The Structure of Scientific Revolutions

Thomas S. Kuhn (1922–96) was the Laurence Rockefeller Professor Emeritus of linguistics and philosophy at the Massachusetts Institute of Technology. His books include The Essential Tension; Black-Body Theory and the Quantum Discontinuity, 1894–1912; and The Copernican Revolution.

The University of Chicago Press, Chicago 60637

The University of Chicago Press, Ltd., London

© 1962, 1970, 1996, 2012 by The University of Chicago

All rights reserved. Published 2012.

Printed in the United States of America

21 20 19 18 17 16 15 14 13 12     1 2 3 4 5

ISBN-13: 978-0-226-45811-3    (cloth)

ISBN-13: 978-0-226-45812-0    (paper)

ISBN-13: 978-0-226-45814-4    (e-book)

ISBN-10: 0-226-45811-3           (cloth)

ISBN-10: 0-226-45812-1           (paper)

ISBN-10: 0-226-45814-8           (e-book)

Library of Congress Cataloging-in-Publication Data

Kuhn, Thomas S.

The structure of scientific revolutions / Thomas S. Kuhn ; with an introductory essay by Ian Hacking.—Fourth edition.

        p. cm.

Includes bibliographical references and index.

ISBN-13: 978-0-226-45811-3 (cloth : alkaline paper)

ISBN-10: 0-226-45811-3 (cloth : alkaline paper)

ISBN-13: 978-0-226-45812-0 (paperback : alkaline paper)

ISBN-10: 0-226-45812-1 (paperback : alkaline paper)

[etc.]

1. Science—Philosophy. 2. Science—History. I. Hacking, Ian. II. Title.

Q175.K95 2012

501—dc23

2011042476

This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

The Structure of Scientific

REVOLUTIONS

FOURTH EDITION

THOMAS S. KUHN

With an Introductory Essay by Ian Hacking

The University of Chicago Press

Chicago and London

Contents

Introductory Essay by Ian Hacking

Preface

I

Introduction: A Role for History

II

The Route to Normal Science

III

The Nature of Normal Science

IV

Normal Science as Puzzle-solving

V

The Priority of Paradigms

VI

Anomaly and the Emergence of Scientific Discoveries

VII

Crisis and the Emergence of Scientific Theories

VIII

The Response to Crisis

IX

The Nature and Necessity of Scientific Revolutions

X

Revolutions as Changes of World View

XI

The Invisibility of Revolutions

XII

The Resolution of Revolutions

XIII

Progress through Revolutions

Postscript—1969

Notes

Index

Introductory Essay

IAN HACKING

Great books are rare. This is one. Read it and you will see.

Skip this introduction. Come back to it if you want to know how the book came into being half a century ago, what its impact was, and the disputes that raged around its theses. Come back if you want one experienced opinion of the status of the book today.

These remarks introduce the book, not Kuhn and his life work. He usually referred to the book as Structure, and in conversation simply as the book. I follow his usage. The Essential Tension is a superb collection of philosophical (as opposed to historical) papers that he published immediately before or soon after Structure.¹ It can be thought of as a series of commentaries and expansions, so it is excellent companion reading.

Since this is an introduction to Structure, nothing beyond The Essential Tension will be discussed here. Note, however, that he often said in conversation that Black-Body and the Quantum Discontinuity, a study of the first quantum revolution launched by Max Planck at the end of the nineteenth century, is an exact example of what Structure is all about.²

Just because Structure is a great book, it can be read in endless ways and put to many uses. Hence this introduction is only one among many possible ones. The book launched a fleet of books about Kuhn’s life and work. An excellent short introduction to the work of Thomas Samuel Kuhn (1922–96), with a different slant from this one, is to be found in the online Stanford Encyclopedia of Philosophy.³ For Kuhn’s final reminiscences of his life and thoughts, see the interview conducted in 1995 by Aristides Baltas, Kostas Gavroglu, and Vassiliki Kindi.⁴ The book about his work that he most admired was Paul Hoyningen-Huene’s Reconstructing Scientific Revolutions.⁵ For a list of all Kuhn’s publications, see James Conant and John Haugeland’s The Road since Structure.⁶

One thing is not said often enough: like all great books, this is a work of passion and a passionate desire to get things right. This is plain even from its modest first sentence on page 1: History, if viewed as a repository for more than anecdote or chronology, could produce a decisive transformation in the image of science by which we are now possessed.⁷ Thomas Kuhn was out to change our understanding of the sciences—that is, of the activities that have enabled our species, for better or worse, to dominate the planet. He succeeded.

1962

The present edition commemorates the fiftieth anniversary of Structure. Nineteen sixty-two was a long time ago. The sciences themselves have radically changed. The queen of the sciences, then, was physics. Kuhn had been trained as a physicist. Few people knew much physics, but everybody knew that physics was where the action was. A cold war was in progress, so everyone knew about the Bomb. American schoolchildren had to practice cowering under their desks. At least once a year towns sounded an air raid siren, at which everyone had to take shelter. Those who protested against a nuclear weapon, by ostentatiously not taking shelter, could be arrested, and some were. Bob Dylan first performed A Hard Rain’s A-Gonna Fall in September 1962; everyone assumed it was about nuclear fallout. In October 1962 there was the Cuban Missile Crisis, the closest the world has come, after 1945, to nuclear war. Physics and its threat were on everyone’s mind.

The Cold War is long over, and physics is no longer where the action is. Another event of 1962 was the awarding of Nobel prizes to Francis Crick and James Watson for the molecular biology of DNA and to Max Perutz and John Kendrew for the molecular biology of hemoglobin. That was the harbinger of change. Today, biotechnology rules. Kuhn took physical science and its history as his model. You will have to decide, after reading his book, about the extent to which what he said about the physical sciences holds true in the teeming, present world of biotechnology. Add in information science. Add in what the computer has done to the practice of science. Even experiment is not what it was, for it has been modified and to a certain extent replaced by computer simulation. And everyone knows that the computer has changed communication. In 1962 scientific results were announced at meetings, in special seminars, in preprints, and then in articles published in specialist journals. Today the primary mode of publication is in an electronic archive.

There is yet another fundamental difference between 2012 and 1962. It affects the heart of the book, fundamental physics. In 1962 there were competing cosmologies: steady state and big bang, two completely different pictures of the universe and its origin. After 1965 and the almost fortuitous discovery of universal background radiation, there is only the big bang, full of outstanding problems pursued as normal science. In 1962 high-energy physics seemed to be an endless collection of more and more particles. What is called the standard model brought order out of chaos. It is unbelievably accurate in its predictions, even if we have no idea how to fit it together with gravity. Perhaps there will not be another revolution in fundamental physics, although for sure there will be surprises galore.

Thus The Structure of Scientific Revolutions may be—I do not say is—more relevant to a past epoch in the history of science than it is to the sciences as they are practiced today.

But is the book history or philosophy? In 1968 Kuhn began a lecture insisting, I stand before you as a practicing historian of science. . . . I am a member of the American Historical, not the American Philosophical, Association.⁸ But as he reorganized his own past, he increasingly presented himself as always having had primarily philosophical interests.⁹ Although Structure had an immense immediate impact on the community of historians of science, its more enduring effects have probably been upon philosophy of science and, indeed, on public culture. That is the perspective from which this introduction is written.

Structure

Structure and revolution are rightly put up front in the book’s title. Kuhn thought not only that there are scientific revolutions but also that they have a structure. He laid out this structure with great care, attaching a useful name to each node in the structure. He had a gift for aphorism; and his names have acquired an unusual status, for although they were once arcane, some of them are now part of colloquial English. Here is the sequence: (1) normal science (§§II–IV—he called these sections, not chapters, for he thought of Structure as more of a book outline than a book); (2) puzzle-solving (§IV); (3) paradigm (§V), a word which, when he used it, was rather uncommon, but which after Kuhn has become banal (not to mention paradigm shift!); (4) anomaly (§VI); (5) crisis (§§VII–VIII); and (6) revolution (§IX), establishing a new paradigm.

That is the structure of scientific revolutions: normal science with a paradigm and a dedication to solving puzzles; followed by serious anomalies, which lead to a crisis; and finally resolution of the crisis by a new paradigm. Another famous word does not occur in the section titles: incommensurability. This is the idea that, in the course of a revolution and paradigm shift, the new ideas and assertions cannot be strictly compared to the old ones. Even if the same words are in use, their very meaning has changed. That in turn led to the idea that a new theory was not chosen to replace an old one, because it was true but more because of a change in world view (§X). The book ends with the disconcerting thought that progress in science is not a simple line leading to the truth. It is more progress away from less adequate conceptions of, and interactions with, the world (§XIII).

Let us look at each idea in turn. Obviously the structure is all too neat. History, the historian protests, just is not like that. But it was precisely Kuhn’s instinct as a physicist that led him to find a simple and insightful all-purpose structure. It was a picture of science that the general reader could pick up. It had the merit of being to some extent testable. Historians of the sciences could look and see the extent to which momentous changes in their fields of expertise did in fact conform to Kuhn’s structure. Unfortunately it was also abused by the wave of skeptical intellectuals who called the very idea of truth in question. Kuhn had no such intention. He was a fact lover and a truth seeker.

Revolution

We think first of revolution in political terms: the American Revolution, the French Revolution, the Russian Revolution. Everything is overthrown; a new world order begins. The first thinker to extend this notion of revolution to the sciences may have been Immanuel Kant. He saw two great intellectual revolutions. They are not mentioned in the first edition (1781) of his greatest masterpiece The Critique of Pure Reason (another rare great book, but not a page turner like Structure!). In the preface to the second edition (1787), he speaks in almost purple prose of two revolutionary events.¹⁰ One was the transition in mathematical practice in which techniques familiar in Babylonia and Egypt were transformed in Greece to proofs from postulates. The second was the emergence of the experimental method and the laboratory, a series of events that he identified as beginning with Galileo. He repeats the word revolution several times in just two long paragraphs.

Notice that although we think of Kant as the purest of scholars, he was living in turbulent times. Everyone knew that something profound was afoot all over Europe, and indeed the French Revolution was only two years away. It was Kant who set in place the idea of a scientific revolution.¹¹ As a philosopher I find it amusing, and certainly forgivable, that honest Kant himself confesses, in a footnote, that he is not in a position to pay attention to the minutiae of historical details.¹²

Kuhn’s first book concerned with science and its history was not Structure but The Copernican Revolution.¹³ The idea of scientific revolution was already very much in circulation. After World War II there was a great deal of writing about the scientific revolution of the seventeenth century. Francis Bacon was its prophet, Galileo its lighthouse, and Newton its sun.

A first point to notice—one that is not immediately obvious on a first skimming of Structure—is that Kuhn was not talking about the scientific revolution. That was quite a different kind of event from the revolutions whose structure Kuhn postulated.¹⁴ Indeed shortly before he published Structure, he had proposed that there was a second scientific revolution.¹⁵ It took place during the early years of the nineteenth century; whole new fields were mathematized. Heat, light, electricity, and magnetism acquired paradigms, and suddenly a whole mass of unsorted phenomena began to make sense. This coincided with—went hand-in-hand with—what we call the industrial revolution. It was arguably the beginning of the modern technoscientific world in which we live. But, no more than the first scientific revolution, did this second revolution exhibit the structure of Structure.

A second point to notice is that the generation preceding Kuhn, the one that wrote so extensively on the scientific revolution of the seventeenth century, had grown up in a world of radical revolution in physics. Einstein’s special (1905) and then general (1916) theory of relativity were more shattering events than we can well conceive. Relativity had, at the beginning, far more repercussions in the humanities and arts than genuine testable consequences in physics. Yes, there was the famous expedition of Sir Arthur Eddington to test an astronomical prediction of the theory, but it was only later that relativity became integral to many branches of physics.

Then there was the quantum revolution, also a two-stage affair, with Max Planck’s introduction of quanta around 1900 and then the full quantum theory of 1926–27, complete with Heisenberg’s uncertainty principle. Combined, relativity and quantum physics overthrew not only old science but basic metaphysics. Kant had taught that absolute Newtonian space and the principle of uniform causality are a priori principles of thought, necessary conditions on how human beings comprehend the world in which they live. Physics proved him totally mistaken. Cause and effect were mere appearance, and indeterminacy was at the root of reality. Revolution was the order of the scientific day.

Before Kuhn, Karl Popper (1902–94) was the most influential philosopher of science—I mean the most widely read, and to some extent believed, by practicing scientists.¹⁶ Popper had come of age during the second quantum revolution. It taught him that science proceeds by conjectures and refutations, to use the title of one of his books. It was a moralistic methodology that Popper claimed was exemplified by the history of science. First we frame bold conjectures, as testable as possible, and inevitably find them wanting. They are refuted, and a new conjecture must be found that fits the facts. Hypotheses can count as scientific only if they are falsifiable. This purist vision of science would have been unthinkable before the great turn-of-the-century revolutions.

Kuhn’s emphasis on revolutions can be seen as the next stage after Popper’s refutations. His own account of the relation between the two is Logic of Discovery or Psychology of Research.¹⁷ Both men took physics as their prototype for all the sciences and formed their ideas in the aftermath of relativity and quanta. The sciences look different today. In 2009 the 150th anniversary of Darwin’s The Origin of Species by Means of Natural Selection was celebrated with great fanfare. With all the books, shows, and festivals, I suspect that many bystanders, if asked what was the most revolutionary scientific work of all time, would very reasonably have answered The Origin of Species. So it is striking that Darwin’s revolution is never mentioned in Structure. Natural selection does enter in an important way on pages 171–72 but only to serve as analogy to scientific development. Now that the life sciences have replaced physics as top dog, we have to ask about the extent to which Darwin’s revolution fits Kuhn’s template.

A final observation: current usage of the word revolution goes far beyond what Kuhn had in mind. This is not a criticism either of Kuhn or of the general public. It does mean that one should read Kuhn attentively and pay attention to what he actually says. Nowadays revolution is pretty much a praise word. Every new refrigerator, every daring new movie, is announced as revolutionary. It is hard to remember that the word was once used sparingly. In the American media (almost forgetful of the American Revolution) the word conveyed more loathing than praise, because revolutionary meant ‘commie.’ I regret the recent debasing of revolution to mere hype, but it is a fact that makes the comprehension of Kuhn a little more difficult.

Normal Science and Puzzle-Solving (§§ II–IV)

Kuhn’s thoughts were really quite shocking. Normal science is, he taught, just working away at a few puzzles that are left open in a current field of knowledge. Puzzle-solving makes us think of crossword puzzles, jigsaw puzzles, and sudoko, pleasant ways to keep busy when one is not up to useful work. Is normal science like that?

A lot of scientific readers were a bit shocked, but then had to admit that is how it is in much of their daily work. Research problems do not aim to produce real novelty. A single sentence of page 35 sums up Kuhn’s doctrine: The most striking feature of the normal research problems we have just encountered is how little they aim to produce major novelties, conceptual or phenomenal. If you look at any research journal, he wrote, you will find three types of problems addressed: (1) determination of significant facts, (2) matching of facts with theory, and (3) articulation of theory. To expand slightly:

1. Theory leaves certain quantities or phenomena inadequately described and only qualitatively tells us what to expect. Measurement and other procedures determine the facts more precisely.

2. Known observations don’t quite tally with theory. What’s wrong? Tidy up the theory or show that the experimental data were defective.

3. The theory may have a solid mathematical formulation, but one is not yet able to comprehend its consequences. Kuhn gives the apt name of articulation to the process of bringing out what is implicit in the theory, often by mathematical analysis.

Although many working scientists agreed that their work confirmed Kuhn’s rule, this still does not sound quite right. One reason Kuhn put things that way is that he (like Popper and many other predecessors) thought that the primary work of science was theoretical. He esteemed theory, and although he had a good sense of experimentation, presented it as of secondary importance. Since the 1980s there has been a substantial shift in emphasis, with historians, sociologists, and philosophers attending seriously to experimental science. As Peter Galison wrote, there are three parallel but largely independent traditions of research: theoretical, experimental, and instrumental.¹⁸ Each is essential to the other two, but they have a good deal of autonomy: Each has a life of its own. Immense experimental or instrumental novelty is simply missed in Kuhn’s theoretical stance, so normal science may have a great deal of novelty, just not theoretical. And for the general public, which wants technologies and cures, the novelties for which science is admired are usually not theoretical at all. That is why Kuhn’s remark sounds somehow wrongheaded.

For a current illustration of what is absolutely right, and also of what is questionable, in Kuhn’s idea of normal science, notice that the high-energy physics most widely reported by science journalists is the search for the Higgs particle. This involves an incredible treasury of both money and talent, all of which is dedicated to confirming what present physics teaches—that there is an as yet undetected particle that plays an essential role in the very existence of matter. Innumerable puzzles, ranging from mathematics to engineering, must be solved en route. In one sense, nothing new in the way of theory or even phenomena is anticipated. That’s what Kuhn was right about. Normal science does not aim at novelty. But novelty can emerge from confirmation of theories already held. Indeed it is hoped that when the right conditions for eliciting the particle are finally established, an entire new generation of high-energy physics will begin.

The characterization of normal science as puzzle-solving suggests that Kuhn did not think normal science was important. On the contrary, he thought scientific activity was enormously important and that most of it is normal science. Nowadays even scientists skeptical of Kuhn’s thought about revolutions have great respect for his account of normal science.

Paradigm (§V)

This element needs special attention. There are two reasons for this. First, Kuhn single-handedly changed the currency of the word paradigm so that a new reader attaches very different connotations to the word than were available to the author in 1962. Secondly, as Kuhn himself stated clearly in his postscript: The paradigm as shared example is the central element of what I now take to be the most novel and least understood aspect of this book (186). On the same page he suggested exemplar as a possible substitute word. In another essay written shortly before the postscript, he admitted that he had lost control of the word.¹⁹ In later life he abandoned it. But we, the readers of Structure fifty years after it was published and after a lot of the dust has settled, can, I hope, happily restore it to prominence.

As soon as the book was published, its readers complained that the word was used in all too many ways. In an often-cited but seldom-read essay, Margaret Masterman found twenty-one distinct ways in which Kuhn used the word paradigm.²⁰ This and similar criticisms prompted him to clarify. The upshot was an essay called Second Thoughts on Paradigms. He distinguished what he called two basic uses of the word, one global and one local. Of the local use he wrote, It is, of course, the sense of ‘paradigm’ as standard example that led originally to my choice of that term. But readers, he said, had mostly used it in a more global way than he had intended, and he continued, I see little chance of recapturing ‘paradigm’ for its original use, the only one that is philologically at all appropriate.²¹ Maybe that was true in 1974, but on this fiftieth anniversary, we can return to the intended use of 1962. I shall come back to local and global but first some recapturing.

Nowadays paradigm, along with its companion paradigm shift, is embarrassingly everywhere. When Kuhn wrote, few people had ever encountered it. Soon it became trendy. The New Yorker, ever alert to and amused by the fashion of the moment, mocked it in cartoons: at a Manhattan cocktail party, a buxom young woman in bell-bottoms says to a balding, would-be hipster, Dynamite, Mr Gerston. You’re the first person I heard use the word ‘paradigm’ in real life.²² Today, it is pretty hard to escape the damn word, which is why Kuhn wrote even in 1970 that he had lost control of it.

Now let’s backtrack. The Greek word paradeigma played an important part in Aristotle’s theory of argument, especially in the book called Rhetoric. That book is about practical argument between two parties, an orator and an audience, who share a great many beliefs that hardly need stating. In English translations the ancestor of our word paradigm is usually rendered as example, but Aristotle meant something more like exemplar, a very best and most instructive example. He thought that there are two basic types of arguments. One kind of argument is essentially deductive, but with many unstated premises. The other is essentially analogical.

In this second basic type of argument, something is in dispute. Here is one of Aristotle’s examples, which many readers will find all too easy to update from the city-states of Aristotle’s time to the nation-states of today. Should Athens go to war with its neighbor Thebes? No. It was evil of Thebes to make war on its neighbor Phocis. Any Athenian audience would agree; it is a paradigm. The situation in dispute is exactly analogous. So it would be evil for us to make war on Thebes.²³

In general: Something is in dispute. One states a compelling example about which almost everyone in the audience will agree—a paradigm. The implication is that what is in dispute is just like that.

In Latin translations of Aristotle, paradeigma became exemplum, which pursued its own career in mediaeval and renaissance theories of argument. The word paradigm was, however, conserved in modern European languages but largely divorced from rhetoric. It tended to have very limited usage, for situations where a standard model was to be followed, or imitated. When schoolchildren had to learn Latin, they were told to conjugate to love—I love, thou lovest, he/she/it loves—as amo, amas, amat, and so on. That was the paradigm, the model to imitate with similar verbs. The primary use of the word paradigm was in connection with grammar, but it was always available as a metaphor. As metaphor it never took off in English, but it seems to have been more common in German. In the 1930s members of the influential philosophy group the Vienna Circle, such as Moritz Schlick and Otto Neurath, were comfortably using the German word in their philosophical writings.²⁴ Kuhn was probably unaware of this, but the philosophy of the Vienna Circle and of other German-language philosophical émigrés to the United States was the philosophy of science on which Kuhn was, in his word, weaned intellectually (9).

Then, in the decade when Structure was maturing, some English analytic philosophers promoted the word. This was partly because the profoundly Viennese Ludwig Wittgenstein had made much use of it in his lectures at Cambridge University during the 1930s. His Cambridge classes were obsessively discussed by those who fell under his spell. The word appears several times in his Philosophical Investigations (another great book, first published in 1953). The first use of the word in that book (§20) speaks of a paradigm of our grammar, although Wittgenstein’s idea of grammar is far more encompassing than the usual one. Later he used it in connection with language-games, an originally obscure German phrase which he made part of general culture.

I do not know when Kuhn first read Wittgenstein, but first at Harvard and then at Berkeley, he had many a conversation with Stanley Cavell, a fascinatingly original thinker who was deeply immersed in Wittgenstein. Each acknowledged the importance, at that moment in their lives, of sharing their intellectual attitudes and problems.²⁵ And paradigm definitely came up as problematic in their discussions.²⁶

At the same time, some British philosophers invented a happily short-lived paradigm-case argument, so named, I think, in 1957. It was much discussed, for it seemed to be a new and general argument against various kinds of philosophical skepticism. Here is a fair parody of the idea. You cannot claim we lack free will (for example), because we had to learn the use of the expression free will from examples, and they are the paradigms. Since we learned the expression from the paradigms, which exist, free will exists.²⁷ So just at the time that Kuhn was writing Structure, the word paradigm was very much in this specialist air.²⁸

The