Pluralism and Responsibility in Post-Modern Science

Citation preview

Pluralism and Responsibility in Post-Modern Science Stephen E. Toulmin Science, Technology, & Human Values, Vol. 10, No. 1. (Winter, 1985), pp. 28-37. Stable URL: http://links.jstor.org/sici?sici=0162-2439%28198524%2910%3A1%3C28%3APARIPS%3E2.0.CO%3B2-I Science, Technology, & Human Values is currently published by Sage Publications, Inc..

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/sage.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Fri May 18 09:08:25 2007

Pluralism and Responsibility in

Post-Modern Science *

A seminar with Stephen E. Toulmin

Stephen Toulmin: Having been, I suppose, the only living person to have studied at the feet of both Dirac and Wittgenstein, I retain a strong sense of the spirit of Aristotle's account of practical wisdom. That is, true progress in practical wisdom can only follow if one goes into a particular field and analyzes the nature of the problems that arise. Only in that way can one begin to understand what is mysterious, what needs explaining, and what needs investigating in the terms that are appropriate for that particular field. The idea that there is a single method for attacking scientific problems for scientists of all kinds does not get us much further than the notion that there is a single set of medical principles that will apply to all cases a physician may encounter, or a single set of moral principles that can give us answers to the human problems with which we find ourselves confronted. There is probably no one method appropriate even to inquiries in a given scientific field at different stages in its development. The recurrence of this case approach or pluralistic approach to practical wisdom as a motif in the history of philosophy is striking. Albert Jonsen of the University of California Medical School has pointed out that Thomas Aquinas, in the Summa Theologica, draws a telling comparison between the tasks that a wise confessor and a prudent physician face when they seek to arrive at some moral or medical diagnosis. Aquinas says that the wise confessor, like the prudent physician, will suspend judgment until he has satisfied himself that he knows all about a particular case with which he is concerned. He will go astray if he goes into a situation with fixed ideas about what the case is and what he said about it. I find the Aquinas point of view linking moral Dr. Toulmin is Professor of Social Thought and Philosophy, University of Chicago, Chicago, IL 60637.

judgment and technical (or clinical) judgment is still very much on the mark. In medicine, the question of how a physician sees a case and manages it, even in its technical aspects, requires that the human implications involved in following one course of treatment rather than another be weighed. So the practice of medicine, even in a technical sense, becomes a moral exercise. And the line between its moral and technical aspects becomes progressively blurred. I want to explore with you today how such considerations bear on differing concepts of scientific progress, why those concepts affect the question of social responsibility, and how they carry over into the activities you support at the National Science Foundation. I have been thinking about the intellectual foundations of science policy at least since Edward Shils commissioned and published a seminal series of articles in Minerva in the early 1960s.' My concerns are not simply a by product of my interests in the philosophical aspect of the theory of science. For the question of scientific choice and the respective roles of the scientist, the administration, the electorate, and the electorates' representatives in Congress, is very much bound up with different ideas about the nature of scientific progress, and is therefore central to the mission and operation of the NSF. To understand something about how and why concepts of scientific progress have changed over the past 300 years, it is useful to reflect on some of the implications of the vast expansion in the scope of scientific inquiry since Laplace asserted at the end of the 18th century that natural philosophers could aspire to become omniscient cal' An edited verslon of a National Science Foundation Executive Roundtable on The Nature of Scientific Progress, held on 8 March 1984 and supported in part by NSF Contract #84SP-0450. The views of the speaker and discussants are their own and do not necessarily reflect the polic~esof the National Science Foundation.

Published by rohn Wiley & Sons for the Massachusetts Institute of Tel:hnology and the President and Fellows of Harvard College CCC 0162-24391851010028-10$04.00 Science, Technology. d Human Values, Volume 10, Issue 1, pp. 28-37 (Winter 1985)

Toulmin: Pluralism and Responsibility in Post-Modern Science culators. This expansion in the scope of scientific inquiry has led to a transformation in how philosophers look at the foundations of science and has led in a direction very different from the view that tried to fit all of the sciences into a classical Newtonian framework. I have referred elsewhere to the fruits of this transformation as "post-modem" science.' In that essay I argue that the "modern" science that developed during the 250 years from about 1650 onward has begun, in the course of the present century, to be superseded by "post-modern" science. As a result, scientists have, in certain respects, broken through bonds and restrictions that were placed on the scientific enterprise by its original founders. What was the character of those initial restrictions? The self-proclaimed new mathematical and experimental philosophers of the 17th and 18th centuries simplified their intellectual tasks by adopting one very particular theoretical stance that came to dominate the approach and worldview of both the 18th-century classical, Newtonian physicists, and the chemists and physiologists who subsequently modeled their investigations on classical physics. Many social scientists today make the mistake of trying to adopt that same stance. It involves a conception of the scientist's situation vis-a-vis his or her objects of study as being detached and external. For the purposes of scientific investigation and theorizing, it was the scientist's duty to place himself and his rational speculations outside the world of nature that was his subject matter. As one result, the scientific worldview that was gradually built up depicted nature as a self-contained, deterministic set of mechanisms from which the processes and forces of humanity were excluded or exempted. The ultimate distillation of that view, as I have already mentioned, was epitomized in Laplace's contention that scientists could, in principle, compile sufficient information about the entities in the universe and the forces between them to determine the precise past and future course of any system. In other words, scientists could aspire to become omniscient calculators, and progress in science could therefore be measured in terms of how near we were to that ideal. During the 20th century, the momentum of scientific change obliged scientists to change their intellectual stance, together with their associated attitudes, methods, and criteria of objectivity and progress. To the extent that scientists cling to

29

the pure spectator approach, they unavoidably narrow the range of topics and activities that can be brought within the scope of science. If, on the contrary, the scope of scientific understanding is to reach beyond the limitations imposed 300 years ago, scientists must learn to see themselves not just as spectators looking on the world of nature from outside, but also as participants with roles to play from within that world. This sort of approach is already characteristic of a number of fields such as psychology and ecology, for example. More generally, the worldview that is in the course of being developed as scientists recognize their roles as participants will, I believe, be correspondingly less mechanistic and deterministic. This picture will reintegrate humanity with naturethe human observer, as agent, with the natural processes that he both studies and influences. The emergence of post-modern science has several implications for scientific activity, for our concepts of scientific progress, and therefore, for science policy. One is that the old positivist idea that all the sciences have to be based on a single set of methods is no longer viable. Another is that since the scientist-as-spectator option is no longer open to us, neither is the assumption that science is value-free or that scientists bear no responsibility for the social consequences of their work. Post-modern science must be increasingly bound up with social, political, and ethical considerations-just as the technical and moral aspects of medicine can no longer be disassociated. Let me give an example of how these considerations bear on the development of the social sciences, and, to some extent, the behavioral sciences. One of the great obstacles to genuine progress in these disciplines, especially in the United States, has been the conception that, in order to attack a set of social problems at a fundamental level, one need only apply a single method or a single set of procedures that can be learned elsewhere. For these disciplines, the "elsewhere" from which this method is learned has not been the examples of scientists in other fields. Rather, social scientists have tried to model themselves on what they have read in philosophy books about what science ought to be. Too often, those accounts have been insufficiently checked against the actual experience of working scientists, who by and large are no longer guided by a strict classical Newtonian approach. As long as the human sciences continue to think that by studying the philosophers of science they

30

Science, Technology, d Human Values-Winter

can get some general characterization of what they ought to be doing and how they can make themselves look respectable in the eyes of their colleagues in the "hard" sciences, they will go astray. Too many people in social science fields used to believe that the only way of demonstrating that they were respectable scientists was to post axiomatic theories. When Thomas Kuhn relieved them of the feeling that t h s activity was essential, they became preoccupied with formulating paradigms which were appropriate for their sciences. This sloganeering, this attempt at a priori methodologizing erects a formidable obstacle to a constructive confrontation between human scientists and the problems that are waiting for them to explore. But during the past 20 years or so, as implicit faith in the inevitability of scientific progress has weakened, some hard scientists have also begun to succumb to making claims of a different nature that I also think have very little to do with scientific understanding or scientific progress. For example, some hard scientists-and many more social scientists-have tried to link scientific progress with the program to improve our capacities for forecasting. One of the defects of the Laplacian image is that it tends to perpetuate the idea that the best measure of our scientific command is our capacity to make predictions. This is a natural error, considering that the best selling point of Newton's planetary astronomy was that it gave us equations with the help of which we could make predictions that appear to be wellgrounded. But the planetary system is a quite exceptional system. No other system of entities in the natural world lends itself to prediction in the same way. A fruitful line of inquiry that has not been sufficiently explored is whether and how the demand for prediction has damaged our conception of what the task of the sciences should be, particularly in the realm of the human sciences. As early as 700 B.C., even before Thales, Anaximander, and Anaximenes were beginning to produce generalizations about the composition of the universe, people knew what procedures were needed to make predictions concerning planetary motion or tidal motion. Our ability to make predictions in actual concrete situations did not become a scientific issue until some time in the 18th century, when the Newtonian account of the planetary system failed in certain respects to do justice to a wellestablished technology about how to predict pla-

1985

netary positions. In fact, the real triumph of Laplace's reworking of Newtonian planetary theory at the very end of the 18th century was his ability to show that Newton's theory led to predictions that were as good as those that were then being made without the benefit of that theory. The lstinction between scientific understanding and technical forecasting capacity is particularly important to understand for the social sciences. For instance, there is a great confusion in the public mind-and also among many economistsabout whether economists should be able to forecast or not. Should economists be able to predict what the rate of inflation is going to be next year? Because the rate of inflation will depend on what a very large number of human beings do between now and next year, there is no way in which economists can give us a rational, well-founded argument that would yield anything but a hypothetical conclusion of the form: Unless we do "XI1' the rate of inflation next year is going to be "Y1' percent.

Susan Cozzens (Policy Analyst, Division of Policy Research and Analysis, NSF): One traditional definition of scientific progress has been cast in terms of explanation, prediction, and control. If an increase in predictive capacity is not an important criterion by which we measure progress, then how do we tell if we are progressing at all? Stephen Toulmin: There is an ambiguity in the word prediction as it has come to be used in discussing the merits of scientific theory. If you ask a working scientist who is affected with a positivist terminology about what he means by saying he is interested in the question of what a particular theory predicts, his answer will be something of the form: If I do "XI" then "Y" is likely to happen. That is, "predict" no longer means "forecast." It no longer means that at 8:33 tomorrow evening, there is going to be an event of such and such a kind. In the years following Newton, the idea grew that what one wanted to do was to extend Newton's program as far as possible. The French encyclopedists Saints-Simon and Laplace, for example were seeking to develop the foundations and methods for a social science in the late 18th century. They competed to see which of them would become the Newton of the social sciences. They thought that, to do so, they had to develop a picture of how the planetary system works. That

Toulmin: Pluralism and Responsibility in Post-Modern Science program was kept alive in the face of increasing evidence of the formidable obstacles confronting by reinterpreting the word "prediction" so that it no longer meant "forecasting."

JosephYoung (Applied Research Coordinator, Division of Behavioral and Neural Sciences, NSF): Your distinction between the technology of forecasting and scientific understanding of the basis of those technologies has important implications for what we do at the NSF. In economics, for example, there are theoretical economists and there are econometricians who, I gather, use very sophisticated, empirical kinds of methodologies to make much more detailed predictions than one can derive from basic theories. In psychology and sociology, one can often make incredibly detailed if/then predictions based on complex regression models that say little about underlying process. When we receive proposals to do that sort of thing, we tend to classify them as "technology" and give them a relatively low priority for funding. We are more positively inclined toward the person who proposes to make modest predictions based on a theoretical underpinning that seems to promise an advance in basic understanding. Perhaps, in the context of your opening remarks about Aristotlets definition of practical wisdom, we ought to take more seriously the attempts to improve forecasting capacities, even if only on an ad hoc basis. Toulmin: I might also argue the converse point. One of the best pieces of evidence that we have achieved scientific understanding in a particular field is the ability to explain why it is so hard to make predictions. We can say that meteorology, for example, is a science because meteorologists can explain what the factors are that make meteorological forecasting comparatively difficult. The fact that we cannot predict the weather is not evidence that meteorology is not a science. On the contrary, the fact that we can explain why it is so hard to make meteorological predictions is evidence that meteorology is a science. Alexander Morin (Director, Division of Research Initiation and Improvement, NSF): I would like to extend our discussion to consider the extent to w h c h philosophers' loss of faith in the classical Laplacian view of prediction has any bearing on the way science is conducted. I contend that, in the daily work of a chemist, an ecologist, or a

31

sociologist, questions of methodological foundations and paradigm shifts simply do not arise. I wonder whether it is realistic to proclaim a philosophical revolution without pointing to what its clear implications are. You seem to see somethng important following from that philosophical revolution, and I would like to know what it is.

Toulmin: In my ideal ethnographic playground, the bar near the University of Chicago, one can raise these kinds of questions and scientists are happy to play with them over a martini. However, I agree that they are not concerned with them back at the lab. These questions actively arise for working scientists at times that Kuhn would call crises-on those occasions when questions about the strategy of scientific inquiry become active questions, as they did in theoretical physics at the turn of the century. If you look at the literature of physics between 1890 and 1910, you will find top-notch physicists expending a great deal of energy discussing these issues. And the reason is because at that point there were genuine doubts about the strategic directions in which the development of the basic ideas of the physical sciences should move. Turning to a more current example, there is still, among professional biologists, real debate about the intellectual place and significance of evolutionary biology. It is similar but perhaps with deeper roots than the controversies that used to go on among different sects of psychology. I happen to be passionately committed to the importance of the kinds of historical methods that Darwin developed, particularly their populational aspects. But this is a contentious method among many prominent molecular biologists. I suspect they dismiss it because they do not really understand the unique aspects and contributions of evolutionary biology. This reductionism, this assumption that there is no real value to an historical, ecological approach, is a continuation of the view about the intellectual mission of science that was born in the heyday of the classical, Newtonian conception of science. It also suggests that some first-rate biologists continue to take a positivist attitude and worry that biology can only be a respectable science if it is modeled after physics. Alexander Morin: Even though your examples from physics and biology are telling in themselves, I am still not convinced that the philosophcal issues

32

Science, Technology, et) Human Values-Winter

associated with abandoning the old Laplacian view and accepting the theorist as a participant in the system he is thinking about affects the conduct of modern science very deeply. But leaving that question, I am also not convinced of the part of your thesis that insists that, given the institutional, social, economic, political, and moral implications of what scientists are now doing, they will inevitably be involved with issues of ethics, politics, and society. Why is there a necessary connection between the philosophical view that the scientist-as-spectator option is no longer viable and the thesis that scientists must henceforth be more socially responsible?

Stephen Toulmin: I admit that there is no necessary entailment between any statement having to do with the spectator-participant relationship in the development of theoretical attitudes, and statements about changes in the political relationship between scientists and the larger society. There are interesting historical associations between these two developments-one internal to the self-conception of what the task of science is, and the other external in the relations between the community of science and the rest of the community. I do think that there is a good deal of conceptual overlap between the circumstances in which these two parallel sets of changes are taking place. For example, it is certainly the case that ecologists cannot go very far towards gaining basic understanding of the environment without considering human implications, although I suppose that particle physicists and physical chemists still can. Among the ecosystems that ecologists must study are ecosystems into which human action itself enters as one of the theoreticallysignificant contenders. Thus, human ecology becomes part of ecology as a science. And questions about the effect of introducing arable farming techniques into high erosion, tropical forest areas, for example, become a s i p f i c a n t part of the agenda of ecology as a science. Robert Wall (Head, Ocean Sciences Research Section, Division of Ocean Sciences, NSF): You can look at these different sciences on the basis of how close society is, in all its aspects, to the extremes of their basic-applied spectra. For all the environmental sciences, even the basic end of the spectrum is very closely tied up with social concerns, certainly as compared with mathematics and physics. Most of the problems at which en-

1985

vironmental scientists in the world community are looking are socially important. The whole broad spectrum is really our raison d'etre.

Stephen Toulmin: That approach can yield useful insights, although the issue is really somewhat more complex. Quite apart from the distinction between "basic" and "applied" is the distinction between "abstract" and "concrete." The historical sciences such as the earth sciences, the ocean sciences, and, to some extent, evolutionary biology are concerned with the here and now in a very concrete way. This perspective is different than for those sciences in which the concern is with what can be said in general, abstract terms about phenomena of a given kind. The historically-oriented fields explore phenomena for their own sake and are concerned with rather concrete issues. H. T. Huang (Program Director, Alternate Biological Resources, Division of Physiology, Cellular and Molecular Biology, NSF): Regardless of whether there is a causal or logical connection between the philosophical basis of science and social responsibility, it seems to me that scientists have been acting in a more socially responsible manner during the past 20 years than they did, say, before World War 11. A case in point is the 1974 Asilomar conference that led to the formulations of guidelines on recombinant DNA research. Since the aftermath of Asilomar created so much alarm among biologists and some members of the public, do you think that what the scientists did at Asilomar was the correct course of action? Are there any alternatives to that type of action in similar situations in the future?

Stephen Toulmin: I do not think that the public alarm that came in the wake of Asilomar could have been prevented, and I t h n k that the scientists had a good sense of what type of alarm might have risen had they not acted. The fact that they took a responsible initiative meant that it was easier to deal with the alarm once it was aroused. If the recombinant DNA people had not raised this issue on their own account, it would have been raised in a much more irresponsible way by other people. In that case, the public alarm would have been much greater and harder to deal with. Because then the molecular biologists could have been presented as being not only dangerous scientists, but mad ones. As it was, they were able to raise the issue of potential risk in their own

Toulmin: Pluralism and Responsibility in Post-Modern Science terms and in a way which made clear the fact that even though they might be dangerous scientists, they certainly were not mad ones. H. T. Huang: Do you feel that Asilomar serves as a useful guide for similar situations? Stephen Toulmin: When I was working with the National Institutes of Health's National Commission for the Protection of Human Subjects of Biological and Behavioral Research, there was a certain amount of discussion about whether continuing institutional machinery should be established as an early warning system for issues of potential risk. It was suggested, for instance, that the National Academy of Sciences might take as one of its charters the task of having a group of people who would monitor science for developments that had a potential for public alarm and for substantive risk. I still think that was an interesting suggestion. But let me broaden the scope of this discussion. One of my feelings about the recombinant DNA initiative was that it was anomalous. The research biochemists were being harrassed at a time when Allied Chemicals was pouring more toxic chemicals into the James River every week than all the recombinant DNA biochemists in the world could possibly produce in 100 years. The question of priorities has to be faced. The discharge of toxic chemicals from industry is a more active threat to human welfare than the recombinant DNA people are likely to pose in a long period of time. Joseph Young: It has always struck me that there was excessive focus after Asilomar on one side of the set of arguments concerning risk. Insufficient attention was paid to the consequences of the lost research time in terms of human welfare. We now see products resulting from recombinant DNA research w h c h are about to be marketed and which have considerable potential for alleviating human suffering. In hindsight, did the Asilomar initiative retard-unacceptably-both scientific and social progress? Stephen Toulmin: The ~ r o b l e mof how to introduce the scientific and social costs of retarding research into any cost-benefit calculus is important. Here is another example in which the problem of balancing risks against the loss of potential benefits comes up in a way that is more acute than the way it comes up with recombinant DNA.

33

That issue has to do with the allegations that the procedures of the Food and Drug Administration (FDA)for testing a new drug are too demanding, and the consequent losses to human welfare excessive. The FDA will point to thalidomide as a counterexample. But recognition of the risks of thalidomide were serendipitous rather than a result of an elaborate testing procedure. Something rang a bell in the head of a particular FDA inspector. It was not the fact that the testing took seven years rather than a year and a half that led her to spot the problem. There are people who argue that, thalidomide apart, Europeans have not been damaged by the fact that their drug-testing procedures are often more expeditious than the drugtesting procedures of the FDA. H. T. Huang: In the thalidomide case, problems arose because the drug companies were not forthright in coming out with evidence of risk. They had the information, but they were not releasing it. Unfortunately, this situation seems to happen time and time again. Joseph Young: Perhaps this example points to a similar and significant facet of the recombinant DNA case. Once the results of scientific research reach the marketplace, a different set of forces operate than at the research stage. In the Asilomar case, the research scientists were involved early in the discussions of implications. By so doing, they reserved for themselves a place in the decisionmaking process. Whereas more often in the applications of scientific results, the research scientists abrogate their decisionmaking power because the bomb does not burst until it's on the market and things are happening that scientists cannot control. In the Asilomar case, the scientists were controlling their own ethcal destiny, whereas in the James River case, a different set of forces were acting. Paul Chapin (Director, Linguistics Program): I am concerned about the scenario that unfolded after Asilomar. Many scientists were so embittered by the results that they swore they would not willingly take such an open, responsible position again. William Blanpied (International Policy Specialist, STIA Directorate, NSF): Some of the prime movers-Maxine Singer, for example-said she would not touch another public policy issue for 100 years.

34

Science, Technology, et) Human Values-Winter

1985

David Baltimore and James Watson have also been very negative about the aftermath of A ~ i l o m a r . ~

Stephen Toulmin: I think those among the Asilomar prime movers who say "never again" are wrong. I think they should still feel happy about what they did. Hopefully, there will always be some who feel that their responsibility for the larger commonwealth transcends their responsibility to science. This affords a nice illustration of a point I like to make about the value of exploring the psychology of scientists. Many people enter science for a quiet life; so they can do the things they passionately care about and will not have to be involved in the messy world of politics, public affairs, public opinion, or dispute. In an important sense, science is a consensus-building activity. For those who have lived in a consensus world, dealing with the messy world of headlines and imputed motives is very disagreeable. Otto Larsen (Senior Associate, Social and Behavioral Sciences, NSF): May I return to a more internalist question stimulated by your opening remarks? While post-modern science as you define it is pluralistic, it seems to me also to entail larger and larger units of scientific concern. Post-modem science seems to be anti-disciplinary. Is that a reasonable conclusion? Stephen Toulmin: I suspect your question is connected with the fact that we find ourselves more and more concerned with enlarging rather than localizing the boundaries of inquiry. It seems to me that, in the study of systems, one must ring in theoretical conceptions from a whole range of disciplines. One cannot really be a good geologist without knowing a good deal of physics and chemistry, for example, and one may have to know many other things outside of science as well. We can tie this up historically with the whole Bacon-Newton dichotomy. Bacon said that if only we pursue certain lines of inquiry in natural philosophy, we shall eventually be in a position to promote human welfare. I do not know how long he thought it was going to take, but in fact it took 300 years. The Professor of Natural Philosophy at Edinburgh, John Playfair, writing in the Encyclopedia Britannica in about 1816, bewailed the fact that, even after 200 years, Bacon's worthy prophecy still had not paid off. During the 300 years or so between Bacon and the late 19th century-when science really did

begin to show some signs of paying off-progress was made toward laying the basis for fulfilling the Baconian program. It was made by abstracting out and creating a division of labor for dealing with those theoretical issues on which progress had to be made before science, as Bacon conceived of it, would be in a position to attack concrete problems concerning human welfare. One of the things that is happening now is that, in fields like fundamental physics, basic chemistry, general physiology, and cell biology, we have reached the point at which-however imperfect our understanding of these different systems-we possess sufficient knowledge to be able to start cashing in some of those promissory notes that Bacon issued in the late 16th century. But to do so, we must move away from a view of science in which each of the abstract disciplines created in the earlier process of the division of intellectual labor remains self-sufficient. We must look more at the concrete issues that arise when we really want to deal with problems within which the Baconian payoff can actually take place and be prepared to use ideas from whatever theoretical disciplines have something to contribute to the solution of those problems.

George Brosseau (Project Manager, Office of Interdisciplinary Research, Engineering Directorate, NSF): You cannot give predominance either to disciplinary science or to the problem-oriented approaches because each is incomplete in itself. Disciplinary and problem-oriented approaches have to coexist. What you are talking about in multidisciplinary and interdisciplinary approaches are studies that look backward from the need and bring the necessary intellectual resources to bear on the solution of specific problems. But the perception of need is a function of what we already know. It is a result of knowledge that is most often accumulated from research in disciplinary fields. When we know enough, then that accumulated knowledge allows us to perceive a social need that we may be able to do something about. By "need" I mean an externally perceived need: the need to build better automation, the need to cure a certain kind of disease. Those needs may not be perceived as such until we know enough about a system to recognize them as needs. If you do not know anything about bacteria, then you do not perceive cures for infectious disease as a need that can be addressed by scientific knowledge. But dealing with such needs does not add significantly to our knowledge base. For that

Toulmin: Pluralism and Responsibility in Post-Modern Science reason, we may not be able to perceive the next set of problems after we have solved the last ones. That is why we must build the knowledge base concurrently. Otto Larsen: Let me ask you to address the unityof-science issue. Do you contend that when the post-modern sciences finally emerge they will be very pluralistic, will not involve a unifylng theory, will not have a great deal of commonality in methodology, and that the agenda of the sciences will be very diverse? Stephen Toulmin: That may well be a legitimate inference to draw from the argument I have presented. Part of my stance about where science is going is that it will be more pluralistic-not chaotic, but pluralistic. In that future, an understanding of the mutual relevance of the different parts of science will have to consist not in understanding how they all should succeed in approximating more or less a common ideal (which was part of the classical view) but in understanding how each of them has come to develop in a way that is appropriate to a particular set of problems that are relevant to it. The question of relevance would be one of topical relevance rather than of analytical relationships. Otto Larsen: Is there anything in the knowledge game itself that can help make choices in the allocation of resources among the several sciences? Can science address the question of priorities, or must that be based on an external set of factors? Stephen Toulmin: The more narrowly you define the priorities question, the more powerful the professional authority of the people who have technical command of the field becomes. The larger the scope of the question, the less professional monopoly the scientist has in the decision. Between mathematics, physics, chemistry, oceanography, and the social sciences, for example, the priorities question seems to be a straightforward political question. The National Academy of Sciences can provide indispensable material for public discussion of issues related to priorities among various disciplines, but the issues themselves probably have to be resolved on political grounds. Michael Polanyi wrote a particularly good paper called "The Republic of Science" as one of a series on scientific priorities that Edward Shls published in Minerva in the early 1960sS4Polanyi argued

35

that the structure of science was a bit like the structure of Switzerland. He likened each individual science to a Swiss canton. Then he asserted that, as in Switzerland, the amount of authority each canton surrenders to any kind of c,entral authority should be minimal. Alvin Weinberg, in another article in ~ i n e r v a , " described Polanyi's metaphor as a bit too crude. He argued that what one finds is that people in neighboring disciplines do have something useful to contribute to judging priorities within a given discipline insofar as some parts of organic chemistry, for example, are more likely to have important implications for biochemistry than others. They are therefore more likely to have implications for genetics. So that to some extent, acccording to Weinberg's argument, one has to allow people other than those who are narrowly concerned with a particular discipline to have a say in determining resource allocation priorities. I agree with Weinberg that the broader the range within which the judgment of value has to be made, the larger the community of people who have to be allowed to have a say within it. But of course both the Weinberg and the Polanyi arguments take government support of science for granted. I find myself talking to a lot of people nowadays who really seem to believe that science exists only by virtue of the fact that public funds are made available to it. But I grew up in a period when all good scientists knew that if they took public money for their research, then their fellow scientists would not trust the results. I always feel a little sad when I hear questions about intellectual priorities in science (or, indeed, about the possibility of doing science at all) sliding over into questions of resource allocation. Otto Larsen: In your scheme, are there different kinds of money? Is the National Science Foundation's money different, perhaps, than other Federal money? Stephen Toulmin: I am prepared to answer that question in either of two contexts. One context is that there should be government agencies prepared to fund science that is not mission-oriented. A second, stronger context is that it is an excellent thing that the government should recognize that the advancement of knowledge for its own sake is a legitimate mission. I would rather see it put the latter way. I still cleave to the argument Vannevar Bush advanced when he maintained that the improvement of human understanding should

36

Science, Technology, 4 Human Values-Winter

be to the late 20th century what the building of a cathedral was in the Middle Ages.

Otto Larsen: George Keyworth says something like that today when he refers to the support of basic research as a national trust. And that thought brings me back to a question that Alex Morin raised: If scientists are unaffected philosophically, how else are they affected and how will the agenda of research be altered by a post-modern science that involves more interaction? What would the NSF do differently from what it is now doing? What are the implications for our procedures? Stephen Toulmin: One way in which these questions become of active importance to the NSF is in those areas in w h c h there really is a continuing dispute about the relative levels of fundamentalness of different kinds of inquiries within a larger field. As I noted earlier, there are still biologists who believe that if you really understand molecular biochemistry, you do not need all this anecdotal stuff about fossils. Therefore, by implication, a government that seeks to support pure science will give a lot of money to molecular biochemistry, and will leave the ecologists and evolutionary biologists to scratch around for a few pence from private foundations that are run by millionaires who have a taste for that kind of eccentricity. Otto Larsen: But if the people who are in charge at the NSF are the ones who maintain the classical conception, those of us who opt for pluralism are going to have problems. Progress in our fields will be deterred by a lack of resources. William Blanpied: The idea that methodologies in different scientific fields do not have to be related to some central norm could have severe implications for some fields at the NSF. Debates in the late 1940s over the issue of whether the social sciences should be explicitly written into the National Science Foundation Act turned in large measure on the question of how "scientific" those disciplines were. That question is still largely unsettled in the minds of many "hard" scientists, as is evidenced by the priorities and budgets for the social and behavioral sciences within the NSF. The fact that the NSF has been able to support a wide variety of science and engineering disciplines has often been justified on the basis of the implicit premise that there is some commonality among those disciplines. If we now proclaim loudly

1985

and publicly that the social sciences are "scientific" but not in the same way as physics, the problems of the social sciences within the NSF might grow even more severe.

Stephen Toulmin: The distinction between pluralism and chaos is centrally important here. Some strict logical positivists have argued that pluralism really does imply chaos. In essence, they argue that if one accepts what Thomas Kuhn says about the character of scientific revolutions, then we will not be able to exclude the astrologers. That is just not true. The crucial point, as far as your procedures at the NSF are concerned, comes at the point of interaction between the program officer and the peer review panel. The task of acting as some kind of holder of the balance of equity between the different groups who enter into this competition for funds is a very important one. William Blanpied: In your post-modern world, what do you suggest should be criteria for differentiating between "real" science of the kind we support at the NSF and other, nonlegitimate claimants to scientific status? Stephen Toulmin: Ultimately one has to agree that those decisions come down to political decisions in the broadest sense. I remember going to a meeting in Vienna of the European governments to discuss science policy. There, even problems of the sheer translation of the word "science" created problems because, for instance, in Germany, the public funding of seminaries was included under the heading of Wissenschaftpolitik. As far as German politics is concerned, you cannot rule out the questions of what seminaries are to be supported. But the question of whether, having a general policy that gave financial support to seminaries, any government would feel it would therefore have to fund the Reverend Moon's Unification Church as well would immediately arouse acute political controversy. This is the point in making a distinction between pluralism and chaos. Politically speaking, a lot of people are unhappy about pluralism because they see it as a slippery slope, an invitation to use tax money to finance some types of extreme, wayout things. They feel so horrid that they have got to go back to nanny. One cannot be completely logical about the problem of separating "real" science from other systems. One simply must have some confidence that pluralism need not imply chaos in the public mind.

Toulmin: Pluralism and Responsibility in Post-Modern Science

And that point brings me back to my opening remarks about why trying to understand the differing conceptions of scientific progress has such an important bearing on the intellectual foundations of science policy, and why I believe such an understanding is so central to what you are trying to accomplish at the National Science Foundation.

2. 3. 4.

Notes 5. 1. Stephen Toulmin, "The Complexity of Scientific Choice: A Stocktaking," Minerva, Volume 2, Number

37

3 (Spring 1964): 343-359; and "The Complexity of Scientific Choice: Culture, Overhead, or Tertiary I n d ~ s t r y ? ,Minerva, ~' Volume 4, Number 2 (Winter 1966): 155-169. Stephen Toulmin, "The Emergence of Post-Modern Science," in The Great Ideas Today-1981 (New York: Encyclopedia Britannica, Inc., 1981 ), pp. 69-1 14. avid Baltimore, "Limiting Science: A Biologist's D Perspective," Daedalus, Volume 107, Number 2 (Spring 1978): 37-46. Michael Polanyi, "The Republic of Science: Its Political and Economic Theory," Minerva, Volume l , Number 1 (Autumn 1962): 54-73. Alvin Weinberg, "Criteria for Scientific Choice," Minerva, Volume 1 , Number 2 (Winter 1963): 159171.