The United States president visits Russia on 6-8 July 2009 at a time of some tension but also promise in the relationship between the two former superpower rivals. The areas of tension that Barack Obama and Dmitry Medvedev face are familiar: missile-defence, the war in Georgia and the post-election crisis in Iran among them. The promise is of cooperation over Afghanistan and, more broadly, on a subject that during the cold-war dominated Washington-Moscow negotiations: nuclear weapons.  A modest reduction in the two countries' nuclear arsenals is on the cards, codified in a replacement of the strategic-arms reduction (Start) treaty of December 1991 which expires in December 2009. This might, just possibly, be the first concrete step towards a nuclear-free world.

In his speech in Prague on 5 April 2009, Obama announced that the United States was committed "to seek the peace and security of a world without nuclear weapons". It was a signal that global nuclear disarmament, an idea once associated primarily with peace-movement activists, had become a core element of US security policy.

"We do not want a world of continued nuclear proliferation", Obama told reporters after the meeting. Soon after, on 25 May 2009, the North Koreans underscored the warning by detonating their second nuclear weapon.

But there is positive momentum as well as worry behind Obama's aspiration, reflected in signs of change among elite opinion that have been building for several years. In January 2008, a group of American elder statesmen of the cold-war period - Henry Kissinger, George Schultz, William Perry, and Sam Nunn - issued their second appeal within two years for the abolition of nuclear weapons. "With nuclear weapons more widely available", they argued, "deterrence is decreasingly effective and hazardous" (see "Toward a nuclear-free world" [Wall Street Journal, 15 January 2008]).

The campaign of these and other luminaries has been gathering pace since 2007 (see "A World Free of Nuclear Weapons" [Wall Street Journal, 4 January 2007]). There are comparable moves in other nuclear-weapon states. In Britain, a group of establishment figures has questioned the country's commitment to building a new generation of nuclear missiles for deployment after the Trident missile reaches the end of its life (see Paul Rogers, "The nuclear-weapons prospect", 4 June 2009).

These proponents of a drawdown in nuclear weaponry are certainly concerned about nuclear proliferation. But in the case of Kissinger and his collaborators, there is also a  coldly "realist" calculation: that in a nuclear-free world, the United States would be the world's pre-eminent military power - one no longer be vulnerable to annihilation by distant enemies.

The constraint

There is strenuous opposition to this view on the grounds that the nuclear disarmers are naïve. They stand accused of ignoring a critical danger: that after the "end" of nuclear weapons, a rogue state would "break out" of its treaty commitments - either by concealing a small stash of weapons during the disarmament process, or by secretly rebuilding a nuclear arsenal. Honest disarmers would, sooner or later, find themselves at the mercy of unscrupulous nuclear cheats.

The problem is obvious - even without taking into account the challenge of addressing disarmament among actual nuclear-weapon states and potential ones (such as Iran). If a state declares that it has 5,000 nuclear weapons, but has hidden a hundred more, the extra weapons confer no strategic advantage.: deterrence remains robust. But, if all nuclear-weapon states are supposed to have disarmed and just one retains a secret one-hundred bomb stockpile - enough to destroy every major city in Europe, Russia and the United States - cheating would matter very much indeed.

Thus, creating verification regimes that can confidently detect treaty violations will pose huge challenges - and the technical capacity to verify compliance with high confidence doesn't yet exist. But arms controllers often forget that there are other effective ways to deter cheating.

No state that contemplated reneging on its disarmament commitments could be certain that its transgression would not be revealed from within. If just one individual refused to go along with the deception, all would be revealed.

This is not just idle speculation. Nuclear and other weapons-of-mass-destruction "whistle-blowing" is a reality - and one that has been evident in authoritarian states as well as democracies. Israel's nuclear-weapons programme had its Mordechai Vanunu, Russia's chemical-weapons plans had their Vil S Mirzayanov, and Saddam Hussein had his defector son-in-law, Hussein Kamel Hassan.

Governments tempted to cheat could have no way of determining in advance whether or not their violations would be revealed from within - especially if the international community offered lucrative rewards for nuclear whistle-blowers. In this possibility lies a measure of deterrence.

Much of the concern about nuclear weapons today focuses on their possible acquisition and use by terrorists. But unlike states, terrorists offer no target for nuclear weapons - they are immune to nuclear deterrence. Nuclear weapons, in other words, are useless against terrorists.

In any case, anxiety about the dangers of "nuclear breakout" derives primarily from the belief that having a nuclear monopoly would confer great military utility on the monopolist. But from 1945 until the first Soviet nuclear test in 1949, America's nuclear monopoly proved powerless to prevent the greatest expansion of the Soviet empire during the entire cold-war period.

The belief that nuclear weapons are a guarantee against conventional attack is another illusion. US nuclear weapons did not prevent Vietnamese revolutionaries from attacking American forces during the Vietnam war in the 1960s-1970s; nor did they deter China from attacking the US during the Korean war of 1950-53. Israeli nuclear weapons did not deter Egypt from attacking Israel in 1973, and Soviet nuclear weapons did not deter the mujahideen resistance in Afghanistan - nor did they prevent a Russian defeat.

The reluctance of states to use their nuclear arsenals arises in part because the years since the bombing of Hiroshima and Nagasaki in August 1945 have witnessed the emergence of a global norm against the use of nuclear weapons. This norm would be strengthened by a global ban on nuclear weapons.

For those sceptical about the effectiveness of normative constraints, the realist response is that any nuclear monopoly gained by a violation of the nuclear-free regime would almost certainly be short-lived. By confronting a "breakout" state, former nuclear powers could quickly reconstitute previously dismantled nuclear-weapons programmes, or start new programmes from scratch. Indeed the fact that all former nuclear-weapons states would have the capacity to recreate their nuclear arsenals quickly would create a system of "virtual deterrence". In this light the reversion to a partially rearmed nuclear world, though highly undesirable, would be no more dangerous than the current situation.

The momentum

In order to determine the balance of risks associated with global nuclear disarmament, the world will have to think about two further issues: the incentives for states not to breakout of a global nuclear-free regime, and what real strategic value nuclear deterrence has in a world where cross-border aggression is increasingly rare.

First, if a global disarmament regime is in place, some of the current motives for nuclear acquisition would no longer apply. In a denuclearised world, states would no longer need nuclear weapons to deter nuclear attacks from other states - which is by far their most important function today. And if nuclear weapons were legally and normatively proscribed for all states, their acquisition would no longer confer prestige or influence.

Second, the wars of the 21st century are overwhelmingly civil wars in which nuclear weapons can have no conceivable role. Interstate war has become increasingly rare because its costs have risen, while its benefits have decreased. As the incentives to wage interstate war continue to decline, a benign dynamic comes to operate: lower levels of (non-nuclear) deterrence will suffice, the case for the "ultimate deterrent" will become less compelling, and the case for a world free of nuclear weapons more so.

The modest advance in nuclear diplomacy agreed by Dmitry Medvedev and Barack Obama may be just that: modest. The dangers represented by states's actual or potential possession of nuclear weapons remain. But the forces pushing in the direction of a nuclear-weapons-free world are at work. Moscow may be just the beginning.