As pressure mounts to reckon with Iran's nascent nuclear program, some strategists are arguing that the United States has run out of alternatives to military action. Many of them are pointing to Israel's 1981 air attack on Iraq's Osirak reactor as a model for action--a bold stroke flying in the face of all international opinion that nipped Iraq's nuclear capability in the bud or at least postponed a day of reckoning. This reflects widespread misunderstanding of what that strike accomplished. Contrary to prevalent mythology, there is no evidence that Israel's destruction of Osirak delayed Iraq's nuclear weapons program. The attack may actually have accelerated it.

Osirak is not applicable to Iran anyway, since an air strike on a single reactor is not a model for the comprehensive campaign that would be required to deal, even unsatisfactorily, with the extensive, concealed and protected program that Iran is probably developing. As the United States crafts non-proliferation policy, it should soberly consider the actual effect of the Osirak attack and the limitations of even stronger air action.

In contrast to a ground war, air power has the allure of quick, clean, decisive action without messy entanglement. Smash today, gone tomorrow. Iraq's nuclear program demonstrates how unsuccessful air strikes can be even when undertaken on a massive scale. Recall the surprising discoveries after the Iraq War. In 1991 coalition air forces destroyed the known nuclear installations in Iraq, but when UN inspectors went into the country after the war, they unearthed a huge infrastructure for nuclear weapons development that had been completely unknown to Western intelligence before the war.

Obliterating the Osirak reactor did not put the brakes on Saddam's nuclear weapons program because the reactor that was destroyed could not have produced a bomb on its own and was not even necessary for producing a bomb. Nine years after Israel's attack on Osirak, Iraq was very close to producing a nuclear weapon. Had Saddam been smart enough in 1990 to wait a year longer, he might have been able to have a nuclear weapon in his holster when he invaded Kuwait.

There are two methods for developing fissionable material for a nuclear weapon. One is to reprocess spent fuel from a nuclear reactor like Osirak into fissionable plutonium. In order to reprocess the fuel from Osirak on a significant scale, the Iraqis would have needed to construct a separate plutonium reprocessing plant. Many laymen commonly assume the effectiveness of the Israeli strike because they mistakenly believe that a nuclear reactor alone can produce explosive material for a bomb. Iraq had made no move toward building the necessary reprocessing facility at the time the Israelis struck the reactor. Without such a separate plant, the destruction of the reactor was practically superfluous.

In fact, a reactor is not even essential for developing a weapon--it is simply one building block for one option. Destruction of Osirak did nothing to impede the separate development project that brought Iraq to the brink of weapons capability less than a decade later. Iraq went on to a fast-paced weapon-development effort by choosing the route toward the enrichment of natural uranium. This is the route that Iran now appears to be taking. Western intelligence did not detect Iraq's enrichment facilities when Saddam Hussein was actively developing a nuclear capability during the 1980s.

If anything, the destruction of the reactor probably increased Saddam's incentive to rush the program via the second route. It is unlikely that Saddam would have been able to develop nuclear weapons much faster through the Osirak reactor--given that he would have had to plan, construct and operate a reprocessing plant--than through enrichment. Israel's preventive strike was not an example of effective delay.

The Israelis' thinking was certainly not mindless. They may have believed that Iraq's building of a reprocessing plant was just a matter of time. It was sensible to assume that Saddam wanted a nuclear weapon and that he intended to use Osirak's output to build one, after a reprocessing capability would have made it possible to extract the plutonium at some point years later. If so, the reprocessing plant, rather than the reactor, would have been the appropriate target.

Israel's strike on Osirak did not pre-empt an imminent threat. The Israelis understood in 1981 that Osirak's threat lay in the future. If the reactor was ever to be bombed, however, it made sense to do so before its construction was complete, since once it was in operation its destruction would have spread radioactivity to the surrounding area. But the strike strategy entailed a provocative action before it was certain that plutonium reprocessing was in fact going to be developed. After the strike, Prime Minister Menachem Begin made a gaffe, for which he later had to apologize, when he claimed the Israeli planes had destroyed a secret underground laboratory in the reactor, forty meters beneath the earth. Had the Iraqis actually constructed such an underground chamber, its existence would have made a case for the attack, since it might have hidden a small-scale reprocessing capability. Some analysts speculated that Begin had suffered a "senior moment", confusing Osirak with Israel's own secret underground facility, which it had constructed long before at the Dimona reactor site to conceal its nuclear weapons program.

Iraq's Osirak-era capabilities were not remotely comparable to Iran's current nuclear program, which is far more advanced. Today's Iran has also been on notice for a long time that it is in the crosshairs of American military planners. It would be surprising if strategists in Tehran have failed to disperse and conceal important facilities in the interests of frustrating U.S. intelligence collection. An American air campaign could easily destroy all identified or suspected nuclear facilities--at least any not located in very deeply buried bunkers--but attack planners could not be sure that all crucial facilities had been hit, because they could not be confident that all had been found.

Desperation or bravado has led some strategists to question that reality. While they recognize that an air campaign would not guarantee full destruction of Iran's nuclear capability or even prevent Iran from rebuilding, they reason that it could at least delay the program. The question remains, then, would a strike that was successful in wiping out a big chunk of Iran's program be more effective than Israel's venture in 1981?

With more to destroy than in Iraq back then, the evolving Iranian program might be more disrupted, but by the same token more hidden capabilities might survive. When it comes to nuclear weapons, the key is not how much capability a preventive attack eliminates, but how much it does not. Unless the Iranians have been extremely negligent, they have not left all of their enrichment capacity in locations accessible to American intelligence collectors and Air Force targeters.

Advocates of an air assault tend to take comfort in the proposition that the destruction of a major portion of Iran's nuclear establishment would set back acquisition of weapons by many years. When asked what to do when Iran picks up the pieces and starts over again, they echo the argument of General Curtis LeMay, who advocated the preventive destruction of China's industry in the early 1960s. When Ambassador Averell Harriman asked LeMay what the United States should do when China rebuilt its capability, he said, "Hit 'em again."¹

Political, diplomatic and military obstacles to taking action in Iran have been well recognized. Strategists who think of themselves as stalwart, steely-eyed and far-seeing regard these obstacles as challenges to be simply overcome or disregarded in order to do what is necessary, even if it is less than a perfect solution. But if bombing known nuclear sites were to mean that Tehran could only produce a dozen weapons in 15 years rather than, say, two dozen in ten years, would the value of the delay outweigh the high costs? The costs would not be just political and diplomatic, but strategic as well. Provoking further alienation of non-Western governments and Islamic populations around the world would undermine the global War on Terror. Inflaming Iranian nationalism would turn a populace that is currently divided in its attitudes toward the West into a united front against the United States. Rage within Tehran's government would probably trigger retaliation via more state-sponsored terrorist actions by Hizballah or other Iranian agents.

The military option that is possible would be ineffective, while the one that would be effective is not possible. The military action that would work--an invasion of Iran--cannot be done, since America's volunteer army has already reached the breaking point in handling missions less challenging than subduing Iran would be. The only means of definitively preventing Iran from acquiring nuclear weapons would be occupying the country forever. This would ensure that the regime we install remains compliant with American judgments about what the country does or does not need for its own security in a dangerous neighborhood. One might note in passing that there is no reason to assume that the reformed Iraqi government the United States is struggling to stand up will not revive a nuclear weapons program if U.S. forces were ever to allow it genuine independence.

What else should Washington do? There is no good answer. The crusade to keep all second-rate powers from acquiring a nuclear weapon--which, we should remember, is now sixty-year-old technology--can succeed in some cases for some time, but it is ultimately a rear-guard action. Regarding both Iran and North Korea--which is probably more dangerous than Iran yet somehow has slipped into second place in proclamations of alarm--the two answers may be unsatisfactory but are less unsatisfactory than other conceivable options.

One answer is the prosaic, two-track strategy of political and economic carrots and sticks. Tehran should be offered diplomatic concessions if it comes back into full compliance with requirements of the Nuclear Non-Proliferation Treaty (NPT), including the more intrusive inspections mandated under the Additional Protocol. Concessions would be a difficult choice, since they would compromise the War on Terror, given Iran's bad record as a state sponsor. But limiting the nuclear threat is a higher priority.

It is also true that this approach is no foolproof guarantee of non-proliferation. Iran could presumably cheat and maintain clandestine weapon-development programs. Still, it would provide the only incentive for Iranian restraint, and inspections would at least complicate and impede concealment of illicit activity.

If, on the other hand, Tehran fully disengages from the obligations of the NPT, the United States should promote multilateral tightening and extension of economic sanctions. Unfortunately, the self-inflicted wound of the invasion of Iraq poisoned the well for convincing fence-sitters in other countries to sign on to such measures. America's credibility regarding the threat posed by Iran has been weakened by its unsubstantiated claims of Iraq's weapons of mass destruction.

The second component--which is unsatisfactory but better than the alternatives--is to replicate the Cold War strategy of containment and deterrence until such time that the regime in Tehran mellows or is replaced from within. Many today forget that Stalin's Soviet Union and Mao Zedong's China were seen as more threatening in both capabilities and intentions than are today's mullahs in Tehran. For reasons remarkably similar to those proclaimed today, alarmed American strategists discussed the option of preventive war against the USSR in the 1950s and against China in the 1960s. Fortunately, the U.S. government rejected those ideas. Then as now, it was risky to tilt towards the hope that steady defensive resistance--rather than aggressive military action--would hold the line until enemies eventually reformed or stood down. Now as then, that risk is uncomfortable but remains the best among bad options.

To some in the Bush Administration, this reasoning smacks of defeatism. Which way will the president tilt? Will the born politician in George Bush guess that another military adventure in the midst of the Iraq fiasco would devastate his standing, given that he is already on the ropes; or will the born-again crusader in Bush wager on bold and decisive use of American muscle, in the belief it would restore public confidence?

Neoconservative zealotry has fallen from grace in the second Bush Administration, and a prudent State Department has risen in influence under Condoleezza Rice. Still, the rationale for military action could appeal to some of the stalwart staffers around Bush, Cheney and Rumsfeld. Many of these men's faith in force was annealed in the crucible of September 11 and remains unshaken by the Iraq experience. They confuse caution as a cover for timidity and believe that disregarding liberal conventional wisdom is a mark of Churchillian courage.

Reliance on containment, deterrence and pressure short of force remains unsettling to Americans who seek closure in conflict and suspect that restraint betrays fecklessness. Force has the allure of apparent decisiveness. But the greatest military philosopher, Carl von Clausewitz, warned, "In war the result is never final." Unless victor and vanquished come to agreement on a peacetime order, peace will not endure. Military action might at best suppress Iran's nuclear ambitions temporarily; at worst, and no less probably, an attack could make them more intense and more dangerous.

¹C. L. Sulzberger, An Age of Mediocrity (New York: Macmillan, 1973).

Richard K. Betts is director of the Saltzman Institute of War and Peace Studies at Columbia University. He was a member of the National Commission on Terrorism.