In the presidential campaign of 1980, Ronald Reagan helped make SALT a four-letter word, all but unmentionable in polite but hard-headed company. Such was his distaste for the Strategic Arms Limitation Talks that had been under way for more than a decade. Yet during his Administration, the quest for an agreement between the United States and the Soviet Union on the size and composition of their intercontinental nuclear arsenals has continued under the new acronym of START, for Strategic Arms Reduction Talks.

As Mr. Reagan nears the end of his presidency, an important part of his legacy is a work-in-progress: a START treaty already in the form of what diplomats call a joint draft text. But that document still contains numerous brackets that indicate points of disagreement. Soviet and American negotiators in Geneva labored hard to remove some of those brackets during this past summer. In late September, Secretary of State George Shultz and Foreign Minister Eduard Shevardnadze are scheduled to meet in New York and Washington, and START will be high on their agenda. While last-minute breakthroughs are still possible, neither side expects a completed treaty to emerge from that meeting.

Time may have run out for the Reagan Administration in arms control, though not for the enterprise itself. Both of President Reagan’s would-be successors, George Bush and Michael Dukakis, have indicated that they would build on the considerable progress that President Reagan, in partnership with General Secretary Mikhail Gorbachev, has made in START during the past few years.

However, in addition to inheriting an unfinished treaty, the next administration will also inherit some unanswered questions about the future of American defense programs. Further progress toward finishing the treaty will almost certainly require progress toward answering those questions.

II

When they met in Washington in December 1987, Ronald Reagan and Mikhail Gorbachev signed a pact eliminating intermediate-range nuclear missiles. The "zero solution" was stunning in its simplicity and boldness, yet it accounted for only about five percent of the firepower in the Soviet and American arsenals. Therefore, while the signing of that agreement was the pretext for the Washington summit, the main business transacted during those three days was an intensive effort to achieve progress in START, which had been proceeding in parallel with the intermediate-range nuclear forces (INF) talks in Geneva.

Both leaders had strong incentives to follow up on their success in INF with a START treaty in 1988. President Reagan wanted to achieve his long-sought goal of "radical," 50-percent cuts in the most dangerous weapons on earth. He wanted to claim that the American buildup in offensive weaponry, along with his cherished Strategic Defense Initiative (SDI), had served to bring the Soviets not only to the START negotiating table but to the treaty-signing table as well.

A strategic arms pact in 1988 would also have helped offset the tribulations afflicting the Administration elsewhere in its foreign policy, particularly in Central America and the Middle East. The more bad news coming out of Nicaragua, Panama, the occupied territories of the West Bank and Gaza, and the Persian Gulf during the winter of 1987-88, the more attractive to the president and his advisers was the prospect of springtime in Moscow.

For Gorbachev, too, a clear-cut diplomatic triumph would be welcome in a year of troubles. He had to cope with an outbreak of ethnic unrest in the Caucasus and a recurrence of labor strife in Poland. In Afghanistan, the Soviet Union was cutting its losses and withdrawing the forces that for nine years had waged a quagmire war against the local, U.S.-backed anti-communist resistance. Gorbachev’s internal reforms were a major preoccupation for him and the rest of the leadership, arousing resistance if not outright opposition at virtually all levels of Soviet society. Perestroika (restructuring) at home required a peredyshka (breathing space) abroad. Summitry would help to that end. Furthermore, reductions in strategic arms of the kind envisioned for START seemed to be consistent with Gorbachev’s political, military and economic priorities.

So at the beginning of the new year—Reagan’s last in office—political will appeared to exist at the highest levels in the White House and in the Kremlin for concluding an agreement. Also present were the main contents of the strategic arms reduction treaty itself.

The basic structure of the prospective START agreement is similar to that of its unratified predecessor, the SALT II treaty of 1979. An interlocking set of ceilings and subceilings was intended in the case of SALT to limit—and is intended in the case of START to reduce—those weapons that the superpowers might use to strike each other’s territory in an all-out war: intercontinental ballistic missiles (ICBMs), submarine-launched ballistic missiles (SLBMs) and long-range bombers. Of particular concern have been those weapons that have the capability of carrying out a preemptive and disarming first strike. The object of arms control is to achieve not just a subtraction in the overall numbers of strategic arms, but preferential cuts in those systems that are most likely to be used to start a war. If those components of the two arsenals are reduced, so will be the danger of war itself.

From the American standpoint in both SALT and START, the principal purpose of strategic arms control has been to constrain Soviet ICBMs, especially the U.S.S.R.’s so-called heavy ICBMs. These are larger than any missiles on the American side, and each is armed with as many as ten warheads or multiple independently targetable reentry vehicles (MIRVs). Its preponderance of highly accurate land-based MIRVs gives the Soviet Union its most threatening advantage over the United States, since those warheads are the putative instrument of a bolt-from-the-blue attack against America’s own ICBMs.

Thus, an important American criterion for judging progress in arms control has been the extent to which the process blunts the cutting edge of the Soviet Strategic Rocket Forces. By that standard, START had already produced the makings of a promising agreement by the beginning of 1988. Twin ceilings of 1,600 launchers (intercontinental bombers, ICBMs and SLBMs) and 6,000 warheads had been agreed upon over a year before, at the Reagan-Gorbachev meeting at Reykjavik in October 1986. The Washington summit of December 1987 yielded a subceiling of 4,900 on all ballistic missile warheads (i.e., both land- and submarine-based). Two additional features were especially welcome to the American side. These were mandated 50-percent cuts in heavy missile warheads, where the Soviets had a monopoly, as well as in Soviet ballistic-missile throw-weight (the cumulative capacity to hurl warheads at the enemy), where they had an advantage.

Another salutary feature of the prospective agreement was a formula whereby bombs and short-range missiles on intercontinental bombers would be "discounted"—that is, treated less stringently than ballistic missile warheads. The discount on bomber weapons constituted an incentive for the two sides to retain forces that were better suited for retaliatory missions rather than ballistic missiles, which pose the threat of a first strike.

To be sure, some crucial matters remained in dispute, such as "counting rules" (how many weapons to attribute to various kinds of launchers) and verification (how one side could assure itself that the other was not cheating). But the broad outlines and the key provisions of the treaty were already established, and even in its unfinished form, START already represented a potential landmark in the 20-year search for "stabilizing reductions."

START appeared to be at least as close to completion in January 1988 as SALT II had been in January 1976, when the secretary of state at the time, Henry Kissinger, traveled to Moscow hoping for a breakthrough that would permit Gerald Ford to sign a treaty before that year’s presidential election. Ford faced a conservative challenge from Ronald Reagan, and some presidential advisers feared that the American concessions necessary to deliver a treaty would hurt Ford’s chances of nomination. Therefore they undercut Kissinger’s effort to advance the negotiations that January. SALT II entered a state of suspended animation until Jimmy Carter became president, and it was not until three years later that a treaty was finally signed.

At the beginning of 1988, START seemed to suffer from no such election-year liability. Public opinion polls showed strategic arms control to be a popular goal, especially given the new atmosphere of optimism in Soviet-American relations and the need for fiscal restraint to cope with the huge federal deficit. It was widely believed that a completed START treaty would only help the Republicans against the Democrats in the fall. Thus there was all the more reason to expect an agreement, perhaps even by early summer.

Moreover, the Washington summit provided the glimmer of a solution to what had become a major conundrum: how to reconcile strategic arms control—and particularly strategic arms reductions—with the Reagan Administration’s determination to develop, eventually test and in due course deploy a large-scale, space-based system of high-technology defenses against ballistic missiles.

III

For 20 years superpower deterrence and diplomacy alike had been based on the morally troublesome, politically controversial but curiously durable proposition that each side had an interest in its adversary’s capability to retaliate against an attack. To be secure, each side must be assured of the ability to survive a first strike. At the same time, however, to be deterred from launching such an attack, each side must remain vulnerable to a second, or retaliatory, blow. The SALT agreements of 1972 and 1979 depended on a tradeoff between strategic offense and defense: under the SALT I Anti-Ballistic Missile (ABM) Treaty of 1972, the United States and the Soviet Union agreed to an open-ended prohibition against nationwide antimissile defenses while they set about first to limit, then to reduce offensive forces.

By the time the Reagan Administration was in office, the assumptions underlying SALT were officially in doubt if not in disgrace. What many experts believed to be the looming vulnerability of the United States to a Soviet first strike undercut American confidence in deterrence and stimulated fresh interest in strategic defense. Conversely, an American quest for near-perfect defense and near-absolute invulnerability under the banner of SDI (announced by President Reagan in March 1983) constituted a powerful incentive for the Soviet Union to increase its offenses rather than reduce them.

Hence there has been in the dynamic of arms control since 1983 a perplexing contradiction between START and SDI. By the time Reagan and Gorbachev met in Washington in December 1987, their governments had been trying for some time if not to resolve that contradiction, then at least to postpone its consequences long enough to achieve progress in START.

Compromise proved difficult. At the center of the negotiation was a Gordian knot of seemingly irreconcilable differences over the future of deterrence. Reagan wanted to make sure that the research, development and testing of SDI would lead to eventual deployment of a system that would revolutionize how global peace was assured. From his standpoint, the sooner R&D gave way to deployment the better. He was determined to prevent what he saw as any Soviet effort to "kill" SDI, such as Gorbachev’s proposal at Reykjavik to limit all SDI research and development to the laboratory.

As for Gorbachev, he was sufficiently eager for a START deal that he adjusted the Soviet position on SDI considerably during 1987. But he never accepted the desirability, much less the inevitability, of a large-scale, fully deployed SDI. He would not insist on killing the program outright, but he clearly wanted to hamper it as much as possible.

Meanwhile many members of the Reagan Administration had their own doubts about the technical feasibility and strategic wisdom of comprehensive, space-based strategic defense in its most grandiose, presidentially sanctioned form. Therefore they found it possible, when Gorbachev came to Washington in December 1987, to make common cause with their Soviet counterparts on compromise language that begged the question of what SDI would turn out to be and how it would be governed by the ABM treaty.

In the communiqué released at the end of the Washington summit, the two leaders agreed on a de facto if not de jure way of updating the long-standing tradeoff between offense and defense while setting aside the hard issues of SDI. The most important, and most contorted, passage in the communiqué was the statement that the two sides would "observe the ABM treaty, as signed in 1972, while conducting their research, development, and testing as required, which are permitted by the ABM treaty, and not to withdraw from the ABM treaty, for a specified period of time." The commas and subordinate clauses left deliberately vague what level of testing would be permissible during the period of nonwithdrawal.

The American officials responsible for what became known as "the Washington finesse" were trying to make a virtue of a necessity. Five years after Reagan’s so-called Star Wars speech, the prospects for the SDI program bore little resemblance to Reagan’s original dream of an impregnable shield that would transform the nature of deterrence by rendering offensive weapons, in the president’s famous phrase, "impotent and obsolete."

A majority in Congress had resisted the Administration’s requests for large funding increases and ruled out SDI tests that would violate the ABM treaty. Secretary of Defense Caspar Weinberger, who resigned shortly before the Washington summit, had failed in his drive for early demonstrations of SDI technology and early deployment of an actual system. No more successful had been the presidential campaign of Jack Kemp, the one candidate who had made all-out advocacy of Star Wars a major theme in his bid to succeed Reagan.

The new secretary of defense, Frank Carlucci, called in General James Abrahamson, the director of the SDI Organization, and told him that from now on he would be a program manager, no longer a salesman. In testimony on Capitol Hill, Carlucci offered assurances that tests scheduled through 1991 would "probably not raise the [ABM] treaty compliance issue."

Suddenly everyone seemed to be looking for ways to scale back the goals and cost of SDI without killing the program entirely. The uniformed military had already closed ranks behind advocacy of a system that might stop about half of the Soviets’ heavy ICBM warheads and only 30 percent of the entire missile force in a massive first strike. Such a defense would, according to its advocates, complicate the Kremlin’s calculations and thereby bolster deterrence, but it was a far cry from what the president had originally wanted and still seemed to want in early 1988.

Senator Sam Nunn of Georgia, the influential Democratic chairman of the Armed Services Committee, proposed something much more modest—what he called an accidental launch protection system, a land-based defense against stray missiles.

Almost no one wanted to be an apologist for the idea of leaving the United States forever entirely naked to nuclear attack. The new fashion in strategic defense, however, was minimalism. One of the oldest, most legitimate, most difficult, and still unanswered questions of the nuclear age was now back in focus: Was there a role for strategic defense in keeping the peace? Would it be possible for one side to deploy defenses in a way that would increase its ability to withstand a first strike without decreasing the other side’s confidence in its own ability to carry out a second strike in retaliation against an attack? Could limited defense be put to work in the service of offense-dominant deterrence in a way that would bolster mutual deterrence and mutual security?

These were genuine questions, quite different from the declaration of faith in the deus ex machina of pure defense that had driven SDI. Even with President Reagan, the ultimate SDI enthusiast, still in office, all eyes were turning to the post-Reagan era. And in the eyes of almost everyone except Reagan, SDI had reached the point that the ABM program had reached 20 years earlier: the order of the day was a "thin" defense, whether against a fraction of the Soviet missile force or against a lone missile fired by mistake, by a rogue colonel or by a Third World madman. Just as the adoption of the "thin ABM" in the late 1960s had laid the ground for the offense-defense tradeoff in SALT in 1972, so the emergence of a loose consensus in favor of a thin SDI in 1987-88 seemed to make possible a similar breakthrough in START.

IV

But it was not to be, at least not by the spring, as Reagan and Gorbachev had hoped. They kept to their timetable for a fourth summit by mid-1988. The meeting was held in Moscow May 29 to June 2. But the president and the general secretary were unable to use the meeting to sign a final agreement in START. Instead, they had to settle for another joint statement. This time it was one that indicated progress only on minor matters while acknowledging that "serious differences remain on important issues."

A number of factors combined to dash the hopes of the two leaders. One problem was that the INF treaty had run into unexpected trouble in the Senate. Some of the opposition came from the predictable quarter of ultraconservative senators like Jesse Helms (R-N.C.). He and other diehards on the right used the INF treaty as a kind of voodoo doll in which they could stick pins in order to inflict pain on the body of arms control generally and START in particular. Seizing on discrepancies among estimates of how many undeployed, uncounted INF missiles the Soviets had, Senator Helms accused the State Department of "cooking the books" to conform with Soviet lies about their missile inventory.

Moderates, too, had their reservations. Senator David Boren, the Democratic chairman of the Select Committee on Intelligence, contended that the government would have to improve its coverage of the Soviet Union with spy satellites to monitor compliance with the treaty.

The Administration’s resulting defensiveness over a treaty that it had already signed and the long delay in its ratification was a drag on progress in START. Not until Reagan had already departed for Moscow in late May did the Senate finally (though overwhelmingly) approve the INF treaty.

The major cause of the long stall in START in 1988 was what amounted to a revolt by the military leaders against their civilian colleagues over the conduct of arms control.

The Joint Chiefs of Staff and their chairman, Admiral William Crowe, had never seen themselves—or been seen by others—as in any sense against arms control. In that respect there was a marked contrast between Crowe and the Joint Chiefs on the one hand and their former boss, Caspar Weinberger, on the other. Weinberger had spent much of his nearly seven years as secretary of defense thwarting, or trying to thwart, the pursuit of negotiable agreements.

In an interview on June 3, 1988, Crowe summed up his differences with Weinberger: "I was never sure he was for an agreement, and I knew I was." Crowe recalled that earlier in the year he had assured Secretary of State Shultz that the Joint Chiefs wanted to do everything they could to "advance the endgame" in START. Officials at the State Department confirmed that they regarded Crowe as essentially friendly to the objective of arms control even when there were differences over tactics and details.

But as the Moscow summit approached during the spring of 1988, those differences, concerning the pace as well as the substance of negotiations, became very sharp indeed. The Joint Chiefs still saw arms control as a useful auxiliary to national security policy. But in their view, national security policy suffered from too much confusion, uncertainty and dissension for arms control to go forward at anything like the forced-march pace that the State Department and White House wanted. The Joint Chiefs were in a position to slow the process in part because the Reagan Administration, wounded by the Iran-contra affair and other setbacks, was limping into its last year in office. They saw an opportunity, and a need, to assert themselves—to insist on a pause in arms control and a reassessment of the nation’s strategy and weapons programs.

The diplomatic process had been under way for most of the last six years, but only recently had the Joint Chiefs of Staff faced up to its ramifications. They could perhaps be forgiven for their delayed reaction. For much of the Reagan presidency—from 1981 through most of 1986—the U.S. military leaders, like many others, had not taken very seriously the possibility that the Geneva arms negotiations would actually produce a strategic arms treaty that would concretely affect American forces. For years START had seemed doomed to stalemate, perhaps even deliberately so in the minds of some hard-liners on the American side. The largely sterile exchange of what were widely considered nonnegotiable proposals between Moscow and Washington appeared to have little consequence for the real world of procurements and deployments on which the military chiefs concentrated their attentions and energies. There had even been a total collapse of the negotiations that lasted for more than a year—from the Soviet walkout at the Geneva arms control talks in late 1983, after the U.S. INF deployment in Europe, until the resumption of negotiations in early 1985.

Not until late 1986 and early 1987 did the Joint Chiefs begin to pay close attention to what was going on in arms control and its near- and middle-term implications for American defense programs. They did not like much of what they saw, particularly in the confused record of the hectic, largely improvisational, and highly ill-considered bargaining that went on between Reagan and Gorbachev at Reykjavik in October 1986. In the dizzying last hours of that encounter, President Reagan embraced the idea not only that strategic ballistic missiles, like the intermediate-range missiles of INF, should be reduced to zero but that this giant step toward nuclear disarmament might be accomplished within ten years.

True, the so-called zero-ballistic-missiles ingredient of the Reykjavik package was contingent on American concessions in SDI that the president was not willing to make (Gorbachev’s "killer" insistence that the research-and-development program be confined to the laboratory). Nonetheless, the willingness of the president to agree, however conditionally, to the elimination of the most powerful of all weapons and the backbone of the American deterrent came as a rude shock to the Joint Chiefs of Staff. They were profoundly doubtful that the United States could assure its security and that of its allies without ballistic missiles.

What was more, they were already fighting an uphill battle in Congress on behalf of a number of new ballistic missile systems (the MX ICBM and the Trident II SLBM). They feared that their selling job would be all the harder now that the commander in chief had endorsed the idea of eliminating all such systems in ten years.

The Joint Chiefs were instrumental in making sure that the zero-ballistic-missiles idea ended up on the cutting room floor during the aftermath of the Reykjavik meeting. They also had qualified reservations about the plan for 50-percent reductions. The Joint Chiefs saw the emerging START treaty as featuring ceilings and subceilings that could indeed be conducive to strategic stability. But whether START in fact turned out to be stabilizing would depend on how the United States chose to fill its allotment under ceilings of 1,600 launchers, 6,000 nuclear charges, 4,900 ballistic missile warheads, and so forth. More specifically, it would depend on what steps the United States took to assure that its reduced number of strategic ballistic missiles was less vulnerable than before to the threat of a Soviet attack.

The military leaders believed that invulnerability meant some sort of mobility for land-based missiles. Stationary missiles in underground silos were simply too easy for highly accurate Soviet warheads to destroy. This was one reason why only 28 of the nation’s newest missiles, the ten-warhead MX, had been deployed in silos. The Air Force was moving ahead with a plan to deploy the MX on railroad cars, but it ran into resistance. Key members of Congress, opposed to MIRVed missiles in general, were pushing the idea of developing a small, mobile, un-MIRVed missile called Midgetman. Its single warhead would permit the deployment of a large number even under the agreed START ceilings. The Joint Chiefs, however, believed that Midgetman, precisely because it would be designed to carry only one warhead per launcher, would turn out to be too expensive. They were concerned that funds for developing it would have to come out of other corners of the defense budget. This in fact happened in the current budget, when Congress took funds for Midgetman out of the MX account.

The worst outcome, Crowe feared, was that enthusiasts for Midgetman in Congress would block any further deployment of the MX, particularly as a mobile missile, while the cost-cutters on Capitol Hill would block the Midgetman, and the United States would end up with no mobile intercontinental missiles at all. Meanwhile, the Soviets would continue their deployment of two mobile systems: the SS-24, roughly comparable to the MX, and the SS-25, similar to the Midgetman.

Partly because that fear was shared elsewhere in the government, since 1985 the Administration had been formally proposing to ban all mobile ICBMs in the START treaty. That ban remained part of the official American position in 1988.

Thus, there were multiple contradictions on the American side over what the United States felt it could and should do in its mobile ICBM program; there were contradictions within the executive branch and between the executive and legislative branches, as well as between U.S. military programs and the American negotiating position in Geneva.

For this reason Admiral Crowe repeatedly warned the president in early 1988 that it would be a mistake even to try to complete a START treaty until the United States had a much clearer idea of what sort of ICBM program made military, economic and political sense.

Various influential voices outside the Administration, notably including Senator Nunn’s, echoed that caution. These concerns seemed to carry the day with the Reagan Administration, dampening hopes for a START treaty in 1988 and providing a rationale for a period of reassessment in U.S. military policy before diplomacy could resume in earnest.

V

While the Reagan Administration was unable to reconcile differences at home over its ballistic missile program, the negotiations with the Soviets were encountering another difficulty. The reduction of strategic ballistic missiles, mobile or otherwise, was not the principal bilateral problem contributing to the slowdown in START. Instead, the weapon that represented what seemed to be an almost insurmountable obstacle to a final agreement by the time of the Moscow summit was a small, low-flying, slow-flying drone. This was the nuclear-armed, sea-launched cruise missile, known as the SLCM (pronounced "slickum"), which could be deployed on both surface ships and submarines.

SLCMs had been around for a long time as a weapon system and also as a challenge to arms control. The Ford and Carter Administrations had haggled over SLCMs, both within the corridors of the U.S. government and with the Soviets. In SALT II an agreement was reached to defer SLCM deployments for three years; after this special protocol lapsed, the United States began some deployments. The United States enjoyed a significant lead in the miniaturized guidance and propulsion systems for all cruise missiles, and partly for this reason the Soviets wanted to limit them severely. Ground-launched cruise missiles were banned by the INF treaty, and START yielded a formula that would permit air-launched cruise missiles. SLCMs, however, seemed to defy compromise and agreement.

Marshal Sergei Akhromeyev, the chief of the Soviet General Staff, was particularly adamant that a START treaty must include stringent limits on SLCMs. At the Reagan-Gorbachev meeting at Reykjavik in October 1986, Akhromeyev, who served as the chairman of the Soviet "working group" on arms control, said that reductions in ballistic missiles would be meaningless and indeed unacceptable if the United States retained the right to surround the U.S.S.R. with nuclear-armed SLCMs. It took more than an hour at Reykjavik to agree on a joint statement saying no more than that the two sides would work toward a mutually agreeable solution.

But not until 1988 did it become apparent that what for years had been a worrisome detail in strategic arms control might, at the eleventh hour, become a treaty-blocker if not a treaty-buster. That occurred in part because, since Reykjavik, START itself had evolved from a largely abstract, propagandistic exercise into a negotiation that finally—and, to many, surprisingly—seemed about to produce an agreement. The more movement there was on other matters, the more serious the issue of SLCMs became.

Max Kampelman, the chief American arms control negotiator in Geneva and the counselor of the State Department, is a distinguished Washington attorney. In an interview on May 12, 1988, he compared what was happening in START to his own earlier experience with last-minute, drawn-out snags in contract law: "After you’ve got the major issues in your pocket, the minor issues become major issues."

SLCMs were a particularly nettlesome example of this phenomenon because the U.S. Navy had already begun deploying nuclear-armed long-range SLCMs. Thus many in the military had acquired a vested interest in the continuation of the program.

Nonetheless, other American military experts argued that the United States should accept stringent restrictions on SLCMS while the Soviets were still interested. The American technological edge would almost inevitably prove temporary, while geographical asymmetries between the superpowers were permanent—and permanently favorable to the Soviet Union. Key American cities and military installations were near the coasts, and therefore easy marks for Soviet SLCMs, while many comparable Soviet targets were deep inland and protected by the most extensive air defenses in the world. A number of NATO officials, particularly from Norway and Canada, made the same point, and Senator Nunn said in an interview on May 24 that he found this view "most persuasive."

Another consideration militated against nuclear-armed SLCMs. By any standard, the United States’ navy was superior to the Soviet Union’s. America could out-fight the Soviet Union on the high seas—as long as the war was confined to conventional forces. However, if the Soviets "went nuclear" in battles at sea, American aircraft carriers and battleships might disappear in mushroom clouds. In that event, the Soviets would have used their naval nuclear weapons as equalizers to compensate for America’s maritime superiority, much as the United States counted on its own nuclear weapons to make up for NATO’s numerical inferiority in conventional land-based forces in Europe. Therefore, from the American standpoint, it was better to raise the so-called nuclear threshold as high as possible in the naval arena; that meant no nuclear-armed SLCMs.

Paul Nitze, the senior adviser to the president and Secretary Shultz on arms control, was the principal proponent within the Administration of the view that possession of nuclear-armed SLCMs by both sides made those weapons "inherent losers" for the United States. As secretary of the navy in the Johnson Administration, he had been responsible for removing nuclear-armed carrier-based aircraft from the so-called Single Integrated Operational Plan—the Pentagon’s top-secret, highly detailed master plan for waging nuclear war against the Soviet Union. Nitze had concluded that carrier-based planes were unlikely to survive long enough to be of any use in a nuclear war. Recalling that decision of a quarter-century earlier, he asked a number of admirals in 1988 whether the U.S. Navy would not be better off with a simple, radical measure—a ban on nuclear-armed SLCMs, as well as on nuclear-armed torpedoes, depth charges, surface-to-air missiles and other such weapons. Each side had roughly 2,000 such "nonstrategic" naval nuclear weapons that would be eliminated by a ban. The only nuclear weapons remaining in the superpowers’ navies would be submarine-launched ballistic missiles.

Nitze hoped that American willingness to do away with nuclear-armed SLCMs would be a central part of what he called a "more forward-leaning" package of new American START proposals that Secretary Shultz could take to Moscow for a ministerial meeting with Shevardnadze at the end of February 1988.

But Secretary of Defense Carlucci and the Joint Chiefs of Staff vetoed the idea. Carlucci regretted that the Pentagon had ever agreed even in principle to put SLCMs on the agenda for START. He and the Joint Chiefs believed that the United States had to preserve the option of deploying ship-to-shore (or "land-attack") nuclear-armed SLCMs as a means of bolstering nuclear deterrence in Europe, especially now that American intermediate-range missiles were going to be removed as a result of the INF treaty. Therefore, they concluded, START should permit each side more than 700 nuclear-armed SLCMs as well as more than 3,000 conventionally armed ones.

That should have meant coming up with a verifiable way of differentiating between the two. SLCMs are small and easy to hide—it is almost impossible to distinguish the nuclear-armed version from the conventionally armed one at a distance—and, despite some vague Soviet offers, neither side would probably be willing to permit the other to conduct on-board inspection of its ships.

At the conclusion of the Washington summit in December, Gorbachev had touched off a minor stir by claiming at a press conference that it was technically feasible for one side to determine from a distance whether the other side’s ships had nuclear-armed SLCMs on board. Some of his scientific advisers had encouraged him in the belief that such remote monitoring was possible. American experts were extremely doubtful. They were sure that it would be possible to shield nuclear warheads so that the rays they emitted could not be detected by the other side’s remote sensors. A number of Soviet military experts shared their skepticism.

The Pentagon’s solution was for each side simply to declare how many SLCMs it planned to deploy. This made a mockery of the long-standing American insistence that each side must be able independently to monitor the other’s compliance with all provisions of an agreement. In his meetings with Gorbachev, President Reagan was given to repeating a Russian proverb, Doverai no proverai ("Trust but verify"), as the watchword for arms control. Yet where SLCMs were concerned, the American position was: Trust, period.

Secretary Carlucci’s way of dealing with this awkwardness was to assert, in a television interview with the Cable News Network on May 30, during the Moscow summit, that SLCMs were not really strategic weapons anyway.

If there was to be a breakthrough in START in time for Reagan to sign a treaty before he left office, there would have to be a capitulation by one side or the other. The United States would have to decide that it could live without nuclear-armed SLCMs after all, or the Soviet Union would have to decide it could live with large, probably uncountable numbers of those weapons aboard American ships after all. Neither outcome seemed likely.

Another option—favored by some arms control advocates on both sides—would be to finesse the SLCM issue in somewhat the way the SDI problem had been finessed at the Washington summit. One possibility would be to set SLCMs aside for the time being, putting them on the agenda of a new, separate negotiation, much as intermediate-range nuclear forces (including various kinds of cruise missiles) had been split off from SALT in the 1970s.

However, through the Moscow summit and into the summer, the Soviet military was not willing to exempt SLCMs from START. Unlike SDI, SLCMs represented a technology that was already available—indeed, a weapon system that was already being deployed. Therefore the Kremlin continued to insist that SLCMs be subject to low, verifiable limits as part of a START treaty.

VI

The impasse over SLCMs might have been resolved had there been more ability and determination on the part of the Administration to forge a consensus about the future of U.S. military programs. As 1988 unfolded, it became increasingly clear that the division within the American government was not so much over how START should deal with a particular issue, such as mobile ICBMs or SLCMs, as it was over how realistic it was to expect, or how prudent to attempt, to have any treaty at all before the Administration left office. Admiral Crowe’s conviction was growing that the difficulty in what he called "getting our own act together" was too deep-rooted for there to be a treaty by June. Meanwhile, on visits to Capitol Hill, Secretary of Defense Carlucci was hearing warnings to the effect that START would encounter even more trouble than the beleaguered INF treaty.

These congressional misgivings about the prospective treaty seemed mild and diffuse compared to the drumbeat of criticism already coming from a number of distinguished private citizens who, as government officials in the 1970s, had been closely associated with SALT but who were now engaged in something close to a preemptive strike against START.

In a newspaper column in mid-January Henry Kissinger warned: "The current nuclear arms control negotiations are too one-sided." He argued that unless the United States restructured its own forces considerably by "going mobile" with its land-based missile force and dispersing its submarine-launched missiles among a larger number of smaller boats than navy programs then envisioned, the American deterrent would end up no less vulnerable to a Soviet first strike than before—and perhaps more so.

Kissinger was, in effect, blaming START for a state of affairs that had come about as a result of unilateral American decisions—or, in the case of mobile ICBMs, indecision. This had also been a habit of critics of arms control back in the days of SALT, when Kissinger was a negotiator and defender of the agreements in question.

For months, the Reagan Administration seemed to be biting its tongue. It concentrated its forensic powers on defending the INF treaty until after that document was safely ratified. Not until June 21, in an article in The Washington Post, did Paul Nitze reply, on behalf of the Administration, to Kissinger and other critics. Nitze contended that they were placing an unfair burden on arms control. It was up to the United States to decide how most sensibly—i.e., least vulnerably—to base its ballistic missiles on land and at sea. "No foreseeable arms control agreement can alone solve" the problem of vulnerability, he wrote. Besides, he continued, because of the drawdown in Soviet warheads and throw-weight, "START would reduce the threat to whatever system we chose" for the basing of American missiles.

There was still, however, a necessity for choice. In their own deliberations on START, the Joint Chiefs of Staff were stressing that the treaty would be stabilizing only if the United States took steps to assure that the 4,900 ballistic missile warheads allowed by the treaty would be deployed in a fashion that assured their ability to survive a first strike.

Reagan’s military advisers came to believe, early in the year, that too many important military choices were left hanging, or were being rushed, for the diplomats to proceed. In late January, just as Secretary Shultz was about to undertake the first in a series of visits to Moscow to prepare the way for the midyear summit, Carlucci warned, during a White House meeting, against going into a "two-minute drill." Shultz replied that the United States would never know what could be accomplished without trying, and Howard Baker, the White House chief of staff, sided with Shultz. The president agreed. Borrowing a phrase from the Winter Olympics then under way in Calgary, he said that the United States should "go for the gold."

But the negotiators needed more than a presidential blessing. A breakthrough required fresh initiatives—on mobile ICBMs, SLCMs and other sticking points—and there were none. When a group of influential senators, including Sam Nunn and Robert Byrd, the Democratic majority leader, visited Geneva in early February, they found the American and Soviet diplomats there discouraged by the treadmill-like quality of the negotiating process. On their return to Washington, the senators called at the White House and conveyed to Reagan and Baker their pessimism about the chances for an agreement in time for the summit. Nunn used the occasion to reiterate his reservations about whether it was wise to proceed toward a START agreement when there was still so much division, uncertainty and confusion on the American side over the future of American strategic offensive programs, including both mobile and cruise missiles.

The president decided that the time had come to lower expectations for the summit and stress that he was more interested in a good deal than an early deal. In an interview with Lou Cannon of The Washington Post on February 25, Reagan said:

I have to tell you that common sense indicates that the time is too limited for us to really think that we could bring a treaty ready for signature to [the summit]. . . . This one [START] is so much more complicated with regard to verification and everything else than the INF treaty, which we were able to bring together. But even that took a few months to do. So we’re not going to be disappointed . . . we’re not at this moment anticipating that it would be ready for signature then.

This statement was read with relief at the Pentagon and dismay at the State Department. Advocates of an all-out push, under the pressure of a deadline, had publicly lost their main sponsor. The prospect for a START treaty in 1988—seemingly so bright after the Washington summit—now seemed bleak indeed as the Moscow summit drew closer.

Presidential impetus is crucial in arms control, even in the best of times. An election year and the final months of a lame-duck administration are not the best of times, especially if the chief executive has never been deeply engaged in arms-control policymaking and, with his eye on the exit, has recently been disengaging more than ever. Such has been the case with Ronald Reagan. He seemed to conclude that while a START agreement would have been welcome, it was not indispensable, as his crowning achievement. Throughout much of 1988, his involvement in START was largely a matter of saying "good point" to those, like Crowe and Carlucci, who wanted to go slow in the negotiations and "good luck" to those, like Shultz and Baker, who wanted to press ahead. In the delicate and complicated machinery of arms control, the brakes tend to work better than the accelerator.

VII

For three months—March, April and May—there was some progress toward measures to verify limits on mobile ICBMs as well as toward a compromise on how to count air-launched cruise missiles. But largely because of its own ambivalence about mobile ICBMs, the United States was still formally holding out for a ban on such systems, and negotiations over SLCMs remained at an impasse. Shultz came to feel that the best the two leaders could do at the summit would be to "take a photograph of where we are" by consecrating the work that had already been done on the joint draft text of a START treaty, leaving it to their subordinates, if not their successors, to remove the remaining brackets indicating points of disagreement.

En route to Moscow, Air Force One stopped in Helsinki for three days to give the president a chance to catch up on jet lag. During the layover, on May 28, the president’s national security adviser, General Colin Powell, gave a briefing for the traveling White House press corps. He spelled out his own hope for the summit, acknowledging what had become the number-one sticking point: "If we make some progress on sea-launched cruise missiles—if not a breakthrough, we might make sufficient progress that we can return to work in Geneva, finish it over the summer, and in the fall at a ministerial meeting or higher, should that come about, we could go to closure on it."

Powell acknowledged that in addition to the remaining points of contention between the United States and the Soviet Union, there was also "a level of negotiations that takes place in Washington within the Administration as we debate out positions and make sure that the positions we come down on to present to the Soviets are acceptable to all of the interested parties in Washington. . . . There are outstanding issues within the Administration."

Both sets of negotiations—between the two governments and within the American government—were grinding to a halt, well short of the finish line that Reagan had hoped for as recently as February.

There was no breakthrough—in fact, no progress at all—on SLCMs at the summit, and only the most modest, inconclusive progress on other, lesser issues. For all the important political symbolism of Ronald Reagan strolling through Red Square arm-in-arm with Mikhail Gorbachev, the theme of nuclear diplomacy at the meeting was in a distinctly minor key.

There was a sense of fatigue on the American side and of impatience on the Soviet side—not just restlessness with the same old disagreements, but eagerness to deal with a new American leadership. After the summit, in July, a number of Soviet spokesmen and officials, including two personal advisers to Gorbachev, remarked privately in Moscow that their side had come to a political judgment: the Kremlin had done nearly all the arms-control business it could usefully accomplish with the Reagan Administration.

VIII

The Soviets may be right in betting that George Bush or Michael Dukakis will move relatively quickly to finish what he inherits in START. They may be right that there will be no return to square one in strategic arms control of the sort that occurred early in both the Carter and Reagan Administrations, infuriating the Kremlin and delaying agreement for years.

In many of their campaign statements about arms control, both Bush and Dukakis have had kind words for the broad outlines of the prospective START deal. The candidates and their advisers have cited strategic arms control as one area of foreign and defense policy where there is likely to be a high degree of continuity once a new administration takes office.

The Soviets may also be right that either Bush or Dukakis will have a relatively easy time shepherding a START treaty to ratification, past the ideologically motivated opposition of the ultraconservatives and the more technical concerns of critics like Kissinger. A new president, after all, will presumably have the benefit of a honeymoon with Congress and public opinion. He would also have a higher level of energy and more willingness to crack heads among his advisers, including his military advisers, than has been the case in the waning months of Reagan’s presidency.

But even if the next president decides that he wants a START treaty along the lines of what has already been negotiated, he will still have to decide what the small print in that treaty should say about mobile ICBMs and SLCMs, and what an accompanying document should say about SDI. Those weapon (and antiweapon) systems are not—or at least should not be—bargaining chips. Their future should not be decided on the basis of arms control tactics. They are matters of defense policy and military strategy, and they should be treated as such.

Nor will the conflict over those issues go away with the inauguration. Admiral Crowe will still be chairman of the Joint Chiefs when Bush or Dukakis takes office in January. If the Democrats maintain control of the Senate, Sam Nunn will still be chairman of the Armed Services Committee.

Implicit in the cautionary advice that Crowe and Nunn have been giving Ronald Reagan for much of this year is a message for the next president as well: before a new administration can pick up where the old one leaves off in START, it should impose some order on the chaos of U.S. official thinking (including that in Congress) about mobile ICBMs; it should decide whether there is a militarily sound mission for nuclear-armed SLCMs; and it should develop a plan for further research and development of strategic defenses that realistically reflects available funding and scientific promise (neither of which is unbounded).

This year has been as instructive as it has been suspenseful. In addition to running the gamut from high hopes after the Washington summit to muted disappointment after the Moscow summit, 1988 has offered an object lesson in the need for the nation’s political leaders to have clearly in mind a coherent, sustainable military program before they can conclude comprehensive and prudent agreements with the other superpower.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • Strobe Talbott is the Washington bureau chief of Time magazine. His latest book, The Master of the Game: Paul Nitze and the Nuclear Peace, will be published by Alfred A. Knopf in October.
  • More By Strobe Talbott