On July 16, 1945, not long after the first atomic bomb was successfully detonated in the desert near Alamogordo in southern New Mexico, Winston Churchill is said to have exclaimed: “What was gunpowder? Trivial. What was electricity? Meaningless. This atomic bomb is the Second Coming in wrath!”

At that time, debate was under way in the United States over whether to unleash the destructive power of nuclear weapons upon the Japanese, and Churchill was directing his cautionary remarks to then-US Secretary of War Henry Stimson. But the decision to use the bomb came quickly in Washington; the country had just suffered a devasting attack on Pearl Harbor. As Stimson later put it, “I felt that to extract a genuine surrender from the Emperor and his military advisers they must be administered a tremendous shock which would carry convincing proof of our power to destroy the Empire.”

More than half a century later, it is difficult to imagine using nuclear arms in the heat of battle, for any reason. Throughout much of the cold war, they were largely justified as instruments of strategic deterrence rather than as weapons usable in combat. Since the end of the cold war new nuclear powers like India and Pakistan have emerged onto the world stage, but the older ones have significantly scaled back their arsenals. Moreover, as the US-led wars in Serbia and Afghanistan have shown, conventional US military force has become so overwhelmingly powerful that Pentagon planners no longer “need” atomic explosives to create the “tremendous shock” required to obliterate hostile regimes.

Yet within the US military establishment, nuclear weapons do not appear to be irrevocably sliding down the path to extinction. Quite the contrary–over the past several years there has been a growing push both within and outside government to make nuclear weapons more “usable,” or pertinent, in a world troubled by terrorism, rogue dictators, crumbling Russian might and ascending Chinese power. Not surprisingly, this push has brought with it a shift in the way some members of the defense community are thinking about nuclear arms and military strategy. The ideas defining atomic weapons–when to use them, how and why–are in flux.

In many ways, the Pentagon’s top-secret nuclear policy review–released to Congress in January but leaked to major media in March–is the culmination of this movement. The review states that countries such as Iraq, Iran, North Korea, Syria and Libya should be added to nuclear targeting plans. It also advocates new, smaller nuclear weapons that would be incorporated into conventional war-making tactics. However, these ideas have long been in the making.

If current policy does not change course, twenty years from now we could experience the following: Rather than pursue the path to total nuclear disarmament, Washington will command a new class of small-scale atomic weapons intended for use on the battlefield. The cold war arsenal will have been substantially reduced, but in case unforeseen threats arise, the deactivated warheads will have gone into storage, rather than been destroyed. Meanwhile, America’s remaining cold war atomic weapons will be targeted not just at Russia but also at an array of developing countries. The conceptual firewall currently separating nuclear weapons from conventional ones will have largely crumbled, and the United States will have openly abandoned its unwillingness to use nuclear weapons against nonnuclear threats.

To see this, one need only examine recent government documents on nuclear policy: the quietly expanding production of components required to build new atomic bombs; the push to resume nuclear testing within a year or sooner; and statements made by top-level Defense and Energy Department officials explaining that our arsenal will be more “responsive” or “capabilities-based” to deal with ever more elusive enemies.

It’s difficult to pinpoint the genesis of this line of thinking within the military community, but one important document is Paul Robinson’s 2001 white paper, “Pursuing a New Nuclear Weapons Policy for the 21st Century.” In this apologia for atomic weaponry, Robinson, director of Sandia National Laboratories, laments that “far too many people (including many in our own armed forces) were beginning to believe that perhaps nuclear weapons no longer had value.” He warns that atomic bombs play an important role in global security and that since they cannot be uninvented, they must be adapted to respond to biological or chemical attacks and targeted at the leaders of rogue states and their arsenals. “I believe that we would desire primarily low-yield weapons with highly accurate delivery systems for deterrence in the non-Russian world,” he writes.

Perhaps the best example of this drive to bring nuclear weapons into the ambit of conventional war-fighting is the move to develop so-called mini-nukes, or precision-guided atomic warheads with yields of five kilotons or less–weapons that could generate explosions smaller than the conventional “daisy cutter” bombs used by US forces in Afghanistan. (The nuclear bomb dropped on Hiroshima had an explosive yield of fifteen kilotons.) Over the past several years, proponents of mini-nukes, who can be found in different parts of government, have defended these weapons with different shades of reasoning, but all appear to agree that the decades-long taboo surrounding the bomb must be reconsidered. “All I’m advocating is public discussion of what the parameters should be as we fight an unconventional war,” says Representative Steve Buyer, an eight-year member of the House Armed Services Committee, who called for the use of tactical nuclear weapons against Osama bin Laden’s Afghan fortifications late last year.

Destroying hard and buried targets, such as bunkers built beneath mountains or tunnels placed hundreds of feet below ground, is the main justification behind the bid to develop new low-yield, earth-penetrating nuclear weapons. American intelligence estimates cited by people who are worried about these threats say there are currently 10,000 hard and buried targets worldwide, and that number is expected to increase over the next decade. Stephen Younger, head of the Defense Threat Reduction Agency, which is part of the Defense Department, has argued that some hard and buried targets can be destroyed with nothing less than a nuclear blast. “A 5-kiloton nuclear explosive detonated on a 30-foot-thick missile silo door will vaporize that door, destroying the missile,” he wrote in an influential study published in June 2000. “A benefit of lower-yield weapons is that the collateral damage sustained by the near-target area may be reduced, an important factor in attacks near urban areas.”

Younger’s ideas were echoed in a similar report by the National Institute for Public Policy, a nonprofit defense think tank, which was published last year and which forms the bedrock of the Bush Administration’s nuclear policy. A number of that report’s authors were brought into government with the Bush Administration, and they have taken influential positions in shaping policy. Last July Adm. Richard Mies, commander in chief of the US Strategic Command, told Congress the NIPP report was a “good blueprint to follow” in drafting future nuclear policy. That same month a government study was convened to explore new ways to destroy hard and buried targets, including the use of mini-nukes.

That government study was sent to Congress last fall. In it, the Defense and Energy Departments explained that they had taken an initial look at how to use nuclear weapons against hard and buried targets, and that they were “investigating potential options and costs” of developing new atomic bombs to deal with these threats. According to the study, “Nuclear weapons have a unique ability to destroy both agent containers and CBW [chemical and biological weapons] agents.”

This January the Pentagon delivered to Congress its top-secret Nuclear Posture Review, outlining a revised US nuclear-weapons policy. The review hews closely to the NIPP report and the work of Robinson and Younger. It states that the country’s nuclear arsenal will be cut, but many warheads removed from deployment will be kept in storage as a hedge against future threats. It also explains that nuclear weapons will be “integrated with new non-nuclear strategic capabilities,” according to Defense Secretary Donald Rumsfeld. Moreover, it suggests that nonnuclear adversaries could now face nuclear retaliation. According to the posture review, atomic weapons must offer “a range of options to defeat any aggressor.”

When the posture review was completed earlier this year, Assistant Defense Secretary J.D. Crouch said in an unclassified briefing that “at this point, there are no recommendations in the report about developing new nuclear weapons.” He then added, “Now we are trying to look at a number of initiatives. One would be to modify an existing [nuclear] weapon, to give it greater capability against…hard targets and deeply buried targets.” Some observers, though, found this highly troubling, because the government has often used the term “modify” as a fig leaf under which new weapons are designed without appearing to violate international commitments. By 1997, under the banner of “modification,” the US military had developed a sort of prototype earth-penetrating low-yield nuclear bomb, the B61-11.

In March, it became clear that Crouch was making just such a semantic distinction. A copy of the classified document was obtained by the Los Angeles Times and later by the New York Times, which indicated that the review “argues that better earth-penetrating nuclear weapons with lower nuclear yields would be useful.” It also said that “new earth-penetrating warheads…would be needed to attack targets that are buried deep underground.”

Just how far along is the government in building such weapons? The Energy Department’s 2003 budget request calls for further studies of a “robust nuclear earth penetrator.” In February Gen. John Gordon, an under secretary at the National Nuclear Security Administration, told the Senate Armed Services Committee that his division of the DOE had already created work groups to develop atomic warheads. “The teams will carry out theoretical and engineering design work on one or more concepts, including options to modify existing designs or develop new ones,” he said. “In some instances, these activities would proceed beyond the ‘paper’ stage and include a combination of component and subassembly tests.”

Although the United States is not a signatory to any treaty preventing it from testing, Washington has imposed a nuclear testing moratorium on itself that is expected to expire in the next two years. Leaks from the classified Nuclear Posture Review indicate that this may be cut to a year or sooner. The government says testing is required to maintain the safety and reliability of an aging nuclear arsenal, but countless experts insist that this is simply not credible. “There is no scientific justification for testing for the safety of our arsenal,” says Joseph Cirincione, a senior associate at the Carnegie Endowment for International Peace. “The only reason you would need new tests is to verify new designs, new types of weapons, period.”

This development parallels a program under way to build facilities for manufacturing plutonium pits, used in a warhead’s trigger mechanism. Last July, funding for a “modern pit facility” was tripled, according to Nuclear Watch of New Mexico, a watchdog group. Yet even the Energy Department has stated that it does not currently believe the degeneration of plutonium pits in the existing arsenal “will become a problem in less than 50 years.” Similarly, in late January the Tennessee Valley Authority approved $3.25 million toward the production of tritium, a gas used in nuclear weapons to make them lighter and smaller.

Both international and domestic law constrict the development of new mini-nukes. If the United States were to build a new atomic bomb, it would offend the 1970 Nuclear Non-Proliferation Treaty. In 1994 Congress explicitly forbade the Energy Department from developing or researching the weapons because “low-yield nuclear weapons blur the distinction between nuclear and conventional war.”

Indeed, the biggest problem with the idea of using new nuclear weapons may be moral rather than legal. As some scientists have pointed out, the notion that a small nuclear warhead could burrow into the earth and destroy a bunker without causing extensive damage above ground is flawed. Robert Nelson of the Federation of American Scientists argues that there is no way an atomic bomb could penetrate the earth deeply enough to contain the explosion, even if its yield were 1 percent of that of the bomb dropped on Hiroshima. Such a bomb would create a fireball that would blast through the earth’s surface, carrying a cloud of radioactive dirt and debris, according to Nelson, who notes that five-kiloton atomic bombs had to be detonated at the Nevada Test Site at a depth of 650 feet to be fully contained–far deeper than any mini-nuke could travel.

“Nuclear weapons as now designed and employed are essentially useless because you cannot cross this threshold between conventional and nuclear without being unequivocal about it: If you use a nuclear weapon, you’re saying, ‘We’re now in nuclear phase,'” explains retired Rear Adm. Eugene Carroll. “Whereas these people say, ‘Well, this new weapon is so little, and we can apply it so precisely, and it’s for such a specific purpose that nobody can believe that we’re being irresponsible or careless or radical in our use of such a wonderful little weapon.’ But the truth is, the first use of nuclear explosives in warfare breaches the firewall, as some people call it, and when we go on beyond that, we’re put at the mercy of the other side, which probably doesn’t have such ‘useful’ or ‘usable’ weapons.”