Review: The Structure of Scientific Revolutions

The Structure of Scientific Revolutions by Thomas S. Kuhn

My rating: 5 of 5 stars

Observation and experience can and must drastically restrict the range of admissible scientific belief, else there would be no science. But they cannot alone determine a particular body of such belief. An apparently arbitrary element, compounded of personal and historical accident, is always a formative ingredient of the beliefs espoused by a given scientific community at a given time.

This is one of those wonderfully rich classics, touching on many disparate fields and putting forward ideas that have become permanent fixtures of our mental furniture. Kuhn synthesizes insights from history, sociology, psychology, and philosophy into a novel conception of science—one which, despite seemingly nobody agreeing with it, has become remarkably influential. Indeed, this book made such an impact that the contemporary reader may have difficulty seeing why it was so controversial in the first place.

Kuhn’s fundamental conception is of the paradigm. A paradigm is a research program that defines a discipline, perhaps briefly, perhaps for centuries. This is a not only a dominant theory, but a set of experimental methodologies, ontological commitments, and shared assumptions about standards of evidence and explanation. These paradigms usually trace their existence to a breakthrough work, such as Newton’s Principia or Lavoisier’s Elements; and they persist until the research program is thrown into crisis through stubborn anomalies (phenomena that cannot be accounted for within the theory). At this point a new paradigm may arise and replace the old one, such as the switch from Newton’s to Einstein’s system.

Though Kuhn is often spoken of as responding to Popper, I believe his book is really aimed at undermining the old positivistic conception of science: where science consists of a body of verified statements, and discoveries and innovations cause this body of statements to gradually grow. What this view leaves out is the interconnection and interdependence between these beliefs, and the reciprocal relationship between theory and observation. Our background orients our vision, telling us where to look and what to look for; and we naturally do our best to integrate a new phenomenon into our preexisting web of beliefs. Thus we may extend, refine, and elaborate our vision of the world without undermining any of our fundamental theories. This is what Kuhn describes as “normal science.”

During a period of “normal science” it may be true that scientific knowledge gradually accumulates. But when the dominant paradigm reaches a crisis, and the community finds itself unable to accommodate certain persistent observations, a new paradigm may take over. This cannot be described as a mere quantitative increase in knowledge, but is a qualitative shift in vision. New terms are introduced, older ones redefined; previous discoveries are reinterpreted and given a new meaning; and in general the web of connections between facts and theories is expanded and rearranged. This is Kuhn’s famous “paradigm shift.” And since the new paradigm so reorients our vision, it will be impossible to directly compare it with the older one; it will be as if practitioners from the two paradigms speak different languages or inhabit different worlds.

This scandalized some, and delighted others, and for the same reason: that Kuhn seemed to be arguing that scientific knowledge is socially solipsistic. That is to say that scientific “truth” was only true because it was given credence by the scientific community. Thus no paradigm can be said to be objectively “better” than another, and science cannot be said to really “advance.” Science was reduced to a series of fashionable ideas.

Scientists were understandably peeved by the notion, and social scientists concomitantly delighted, since it meant their discipline was at the crux of scientific knowledge. But Kuhn repeatedly denied being a relativist, and I think the text bears him out. It must be said, however, that Kuhn does not guard against this relativistic interpretation of his work as much as, in retrospect, he should have. I believe this was because Kuhn’s primary aim was to undermine the positivistic, gradualist account of science—which was fairly universally held in the past—and not to replace it with a fully worked-out theory of scientific progress himself. (And this is ironic since Kuhn himself argues that an old paradigm is never abandoned until a new paradigm takes its place.)

Though Kuhn does say a good deal about this, I think he could have emphasized more strongly the ways that paradigms contribute positively to reliable scientific knowledge. For we simply cannot look on the world as neutral observers; and even if we could, we would not be any the wiser for it. The very process of learning involves limiting possibilities. This is literally what happens to our brains as we grow up: the confused mass of neural connections is pruned, leaving only the ones which have proven useful in our environment. If our brains did not quickly and efficiently analyze environmental stimuli into familiar categories, we could hardly survive a day. The world would be a swirling, jumbled chaos.

Reducing ambiguities is so important to our survival that I think one of the primary functions of human culture is to further eliminate possibilities. For humans, being born with considerable behavioral flexibility, must learn to become inflexible, so to speak, in order to live effectively in a group. All communication presupposes a large degree of agreement within members of a community; and since we are born lacking this, we must be taught fairly rigid sets of assumptions in order to create the necessary accord. In science this process is performed in a much more formalized way, but nevertheless its end is the same: to allow communication and cooperation via a shared language and a shared view of the world.

Yet this is no argument for epistemological relativism, any more than the existence of incompatible moral systems is an argument for moral relativism. While people commonly call themselves cultural relativists when it comes to morals, few people are really willing to argue that, say, unprovoked violence is morally praiseworthy in certain situations. What people mean by calling themselves relativists is that they are pluralists: they acknowledge that incompatible social arrangements can nevertheless be equally ethical. Whether a society has private property or holds everything in common, whether it is monogamous or polygamous, whether burping is considered polite or rude—these may vary, and yet create coherent, mutually incompatible, ethical systems. Furthermore, acknowledging the possibility of equally valid ethical systems also does not rule out the possibility of moral progress, as any given ethical system may contain flaws (such as refusing to respect certain categories of people) that can be corrected over time.

I believe that Kuhn would argue that scientific cultures may be thought of in the same pluralistic way: paradigms can be improved, and incompatible paradigms can nevertheless both have some validity. Acknowledging this does not force one to abandon the concept of “knowledge,” any more than acknowledging cultural differences in etiquette forces one to abandon the concept of “politeness.”

Thus accepting Kuhn’s position does not force one to embrace epistemological relativism—or, at least not the strong variety, which reduces knowledge merely to widespread belief. I would go further, and argue that Kuhn’s account of science—or at least elements of his account—can be made to articulate even with the system of his reputed nemesis, Karl Popper. For both conceptions have the scientist beginning, not with observations and facts, but with certain arbitrary assumptions and expectations. This may sound unpromising; but these assumptions and expectations, by orienting our vision, allow us to realize when we are mistaken, and to revise our theories. The Baconian inductivist or the logical positivist, by beginning with an raw mass of data, has little idea how to make sense of it and thus no basis upon which to judge whether an observation is anomalous or not.

This is not where the resemblance ends. According to both Kuhn and Popper (though the former is describing while the second is prescribing), when we are revising our theories we should if possible modify or discard the least fundamental part, while leaving the underlying paradigm unchanged. This is Kuhn’s “normal science.” So when irregularities were observed in Uranus’ orbit, the scientists could have either discarded Newton’s theories (fundamental to the discipline) or the theory that Uranus was the furthest planet in the solar system (a superficial fact); obviously the latter was preferable, and this led to the discovery of Neptune. Science could not survive if scientists too willingly overturned the discoveries and theories of their discipline. A certain amount of stubbornness is a virtue in learning.

Obviously, the two thinkers also disagree about much. One issue is whether two paradigms can be directly compared or definitively tested. Popper envisions conclusive experiments whose outcome can unambiguously decide whether one paradigm or another is to be preferred. There are some difficulties to this view, however, which Kuhn points out. One is that different paradigms may attach very different importance to certain phenomena. Thus for Galileo (to use Kuhn’s example) a pendulum is a prime exemplar of motion, while to an Aristotelian a pendulum is a highly complex secondary phenomenon, unfit to demonstrate the fundamental properties of motion. Another difficulty in comparing theories is that terms may be defined differently. Einstein said that massive objects bend space, but Newtonian space is not a thing at all and so cannot be bent.

Granting the difficulties of comparing different paradigms, I nevertheless think that Kuhn is mistaken in his insistence that they are as separate as two languages. I believe his argument rests, in part, on his conceiving of a paradigm as beginning with definitions of fundamental terms (such as “space” or “time”) which are circular (such as “time is that measured by clocks,” etc.); so that comparing two paradigms would be like comparing Euclidian and non-Euclidian geometry to see which is more “true,” though both are equally true to their own axioms (while mutually incompatible). Yet such terms in science do not merely define, but denote phenomena in our experience. Thus (to continue the example) while Euclidian and non-Euclidian geometries may both be equally valid according to their premises, they may not be equally valid according to how they describe our experience.

Kuhn’s response to this would be, I believe, that we cannot have neutral experiences, but all our observations are already theory-laden. While this is true, it is also true that theory does not totally determine our vision; and clever experimenters can often, I believe, devise tests that can differentiate between paradigms to most practitioners’ satisfaction. Nevertheless, as both Kuhn and Popper would admit, the decision to abandon one theory for another can never be a wholly rational affair, since there is no way of telling whether the old paradigm could, with sufficient ingenuity, be made to accommodate the anomalous data; and in any case a strange phenomena can always be tabled as a perplexing but unimportant deviation for future researchers to tackle. This is how an Aristotelian would view Galileo’s pendulum, I believe.

Yet this fact—that there can be no objective, fool-proof criteria for switching paradigms—is no reason to despair. We are not prophets; every decision we take involves risk that it will not pan out; and in this respect science is no different. What makes science special is not that it is purely rational or wholly objective, but that our guesses are systematically checked against our experience and debated within a community of dedicated inquirers. All knowledge contains an imaginative and thus an arbitrary element; but this does not mean that anything goes. To use a comparison, a painter working on a portrait will have to make innumerable little decisions during her work; and yet—provided the painter is working within a tradition that values literal realism—her work will be judged, not for the taste displayed, but for the perceived accuracy. Just so, science is not different from other cultural realms in lacking arbitrary elements, but in the shared values that determine how the final result is judged.

I think that Kuhn would assent to this; and I think it was only the widespread belief that science was as objective, asocial, and unimaginative as a camera taking a photograph that led him to emphasize the social and arbitrary aspects of science so strongly. This is why, contrary to his expectations, so many people read his work as advocating total relativism.

It should be said, however, that Kuhn’s position does alter how we normally think of “truth.” In this I also find him strikingly close to his reputed nemesis, Popper. For here is the Austrian philosopher on the quest for truth:

Science never pursues the illusory aim of making its answers final, or even probable. Its advance is, rather, towards the infinite yet attainable aim of ever discovering new, deeper, and more general problems, and of subjecting its ever tentative answers to ever renewed and ever more rigorous tests.

And here is what his American counterpart has to say:

Later scientific theories are better than earlier ones for solving puzzles in the often quite different environments to which they are applied. That is not a relativist’s position, and it displays the sense in which I am a convinced believer in scientific progress.

Here is another juxtaposition. Popper says:

Science is not a system of certain, or well-established, statements; nor is it a system which steadily advances towards a state of finality. Our science is not knowledge (episteme): it can never claim to have attained truth, or even a substitute for it, such as probability. … We do not know: we can only guess. And our guesses are guided by the unscientific, the metaphysical (though biologically explicable) faith in laws, in regularities which we can uncover—discover.

And Kuhn:

One often hears that successive theories grow ever closer to, or approximate more and more closely to, the truth… Perhaps there is some other way of salvaging the notion of ‘truth’ for application to whole theories, but this one will not do. There is, I think, no theory-independent way to reconstruct phrases like ‘really there’; the notion of a match between the ontology of a theory and its ‘real’ counterpart in nature now seems to me illusive in principle.

Though there are important differences, to me it is striking how similar their accounts of scientific progress are: the ever-increasing expansion of problems, or puzzles, that the scientist may investigate. And both thinkers are careful to point out that this expansion cannot be understood as an approach towards an ultimate “true” explanation of everything, and I think their reasons for saying so are related. For since Popper begins with theories, and Kuhn with paradigms—both of which stem from the imagination of scientists—their accounts of knowledge can never be wholly “objective,” but must contain an aforementioned arbitrary element. This necessarily leaves open the possibility that an incompatible theory may yet do an equal or better job in making sense of an observation, or that a heretofore undiscovered phenomenon may violate the theory. And this being so, we can never say that we have reached an “ultimate” explanation, where our theory can be taken as a perfect mirror of reality.

I do not think this notion jeopardizes the scientific enterprise. To the contrary, I think that science is distinguished from older, metaphysical sorts of enquiry in that it is always open-ended, and makes no claim to possessing absolute “truth.” It is this very corrigibility of science that is its strength.

This review has already gone on for far too long, and much of it has been spent riding my own hobby-horse without evaluating the book. Yet I think it is a testament to Kuhn’s work that it is still so rich and suggestive, even after many of its insights have been absorbed into the culture. Though I have tried to defend Kuhn from accusations of relativism or undermining science, anyone must admit that this book has many flaws. One is Kuhn’s firm line between “normal” science and paradigm shifts. In his model, the first consists of mere puzzle-solving while the second involves a radical break with the past. But I think experience does not bear out this hard dichotomy; discoveries and innovations may be revolutionary to different degrees, which I think undermines Kuhn’s picture of science evolving as a punctuated equilibrium.

Another weakness of Kuhn’s work is that it does not do justice to the way that empirical discoveries may cause unanticipated theoretical revolutions. In his model, major theoretical innovations are the products of brilliant practitioners who see the field in a new way. But this does not accurately describe what happened when, say, DNA was discovered. Watson and Crick worked within the known chemical paradigm, and operated like proper Popperians in brainstorming and eliminating possibilities based on the evidence. And yet the discovery of DNA’s double helix, while not overturning any major theoretical paradigms, nevertheless had such far-reaching implications that it caused a revolution in the field. Kuhn has little to say about events like this, which shows that his model is overly simplistic.

I must end here, after thrashing about ineffectually in multiple disciples in which I am not even the rankest amateur. What I hoped to re-capture in this review was the intellectual excitement I felt while reading this little volume. In somewhat dry (though not technical) academic prose, Kuhn caused a revolution still forceful enough to make me dizzy.

View all my reviews