In response to:

Scientists as Servants from the June 28, 1979 issue

To the Editors:

In re David Joravsky’s treatment—one can hardly say review—of The Physicists [NYR, June 28], I regret that he ignored the content and distorted the stance of the book. Yet more distressing was the degree to which he twisted the historical record and, worse, utterly confused difficult questions of science and public policy. I do not wish to defend The Physicists, here, but I think it important to examine the grave issues of science in the American democracy that Joravsky has so beclouded.

Joravsky’s confusion is evident in his misreading of the scientific community’s political elitism. He takes this elitism to mean a preference among scientists for power to determine such major public policy issues as nuclear weapons. In fact, the elitism in question has centered on scientists’ frequent demand for federal research support kept free of political interference by insulation from direct governmental control. Over the years, political elitism has thus pitted the scientific community against fundamental principles of democratic government, with markedly varying outcomes for public science policy.1

Joravsky’s misreading of political elitism appears to suit a polemical purpose—to show that the elitists, for all their snobbishness of attitude, enjoy merely the illusion of power, that they are in reality politically impotent “technical servants.” After all, was not the Franck report’s plea against the unannounced atomic bombing of Japan as fruitless as Leo Szilard’s journey to James F. Byrnes for the same purpose? But Joravsky ignores the past political influence of scientists, e.g., in shaping public land and conservation policy, and overlooks the role scientists played more recently in bringing about a nuclear test ban treaty, environmental protection, and the demise of the anti-ballistic missile system.

Historically, the scientific community’s political potency has varied with disciplinary and institutional interests as well as with the changing political posture of its leadership.2 Instead of treating these historical considerations with thoughtfulness and subtlety, Joravsky wields a bludgeon, ignores the long political history of American science, and narrows his discussion primarily to nuclear policymaking in World War II. If there the Franck group was frustrated in the decision-making process, then democracy itself, according to Joravsky, must mean the subjection of scientists to “top bosses,” a “charmed circle of Policymakers” from which dissenting scientists are excluded.

Joravsky does not mention that Secretary of War Henry L. Stimson appointed a nuclear policy committee that included three accomplished scientist/administrators deeply involved with the Manhattan Project. He does discuss this committee’s separate Scientific Panel, consisting of Oppenheimer, Fermi, Lawrence, and A.H. Compton, which on June 16, 1945, took up the Franck report. The Panel had previously considered whether a harmless technical demonstration of the bomb might be devised that would convince the Japanese to surrender. Now it reiterated the conclusion that no such demonstration would match a surprise atomic attack for bringing the war to a speedy end. The Panel also observed that, in contrast to the Franck group, many other Manhattan Project scientists supported the military use of the bomb to save American lives and end the war quickly.

Inexplicably Joravsky asserts that the Panel “smothered Franck’s arguments with commentary.” Besides ignoring the problem of a technical demonstration and the division of opinion among atomic scientists, he omits the Panel’s—and the nuclear policy committee’s ultimate—endorsement of the Franck report’s urgent point: that the Soviet Union should be informed of the pending use of the bomb and invited to join the United States in the prevention of a postwar nuclear arms race.3 However much one may deplore Hiroshima, the full record hardly supports Joravsky’s claim of the exclusion of knowledgeable scientists from the decision-making process. And given the disagreement of these scientists on the use of the atomic bomb against Japan, that the decision went against the group with whom Joravsky agrees hardly demonstrates the political impotence of scientists as such.

Obviously what I mean here by political influence is the power to shape policy by participation in the process of decision. What Joravsky understands as political power is difficult to discern, unless he means that when a Szilard speaks, policy must follow. Since when Szilard did speak, Hiroshima followed nevertheless, Joravsky discounts the legitimacy of democracy. To guarantee the benign use of science, perhaps he would vest power in a court of right-thinking scientists. If so, who would decide what is right-thinking? And how are the right-thinkers to be checked, particularly if they turn wrong-headed? Ancient questions, to be sure, but to ignore them while denouncing the political processes of science in a democratic state seems intellectually self-indulgent.

To delineate the relationship of scientists to democratic power is not, as Joravsky insists, to anoint with rectitude scientists who serve that power. Does the point require belaboring that moral responsibility rests upon those scientists who supply the instruments—e.g., atom bombs—of national policy? Not unless one wishes to insist, as Joravsky seems to, that any scientist who contributes to the national military arsenal is ipso facto immoral.


In the late 1930s scientists took moral responsibility to mean bolstering the national defenses of the Western democracies against the aggression of fascist dictatorships. Contrary to Joravsky, Szilard proposed a conspiracy of silence among nuclear physicists, not to prevent any atomic bombs, but to block the development of a Nazi bomb. 4 Does Joravsky wish to attack as morally irresponsible the physicists whose radar helped defeat the German Air Force in Europe or whose proximity fuzes helped keep the kamikazes at bay in the Pacific? To be sure, he implies that on the atomic bombing of Japan there was only one morally responsible position to take. But he fails to explore whether that tortuous issue might have divided well-intentioned, moral men.

The moral choices were difficult in 1945; they have grown more so since, given the belief widely held not only among physicists that moral responsibility has entailed the double commitment of discouraging a nuclear arms race and simultaneously maintaining the national defense. Have physicist thus subjected themselves to conflicting imperatives, to moral schizophrenia? Can they realistically expect that by supplying weapons with one hand they might temper the arms race with the other? Complicated questions, these. More than anything, Joravsky’s essay makes clear that angry polemicizing does nothing to advance the search for answers.

Daniel J. Kevles

California Institute of Technology

Pasadena, California

To the Editors:

I think you owe me a review. Not a perceptive, penetrating, sparkling review—that might be difficult to come by—just a plain old review that says what is in a book and how good it is.

David Joravsky whom you hired to do the job [NYR, June 28] certainly made a mess of it. It is obvious that the more complex arguments of my books [Science in a Free Society and Against Method: Outline of an Anarchistic Theory of Knowledge] are beyond him. So he concentrates on asides and personal remarks, picks up a line here, a word there, and uses them as starting points for digressions and aggressive fantasies of his own. We have a stream of consciousness, we do not have a review.

For example, DJ spends more than half of his space on raindances, witchcraft, Chinese medicine. But raindances etc. occupy only three pages (out of 520) and my remarks on them are asides, not part of the main story. The reader never hears what the main story is and how the asides fit into it.

DJ has great difficulty recognizing an argument or the need for one. According to him I say that traditional medicine is perfect and Western medicine without merit. But I only claim that traditional forms of medicine have achievements not found elsewhere, that these achievements are often removed either by institutional means or by an appeal to “method” and that a more tolerant attitude (or, in the case of institutional intolerance, state intervention) is needed to guarantee the advance of medicine as a whole. DJ’s attempts to show the failures of traditional medicine and the successes of science are quite irrelevant to this claim.

DJ “readily” admits that science is not governed by an overriding rationality. That is nice of him but hardly to the point. What we need to improve our understanding of science are not “ready” admissions, what we need are arguments. I provide arguments. Yet DJ says that I preach arbitrariness and chaos and get “infuriated” when specific procedures are introduced. Now if that is my view, then why do I discuss Copernicus, Galileo, Einstein in such detail and over so many chapters? Why do I invite scientists to follow their example? Why do I devote an entire chapter to Aristotle, the arch logician, and defend him against critics? All these scientists proceeded in a reasonable manner, none of them recommended or practiced chaos, all of them were most interested in specific methodological suggestions. Their work looks like chaos only when viewed through the spectacles of an abstract rationality (just as Shakespeare looks like chaos to the followers of Castelvetro and Corneille). All this is explained in great and often tedious detail in my books. DJ does not notice.

As usual superficiality and incompetence go hand in hand with authoritarian posturing: astrology, witchcraft, Homeric common sense are “obsolescent.” I wonder how deeply DJ has thought about this matter. Most likely he has not thought at all but simply repeats scientific gossip. Now I have examined the gossip and have criticized it. Discussing the rise of Western rationalism I pointed out that the transition created more problems than it solved, that most of the problems are still with us, that they do not occur in Homer, that Aristotle was aware of this advantage and therefore adapted philosophy to common sense. I also show how “obsolescent” points of view can be turned and often were turned into instruments for the examination of modern theories and I give examples where such an examination shows them to be superior. There is a whole chapter on the incompetence of modern scientific objections to astrology. What is DJ’s reaction? I have already described it: Homer, astrology, witchcraft are “obsolescent.” One might as well talk to a Barbie Doll.


There are a few things DJ seems to have read, or heard about, but he doesn’t quite know what to do with them. Like a sophomore who has just absorbed some new facts now calls everybody who does not constantly talk about them an ignorant fool DJ accuses me of an “astonishing provincialism” because I deplore the prevalence of the sciences and fail to mention the humanities. But if he had studied the matter a little more closely he would have learned that the humanities are remnants of points of view he himself calls “obsolescent,” that they changed since the scientific revolution, that by now most of them have been voided of ontological content (in theology this is called demythologization) and can therefore no longer be used as critics of the scientific enterprise.

DJ seems to notice that not all is well with his review and he produces an apology: “I am not sure when he is clowning, or crying, or philosophizing.” This is a most interesting statement. It admits, what is obvious anyway, that DJ does not know how to recognize an argument. In a word—he admits that he is incompetent to review books of a certain complexity. Not that my books are all that complex; the rhetoric is restricted to small islands and the facts and arguments are announced and summarized at the beginning of each chapter, in a separate, italicized introduction. Yet DJ has not found a single one of them.

Reading the review I was occasionally struck by a sense of unreality: surely an intelligent man cannot willingly write such puerile trash; surely he must have been paid, or otherwise encouraged to do a hatchet job. This would not have been the first time that an intellectual or pseudo intellectual sold his typewriter to a powerful bidder. Whatever the cause—it is somewhat surprising that a reputable journal like yours saw fit to publish the result. But, having decided to pay attention to my book you now definitely owe me a review.

Paul Feyerabend

Department of Philosophy

University of California

Berkeley, California

David Joravsky replies:

I did not misuse Kevles’s book. I used it as the occasion for an essay on his principal theme: the social role of physical scientists in industrial nation-states. Since his was only one of several books under review I had to be very selective in reporting what he covered. Perhaps I should have noted that he covered a great deal, that his book is a massive compilation of historical information, and will probably be regarded for a long time as the standard reference on the history of American physics. Because it is a lasting, important book the national self-congratulation that fills its pages is especially deplorable. The most painful issues in the social role of scientists are evaded or smoothed over, as they are once again in his reply. Secondary issues, the politicking and the status ranking that attend the getting and spending of research funds, are examined at great length under the mystifying title, “Conflict between democracy and elitism.” (All the particular topics referred to in note 1 of his reply are of that sort. I still can’t see how the pork-barrel politics of science, and status ranking within the scientific establishment, are evidence of elitism in conflict with democracy.) My review article called attention to the more serious issues that are scanted in Kevles’s history. They can be summed up as the specialization of function which assigns to scientists and engineers the job of enhancing the power of industrial, military, and political leaders. Ideological justifications do not change that specialization of function or reduce the enormous danger that is inherent in it.

The most striking illustration is the development of nuclear weapons. Whether scientists lived in the US or the USSR or Nazi Germany, they worked hard to provide their leaders with nuclear bombs. The overwhelming majority of those workers, even in the US with its democratic traditions, ignored the military and political consequences of their invention, even in private discussions among themselves. That evidence of their essential “unconcern” was established by interviews and reported in Alice Kimball Smith’s A Peril and a Hope: The Scientists’ Movement in America, 1945-1947 (MIT Press, 1970, pp. 60-61.) Her sensitive book focuses on the small minority of “concerned” scientists who tried in vain to dissuade their leaders from the absolute conviction that they could start and win the nuclear arms race with a single murderous stroke.

Kevles’s history almost entirely ignores both the “unconcern” of the majority and the “concern” of the minority. His eye is worshipfully fixed on the chosen few, the seven scientists invited by government leaders to sit on advisory panels. I call that bureaucratic co-optation, not democracy, but names matter less than basic patterns of behavior. Bureaucratically co-opted scientists rarely advise restraint in the expansion of national power. When they do, as Oppenheimer’s committee did on the H-bomb decision in 1949, they are usually brushed aside. There have been a few exceptions, such as the ban on atmospheric testing, but they do not change the pattern, which is inherent in our system no less than the Soviet system. Scientists and engineers endlessly increase the power of national leaders to blow up the world. The most vivid history of the deadly process is Herbert York’s The Advisors: Oppenheimer, Teller, and the Superbomb (W.H. Freeman, 1976).

Kevles waves the flag and asks if I know something better than the American way of dealing with such problems. I’ll wave the flag right back, reminding him that this country was begun by fractious individuals who lectured their rulers on self-evident truths, as Leo Szilard tried to do. (See Spencer Weart, Scientists in Power, pp. 76-77 and 97, for evidence that Szilard’s agitation in the 1930s had broader purposes than prevention of a Nazi A-bomb. He wanted organized scientists “to exercise some measure of influence over a socially dangerous development.”) I wish that more scientists and engineers would struggle against the narrow, subordinate role that modern society increasingly assigns to them. I cling to Jefferson’s notion of democracy as perennial conflict between rulers and ruled. Historians like Kevles, who nourish complacent trust in the powers that be, seem to me inherently undemocratic.

To Feyerabend, my anarchistic foil to democratic Kevles, I offer no tardy acknowledgment of merits that my review neglected to mention. On the contrary, I was too generous when I wrote that he “readily convinces me” of his central negative thesis: the incongruence between the universal methodologies constructed by philosophers of science and the patterns of discovery revealed by historians of science. In fact, Feyerabend did not convince me. Substantial scholars did, such as Burtt, Koyré, and Kuhn. Feyerabend’s chapters on Copernicus and Galileo present some of their findings, in a slapdash, sensationalist way. “Galileo the mountebank” used “deception,” “trickery,” and outright “lying” to promote views he knew he could not prove by rational argument with available evidence; and that’s the way that science develops. I passed over such nonsense in magnanimous silence, to get at the central argument: since philosophy of science has failed so far to fit the patterns of discovery revealed by the history of science, it must always fail, unless we acknowledge “Anything goes” as the only universal methodology. Thus in book one; in book two Feyerabend backs off from “Anything goes.” Almost anything goes; many different methods work in the cognitive process. But he evades the obvious question; Which? Which methods show persistent patterns of success, and which of failure, in this, that, or the other field of cognitive endeavor?

I tried to get past that evasiveness by unpacking the implications of some examples he gave to illustrate cognitive success outside the realm of modern physical science. (Outside, because I think that physics cannot be the model for all knowledge.) Now Feyerabend disclaims his own examples. Except for astrology, to which he devoted a chapter. There he takes a stand, asking what authority I have for declaring astrology obsolescent. I’ll pick you, Feyerabend; you can serve as well as anyone to prove such an obvious truth. Your chapter on astrology is a diatribe against modern scientists for dismissing astrology out of hand, and against most astrologers for disgracing the subject by their crudity. That is clear evidence of obsolescence in two different communities. Evidently you have some more authoritarian criteria of truth than your declared respect for the evolving opinions of diverse communities. These criteria require explanation. If the state corrects “institutional intolerance” by forcing schools to establish courses in astrology, will the state prohibit the crude majority of astrologists from teaching those courses? That would be two gross violations of “Anything goes.”

Similarly with the example of traditional Chinese medicine, which recurs obsessively, in part because acupuncture made Feyerabend’s urine clear after scientific medicine had failed. (Fastidious people may object to such “personal remarks,” but Feyerabend put them in his book as part of the argument.) I asked him to take the example seriously, and tell us what theoretical framework enabled traditional Chinese physicians to discover a method of anesthesia that scientific physicians admit they cannot explain. Most of all I asked him to do the philosopher’s job and explain what it is in anybody’s theoretical framework that facilitates or inhibits recognition of inability to explain. I think it obvious that scientific medicine has been replacing traditional medicine all over the world because it is more systematically self-critical, strives more rigorously to distinguish successes from failures, and therefore achieves more successes.

To all that Feyerabend now replies; irrelevant. Since some good is done by traditional as well as scientific medicine, let us not make invidious comparisons of successes and failures. Let us be tolerant and use state intervention against the “institutional intolerance” that has been suppressing traditional medicine in favor of scientific. No point in protesting that comparative rates of success and failure are the reason for such intolerance; Feyerabend has declared such comparisons irrelevant. And no point in asking what knowledge the state should require in the licensing of physicians. If “Anything goes,” there will be no licensing. If “Almost anything goes,” we still have no right to ask for rules; they would create “institutional intolerance.”

I treasure Feyerabend’s pronouncement on the humanities: since the scientific revolution they “have been voided of ontological content.” It is not only the clumsy ambiguity of the verb that delights me, or the graceful circularity that nevertheless emerges. (How do we know that the passage of physics voided the humanities of ontological content? Because ontological content is conveyed only by physics.) I am also delighted by the unconscious testimony he has given for the final point in my review: “He imagines that he has escaped from the cage of scientism when he is only shaking the bars.” Most of all I thank Feyerabend for adding a perfect specimen to my collection of resonantly empty barrels, to be rolled out whenever I must explain what is meant by scientism.

This Issue

October 11, 1979