Social scientists have long used systems theory to point out links that are sometimes overlooked. Such a link now leads to the suggestion that to deal with the crisis in Pakistan and the mounting difficulties in Afghanistan, one should go to, of all places, India. The main reason the Pakistani army is reluctant to take on the Taliban, who threaten to overrun the country, is that the army considers India its enemy.

The main source of tension between the two countries is the question of who will control Kashmir. At first, any suggestion of solving this tricky problem in order to get the Pakistani army to refocus would seem truly far-fetched, if not visionary to the nth degree - like asking someone to improve the climate by first stopping earthquakes. Dealing with the internal tensions in Pakistan is difficult enough; add to the mix a conflict that has been simmering for decades and seems to resist any solution, and your way seems completely blocked.

The good news is that secret negotiations between India and Pakistan about Kashmir have been taking place. Those now need to be restarted, supported, and accelerated. In the process should we also dump the obsolete notion that India should be made to "balance" China, and we may then be on our way to a more stable region in this much contested part of Asia. Both steps require some elaboration.

India and Pakistan have fought three wars and come close to blows on several other occasions, mainly over Kashmir. The Pakistani and Indian armies' training, formations, and alignments are focused on deterring each other, and are largely concentrated on their mutual border. The terrorists who recently attacked Mumbai, and before that the Indian Parliament and markets in New Delhi, see themselves as Kashmiri freedom fighters. It is hard to imagine major repositioning of the troops involved and changes in the strategic planning of both countries until the Kashmir issue is resolved.

A large variety of suggestions have been made about the way to determine the future of Kashmir. Some favor that the people of the area be allowed to vote on which nation they wish to join, as was prescribed by the UN Security Council after the partition of India; others favor dividing the area between the two nations; still others favor turning it into an area which enjoys a large measure of independence. The secret negotiations that took place between India and Pakistan during Pervez Musharraf's tenure as president appear to have focused on the idea of 'making borders irrelevant,' that is, granting Kashmir a great deal of autonomy to avoid either country having to lose face by giving up their claims to control. Since General Musharraf stepped down these negotiations seem to have stopped. The Obama Administration would do well to apply here the same system-wide approach it is revealing in the Middle East: follow the links. In this case, encourage, cajole, and help Pakistan and India to move toward settling the Kashmir issue. Eliminating this source of terrorism and war, and the reason both nations are keen to maintain nuclear arms, is of course worthy in its own right. It will also pay off handsomely in getting the Pakistani army to focus on the Taliban, including both those who threaten the government at home and those who cross the border from Pakistan to Afghanistan.

Obama seems not to shy away from complexities. In this case, dealing with India must take into account two more related matters. One involves a Bush Administration-backed deal with India which allows US companies to sell it nuclear fuel and other nuclear technologies to be used only in India's civilian reactors. However, the net effect it that this deal allows India to shift from civilian to military use the limited uranium it has from other sources. (Bush assumed that it is acceptable for 'good' governments--like that of India and Brazil--to develop their nuclear industries, while only those of 'bad' governments, such as Iran and North Korea, need to be reined in.) However, as we now fear in Pakistan, government can change overnight, and above all, if the norm against proliferation is to have a chance to take hold, the fewer nations that go down this road, the greater the chances of enforcing a non-nuke global regime. This means is that even if the US can't get India and Pakistan to give up their nukes, at this stage at least, it is ill-advised to help them increase their armed nuclear program. Hence the Obama Administration should renegotiate the agreement to provide nuclear fuel and technologies to India, and both India and Pakistan should be encouraged to scale back on the nukes and join the NPT, as their conflicts are resolved.

All this assumes that one drops the obsolete notion that nations can be used to balance each other and that a militarily powerful India somehow will serve to countervail a rising China. China is focused on its economic development, and its military has a very long way to go before it will pose a threat to the United States, even if there was reason to believe that it would seek such a confrontation. And if China has such intentions, whether India has a few extra nukes or not would hardly matter, given the size of the American arsenal. In short, to the extent that the US has any influence on India, it can be used to scale back its nuclear program without any discernible loss to American strategic interest.

I will be the first to grant that this kind of social science thinking leads one to try to tackle some very tough and persistent international problems as a way to treat some very acute local problems inside Pakistan and in Afghanistan. In that sense, it is truly far-fetched. However, these pestering international conflicts call for resolution anyhow, and if they can at least be curbed, such progress would help on the local fronts. Dealing with the Taliban may entail going to Kashmir, New Delhi, and Islamabad, and forgetting for now about China.

Amitai Etzioni professor of international relations at The George Washington University and the author of Security First (Yale 2007). For more, go here.