One thing is at least clear from the success achieved by Operation Mushtarak against the Taliban in the most troubled parts of Afghanistan: if the NATO forces and their Afghan allies move on with the offensive they could always achieve their military objectives.

The question is whether military success translates into durable political gain?

Regarded as a half-forgotten low intensity war only a year ago, the conflict in Afghanistan is emerging as a key issue as President Barack Obama tries to develop a global strategy for the Greater Middle East. Obama has almost doubled the number of US combat troops in Afghanistan.

In intervening in Afghanistan in 2001, the US had three key interests.

The first was to show to friend and foe alike that it could not be attacked with impunity. Prior to 9/11 the US had suffered a series of terrorist attacks, from the seizure of its diplomats as hostages in Tehran to the mass murder of 241 Marines in Beirut, and the first attack on the World Trade Center in 1993, but had not hit back in ways that might have dissuaded future aggressors. By declaring war on Afghanistan, the US was making up for that lacuna, reassuring its friends and warnings its foes. The idea was that if "we did not go to the Hindukush, Hindukush will come to us."

The US had a second interest in invading Afghanistan: finding and destroying the bases from which terror had been exported to America and, where possible, capturing or killing the masterminds.

The United States’ third interest was to help Afghans replace the Taliban with a government of their choice in the hope that it would prevent the re-emergence of terrorist bases.

By 2005, all those objectives had been achieved.

The US would have been able to declare victory in Afghanistan and start reducing its military footprint in preparation for disengagement. However, the Bush administration could not contemplate such a course for domestic political reasons.

It is also important to emphasise a fact often ignored in the debate on Afghanistan. The Taliban regime had never been an explicitly anti-American outfit.

Today, the best structure for Afghanistan is that of a loose federation in which its 18 ethnic and religious communities enjoy full economic, cultural and administrative autonomy. However, the system developed in Afghanistan since 2002, has gone in the exact opposite direction. The Afghan president today has powers that no Afghan king ever dreamt. The problem is that these powers cannot be used without provoking violent resistance from a majority of Afghans. And such violence cannot be dealt with except by force. President Karzai, a member of a minor Pushtun tribe, and lacking a constituency of his own, cannot master the force needed to impose central government control throughout his unruly land. He has been trying to do so by relying on American power. The result is that the US has been sucked into Afghan politics as just another tribe albeit one that has greater firepower than the others.

NATO has around 90,000 troops in Afghanistan, a country the size of California. Of these, at least a third won't fight because of caveats imposed by their governments. That makes a division of labor imperative. Those NATO units that do not wish to fight must be enlisted to relieve the pressure on those ready to fight by shouldering tasks such as policing, patrolling, helping with construction work, and intelligence.

Even then, General McChrystal, the US commander in Afghanistan, might find it hard to go after the insurgents in a big way with just 60,000 or so troops. Taking into account the need for rotation as well as logistical and administrative duties, the general might not have enough troops for a proper search-and-destroy strategy, let alone his "seize-and-hold" scheme even if we add British, Canadian and some French units. If President Obama intends to win his "war of necessity", he would have to increase the number of US troops for a fight that might take two or three more years. Even then, General McChrystal must realize that his version of the "surge" might not secure all the troops needed.

He would have to find allies inside Afghanistan, just as Petraeus did in Iraq.

There are over 150,000 armed former Mujahideen, waiting on the sidelines to see how the wind turns. The Taliban never directly controlled the whole of Afghanistan and do not have enough popular support to govern the country. They spread their rule, often nominal, by bribing the Mujahideen.

According to a proverb, one cannot buy an Afghan but one could always hire him. The policy of shunning the former Mujahideen, and branding their leaders as "warlords", may sound chic in intellectual salons, but is counter-productive in real life. Then there are some 50,000 armed private "security professionals" that, provided they are deployed in the context of a broader strategy, could be used more effectively.

We also have the 180,000 or so members of the new Afghan army and police. Often, these men draw their salaries but spend their time doing the crosswords or at best directing the traffic in Kabul. According to experts, a third of the new Afghan army is reliable and competent. Embedding them with NATO forces could give them a role in taking the war to the insurgents.

The US and its allies managed to create an army of over 600,000 men in Korea in just 18 months because they mobilized the resources necessary. The failure to achieve something similar in Afghanistan is largely due to a shortage of resources.

The drug smuggling rings have 15,000 armed men, often cooperating with the Taliban whose own strength may be 20,000. Smaller insurgent groups, such as Hekmatyar's Hizb Islami (Islamic Party) may command a further 5,000 armed men. Hekmatyar, who worked for the CIA for years but has been partly financed by Iran since 1996, recently made it clear that he was open to offers. In a hierarchy of operations, the Taliban is the top target. This could mean making tactical alliances even with some unsavory armed groups, and buying others. Afghanistan has a military problem that needs a military solution. US strategists are beginning to realize that. This war could and must be won. There is no need to panic and cry for an "exit strategy" even before there has been any major fighting.

The insurgent groups control only 11 of the 362 districts, accounting for less than one per cent of the country's population. Most of their principal bases are in neighboring Pakistan and, in the case of the Hekmatyar group, Iran. And, yet, Washington is all abuzz with the "f" word, for what many see as looming failure in Afghanistan.

Since Karzai's legitimacy has been questioned as a result of the disputed presidential election, his second-term must be endorsed by the Loya Jirgah (The High Assembly). This is an institution representing Afghanistan's different ethnic groups, tribes and religious communities. The new constitution recognizes the Loya Jirgah (or shura) as a kind of constituent assembly and thus the nation's highest legislative organ. The same session of the Loya Jirgah could debate and approve the constitutional amendments needed. The presidential system introduced in 2002, does not reflect Afghanistan's diversity. Constitutional amendments are needed to reduce the powers of the president, create a post of prime minister, install a system of parliamentary democracy, and devolve power to the regions.

Today, Afghanistan is not lost just as Iraq was not lost three years ago when the defeat industry was in full gear in Washington.