The underbelly of the Afghan war has been exposed by the war logs recently released through WikiLeaks. Not so long ago, Afghanistan was seen as the "good war", in comparison to the far more controversial adventure in Iraq. Now, light has been shone into the shadows of Nato's conduct in a war that is struggling to find direction.

A reassessment of our presence in Afghanistan must go back to basics to understand the continued failure to settle on an exit strategy from the country.

First, we should be clear in our understanding of conditions in Afghanistan prior to the 2001 invasion. Over a million civilians had died during the ten years of Soviet occupation that ended in 1989. The next 11 years witnessed a fluctuating civil war. US-led Nato forces picked a winner by providing huge amounts of firepower to drive the Northern Alliance in to Kabul (though they were beaten to it by John Simpson).

The Taliban's senior leadership, aided by a two-faced Pakistani strategy and military incompetence on the part of the US (as typified by "Operation Anaconda"), were able to flee into the Pakistani tribal areas in Helmand and Kandahar Provinces.

Over the following years, the imported exile Hamid Karzai failed to unite the country behind a top-down central government. Meanwhile, warlords who had switched sides during the invasion cemented their control over various fiefdoms, showing little interest in surrendering power to Kabul.

That the "new Afghanistan" was largely rotten at its core was ignored by an administration in Washington that quickly shifted its focus to invading Iraq, stalling efforts to transform the region. With the return of the realists to US politics came implementation of the "surge", which revived the narrative of success in Iraq, despite that country remaining deeply fractured and suffering a violent political inertia.

High-risk strategy

In Afghanistan, the failure of the Karzai government allowed the Taliban to return. Barack Obama doubled down on reviving the war in Afghanistan, speeding up the withdrawal from Iraq while bolstering Afghan troop numbers under the leadership of General Staley McChrystal.

A new counter-insurgency strategy (Coin) looked to buy off the "accidental guerrillas" in Pashtun areas by incorporating them into an army that the Afghan state cannot sustainably afford, and whose ethnic and tribal loyalties are constantly contested. The critical flaw is with the legitimacy of this effort. In Afghanistan, we should be very clear that we are training an Afghan army to kill Afghans in order to protect Afghans. Incidents such as the killing of three British soldiers by an Afghan soldier they were training are simply tragic reminders of the short-term problems inherent in such a strategy.

Repellent as we may find them, the Taliban appear to have a more coherent ideology than Karzai's government. As the Coin expert David Kilcullen has written, "most Afghans historically had little interaction with the central state". Their anti-occupation rhetoric falls on the ears of a population all too aware of Nato's disregard for their lives, as now exposed by the WikiLeaks documents. How can we say that we are in Afghanistan to protect Afghans, when we don't allow them the basic dignity of an independently verified body count?

The mission in Afghanistan relied on a high-risk strategy which predicted that democracy would bestow legitimacy on a foreign military occupation. And yet, today, the US finds itself fighting the longest war in its history; the most recent two months have been the deadliest since the initial invasion. Unless we realise how we got to where we are today, any future policy is doomed to repeat the same mistakes.

James Denselow is a writer on political and security issues affecting the Middle East, and is based at King's College London.