigm is not merely a theory, although it may contain one or more. It is not merely a set of instruments, though it may incorporate some. Nor is it a set of metaphysical or epistemic assumptions, though it may explicitly or implicitly incorporate some of those too. A paradigm is, instead, a concrete artifact or model investigation that produces a particular way of seeing phenomena and which invites elaboration and extension.

   Once a community of practitioners forms around a paradigmatic work, the process of “normal science” commences. Since the paradigm both defines what constitutes interesting problems as well as the strategies and instrumentation that should be used in solving them, normal science is basically a puzzle-solving activity, and normal scientists can be viewed as sophisticated puzzle solvers who have been given both the rules they must follow and the problems they must solve. The progressive character of science follows from normal science rapidly expanding the number of problems solved as scientists elucidate the implications of the paradigm they share. Occasionally, however, the puzzle solving leads to a result that violates paradigm-induced expectations, as when the results of the Michaelson-Morley experiment, expected to determine a relatively large velocity for the Earth’s motion through an electromagnetic ether, suggested that that value was zero, or when the ‘black-body’ problem in physics refused all forms of solution. Kuhn called these novel or unexpected results anomalies.

   At first, anomalies will be ignored, and elaboration of the paradigm will continue unabated. But if the number of anomalies increases substantially, or if they occur in connection with problems that are considered particularly important and central, scientists begin to question aspects of the dominant paradigm. Eventually, a new paradigm will be offered which: (1) resolves many or all of the anomalies of the old paradigm, (2) is capable of solving most or all of the other problems solved by the old paradigm, and (3) suggests an extended domain of problems to be solved. Then begins a “revolutionary” period in the science, during which adherents of the old and the new paradigms compete for converts. In general, proponents of the old paradigm will be unable to change their way of seeing the world, so the new paradigm must gain its support from newer recruits to the field. In time, supporters of the old paradigm will drop away, and followers of the new paradigm will usher in a new period of normal science in the field. All truth,” writes Arthur Schopenhauer passes through three stages. First, it is ridiculed, second, it is violently opposed, and third, it is accepted as self-evident.  Humorously, it has been summarized as: 1) Preposterous! 2) Oh, let’s indulge it and prove just how wrong it is, and 3) I knew it all the time, it had to be true!

   Since the new paradigm will almost certainly solve nearly all of the problems solved by the old, and since it will also have an extended domain of applicability, the change of paradigms is likely to be seen as progressive. Nothing, however, about the process of paradigm shifts guarantees that the process will produce knowledge that comes increasingly close to some perfect knowledge of any “real” natural world. The process of new paradigms replacing old ones is likely to be unending.

   Much of the criticism of Kuhn’s paradigm theory has focused on the claim that different paradigms must be psychologically and sociologically exclusive so that a mature science cannot exist when two or more paradigms coexist. In contrast, Imre Lakatos (1922–1974), for example, argued that sciences are characterized by the existence of robust “research programmes” that may run in parallel without serious conflict. Other criticisms focus on the notion that ordinary science, which had long been characterized as focusing on attempts to test or falsify theories, is instead a relatively routine and intellectually conservative enterprise. This aspect of the paradigm concept has evoked the hostility of many working scientists, who resent the implication that their work is narrow and routine.

   From the perspective of the social studies of science, Kuhn’s paradigm concept has been of considerable importance because it focuses attention on the social process of consensus formation in the sciences as the central feature of scientific activity. In so doing, it draws attention to the scientific community’s mechanisms for enforcing conformity, processes such as peer review for purposes of funding and access to publication.


Face Vase