Genius in the Shadows: A Biography of Leo Szilard, the Man Behind the Bomb

[This article is reprinted, with permission, from American Journal of Physics, September 1993.]

This delightful biography by Bill Lanouette brings Leo Szilard out of the shadows today: the time of the finale of the Cold War, a time foreseen by Szilard in his novelette, The Voice of the Dolphins. It is high time for a comprehensive look at Szilard, who was involved simultaneously in the first steps in building the original nuclear weapons and in the creation of international regimes to control the nuclear genie he had let out of the bottle. Lanouette's book does justice to the man who propelled the world across the nuclear Rubicon with the Einstein-Szilard letter to Roosevelt (p. 205): "Some recent work by E. Fermi and L. Szilard, which has been communicated to me in manuscript, leads me to expect that the element uranium may be turned into a new and important source of energy in the immediate future--. This new phenomenon would also lead to the construction of bombs--[which] might very well destroy the whole port together with some of the surrounding territory."

By obtaining a first access to Szilard's correspondence with his wife Trude, his brother Bela, and many others, Lanouette has been able to fully expose the two competing sides of Szilard's actions. The inherent conflict between building bombs and controlling bombs made Szilard's unique, somewhat frantic personality all the more chaotic. Lanouette shows that Szilard, the unemployed dreamer, was the creative force in the nuclear shadows of Einstein and Fermi, and also the instigator of today's arms control process. This is a humane book about Szilard the person, related through countless anecdotal stories; it is not a formal history of the Manhattan Project.

Szilard was an intuitive applied physicist whose thumb prints are on applied technologies (thermal reactors, breeder reactors, atom bombs, electromagnetic pumps, electron microscopes, and information theory) rather than on fundamental science. Lanouette shows that Szilard was politically very astute. He realized early on that Hitler could get the bomb and that a nuclear arms race would follow the first nuclear weapons. Szilard, driven by these political concerns, acted boldly: he secretly patented the nuclear chain reaction in 1934; he tried to get other physicists not to publish nuclear data, but after they published, he too published; he wrote three Einstein-Szilard letters (two to begin the Manhattan Project and one to slow it); he created much of the Franck report, which called for an initial demonstration explosion over Japan rather than city destruction; he lobbied Congress for civilian control of the atom and various other arms control matters; he helped organize the Federation of Atomic Scientists, the Pugwash meetings, and the Council for a Livable World; and he initiated the creation of the Salk Institute to study both science and its impacts.

Szilard was a most difficult person to work with. He overslept in the morning, then soaked for hours in the bath tub, only to arrive at work at noon with new suggestions for others to carry out. He didn't like to get his hands dirty, and he continually argued with the conventional wisdom of the day. With increasing anti-Semitic instability in Europe, his unpredictable personality became all the more frantic, but purposeful. Szilard, who couldn't be constrained to settle down until the last year of his life, always kept two packed bags, ready to hit the road. Szilard converted his frenetic energy into results. Lanouette argues that Szilard was the first to conceive of the bomb and among the first to initiate arms control because he was so contrary and difficult. His example implies that all large projects, such as the Strategic Defense Initiative, need contrary, but honest, nay sayers with social conscience to point out possible problems.

Genius in the Shadows answers many questions, such as: Who really conceptualized the December 1942 reactor in Chicago? Lanouette points out that it was Szilard, not Fermi, who first understood nuclear chain reactions and first designed the nuclear reactor. Lanouette states (p. 178) that "[Fermi] failed to recognize the importance of this news [of the fission process] and failed even to mention what he had heard to his Columbia colleagues. Fermi was so typically cautious, in fact, that as the grave consequences of fission became apparent to others around him, he repeatedly denied their significance." Fermi was clearly the group leaders, but Szilard was the creative prophet.

Who first conceptualized verifiable arms control agreements? A month after Hiroshima, Szilard addressed the Atomic Energy Control Conference. Lanouette writes (p. 283): "Szilard had the last word at the conference--and the first on record about the touchy topic of verification--when he said a necessary first step would be to 'guarantee immunity to scientists and engineers everywhere in the world in case they should report violations of the [arms-control] agreements'." Later Szilard recommended supplementing his immunity for whistleblowers with $1 million rewards. These ideas sound good today; wouldn't whistleblowers in Iraq, North Korea, India, Pakistan, Israel, South Africa, and several other states have been useful? The special and challenge inspections of today's arms control treaties are further extensions of Szilard's suggestions of September 1945.

Because Szilard had the courage to work against the misuses of science, the American Physical Society and its Forum on Physics and Society have given the Szilard Award for "outstanding accomplishments by physicists in promoting the use of physics for the benefit of society in such areas as the environment, arms control, and science policy." Since 1974 the award has been given to many prominent physicists, including Richard Garwin, Hans Bethe, Wolfgang Panofsky, Andrei Sakharov, and Jack Gibbons. Upon receiving the Szilard Award, Gibbons, the former director of the Congressional Office of Technology Assessment and the present Science Advisor to President Clinton, stated that "Szilard should be the patron saint of OTA!" Many of us fondly remember Trude Szilard, who joined in the early years in presenting the Szilard Award at the Washington APS meetings. More recently, the Szilard award has been enhanced by adding a sculpture of a dolphin which travels from winner to winner.

David Hafemeister California Polytechnic State University
San Luis Obispo, CA 93401

Was the nuclear arms race more an issue of technology or of political perceptions? Was it Newton or was it Freud? Patrick Glynn concludes that the arms race was both technical and political-psychological. Since the marginal utility of nuclear weapons rapidly diminishes with greater numbers of weapons, ultimately it was more political than technical. I agree with the author in that conclusion, but then we part company. Glynn's book clearly lays out the "might makes right" point of view and is a useful reference to understand the basis of the far right's opposition to arms control. Indeed, Closing Pandora's Box was recently described in a Washington Post book review as "the most cogent and informative history of the Cold War, of arms control and of US strategic policy to have emerged from the conservative mind." Despite being "imaginatively researched and elegantly written," the book's weakness, and the reason it fails to accurately present the development of nuclear arms, lies in the author's pre-established and unswerving belief in nuclear arms build-up as the best approach to nuclear arms elimination.

Glynn's main political thesis is that tough, noncompromising nations will prevent wars and end the nuclear arms race. Glynn asserts that arms races don't beget wars, but rather political firmness prevents them. He has strong horses in his stable with the failure of Chamberlain in Munich, and more recently the 1981 Reagan "zero option" for the INF Treaty which far surpassed the Nitze "walk in the woods." Any comparison of Munich and the winding down of the cold war will stretch credibility, particularly because of the differences made by Mikhail Gorbachev. Moreover, at the technical level, Glynn fails to understand the difference between nuclear weapons and conventional weapons, and at the political level he fails to understand the calming effects of confidence-building measures and three decades of almost constant negotiations. He properly takes into account some political factors, for example to appear to act too softly will encourage your adversary, but he fails to understand that ultimately there will have to be some form of arms control. Even a greatly weakened Russia will want to have its nuclear deterrent, and if not Russia, then certainly the other nuclear weapons states should be engaged in the arms control process. And what about treaties that are clearly in our interest such as the Nuclear Nonproliferation Treaty? Doesn't our failure to adhere to a comprehensive test ban weaken the NPT/IAEA regime? Arms-control regimes have to reflect the interests of other nations, rather than prescribing terms from Washington.

Some of my disagreements with Glynn are:

--Glynn didn't want the Soviets to have permissive action links (PALs) on their nuclear weapons if we had to give them some broad, very basic information. He exaggerates the "divulgence of highly classified information" (p. 201) during the passage of descriptive concepts on PALs to the Soviets. Does he want accidental launches?

--Glynn criticizes McNamara for encouraging the Soviets to harden their missile silos because it would "encourage an increase in US vulnerability to a Soviet second strike, in the supposed interests of assuring mutual stability" (p. 203). Does he want the US to attack first? Would he rather have the nervous Soviets after the Cuban missile crisis mistakenly launch to protect their missiles from a dreamed US attack?

--Glynn exaggerates the Soviet buildup by overstating the capabilities of the Soviet's "five new ICBMs, a new SLBM, and four new types of ballistic missile submarines" and asserts that "the Soviets would pursue a first-strike capability against the American ICBM force" (p. 215-6). But it is clear that one cannot have a first strike without being annihilated in the process. In addition, the technical data and all reasonable exchange-model war games show that, except in the case of the SS-18, US systems have always been far superior to their Soviet counterparts. The lethality of the SS-18 was less than that of the MX or Trident D5/W88, but there are twice as many SS-18 warheads.

--The 1960s ABM system was decommissioned at Grand Forks because it wasn't capable or cost-effective. Nonetheless, Glynn thinks that it would have decreased offensive forces.

--His concern about "inequalities in the SALT I agreements" (p. 234) seems to ignore our MIRVed, more accurate survivable submarines.

--By stating that "--partly owning to their special fear of SDI, the Soviets after 1985 were never at the point where they could afford to reverse course and revert to confrontation" (p. 339), Glynn overstates the Soviet's fear of SDI. The Soviet MIR book and Roald Sagdeev, director of the Soviet space program, told Gorbachev the same things that the APS SDI study said to the US, namely that directed-energy weapons in space won't work on a cost-effective basis, and probably not for other practical reasons.

In spite of his strong political bias, Glynn does document many cases in which being intransigent and stubborn won out. However, there is more to the story than Ronald Reagan. For example, he seems to ignore the real role of Mikhail Gorbachev, who received ideas from the American arms control community that pointed out that minimum deterrence, viable on-site inspections, a comprehensive test ban, and de-MIRVing would reduce the chances for nuclear war. Gorbachev knew he needed to save money for his tattered system and that preventing nuclear war with the US was good for mankind. Without a Gorbachev, who gave up 3-to-1 advantages in both the nuclear INF and conventional CFE treaties, the nuclear world could have spun out of control. The debate over START in the Senate tried to set the record straight: The Bush Administration pointed out that to be firm was successful. However, Senator Moynihan countered by showing that he had written as early as 1979 that "The truth is that the Soviet idea is spent. It commends some influence in the world, and fear. But it summons no loyalty. History is moving away from it with astounding speed --. It is as if the whole Marxist-Leninist ethos is hurtling off into a black hole in the Universe." In other words, communism was a religion that was followed by its first and second generations, but beyond that it would collapse under its own weight. In Moynihan's view, we did the right thing by negotiating on arms control, killing some time, and waiting for the system to collapse. Historians will of course continue to debate this issue, but certainly both sides of the debate are not to be found in Glynn's book.

If you are teaching a course on the arms race, I would recommend Glynn's book only as a library reference. He tells some good stories, but with a heavy bias and some inaccurate conclusions. Although it might spark some good debates in your classes, "caveat emptor."

David Hafemeister
Professor of Physics
California Polytechnic State University
San Luis Obispo, CA 93401

One of the many legacies of World War II was a decision that government would take primary responsibility for supporting scientific research, in order to enhance national defense, economic growth, and public health. In Science Funding: Politics and Porkbarrel, Joseph P. Martino contends that it is time to rethink that decision. His conclusion is based on an extensive examination of the current research scene and a look back at two centuries of science funding, stressing the American experience. The author, who is a senior research scientist at the Research Institute of the University of Dayton, has written two previous books: Technological Forecasting for Decision Making and A Fighting Chance: The Moral Use of Nuclear Weapons.

Martino surveys the many pathologies that have emerged in our science funding system, such as Congressional earmarking of funds for specific locales; the bias toward big science; political and bureaucratic micromanagement; funding instability and red tape; the struggle over indirect costs; biases inherent in peer review; and the organization of scientists into a special-interest lobby. He also explores many alternatives to government funding--alternatives which, having contributed historically to scientific research, were subsequently crowded out by the sheer weight of federal largesse. What makes Martino's critique compelling is its grounding in public-choice theory, which clearly indicates that the problems afflicting our research system are not abnormalities, not mere mistakes correctable by better management, but are rather the inevitable consequences of government funding.

The view has long been prevalent in economics that market processes frequently produce harmful, or insufficiently beneficial, results. Government, as the public interest personified, should step in to remedy "market failure." For example, private firms have insufficient incentive to carry out all the research that would benefit society as a whole. Therefore, government must make up the difference.

Public choice theory points out that politicians and government bureaucrats have their own self interest, which may not coincide with the public interest. Bureaucrats tend to seek more power and perks, a bigger budget, and more prestigious missions for their agency. Politicians are in the business of getting elected. Distorted political incentives arise from voters' fragmentary knowledge of government activities and the consequences thereof--an unavoidable situation, given the enormity of government expenditures and regulations in present-day society. The political image of doing good diverges from the reality. Political profit is available from porkbarrel projects, from government actions having concentrated benefits and diffuse costs, from programs whose conspicuous benefits are outweighed by inconspicuous side effects. Simply stated, there is such a thing as government failure. The results of market processes should be compared not with the public interest personified but with the results of real-world government.

By now, science funding is showing all the classic symptoms of government failure. Congressional earmarking, also known as porkbarrel science, is only the most blatant example. With peer review directing academic research money primarily to the wealthy Northeast and West Coast, the wonder is that politicians from elsewhere refrained as long as they did from grabbing some of that money.

An example of the form-over-substance nature of political priorities is found in congressional micromanagement of the National Institutes of Health, through proliferation of highly-specialized new medical-research institutes. These score points with various medical lobbies and project an image of positive action, but they do not necessarily increase the overall NIH budget, and the resulting distortion of funding allocations probably diminishes the net effectiveness of medical research.

The symbiotic relationship that has been developing between big government and big science should come as no surprise. Science mega-projects acquire large constituencies from outside science, gratify bureaucratic imperatives, and lend themselves to rhetoric invoking nationalistic pride. Deep divisions have appeared within the scientific community regarding such massive undertakings as the Superconducting Supercollider and the Human Genome Project, as many scientists express concern over the displacement of smaller-scale research offering a higher return in benefits to society.

There is little or no payoff to funding agency administrators if a risky piece of funded research is successful, only probable recriminations if it is not. In addition, peer review works to rigidify disciplinary boundaries. The result, according to Martino, is a conservatism which probably inhibits innovativeness and appears to be getting worse.

Government subsidy of any activity creates a special-interest group with the incentive to organize and lobby for additional benefits. Martino says this much, but he seems reluctant to spell out the implications which experience has by now fully confirmed.

Government involvement may stem from the best of intentions, but in the subsequent political process it is inevitable that the priorities of the subsidized group will to some degree be advanced at the expense of the public interest. Research has proved no exception to the rule. Government has squandered priceless resources on the pursuit of technological and scientific exotica, from space shuttles to particle accelerators, from quasars to quarks. The point is not that manned space travel will never be practical or that today's esoteric research will never find application. The point is one of timing: Just because it is physically possible to do something does not mean that now is the time to do it. Premature efforts divert intellectual and financial resources that could have been applied more directly toward building a wealthier, technologically more competent society, one which could more comfortably carry out research that had barely been possible before.

The usual argument for government funding of research--that the market would spend too little--is not merely wrong but misconstrues the issue entirely. It is an advantage of the market that private investors cannot afford to waste money on the scale that politicians customarily do. Yet, even if government were peopled by saints, they could not know how much money to spend on science or how to spend it. Martino mentions this latter point but does not place it in context within economic theory. It is not just a matter of perverse incentives or behavior. Government is inherently inept at allocating resources because it cannot utilize the store of dispersed, unarticulated knowledge in society. The relevant insights in this regard were provided by the Austrian economists Ludwig von Mises and F.A. Hayek, (1) long before public choice theory was developed.

In the absence of government funding, philanthropy would undoubtedly play a larger role. "Big science," circa 1900, was astronomy, and its was funded by philanthropy. Even today, one can only speculate whether the scientific returns from the Space Telescope will match those of the privately-funded Keck, which costs twenty times less. Howard Hughes Medical Institute spends $100 million a year on medical research, but NIH apparently regards this as some sort of threat, rather than a philanthropic helping-hand in the fight against disease.

Industrial research has been making important contributions to basic as well as applied science throughout this century. Radio emission from the Galactic center and the cosmic microwave background were both discovered at Bell Labs in the course of applied research. Martino cites evidence that not only has industrial research paid off for private firms, but it paid off more when the research was financed by the firms themselves rather than by the government. Consortia of several firms could carry out more research, probably with a larger ratio of basic to applied; until recently, anti-trust hysteria was an insurmountable obstacle to that option. Government fiscal and monetary policies that discourage savings and drive up interest rates also discourage privately-funded industrial research, because research is an investment that takes time to pay off.

Martino has issued a courageous and persuasive challenge to the prevailing view that scientific research requires massive government support. His book deserves a wide audience.

1. F.A. Hayek, The Fatal Conceit: The Errors of Socialism (University of Chicago Press, 1988).

Allan Walstad
Physics Department
University of Pittsburgh
Johnstown, PA 15904