Theory and Reality

by Peter Godfrey-Smith

To close this chapter, I will describe an example that is very far from Kuhn's usual cases-an example that is on the periphery of science, but which I think illustrates Kuhns insights well. During the z98os and 199os, there was a lot of excitement about a new field known as "Artificial Life." The aim was to use computers to model the most basic features of living systems, in such a way that it might eventually be reasonable to say that the artificial systems were alive. I went to several "Alife" conferences during this time and watched the field develop. At the time that I write this, the field seems to have ground to a halt. Maybe it will revitalize. But the failure of the field in recent years seems to me to involve some very "Kuhnian" reasons.

  During the heyday of the movement, two or three pieces of work appeared that were strikingly successful. Perhaps the most impressive of all was Tom Ray's Tierra project, in which Ray was able to create open-ended evolution among self-replicating programs in a computer (Ray 199z). Work by Chris Langton and Steven Wolfram on the mathematical analysis of "cellular automata," simple systems in which local interactions between elements give rise to global, self-sustaining patterns, might be another case. And closer to the borderline with mainstream biology, there was Stuart Kaufmann's work on "the origins of order" in complex systems (Kauffman 1993).

  All of this was impressive work, and it pointed the way forward to a consolidation of what these imaginative individuals had done. But the consolidation never happened. At each conference I went to, the larger group of people involved all seemed to want to do things from scratch, in their own way. Each had his or her own way of setting up the issues. There was not nearly enough work that built on the promising beginnings of Ray and others. The field never made a transition into anything resembling normal science. And it has now ground to a halt.

  Another reason for the breakdown also relates to Kuhn. The field of Alife suffered from a kind of "premature commercialization." It was realized early on that some of the work had great potential for animation and other kinds of commercial art. At some Alife events, the climax of each talk seemed to be not some new theoretical idea, but a dramatic video. (I even heard speakers half-apologize before the video, as if they knew this was somehow not the right emphasis for their work.) For Kuhn, science depends on the good normal scientist's keen interest in puzzle-solving for its own sake. Looking outside the paradigm too often to applications and external rewards is not good for normal science.

  Further Reading

  Lakatos and Musgrave's collection Criticism and the Growth of Knowledge (1970) contains an excellent set of essays on Kuhn. A more recent edited collection is Norwich, World Changes (1993).

  Kuhn's collection of essays The Essential Tension (1977b) is an important extra source. Kuhn also wrote two historical books (1957, 1978). His later essays have been collected in The Road since Structure (z.ooo).

  Levy 1992 is a readable survey of Alife work. Many of the best Alife papers may be found in the collection Artificial Life II (Langton et al. 1992).

  I have argued so far only that paradigms are constitutive of science. Now I wish to display a sense in which they are constitutive of nature as well.

  THOMAS KUHN, Structure

  "Look," Thomas Kuhn said. The word was weighted with weariness, as if Kuhn was resigned to the fact that I would misinterpret him, but he was still going to try-no doubt in vain-to make his point. "Look," he said again. He leaned his gangly frame and long face forward, and his big lower lip, which ordinarily curled up amiably at the corners, sagged. "For Christ's sake, if I had my choice of having written the book or not having written it, I would choose to have written it. But there have certainly been aspects involving considerable upset about the response to it."

  JOHN H O R G A N, The End of Science

  6.1 Considerable Upset

  The most famous, most striking, and most controversial parts of Kuhns book were his discussions of scientific revolutions. They are the topic of this chapter. Why have two chapters on Kuhn? One reason is the continuing importance and great subtlety of his book. Another is that while the discussions of revolution are the most famous parts of the book, Kuhns analysis of normal science is just as important-and perhaps of more enduring significance. Sometimes it gets lost in the excitement about revolutions: hence chapter 5.

  Kuhn argued that some periods of scientific change involve a fundamentally different kind of process from what we find in normal science. The revolutionary periods see a breakdown of order and a questioning of the rules of the game, and they are followed by a process of rebuilding that can create fundamentally new kinds of conceptual structures. Revolutions involve a breakdown, but they are essential to science as we know it. They have a "function," Kuhn often said, within the totality of science. The special features we associate with science arise from the combination and interaction of two different kinds of activity-the orderly, organized, disciplined process of normal science, and the periodic breakdowns of order found in revolutions. These two processes happen in sequence, within each scientific field. Science as a whole is a result of their interaction and of nothing less.

  Kuhn seemed to divide science into units with strange boundaries between them. Looking within a period of normal science, you can easily distinguish good work from bad, rational moves from irrational, big problems from small problems, and so on. Progress is evident as time goes by. But all this ends with a revolution. In a scientific revolution, as in a political one, rules break down and have to be rebuilt afresh. If you look at two pieces of scientific work across a revolutionary divide, it will not be clear whether there has been progress from earlier to later. It might not even be clear how to compare the theories or pieces of work at all-they may look like fundamentally different kinds of intellectual activity. The people on different sides of the divide will be "speaking different languages." In the climax of his book, Kuhn says that workers in different paradigms are living in different worlds.

  6.2 Revolutions and Their Aftermath

  A revolution is a kind of discontinuity in the history of a scientific field. The suggestion that science has two modes of change, one of them dramatic and abrupt, does not itself have big consequences for philosophy, though it is interesting. The big issues depend upon what the two modes of change are like. And here there are two sets of issues that caused intense discussion. The first is how revolutions occur-what goes on within them. The second has to do with the relations between what we have before and what we have after a revolution.

  How do revolutions occur? We finished the previous chapter describing the transition from normal science to crisis. In Kuhn's story, large-scale scientific change usually requires both a crisis and the appearance of a new candidate paradigm. A crisis alone will not induce scientists to regard a large-scale theory or paradigm as "falsified." We do not find pure falsifications, rejections of one paradigm without simultaneous acceptance of a new one. Rather, the rejection of one paradigm accompanies the acceptance of another. But also, the switch to a new paradigm does not occur just because a new idea appears which looks better than the old one. Without a crisis, scientists will not have any motivation to consider radical change.

  All of Kuhn's claims about what follows what in scientific change tend to be qualified; he is describing the central and characteristic patterns of change, not every case without exception. But the idea that revolutions generally require crises raised some hard historical issues. Was there a crisis in the state of astronomy before Copernicus, or in biology before Darwin? Was there a state of disorder following an earlier period of confident work? Maybe. But taking another biological example, if the appearance of genetics as a science around 1900 was a revolution, it is very hard to find a crisis in the work on inheritance that preceded it. (Maybe Kuhn would regard this as a transition from preparadigm science to normal science, though that could not be said about most of biology around that time.) Some other twentiethcentury revolutions, such as the molecular revolution in biology, seem even less crisis-induced.

  I
n his 1970 "Postscript" to Structure, Kuhn qualified his claims about the role of crisis (181). He still maintained that crises are the "usual prelude" to revolutions. But even that claim is controversial. Kuhn's emphasis on crises sometimes seems driven more by the demands of his hypothesized mechanism for scientific change than by the historical data; Kuhns story demands crises because only a crisis can loosen the grip of a paradigm and make people receptive to alternatives.

  Suppose we do have a crisis, a period full of confusion and strange guests in the philosophy department. Then a new candidate paradigm appears, precipitating a revolution. Using my distinction from the previous chapter, what initially appears is a new paradigm in the narrow sense, an achievement that begins to inspire people and seems to point the way forward. More specifically, what is usually involved is that the new work appears to solve one or more of the problems that prompted the crisis in the old paradigm. The sudden appearance of problemsolving power is the spark to the revolution. Kuhn did not think these processes could be described by an explicit philosophical theory of evidence and testing. Instead, we should think of the shift to a new paradigm as like a "conversion" phenomenon, or like a gestalt switch. Kuhn also argued that revolutions are capricious, disorderly events. They are affected by idiosyncratic personal factors and accidents of history.

  One reason for the disorderly character of revolutions is that some of the principles by which scientific evidence is assessed are themselves liable to be destabilized by a crisis, and they can change with a revolution. Kuhn did not argue that traditional philosophical ideas about how theories should relate to evidence are completely misguided. He made it clear in his later work that there are some core ways of assessing theories that are common to all paradigms (1977c, 32I-zz). Theories should be predictively accurate, consistent with well-established theories in neighboring fields, able to unify disparate phenomena, and fruitful of new ideas and discoveries. These principles, along with other similar ones, "provide the shared basis for theory choice" (3 zz). (I should note that some commentators think these later essays change, rather than clarify, the views presented in Structure.)

  But Kuhn thought that when these principles were expressed in a broad enough way to be common across all of science, they would be so vague that they would be powerless to settle hard cases. Also, these goals must often be traded off against each other; emphasizing one will require downplaying another. Within a single paradigm, more precise ways of assessing hypotheses will operate. These will include sharper versions of the common principles listed above, but these sharper versions will not really be explicit "principles" anymore. Instead they will be more like habits and values, aspects of the shared mind-set of normal scientists imparted to them by their common training and common activities. There will also be some variation within normal science in how these principles are understood and acquired-Kuhn came to see this diversity as a strength of scientific communities as well. But the most important point here is that these sharper, more definite ways of assessing ideas are liable to change in the course of a revolution. In the next section I will give an example of this phenomenon.

  So we have two kinds of scientific change in Kuhn's picture, neither of which is what empiricist philosophies of science might have led us to expect. Change within normal science is orderly and responsive to evidence-but normal science works via a closing of debate about fundamental ideas. The other kind of change-revolutionary change-does involve challenges to fundamentals, but these are episodes in which the orderly assessment of ideas breaks down. Displays of problemsolving power have a key role in these fundamental transitions between paradigms, but the shifts also involve sudden gestalt switches and leaps of faith.

  In Kuhns treatment of revolutionary change, the distinction between descriptive and normative issues is very important. Kuhn uses language that suggests that not only are revolutions bound to happen, but they have a positive role in science. They are part of what makes science so powerful as a means for exploring the world (a "supremely efficient instrument" [1996, 169]). Different interpreters have very different reactions to this kind of talk. Some regard it as colorful and not essential to Kuhn's general message. I have the opposite view; I think this is central to Kuhn's overall picture. Science for Kuhn is a social mechanism that combines two capacities. One is the capacity for sustained, cooperative work. The other is science's capacity to partially break down and reconstitute itself from time to time. When a paradigm runs out of steam, there is nothing within the community that could reliably give science a set of directions for orderly movement toward a new paradigm. Instead, the goals of science are best served at these special times by a disorderly process, in which even very basic ideas are put back on the table for discussion, and a new direction eventually emerges from the chaos. This sounds strange, but I think it was Kuhn's picture.

  6.3 Incommensurability, Relativism, and Progress

  Kuhn said that revolutions have a "non-cumulative" nature; this is essential to his claims about the large-scale historical patterns in science. There is no steady buildup of some useful commodity like truth as science goes along. Instead, according to Kuhn, in a revolution you always gain some things and lose some things. Questions that the old paradigm answered now become puzzling again, or they cease to be questions. So we might want to ask, Do we usually gain more than we lose? In at least the middle chapters of his book, Kuhn seems to think there is no way to answer this question in an unbiased way (1996,109, 11o). Of course it will feellike we have gained more than we've lost, or we would not have had the revolution at all. But that does not mean that there is some unbiased way of comparing what we had before with what we have after.

  This question connects us to one of the most famous topics in Kuhn's work, the idea that different paradigms in a field are incommensurable with each other.

  What does "incommensurable" mean here? Most literally, it means not comparable by use of a common standard or measure. This idea needs to be carefully expressed, however. Two rival paradigms can be compared well enough for it to be clear that they are incompatible, that they are rivals. And people working within any one paradigm will have no problem saying why their paradigm is superior to the other, by citing key differences in what can be explained and what cannot. But these comparisons will be compelling only to those inside the paradigm from which the claim of superiority is being made. If we look down "from above" on two people who work within different paradigms who are arguing about which is better, it will often appear that the two people are talking past each other.

  There are two reasons for this-there are (roughly speaking) two aspects of the problem of incommensurability. First, people in different paradigms will not be able to fully communicate with each other; they will use key terms in different ways and in a sense will be speaking slightly different languages. Second, even when communication is possible, people in different paradigms will use different standards of evidence and argument. They will not agree on what a good theory is supposed to do.

  First let us look at the issues involving language. Here Kuhns claims depend on a holistic view about the meaning of scientific language. Each term in a theory derives its meaning from its place in the whole theoretical structure. Two people from different paradigms might seem to use the same word-"mass" or "species"-but the meanings of these terms will be slightly different because of their different roles in the two rival theories.

  Here I said "slightly different." Kuhn insisted he had a moderate view. Some critics have argued that a holistic view of meaning really has no way to make sense of these differences of degree, because it is not possible to say whether two terms have a "similar" role within two very different theoretical networks (Fodor and LePore 199z). So the critics argue that when holists about scientific language talk about "partial" communication and "slight" differences in the meanings of words, they are bluffing in order to hide the impossibly radical nature of their views.

  Neither the holists nor anyone else has had muc
h success in developing a good theory of meaning for scientific language. This is a confusing and unresolved area. However, a different kind of criticism of Kuhn is possible here. If incommensurability of meanings is real, as Kuhn says, then it should be visible in the history of science. So those who study the history of science should be able to find many examples of the usual signs of failed communication-confusion, correction, a sense of failure to make contact. Although I am not a historian of science, my impression is that historians have not found many examples of failed communication in crucial debates across rival paradigms. Scientists are often adept at "scientific bilingualism," switching from one framework to another. And they are often able to improvise ways of bridging linguistic gaps, much as traders from different cultures are able to, by improvising "pidgin" languages (Galison 1997). Scientists often willfully misrepresent each other's claims, in the service of rhetorical points, but that is not a case of failed comprehension or communication.

  The other form of incommensurability is much more important. This is incommensurability of standards. Here Kuhn argued that paradigms tend to bring with them their own standards for what counts as a good argument or good evidence.

  This topic was introduced in the previous section. There I said that Kuhn thought that although all scientific work is responsive to some broad principles of theory choice, the detailed standards for assessing ideas will often be internal to paradigms and liable to change with revolutions. For Kuhn, "paradigms provide scientists not only with a map but also with some of the directions essential for map-making" (1996, 109).

  One of Kuhns most interesting examples of this phenomenon involves the role of causal explanation. Should a scientific theory be required to make causal sense of why things happen? Should we always hope to understand the mechanisms underlying events? Or can a theory be entirely acceptable if it gives a mathematical formalism that describes phenomena without making causal sense of them? A famous example of this problem concerns Newton's theory of gravity. Newton gave a mathematical description of gravity-his famous inverse square law-but did not give a mechanism for how gravitational attraction works. Indeed, Newton's view that gravity acts instantaneously and at a distance seemed to be extremely hard to supplement with a mechanistic explanation. Was this a problem with Newton's theory, or should we drop the demand for a causal mechanism and be content with the mathematical formalism? Would it be scientifically acceptable to regard gravity as just an "innate" power of matter that follows a mathematical law? People argued about this a good deal in the early eighteenth century. Kuhns view is that there is no general answer to the question of whether scientific theories should give causal mechanisms for phenomena; this is the kind of principle that will be present in one paradigm and absent from another.