In 1946, nuclear strategy was fixated upon the weapon being a war-winning instrument that would end long and bloody conventional wars. The Joint Chiefs' assessment of 1945 had warned Truman that the invasion of Honshu (Japan) would cause 286,000 US casualties. Two atom bombs were chosen instead. For 22 years Pentagon strategists have convinced US presidents that large nuclear arsenals were meant to either terminate war or to respond to a massive conventional superiority. Osgood and Kissinger proposed that limited nuclear wars could be won with tactical nuclear weapons. War gaming and computer simulation changed all that, assisted by nuclear impotence at the Berlin airlift, at Dien Bien Phu, in Korea and Vietnam. Nevertheless, the obvious nuclear impotence in Kashmir confuses and disappoints many.

War gaming in the late '60s showed that if control of the nuclear trigger was decentralised to subordinate commanders to fire nuclear weapons in the central NATO-USSR front, 21,000 tactical weapons would be fired in 48 hours, catapulting limited war through the strategic stage to a nuclear holocaust. Kashmir will be no different. The rationale behind nuclear strategy today is a mystery to lay Indians, which is understandable, but it is an area of darkness to most members of the strategic hierarchy on both sides of the border. There are probably only two libraries in India that can muster more than 50 relevant books on the subject and certainly none of the declassified official studies done in Washington and London. The signals from Pakistan, post Chagai, show wide variations in strategic comprehension. Nawaz Sharif's reference to a hitherto "existential" deterrence indicates that someone there understands strategy. Gohar Ayub clearly has both feet in 1946 and brings back the oft-repeated Hollywood thriller line, "It's not the professional killer, but the amateur that frightens me...."

Most of the arbitrariness, guesswork and emotion was removed from nuclear strategy by McNamara's team of operations analysts, who defined the limiting parameters of assured destruction—and hence mutual deterrence. Any link between conventional war and nuclear weapons was forever severed. The number of weapons required would be justified on the basis of a simulated exchange. This number would vary depending on whether the weapons were counter-value or counter-force. (The former are inaccurate, and have large kilotonnage since they can only destroy population centres. The latter are accurate, smaller and target hostile nuclear weapon sites).

It follows that a counter-force arsenal is ideal for a first strike and contemplating a first strike makes much sense if the enemy had only counter-value weapons which could be destroyed in a first strike. Indeed, political responsibility demands that the primary objective of nuclear strategy is to protect one's own people—killing enemy civilians is unjustified, unethical and bad strategy. Avoiding bad strategy and protecting one's own people now locks both Delhi and Islamabad to a common goal, the desire for which will transcend current animosities. The problem is that nuclear escalation is inevitable until technology, as known today, stabilises arsenals at mutually accepted levels. That level is the nuclear submarine-launched ballistic missile. Pakistan's bravado cannot conceal its inability to explode sub-critical weapons, a necessary step to achieve non-explosive testing, so that nuclear arsenals can be upgraded, after signing CTBT—as the US is doing. Pakistan can be placed in inferior nuclear asymmetry should India choose to do so—and this could bring them to the negotiating table.

In establishing a deterrent, there are some "iron laws"—accepted by all nations who blundered through the early years of calamitous nuclear instability. A nuclear exchange and a nuclear holocaust is most easily caused by misperception. To avoid triggering the wrong response, deterrence communication must be established, and it must be between the people who carry the trigger (the coding devices to arm the warheads).

Prior to Pokhran, India and Pakistan flouted every iron law that exists. They built and possessed bombs in secret, they had ambiguous chains of command, the triggermen were clearly unidentifiable, there were no deterrent communications and the delivery systems were fragile. Not surprisingly, the world saw South Asia as a cauldron of nuclear instability. To fan the flames came US scholars like Perkovich who concluded that the Indians had invented a new strategy called 'nuclear ambiguity'. Indian Babudom was overjoyed, to find a name for sitting on their hands. In contrast is the "posture statement", publicly promulgated to Moscow by McNamara after capping the American nuclear programme. Delhi may have to carry Islamabad into stabilising deterrence, but this is not impossible since much of the logic of deterrence is incontestable mathematics.

The road to deterrence is hard and unpleasant, but sanity lies on the other side of the hill. It is easy to cap two competing nuclear arsenals when deterrence is established. From deterrence it is a short journey, mostly downhill, to arms limitation, and from there, onwards to arms reduction. If the race for deterrence in South Asia is not to run berserk as it did in Washington and Moscow, the Indian nuclear posture statement should be published in a white paper. It must be followed swiftly by arms limitation talks with Pakistan and China.

Nuclear weapons are not secret weapons, meant to surprise the enemy on the battlefield. To deter, they must be made public. For this to happen, those in the hierarchy who have so far converted national power and national strategy into a byzantine struggle for individual or departmental success in Delhi's bureaucratic turf battles, must be made to desist. Nuclear strategy can be written and executed only with the highest levels of humility to withstand the corruption of the worst kind of power there is.

(A retired rear admiral, the writer specialises in defence, strategic affairs)