The Present Scene

With Mr. Gorbachev’s offer to conclude a treaty that would oblige the USSR and the US to withdraw all the intermediate-range nuclear missiles that now face each other in Europe, a first step has been taken to end the confusion that follows what was widely regarded as the failure of the Reykjavik meeting. The offer is forthright and unambiguous. As the official text makes clear, it is made not just in the name of the General Secretary but specifically on behalf of “the Soviet leadership and the country’s Defense Council.” Second, it is made without prejudice to the separate existence of the nuclear arsenals of the UK and France, an issue that helped lead to the breakdown of the earlier Theater Nuclear Force negotiations. Third, the USSR is prepared to begin talks immediately with a view to reducing and finally eliminating “other shorter range theater missiles.” And finally, the offer is not conditional on the US abandoning its program of research and development on SDI.

In short, the USSR has now fully accepted the “zero-zero option” for intermediate-range missiles in Europe, which it rejected only a few years ago, subject, of course, to mutual agreement to the fine print that would deal with monitoring and verification. The Reagan administration has welcomed the offer, and optimism is riding high despite some hesitant noises from the partners of the US in NATO. But the President’s Strategic Defense Initiative (SDI), with its implied threat that the USSR’s intercontinental ballistic missiles could be rendered “impotent,” remains an obstacle to progress in arms control in the “strategic” sphere.

It was never of course in the cards that any Russian nuclear warheads would be rendered “impotent” during the first four years of President Reagan’s “strategic defense initiative.” Year five therefore begins with the United States as wide open to nuclear devastation as it ever was. A single Russian megaton could still eliminate in a flash half a million, perhaps even a million, American citizens, just as easily as it would have done the day that Mr. Reagan first stepped into the White House. What is more, only an extremely gullible chief of staff or the most naive scientist would dare encourage the President to think that the situation might change during the remaining two years of his presidency. There may be, of course, unscrupulous R&D salesmen around who are prepared to give firm assurances that an impregnable defensive screen could be placed over the United States in, say, the next ten, or twenty, or even fifty years, but that is a different matter.

The President’s November 1985 meeting with General Secretary Gorbachev in Geneva achieved little or nothing. Their talks about arms control foundered on the rocks of SDI, as apparently did last October’s discussions at Reykjavik. In the months between the two meetings, the Russian leader’s views about the likely consequences of a search for a space-based ABM system had not changed. However fanciful the President’s dream, to the Russians SDI is a dangerous spur to the intercontinental arms race and, if pursued beyond what Mr. Gorbachev called the laboratory stage, an enterprise that will inevitably violate the 1972 ABM Treaty. In Russian eyes, a future American administration, in an aura of self-deception encouraged by a belief that the United States was protected by a space defense system, might even feel it could afford a “first strike” against the USSR—in the same way as those Americans who man the High Frontier, as one pro-SDI lobby is called, fear that this is precisely what the Russians always had, or may still have, in mind.

Understandably, therefore, Mr. Gorbachev has been unwilling, or unable, to reconcile proposals about “deep cuts” in intercontinental nuclear weaponry with the President’s insistence that work to develop a space-based ABM system not only has to continue but must be carried on in outer space. Nor did the General Secretary find comfort in Mr. Reagan’s promise that he would not deploy a space-based defensive screen for ten years. Arms control and the deployment of weapons in space—possibly nuclear weapons—simply do not go together. Indeed, immediately before the Geneva talks of December 1985, Caspar Weinberger had made it plain to the President that were the USSR to embark upon the development of an SDI type of space-based defensive system, the United States would immediately find itself constrained to add to its strategic nuclear forces.

Be that as it may, and SDI or no SDI, this is precisely what the President had already indicated he was about to do. He was, he said, no longer going to regard the US as bound by the limits on nuclear missiles that had been set by the unratified SALT II treaty. To add to the confusion, not only of the Russians but also of America’s allies, and no doubt of many American citizens, Mr. Weinberger and General James Abrahamson, the director of the SDI program, told the President in December 1986 that so much more progress has been made in the SDI research program than had been anticipated—presumably progress made in the preceding two or three months—that he should now give the order to prepare for deployment of a first-stage SDI system. At the very least, it is being argued, Congress should be persuaded to agree to a “broad interpretation” of the 1972 ABM treaty and to vote for a development program of such a size that no future president or Congress would be able to bring it to a halt.


Arms control negotiations have rarely been distinguished either by their consistency or by the mutuality of their purpose. Pacific words cost nothing. Reciprocity inevitably recedes into the mists of suspicion. SDI has made the situation worse than it has ever been. The President may talk to Mr. Gorbachev about deep cuts in strategic forces; that is his affair. The Pentagon has a different point of view about the enemy’s motives and intentions. For the Pentagon, it was up to the Russians themselves to decide how long their unilateral moratorium on nuclear testing was to last; that was not something that affected the Pentagon’s plans. The Russians could say that they would agree to on-site inspection, even of chemical plants that are suspected to be producing chemical warfare agents. Let’s wait and see, says the Pentagon.

If it so wishes, the USSR can uphold the SALT treaties. It can insist upon the strict interpretation of the 1972 ABM Treaty—even to the extent of dismantling the disputed Krasnoyarsk phased-array radar station, now that doubts have been cast on its legality, provided of course that the United States does not proceed with the replacement of the obsolescent radar installations at Thule in Greenland and Fylingdales in England. (The replacement of these installations, according to such highly respected and politically neutral figures as General Brent Scowcroft and William Perry, the director of defense research and engineering under the Carter administration, would be a clear transgression of the 1972 treaty.) The USSR may make these offers and do these things, but come what may, slowing down SDI is not something the Pentagon would advocate or a price the President seems willing to pay in order to salvage more from the confusion of Reykjavik than has now come within reach.

The President and the Pentagon can always fall back on the argument that Russian olive branches are covered with thorns, and that the forbearance that the Russians have recently been displaying is merely propaganda designed to shake the Western alliance. Here, however, one imagines that the Pentagon view would also be that the US has little to fear since America’s Western allies are far too dependent on the US to oppose effectively any US line.

Well, can they still be relied upon to follow? The answer today is significantly qualified both by the conflicting messages that have been coming out of the State Department and the Pentagon, and by the decline in the prestige and authority of the White House following the Iran-contra revelations. The Tower Commission’s report has done much to clarify the situation, as has the President’s gradual admission that the responsibility for what happened in the Iran-contra affair was his. But the world still awaits the outcome of further congressional and legal inquiries about his conduct and that of his immediate staff. Duplicity and illegal action in high places have undoubtedly shaken the capitals of Europe. Christoph Bertram, the influential diplomatic correspondent of Die Zeit, has written that the shock to the presidency, and the ascendancy of the Democratic party in both houses, provides Europe with an opportunity to be less subservient, and to make its own views heard.1

Obviously Europeans will now have to listen to Congress as well as to the White House, State Department, and Pentagon to discover what the voice of America is saying, especially at a time when a Democratically dominated Congress is not only trying to ferret out the truth about nefarious arms deals, but is also examining the President’s trillion dollar budget, of which almost a third is assigned to defense. They will be looking hard at the request for a 60 percent increase in funding for SDI, about which a large section of Congress is already skeptical. Searching questions are certain to be asked about what the billions already spent on the “initiative” have bought, and what is likely to emerge from the new billions now demanded in order to keep alive what is said to have become the Pentagon’s biggest weapons project ever. The economic and political consequences of the President bending to the pressures exerted by the Pentagon will be rigorously studied from every point of view. And congressional eyes are not likely to find the high-tech era as glamorous today as it appeared before the Challenger and Chernobyl disasters, and before the more recent failures of a Titan-34D rocket and of the usually reliable Delta. The year 1986 was a bad one for high-tech, and its disappointments will affect the budgetary debates that will enliven 1987. Budget and trade deficits on a huge scale, as seem likely, along with a soaring national debt, spell trouble not only for SDI, but for all defense appropriations.



Progress in the SDI

President Reagan’s defense initiative has generated so much discussion from the point of view of its strategic and political implications that at times one is inclined to forget that we still do not know whether his dream of a defensive screen over the United States could ever become a reality, even if the technical and scientific community of the United States were to be provided with all the resources it could ever demand, and given all the time it could ever want. History will have some nasty things to say about the lost opportunities of securing some worth-while post-Reykjavik arms control agreements on intercontinental ballistic missiles if the R&D program for SDI continues indefinitely on a “suck it and see” basis.

The announcement in December by Caspar Weinberger and Assistant Secretary of Defense Richard Perle that so much unexpected progress has been made in SDI research as to justify preparing for partial deployment of a space-based ABM system, carries little conviction and is hardly likely to resolve any doubts. Neither of them is an engineer or scientist, and at best their words mean no more than that this is something they were told. Nor does either have the background or experience even to judge whether their informants or their critics are to be trusted. Even the Joint Chiefs believe that “a deployment decision is premature,” since the “military utility” of what is proposed has not been properly assessed.2 The scientists and engineers who have been using the money that Congress has so far voted for SDI have said relatively little about the progress that they have made. Nor have they provided reassuring answers to the host of technical questions that have been raised by highly respected and prominent physicists and engineers who have the experience that would be needed to justify their critical observations.

In Weapons in Space, one of several excellent books about SDI that appeared in 1986, John Toomay, a retired Air Force general who had been a member of the Fletcher Panel—the body that had given the technical green light to the President’s initiative—makes a powerful but at the same time only a restrained case for pursuing an SDI research program. In the concluding paragraph of his essay, he reminds us that the Fletcher Panel provided no guarantee that the SDI research program would succeed. Its members did little more than agree that the technologies that would be called upon were sufficiently well understood to permit the “sketching out” of “broad systems concepts for a multilayered defense.” The “crucial technological issues,” Toomay adds, “certainly cannot be resolved now by polemics.” They will “take years of research, engineering, and analysis to resolve.”

The difficulty for someone who tries to follow the argument from the sidelines is that nearly all one reads from the pro-SDI technical camp is a form of polemic, of assertions that all is going well, and that whatever the professional distinction of those who criticize, and whatever they may have done in the past to help build America’s defenses, when they cast technical doubts on SDI they are simply being misguided, and almost guilty of an unAmerican activity. Last autumn General Abrahamson delivered a speech in Paris3 whose purpose, he said, was to provide an exact account of the present “status” of the SDI program. He told his audience that what gives him “confidence” are things like “new nuclear-hardened infrared detectors” that are “coming into existence,” and instruments like rail-guns and electromagnetic cannons that are “coming together.”

Most of the general’s address, insofar as one is able to find strategic and political meaning in his tangled prose, seems to boil down to the simple homily that “it is better to save lives than to avenge them.” I have read other statements by General Abrahamson, but so far I have failed to find in any one of them precise scientific answers to the technical queries about SDI that have been raised, for example, about the likelihood that Russian ballistic missiles could be destroyed before they penetrated the atmosphere, i.e., in their “boost phase.” Repeated warnings to his audience or his readers that they should not heed misguided scientists who do not know as much as the men working on SDI in Livermore or Los Alamos are simply not good enough.

A somewhat more readable pro-SDI tract was issued some three months before General Abrahamson’s Paris speech by an organization calling itself the National Strategy Information Center. Under the title The Intelligent Layperson’s Guide to Star Wars, it was organized in the form of answers to sixteen questions about what SDI is setting out to achieve. Again, the answers evaded any direct reply to the technological criticisms that have been made about the likelihood that the requirements of an SDI system could ever be met. Nothing substantial is said, for example, about the potentially devastating effects of Soviet countermeasures, such as antisatellite weapons, or about the grave doubts that have been cast on the feasibility of developing a reliable computer network to control an SDI system. What the “intelligent layperson” gets instead is merely a series of hopeful assertions. Most other pro-SDI tracts or statements that I have read follow the same line, for example those issued by the organization High Frontier, and one or two are all but incomprehensible.

For example, in a recently published “question and answer” interview with General Abrahamson’s assistant, Brigadier General Malcolm O’Neill, we read4 that because the advent of ballistic missiles led to a reduction in the numbers of manned bombers, ballistic missile defenses should in the years ahead lead to a reduction in the number of ballistic missiles. I fail to understand how any SDI spokesman could make such a statement, which on a moment’s reflection can be seen to be logical nonsense. Aircraft and rockets are interchangeable for the purpose of “delivering” warheads. Neither can be interchangeable with a defensive system that is designed to destroy them, any more than the expression of a hope that the world will one day be free of nuclear weapons can be regarded as a statement of strategy.

That all is not in fact going as smoothly inside the SDI scientific camp as its proponents make out also became apparent in 1986. Roy Woodruff, who was nominally leader of Livermore’s X-ray laser project, resigned his position during the year because of arguments with colleagues about the way the work was progressing.5 According to the newspapers,6 his departure was followed not long after by that of Peter Hagelstein, the young physicist who had been described as the “brightest star” of the Livermore SDI team.7 There were also rumors that demonstrations that had been laid on to reveal the success that had been achieved in the development of “components” which might one day become part of an SDI space-defense system, and which had been designed to impress military and congressional dignitaries, had been “embellished,” some said even falsified.

Obviously those in charge of the SDI program of R&D need to keep secret technical details about the way some of the work is proceeding, but the informed technical criticisms that have been made are not of a kind that, if answered as precisely as they are put, would give the USSR information it did not already have, or reveal views that it had not already entertained. Differences of opinion about the strategic and political consequences of pursuing SDI are to be expected, but there should surely be a greater measure of agreement about the scientific and technical considerations that are the basis of the concept of SDI.

A report that has recently been issued by the George C. Marshall Institute8 is not likely to help. It appears under the joint authorship of John Gardner,9 Edward Gerry (both of whom had also been members of the original Fletcher Panel), Robert Jastrow, William Nierenberg, and Frederick Seitz, the latter two being wellknown among scientists. The paper firmly states that were a decision to deploy taken now, that is to say in 1987, “deployment of the full defense can begin in 1994.” In a separate statement attached to the report, Jastrow and Gerry declare that they would go for a system of two thousand small satellites with “a handful of kinetic kill vehicles (KKVs) on each.” (By a “handful” they meant five; and by KKV they mean a modified anti-aircraft rocket which finds its target through a heat-seeking sensor in its nose.) Indeed, they would prefer a larger number of small satellites each armed with one KKV, but then “the whole system becomes extremely expensive.”

This is the same Jastrow who two years ago scathingly dismissed the views of Hans Bethe and Richard Garwin when they suggested that an SDI system might well call for the deployment of very large numbers of space platforms. Early in 1985 Jastrow also declared that the job could be done with 100 satellites each carrying 150 kill weapons.10 The system now proposed in the Marshall Institute paper would, it is said, be “approximately 93 percent effective against a threat cloud of 10,000 warheads and 100,000 decoys,” and could be bought for $121 billion. The authors also tell us that they could find no justification for the estimates of $1 trillion or more that have been made as the likely cost. But they do not provide any justification for their own estimate of $121 billion—remarkably precise even in the accountancy of billions. They dismiss statements that the system could not be deployed until the twenty-first century, and that the “launch capability” that would be called for to put their system into space is not yet available.

I have no doubt that there will soon be calls asking for the basis for what are really a pack of assertions, presented as conclusions and derived from probability calculations starting from speculative assumptions, and which dismiss lightly the vulnerability to Russian countermeasures of the whole proposed system, and particularly of the ten sensor satellites on which it would depend. The authors also seem to entertain the extraordinary view that the Russians would contemplate a first strike at US landbased missiles—and, one presumes, other targets as well—without taking into account the retribution they would suffer from the vast seaborne retaliatory force of the US. As I studied the report, I kept asking myself how the two of the five by whom it is signed and whom I know could have put their names to a document that they would have rejected out of hand had it been submitted as a scientific paper to a reputable scientific journal.

Not surprisingly, there have been suggestions that the appearance of the Marshall Institute report in December 1986 had something to do with the calls for early SDI deployment that were made at about the same time. I suspect, however, that as a piece of presumed scientific support for SDI it is going to backfire. Lowell Wood, who is in charge of Livermore’s SDI O Group, and who is well-known as an enthusiastic SDI researcher, declared in public debate in October 1986, only two months before the publication of the Marshall Institute report, that he had not come “to assert the technical feasibility of strategic defense. To do so,” he continued,

I believe would be intellectually dishonest. Whether or not strategic defense will be technically feasible a half-dozen years hence will become generally known only a half-dozen years hence. And anyone who presumes to tell you now what will be true that far away in this complicated area is frankly a confidence man. If he isn’t reaching for your wallet he probably wants your vote or your political contribution, which is a more popular form of theft against which the law provides no protection.11

I am sure that these words from so prominent a scientific advocate of SDI will return to haunt the five Marshall Institute writers who have so lightly called for an immediate decision to deploy.

A year ago there appeared to be a unified camp of scientific supporters of SDI—most of them men who were working on the research program. Schisms that may have been hidden have now surfaced, for it is not only Lowell Wood who has distanced himself publicly from enthusiasts like John Gardner and his four co-authors. It happens that Gerold Yonas, another ex–Fletcher Panel member, and until recently the scientific director of the SDI office in the Pentagon, also contributed a chapter to Weapons in Space. He, too, tells us that the members of the Fletcher Panel started out without any single view about what was technologically possible. A “bastion of technological optimists” put their faith in directed energy weapons, such as chemical or X-ray lasers, that would dispose of enemy missiles during their boost phase.

Others, whom Yonas calls “the strong bastion of doubters,” were concerned by a fear that there were “a myriad of possible methods to defeat any defense.” Both “bastions” agreed that since the enemy’s missiles could be launched either from land or from the deep seas or from the skies, an SDI defense system would have to be based on satellites which, as they orbit the earth, would immediately have to detect by means of “sensor systems” the launch of any enemy missiles; the defense satellites would just as immediately activate an immensely complex communications “command” computer network that in a split second would direct with absolute precision “kill-rays” at the attacking warheads—which may be accompanied by thousands of decoys.

If for one or other reason enemy missiles could not be destroyed in their boost phase, before they had discharged into space their “MIRVed” warheads and decoys, the chance of providing any significant defense of the American population would be infinitesimal, since an effective solution to the problem of destroying thousands of warheads as they move through space within a cloud of decoys is nowhere in sight. Indeed, the only part of SDI which in theory is technically realizable today is the destruction of warheads after they have reentered the atmosphere, a capacity that from the point of view of the defense of American cities means very little. Writing only a year ago, Yonas thus makes no secret of the totally speculative nature of SDI when viewed from the standpoint of the working technologist.

Another contributor to Weapons in Space is Charles Zraket, a man who speaks from a background of unrivaled experience in the design of complicated weapons systems. His view about the prospects of SDI becoming an operating complex is gloomy in the extreme. As he puts it,

The planning of any ballistic missile defense must consider the best methods of developing, deploying, and testing such a system in stages, through a transition period that may last for decades. As the Soviet threat and the international political environment change, our defense systems may be perpetually in transition. Since a poor defense might actually undermine our security rather than minimally enhance it, we must be sure at every point in the transition that each defense component we have deployed functions well enough to promote stability in crises and military postures.

George Rathjens and Jack Ruina, both MIT scientists who are also highly experienced in the design of high-tech weapon systems, agree fully with this statement. Indeed, the chapter that they contribute to Weapons in Space is even more explicitly pessimistic about the possibility of ever devising an effective SDI defense system. Moreover, as they emphasize, the attempt to devise one can only destabilize the prevailing state of mutual deterrence. As they put it, “almost every effort by one superpower to improve its defense capabilities can be expected to provoke improvement in its adversary’s offense capabilities for as long as defensive developments do not significantly alter the current offensedominated situation.”

Another broad review of the technological and security implications of SDI that is well worth studying has been recently issued by SIPRI, the government-sponsored Stockholm International Peace Research Institute. It is a record of an international conference that was held in July of 1985, and that has been brought up to date by its editor, Bhupendra Jasani, a senior member of the SIPRI staff, who provides an objective overview of the technological and strategic aspects of space weaponry and of their significance to the arms control process. The main part of the book consists of twenty-one papers that were delivered at the meeting, a few of them by authors whose names have become familiar in the SDI debate in the United States. Four of the others are from the USSR, their authors including Yevgeniy Velikhov, the physicist who is now a vice-president of the Soviet Academy of Sciences, and who is well known internationally.

As would be expected, the SIPRI volume is wider ranging than Weapons in Space, since much of it concerns the likely repercussions of SDI on the Western alliance and on international security. In his introductory essay, Jasani reminds us that the distinction between defensive and offensive space weapons, that is to say, the distinction between the components of an SDI defensive system and those of an antisatellite (ASAT) system, is becoming increasingly murky. Were such systems ever deployed, and a communications satellite belonging to one country damaged by another’s ASAT weapon, it might prove impossible to determine whether what had happened was deliberate or caused by an accident—so that the incident could inadvertently become a casus belli. Jasani also warns that whatever success might attend the search for a space-based defensive system, it is likely to be more effective in dealing with ballistic missiles of intercontinental range than with shorter range “tactical” missiles that might be used in a European war. Consequently what he calls “splits” could occur within the Western alliance.

SIPRI’s analysis is well complemented by the contributions to Empty Promise, a follow-up to The Fallacy of Star Wars, which was published in 1984, also under the auspices of the Union of Concerned Scientists. Empty Promise is edited by John Tirman, and of the other eight contributors only Peter Clausen and Richard Garwin participated in the writing of the earlier volume. Garwin’s chapter brings up to date his account of the countermeasures that the Russians could devise in order to defeat a space-based defense system, and adds emphasis to the several critiques with which he has already assailed the entire concept of SDI.

As he and others, including Andrei Sakharov, have pointed out, a basic issue is that the technical problems that would need to be solved in order to render an SDI defense system “impotent” are less difficult to overcome than those to which answers would need to be provided before an effective SDI system could be either designed or deployed. In particular, it would be less difficult to devise an effective way of dealing with what is essentially the only novel aspect of President Reagan’s ABM dream—that of destroying Russian missiles as they rise from their silos or from submarines. Among such countermeasures, for example, could be “fast-burn boosters”—missiles propelled by such high energy fuel that they could release decoys and warheads before space-based rockets could attack them. The Washington Post recently quoted General Abrahamson as saying in 1985, “You probably could not handle the fast-burn booster.”12

Garwin’s conclusions are given support in a chapter provided by John Tirman and Peter Didisheim. As Jasani does in the SIPRI volume, and Zraket in Weapons in Space, these two writers also show just how fragile the whole SDI concept becomes when considered in relation to developments that would be bound to occur in ASAT weaponry were SDI development to proceed. We have to remember, as they put it, that “unrestricted ASAT competition, fueled by its Star Wars parent, will bring new and deeply provocative weapons to the superpower confrontation with no compensating virtues.” And this, as Tirman and Didisheim say,

is the fundamental paradox—and irony—of the Strategic Defense Initiative. Not only do antisatellite weapons, those of today and those that will be created within the SDI, fatally threaten the Star Wars armada of the twenty-first century, but they will threaten the vital military satellites of the twentieth. This is, indeed, a real and perilous “window of vulnerability” that, once opened, will be extremely difficult to close.

Many who have thought hard about the concept of SDI have been driven to the same conclusion. It is sharply emphasized in a “staff report” @f13@cthat was submitted to senators William Proxmire, J. Bennett Johnston, and Lawton Chiles in March of 1986. If SDI were to proceed to the stage when a supposedly defensive system dependent on space satellite platforms could be deployed it would be inevitable that a separate antisatellite system poised to destroy it would have been developed by the USSR, a situation which the writers of the staff report wryly describe as hardly conducive to “a stable environment for the future.” They also tell us “scientists at the Sandir Laboratory who have been intensively studying this question have come to the conclusion that space-based boost phase defenses can never be made survivable unless by Treaty.”

There is also the disquieting prospect of orbiting space platforms carrying destructive devices that could incinerate cities probably more easily than they could destroy Russian missiles@f14@c—which indeed was the thought behind one notion of a strategic nuclear weapons system that was being discussed way back in the immediate post-Sputnik era in the Fifties. Without revealing their sources, those two fiery commentators, Rowland Evans and Robert Novak, have proclaimed that some US spy satellites have already been put out of action by Russian ASATs.15 If the nuclear arms race continues at its present rate, we shall not be “blundering into disaster,” as Robert S. McNamara entitles his new book (which I will discuss in a second article). We shall have sought it deliberately.


The task of the congressional committees that have to decide about the technicalities of SDI, and about early deployment in particular, is clearly formidable. How do they, for example, take a view about the possibility of ever devising the communcations-computer-command network that the system would demand, a network that would link together the thousands of components—sensors, command and control switches, destructive devices—that a space-based defense system, whatever its configuration, would call for? As I have said, the Fletcher Panel did not disguise the difficulties, and Dr. David Parnas early on resigned from the official SDI panel that was dealing with this part of the whole research program because, as a computer specialist, he could not see the problem ever being solved.

In the hearings on SDI in December of 1985, Dr. Solomon Buchsbaum, executive vice-president of AT&T’s Bell Laboratories, testified in a contrary sense before the Subcommittee on Strategic and Theater Nuclear Forces of the Senate Committee on Armed Services. He believes that the necessary computer network could be designed, built, tested, and deployed because, as he sees it, an equally complex system already exists in the US Public Telecommunications Network.

The reference to the US telephone network as a model equal in complexity to what SDI would require has since been challenged by Karl Dahlke,16 who is engaged in the design and maintenance of the very AT&T telecommunications system that Dr. Buchsbaum gave as the basis of his conviction that an SDI network could be devised. Mr. Dahlke draws attention to the fact that “widespread fraud illustrates the vulnerability of our nation’s phone system.” As he sees it, a space-based network could be more easily disrupted by an intelligent enemy and would be far more difficult to repair than a ground-based system. So far as reliability is concerned,

despite our best efforts, the software that controls the telephone network has approximately one error for every thousand lines of code when it is initially incorporated into the phone system. Extensive testing and simulation cannot discover these errors. If SDI contains ten million lines of software (a credible estimate), and its quality is comparable to the telephone network, we can expect ten thousand errors embedded in its software when the Soviets attack. It only takes a few disastrous software errors (one per layer) to cripple any multi-layer SDI implementation.

The same topic is dealt with in an essay in Empty Promise by Greg Nelson and David Redell, another two highly experienced computer scientists. As they put it,

The SDI is inherently dependent on its computer system, without which any collection of exotic weapons, sensors, and strategies would be useless. The required computer system would be far more complex than any previous computerized weapons system and would require software well beyond the current state of the art. Given the magnitude of this software engineering challenge, there is a significant chance that the development effort would simply fail to produce a deployable system at all.

Criticisms of this kind cannot be dismissed as the murmurings of misguided scientists. To help answer them, the SDI organization set up a study group, now known as the Eastport Study Group, to consider the computer network that would be needed to support what it called the “battle management” of a space-based ballistic missile defense (BMD) system. Its chairman was Dr. Danny Cohen of the University of Southern California, who had earlier testified with Dr. Buchsbaum that what SDI needed could be provided. However, the published summary of the group’s report17 is essentially a statement of problems to be solved, with little indication of what has already been achieved.

What is really important is that the report makes it abundantly clear that those who are actually trying to design what is essentially the “nervous system” of a space-based ABM system are doing so without any idea of what the whole organism may eventually look like, if indeed one ever materializes. The SDI computer specialists are not working to any clearly specified military operational requirement. In mild understatement, the Eastport Group writes that “in a process where the threat and the ability of future technologies to cope with the threat are both unclear, the development of weapons and C3 [i.e., the network for command, control, and communications] takes on a greater aura of risk.” The “aura of risk” would seem to be equivalent to the likelihood that a coherent picture could emerge from some giant jigsaw, the separate bits of which were being independently cut out by an army of carpenters each working in isolation.

If, miraculously, the SDI components could be fitted together into a coherent system, a new problem would emerge. How much would it cost to put the bits together and put them into space? The Marshall Institute report stated that all the necessary rockets were available to deploy a system weighing approximately two million pounds now, and at an acceptable cost. General Abrahamson has been reported as saying that the cost of shooting into a low space orbit the satellites that the SDI system would demand is critical.18 He estimates the present price to be about $1,500 a pound of material, and this figure he says would need to be reduced to between $100 and $150 a pound to make the whole system affordable (the lifting cost could otherwise run into tens of billions of dollars for a system of space platforms whose cumulative weight would add up to a total of thousands or of tens of thousands of tons).

Dr. Robert Mozley of the Stanford Linear Acceleration Center has made a close analysis of this particular problem, using both NASA launcher costs—including shuttle costs—and those of other launcher organizations, American and foreign.19 His conclusion is that the present real cost for lifting SDI space platforms into orbit is about twice General Abrahamson’s quoted figure. He does, however, believe that the figure might be reduced to General Abrahamson’s estimate of $1,500 by the middle of the next decade, given a new effort “to improve large expendable rockets.” The possibility of getting the figure down to the general’s “economic” level of between $100 and $150 is, however, remote in the extreme, since the likelihood that it could fall that far is dependent on the successful outcome of a great deal of basic research that presumably has not even been started. In the meantime the SDI Organization has included $110 million in its request for $2.8 billion more than its budget allowance for last year in order to develop the technology for a heavy-lift launch vehicle. Whatever else it may relate to, the existence of this proposed “item” of R&D hardly helps to impart confidence in the view expressed by the Marshall Institute report that the “launch capacity” for a full SDI system is already available.

Apart from the financial cost of the SDI program and, even more important, its diplomatic and political costs, there is another cost that needs to be borne in mind, one that could be called the scientific cost. In a chapter that he contributed to Empty Promise, Jonathan Tucker writes that about 40 percent of all American scientists and engineers are already working in defense-related research, and that more and more American universities are now accepting grants from the SDI office in the Pentagon. Indeed, he tells us that the Pentagon is now the source of as much as 16 percent of all the support that the federal government provides for academic science. This is a role that the Pentagon can well afford since military R&D accounts for almost 75 percent of the resources that the federal government provides for all R&D.

Against the background of the general decline that is occurring in the support of academic science, Mr. Tucker is therefore right to warn that the increasing amount that comes from the Pentagon is highly significant. The universities now receive a total of more than $200 million a year from the SDI program alone. Clearly the Pentagon is diverting a major slice of university research into narrow and highly specialized channels—a development that does not bode well for America’s cultural and economic future. Harvey Brooks, the well-known physicist who over the years has served White House and Pentagon scientific advisory committees at the highest level, has rightly asked whether “the integrity and credibility of the US technical community” can “survive the politicization of such a large fraction of the scientific and technical effort of the country.”20But it is not only the academic community of the United States that is being tempted by Pentagon money. SDI recruiting sergeants are at work in universities in many parts of the world.

Moreover, the challenge so lightly thrown by President Reagan at the American academic and industrial communities was also indirectly addressed to the Russians. The USSR Academy of Sciences has already published its own primer to show how an American space ABM system, whatever shape it takes, could be neutralized.21

It is said that only the President and Mr. Weinberger still believe that a space shield that would make American citizens invulnerable to a nuclear onslaught is a real possibility. I have already quoted General Abrahamson’s admission that the US “probably could not handle” a Soviet fast-burn booster, were the Russians to develop one, as they say they can; and such a Soviet countermeasure would make the critical first phase of an SDI defense unfeasible. Others working on the SDI program have said the same thing. Secretary Shultz, in a carefully worded address that he delivered in Chicago on November 17, provided an official statement about SDI that appeared both loyal to the President and at the same time realistic.22 Even were vast reductions to be made in the nuclear arsenals of the two superpowers, he declared, even if both eliminated all ballistic missiles, the United States would still need an insurance policy, i.e., an ABM system, “to hedge against cheating and other contingencies.” And here the President’s program for defenses against ballistic missiles could, he added, “be a key part of our insurance.”

But which part? Echoing Paul Nitze, he also declared that “we judge defenses to be devisable only if they are survivable and cost-effective at the margin.” What exactly does this mean? In the light of such facts as are available, does it mean what Robert McNamara has styled SDI-1—the President’s original objective? Or SDI-2, a program that aims at devising a partial or “leaky” space screen, at the same time as the US strategic arsenal is improved? Or What? And would the goal of SDI-2, if that becomes US policy, be as objectional to the USSR as is SDI-1, and not a hindrance to the “deep cuts” in strategic forces that were also spoken about at Reykjavik? Mr. McNamara provides his own answers to questions such as these, as do the authors of the books I will consider in a second article.

March 11, 1987

This Issue

April 9, 1987