In January 2002, when President Bush named Iraq, Iran and North Korea as the first targets in his 'global war against terror' - the putative 'axis of evil'--few noticed a curious omission. Pakistan was not on the list.

The targeted countries - we were told - sought weapons of mass destruction. In truth, Iraq and Iran were targeted because they stood in the way of Israeli ambitions - and they had oil. 

Although Pakistan has been unlucky in oil, it could make stronger claims as a target for American and Israeli ire. It is the only Muslim country with nuclear weapons, a nuclear proliferator, the Taliban's chief patron, and a sponsor of jihadis in Kashmir. 

Why, then, did the US not target Pakistan? 

Six years later, this question is not less pertinent: and for two reasons. After being stalled by the Iraqi resistance, US plans for war against Iran are again gathering steam. If Iran is such a tempting target, why not take a few potshots at Pakistan also? 

In addition, since their rout in Afghanistan, bands of Muslim 'extremists' have found safe havens in Pakistan's northern districts, as well as Quetta and Karachi. More ominously, last July, the Taliban challenged the authority of the state in Pakistan's capital. 

Yet, there has been little talk in Washington or Tel Aviv about adding Pakistan to the 'axis of evil.' This is the Pakistani paradox.

This paradox has a simple explanation. In Pakistan, the US had effected regime change without a change of regime. Almost overnight, following the attacks of 9-11, the US had drafted the Pakistani military to wage war against Muslim extremists. The US had gained an army: and Pakistan's military dictators had gained longevity.

Yet, could the Pakistani military deliver on its promise to fight the Taliban and Al-Qaida? At first, it appeared that it was succeeding. General Musharraf boasted that Pakistan had collected $50 million in exchange for extremists handed over to the US. 

These losses, however, did not deter the extremists from regrouping; and before long they were attacking NATO forces in Afghanistan from bases inside Pakistan. As NATO casualties rose, the US ratcheted its pressure on Pakistan. And by August 2004, Pakistan had deployed 100,000 troops to guard its frontier with Afghanistan. 

The extremists now began targeting Pakistani troops. In September 2006, in the face of rising losses, Pakistan pulled out its troops from Waziristan in return for a Taliban promise not to mount attacks from bases in Pakistan. It was an improbable truce. 

In reality, the Taliban had 'liberated' Waziristan.

The US was unhappy about the truce. And with good reason: Taliban attacks in Afghanistan began to rise after the truce. Since then, US has been ratcheting its pressures on Pakistan to hunt down the extremists operating out of bases along its northern frontier. 

According to the Newsweek of Oct. 8, the Pentagon is now demanding that General Musharraf "turn much of Pakistan's military into a counterinsurgency force, trained and equipped to combat Al-Qaeda and its extremist supporters along the Afghan border."

This Latin American approach to counter-insurgency is not likely to work in Pakistan. Their military juntas were firmly rooted in the elites and middle classes, set apart from the leftist insurgents - mostly Amerindians or Mestizos - by both class and race. The boundary between the adversaries in Latin America was firmly drawn.

In Pakistan, the insurgents are Muslim nationalists. They are drawn mainly from Pashtun peasants, but they enjoy broad support among the peasants as well as the middle classes all over Pakistan. 

On the other side, about a fourth of Pakistan army consists of Pashtuns; and mid- and low-ranking officers are middle-class in their origin and orientation. Only the top military brass identify firmly with the elites. 

In Pakistan, the boundary between the opposite camps is not as firmly drawn as in Latin America. As a result, as Pakistan's army escalates the war against its own people, this boundary has been shifting, shrinking the support base of the military elite. 

If this is the irreversible dynamic behind the US-inspired counterinsurgency, it is unlikely that Pakistani elites can long sustain their decision to fight America's war against Muslim nationalists. 

Recent events support this prognosis. As the military has escalated its offensive, its reputation has plummeted. Hundreds of soldiers have surrendered or, more likely, defected. General Musharraf has rescinded corruption cases against Benazir Bhutto to court her party; but this has eroded the standing of her party. 

How is this 'civil war' likely to end? In one scenario, at some point, an alliance of Muslim nationalists - the fighters and their allies in the army and civil society - will enforce their own regime change, and create an Islamist Pakistan. 

This will end the civil war, but not Pakistan's troubles. Instantly, US and Israel will clamor for a regime change of the hard variety: through covert operations, air strikes, invasions, and civil wars.

As these events unfold, the US may well decide to start a war against Iran. This can only advance the timetable for an Islamist take-over in Pakistan. When that happens, the US and Israel will be engaged in a major war along an Islamic arc stretching from Lebanon to Pakistan -and perhaps beyond, to the north and the east.

Is this the 'clash of civilizations' that the Neocons had advocated - and have worked so hard to advance? Over the past century, the nations that initiated the two major wars eventually came to regret them. Is it likely that this history may repeat itself?

Once begun, the course of wars cannot always be foretold. Germany, Japan and Italy learned this lesson the hard way. With some wisdom, the US and Israel could learn this lesson the easy way - from the mistakes of belligerent nations before. Even now, it may not be too late to take this lesson to heart, and avoid a major war that promises to be catastrophic for all sides. 

M. Shahid Alam is Professor of Economics at Northeastern University, Boston. he is the author of Challenging the New Orientalism (North Haledon, NJ: IPI, 2007). He may be reached at alqalam02760 at yahoo dot com.