For the common man, the armed forces are meant to win wars. For the military thinker, the armed forces spend most of their time preventing war; although they sometimes get it wrong even after a half-century of trying. Preventing war is a serious subject in military colleges, as countries often require their soldiers to threaten another government. If the threat is conveyed badly, war may result at one extreme; at the other, the victim government might just ignore the threats. This twilight zone which the politico-military apparatus often traverses is called 'managing a crisis' in military jargon. The two parameters of failing to manage crises are either war, or status quo. India requires Musharraf to take a step he is loath to take—close down the camps and disperse the jehadis. Compelling him to do so requires coercion, and the pain of coercion must be greater than the pain of dispersing the jehadis—at least for awhile. For the Indian state, this is a new game with the foreign office, armed forces, intelligence agencies and strategic thinkers joint players. If past incidents are a guide, the unarguable conclusion is that escalation and de-escalation in a crisis must be progressive—that is to say, step by step. The reason is that anxiety and tension created on the victim government by a series of ascending (or descending) steps are much more intense than if a large threat is conveyed or removed too suddenly. After December, the escalation was done much too rapidly as was the withdrawal of the navy, a service which could have exploited Pakistan's single commodity vulnerability grievously.
Now, the de-escalation. This has to occur mutually and for the government to make a giant jump down the escalatory ladder after Musharraf closes all camps would be disastrous. Musharraf has to survive if he is to rein in the jehadis until the October elections. We are suddenly on the same side. If the two countries are to de-escalate sensibly, there must exist in Delhi a de-escalating model. It needn't be perfect, but it is a reference point from which modifications can be made. A sample model is shown below. Very often, each government is willing in principle to take the next step but feels the last descending step taken by the other is too little. If the de-escalation isn't to halt completely, a quick exchange and promise to go down further may break deadlocks. A neutral third party facilitator may be handy here, assuaging insecurities.
De-escalation is a matter of taking steps within time limits. The mutuality of de-escalation is preserved by both the generosity of the step taken as well as the speed of taking it. In a large democracy like India, with a number of heavyweights playing the game, it is impossible for anyone in Delhi to coordinate this sensitive exercise without working off a model. How much of it is revealed to the third-party facilitator is a matter of trust and choice. Frequently, what Delhi thinks frightens Islamabad actually makes them yawn, while they are perplexed why some terrible step obvious to them doesn't occur to South Block. A facilitator is likely to know the mindset in each capital. The revelation from Pakistan is that Musharraf's first U-turn on Kashmir has the support of the silent majority. What can we give him to keep him going in the same direction, so that he gives us peace in October. Make a model and leak it to Musharraf. If he doesn't think our model is so hot, let's take a look at his.