Source: Carnegie

Originally published in the January 2002 edition of Prospect.

To describe Afghanistan as medieval is an undeserved compliment to its dark-age tribal structures. Even in the 1950s modernization didn't touch people outside the cities. Western-backed democratic statebuilding is not an option; the best we can hope for is an era of unsupervised peace.

To adapt 1066 And All That, the Bush administration's approach to postwar Afghanistan might be described as right but repulsive. There is something deeply disagreeable about the indifference of some senior US officials to humanitarian relief, and their apparent readiness to wash their hands of the country as soon as al Qaeda and the Taliban have been destroyed. After all, the US must bear a large share of responsibility for the Afghan debacle of the past 25 years.

If the US simply pulls out and moves on to bomb Somalia or elsewhere, it risks repeating the original mistake it made in Afghanistan once the goal of destroying the Soviet client state was achieved. For there to be any chance of a stable postwar order the US, like the rest of us, will have to provide substantial aid and US airpower will be necessary to back up a UN force in Afghanistan, if that force is to be taken seriously by the heavily-armed groups on the ground.

On the other hand, the Bush administration is essentially correct on three points: that a UN force should not contain US ground troops; that it should limit itself to providing security for Kabul and to a lesser extent the main roads; and that great caution should be adopted over grandiose projects of internationally directed democratic state-building.

The melancholy truth is that for the foreseeable future, Afghanistan cannot be governed, either by the "international community" or by Afghans themselves. It can at best be managed to prevent a renewal of conflict and give ordinary Afghans the chance to restart basic economic activities with some security-trade, markets, the restoration of irrigation systems in the countryside and so on.

Here, Britain's imperial experience may be of value. While 19th-century British attempts to occupy Afghanistan were disastrous, the attempts after 1880 to manage relations with the Afghan tribes of the North West Frontier were more successful; one reason being that the British eschewed direct administration, let alone attempts to reform Pashtun society.

The tribal areas of the North West Frontier were not managed by the Indian Civil Service (ICS), which administered the rest of British India, but by the Indian Political Service (IPS), a quasi-diplomatic corps which managed relations with the Indian princes, Afghanistan, Nepal and other dependent or semi-dependent local states. Rather than trying to govern, the IPS proceeded by bribes, negotiation and the sanction of armed force, whether in small-scale raids to capture local troublemakers, or full-scale military expeditions (including the early use of airpower) like those in Waziristan in 1936-37. This indirect approach has been continued by the Pakistani state.

These historical lessons risk being lost in a surge of British and western missionary zeal, now cast in terms of democracy, human rights and nation-building. In an Afghan context, most of these plans are bound to fail. And, as in Somalia and elsewhere, they could even make things worse. If mishandled, massive international aid directed through an Afghan central government could itself become a source of conflict, as regional groups and warlords would be tempted to fight for control of the government and the city of Kabul. On the assumption that any national government would be extremely weak militarily, there are arguments for not strengthening it in other ways.

Looking at Afghanistan, the reasons for skepticism about modern state-building are all too evident. The word "medieval" has been used by the western media as an insult when describing Taliban rule. In fact, to describe Afghanistan (and much of Africa) as medieval is an undeserved compliment. Medieval is Afghanistan on a good day. A medieval system-in the west European or Muslim sense-is what we should be trying to help establish; because we will not be able to implement the modernizing projects promoted by the NGOs and the Afghan émigré elites.

What much of Afghanistan has seen over the past generation bears more resemblance to the dark ages: chaotic, morally unrestrained warfare between ethnic groups, tribes and warlords, with a catastrophic effect on economic activity and the only discipline provided by fanatical religion. By contrast, the middle ages in western Europe and (until the Mongol invasion) the Arab world produced great cultural achievements-greater than those of the modern west-giving birth to the international commercial and financial networks which laid the basis for the industrial revolution. Moreover, in both the Christian and Muslim worlds the extent and ruthlessness of wars were moderated by religious ethics.

History is a long business. It can be pushed, but it cannot be rushed off its feet-not without disastrous consequences. For example, to make the past century or so of African history explicable in a British context, you should imagine that the Emperor Claudius and his legions had landed in 1st century iron-age tribal Britain with machine guns, telephones and the rest of 19th and 20th-century technological society.

Would Britain then have experienced an industrial revolution in the 2nd century AD, from such incompatible elements of state and society? Yet that is what we expect from much of the world. The east Asian example does not prove that such miracles are possible. The Chinese state and commercial traditions were old before Londinium was founded. Mid-18th-century Japan-over 100 years before Commodore Perry landed-had a literacy rate (including women) of over 40 per cent, far higher than that of England or most of western Europe at that time.

Under the reign of Amir Abdur Rahman from 1880 Afghanistan did begin a state-building process comparable to that which commenced in France 700 years previously, under King Philip Augustus, or arguably 1,200 years ago, under Charlemagne. Local baronies were destroyed or their holders incorporated into state service; tribes and ethnic minorities were subjugated and their traditions progressively destroyed; the independence of religious figures was ended and the recalcitrant were killed. Over the centuries in France, this process went through several cycles of state failure and recovery, often worsened by foreign (usually English) intervention. Though each advance when it failed left something behind, this was not evident at the time. Only with the reign of Louis XIV in the later 17th century did the unified modern state establish itself firmly enough to survive future revolutions, coups and outside conquests without disintegration.

The fiendish methods used by Abdur Rahman to subjugate his tribal, ethnic and religious opponents horrified Europeans and left a legacy of bitterness among his own subjects. The alienation of so many of these from the modern Afghan state is one reason for the disasters of the past generation. The Amir himself was so hated that, on his death, he could not be buried in the royal cemetery, because of threats to dig him up and cut him in pieces.

But were Rahman's atrocities worse than the sins of European state-building, especially when faced by "savage" tribal resistance? I doubt it. The Duke of Cumberland, who suppressed the Scottish clans in the mid-18th century, was not called "Butcher" for nothing. Until the later 18th century, the punishment for treason in Britain was to have your bowels torn out and burned in front of you, then to be sawn into pieces and those pieces displayed around London.

Of course, modern tribal and semi-tribal societies are different from those of previous eras, in that successful examples of western and east Asian state-building exist as a model and a spur. But Afghanistan is an example of how spurring a country in this way can lead to disaster; when aspects of the modernisation of the state get far ahead of progress in the economy and society, and when state rhetoric raises unrealistic expectations and ambitions. In this case, sections of the younger elites compared Afghanistan's backwardness with the great (though flawed) progress of Soviet central Asia. The result was a disastrous infatuation with communism.

This danger of flawed modernization is especially strong when the state is itself not only culturally alien, but the creation of foreign conquest. For while "Afghanistan" is not an artificial creation of western colonialism, like Sierra Leone or Kenya, it is neither an old state or wholly a creation of "Afghans"-a people who, as an 18th-century traveler recorded, "had no name for their country." From the mid-18th century to 1978, most of the territories now comprising "Afghanistan" were under the rule of members of the Abdali (later renamed Durrani) clan of the Pashtun ethnos. However, this "rule" was the loosest form of hegemony and the frontiers of Durrani dominion fluctuated wildly, at times covering much of the north-west of the Indian subcontinent.

"Afghanistan" as a state was a creation of the end of the 19th century. Its borders were determined by the British empire, and like most colonial borders, reflected no historical or ethnic logic. The northern border marked the extent to which Britain was prepared to see the Russian empire advance. To the west, Britain forced limits on the territory of Persia, viewed by London as a potential Russian client state. The southern and eastern borders of Afghanistan were the limit to which the British Indian empire felt it necessary and safe to advance itself.

Within those borders, a state with modern trappings was created by a confluence of British geopolitical interest and the rule of Abdur Rahman, from 1880 to 1901. The so-called "Iron Amir" was a competent and ruthless ruler, but he would have achieved little without money and weapons from the British, who were anxious to build up Afghanistan as a buffer against the Russians. While internally Afghan government was largely unconstrained, externally Afghanistan was in effect a British protectorate, allowed to conduct formal international relations with the British Indian empire alone. It only achieved the status of a fully independent country in 1919.

Had the modern Afghan state under Abdur Rahman's successors greatly developed the country and benefited the population, then the resentment against their rule would have faded. It failed to do so. But compared to what followed, the royal state (replaced by a republic when the King's cousin, Sardar Daud Khan, seized power in 1973) was a paradise of peace, tolerance and development. Reading about Afghanistan in the 1950s and 1960s, with its US and Soviet industrial developments, irrigation and electrical projects, women politicians and western tourists, and comparing it to 1990s Afghanistan is (as in parts of Africa) a terrifying lesson in how far and fast demodernisation can go.

But one must be careful. Modernization and development touched few Afghans outside the cities. The great internationally-funded hydroelectric project on the Helmand river may have hurt more local farmers than it helped. Measures to improve the lot of women were resented by many people, insofar as they were aware of them.

When I entered the Mujahedin-controlled Pashtun areas of Afghanistan in 1988, I was struck by the completeness with which every element of the state had been swept away-by contrast with more developed countries then in a state of civil war, like Lebanon. One reason for this was the extent to which the Pashtun tribesmen had hated the modern state-"with all its works, and all its empty promises."

The promises had indeed been empty. In modern Europe (not necessarily the US) the state is so much part of our lives that we hardly see it and its impact feels benign. But it took centuries of suffering before this came to be. In the words of one expert on royal Afghanistan, the bulk of the population's conception of the state was "forced settlement, forced labour, military service, taxation and bureaucratic haphazardness." This would be equally true of France or western Europe in the 18th century. In Afghanistan, this contrasted not only with the world of tribal anarchy and independence but also with the old, loose, Islamic order that suited that world pretty well.

As in other "developing countries," the royal state's one success led to disaster. Although the modern education system was limited to a fraction of the population (and a smaller proportion of women), it was enough to create a mass of educated graduates and junior bureaucrats and officers for whom no well-paying jobs could be found. This produced the communist revolution of 1978, essentially an attempt to relaunch the state's modernizing program by a return to the methods of Abdur Rahman.

Like Abdur Rahman's program, that of the communists depended on subsidies and weapons from an outside protector, in this case the Soviet Union. As in Rahman's time, this sparked resistance from religious, ethnic and tribal groups; but this resistance triumphed, and between 1978 and 1992 overthrew the communist regime and then the Afghan state itself, first in the mountains and then in Kabul and the other main cities. But the resistance proved incapable of replacing this state with a unified authority, except-after a period of violent chaos-in the pathological form of the Taliban

The reasons for this catastrophe were twofold. First, the Soviet Union failed to learn the lesson of the British empire and occupied Afghanistan in support of their communist clients, thereby uniting both religious and nationalist feeling against them. Second, because of the cold war and regional factors the rebels gained military and financial aid from the US, Pakistan, Saudi Arabia and Iran. The enmity between these countries then helped fuel the subsequent civil wars between the Afghan resistance forces.

The difficulty of creating an Afghan state from below, or by mass consent, has been hugely complicated by the region's ethnic make-up. There are parallels between Afghanistan and Somalia, but the former makes the latter look simple. Somalia is what Afghanistan would be if it contained only Pashtuns, if the majority of Pashtuns lived in Afghanistan, and if neighboring states had no great interest in intervening in the country.

Afghanistan, by contrast, suffers from a set of interlocking, intractable obstacles to the creation of basic order of the Somali type, let alone a modern state. The original "state-forming ethnos," the Pashtuns ("Afghan" and Pashtun originally seem to have been synonyms), make up fewer than half of the total population of 24m, with the rest divided between a range of different nationalities. Tajiks, Uzbeks, and Hazaras (Shi'as of Mongolian descent) are the largest but many smaller ones play key roles in their areas.

Historically, relations between the Pashtuns and the other nationalities have varied. Enormous resentment was caused by Rahman's use of mainly Pashtun armies to subdue other nationalities, and by his settling of Pashtun colonists on their lands. (The area around Kunduz in northern Afghanistan contained many of these settlements, which helps explain why it was one of the last Taliban strongholds to fall to the Northern Alliance forces.)

For most of the following decades, relations were fairly peaceful, though tensions between Pashtuns and non-Pashtuns were reflected in the savage rivalry between the two factions of the communist movement, the Parcham and Khalq. The Khalq represented the Pashtuns and revealingly was the more radical, anti-establishment force. For, like the Turks of the Ottoman empire before AtatÜrk, though the Pashtuns were the "state-forming ethnos," they and their language were not favored by the ruling dynasty, which was Pashtun but spoke Persian.

Relations between the nationalities have worsened as a result of the events of the past ten years. Shock at the seizure of power in Kabul in 1992 by forces representing the national minorities explains some of the subsequent vicious behavior by Pashtun forces. There has also been intermittent warfare between other nationalities. The Shi'a Hazara of the central highlands have been despised and maltreated by just about everybody, both as heretics and as a source of migrant unskilled labor. None the less, the Pashtun perception of grievance and oppression is very strong, and will not be diminished by the presence of a Pashtun from the old royal elites-Hamid Karzai-as prime minister of the planned interim administration agreed in Bonn. What the Pashtuns will see is that the ministries possessing armed force are in the hands of the Northern Alliance, and that-as I write in early December-their forces are in actual control of Kabul.

The Pashtuns have serious limitations as state-builders arising from the fact they are not a majority, and, historically, have had poor relations with most of the other nationalities. Moreover, the Pashtuns also suffer from internal obstacles to political cooperation and state-formation. Tribal societies like the Pashtuns, the Somalis, the Berbers or (to a lesser extent) the Chechens are by their nature unfitted to act as the basis for the formation of modern states, the needs of which are in direct opposition to their traditions of "ordered anarchy" and tribal codes of individual and family honor, defended by force.

Pashtunwali, the Pashtun ethnic code, has its noble aspects, but this is the nobility of the tribal dark ages. It mandates a pathological commitment to honor and revenge, decrees the bearing of arms and use of them as a central aspect of manliness, and encourages a hysterical vying for supremacy among individual males, including those from the same extended family. "Fear your cousin as you would an enemy," goes the Pashto saying. Given such traditions, it is unsurprising that Afghanistan's only experience of a freely-elected parliament, from 1965-1973, was hopelessly undermined by the political and personal feuding, corruption, ambition and violence of its deputies.

Elsewhere in the world, anarchic traditions like those of the Pashtuns have been overcome by modern nationalism; and in the 1930s and 1940s, there were indeed moves in that direction in the Pashtun areas of British India. These were orchestrated by the so-called Redshirts, a Pashtun nationalist party in anti-British alliance with the Indian National Congress.

But after partition in 1947, this tendency was crippled by the border running through the Pashtun lands which left the majority of Pashtuns in Pakistan. (Today, Pashtuns make up some 38 per cent of Afghanistan's population and 10 per cent of Pakistan's.) The new Pakistani state saw nationalism among its Pashtuns as a threat, and tried to stifle it. On the other hand, many Pashtuns were drawn into the service of that state by considerations of personal and family advantage (for all its drawbacks, Pakistan has always been far larger and more developed than Afghanistan) and by considerations of Islamic loyalty against the heathen Hindus of India.

It should also be said-given a tendency to blame Pashtun culture for all Afghanistan's ills-that the values of other ethnic groups are not much better. The national sport, Buzkashi (see photo, p25), often used as a metaphor for Afghanistan's violent anarchy, is not a Pashtun game but an Uzbek and Turkmen one. In his work on the game and its meaning before the Soviet occupation, Buzkashi: Game and Power in Afghanistan, G Whitney Azoy notes of the society which gave it birth that "virtually everyone is a potential predator." His book describes the governor of Kunduz in 1977 judging an acrimonious game of Buzkashi with the assistance of a sawn-off shotgun, partially covered by a green velvet cloth.

An understanding of this history is of crucial importance for designing policies for aid and state-building in Afghanistan. Many representatives of the westernized Afghan elites, who have spent the past generation in exile, are engaged in telling fairy stories about pre-communist Afghanistan to naïve western aid donors. The tenor of these stories is that Afghanistan was a successful secular modern state with a strong civil society; that elements of this former state and society remain below the surface; and that by implication this state could be recreated with help from the west.

The dangers of the "international community" being led astray are increased by the unprecedented amounts of money being earmarked for Afghanistan. This is generating a gold rush in the world of international NGOs. These are often worthwhile institutions, but they are also a class, with class interests; and they want a piece of the action. Given their prejudices and their lack of knowledge of Afghanistan, they may depend on English (or French, or German) speaking émigrés, who also have their own interests and perceptions and who have a vested interest in encouraging the most ambitious agendas possible.

At the same time, the international community lacks dedicated, expert cadres of the IPS type, willing to dedicate their careers to managing the Afghans. I certainly would not volunteer. My grandmother was the widow of a British official in India, and lived there most of her life-without electricity, let alone air conditioning, in extreme loneliness and isolation and in continual danger of being killed by nationalist terrorists, a death which befell her sister and stepmother (this was Bengal and Bihar from the 1910s to the 1930s). I know I am not half the person she was and I tailor my ambitions accordingly. As for the bushy-tailed postgraduates who run western "democratization" programmes-if they try to reform Afghanistan, they are liable to end up as squirrel kebab.

The UN's ability to impose political direction risks being overwhelmed by a flood of money and aid groups with narrow ideological agendas. The effect could be waste, corruption and local cultural irritation- and, more importantly, the loss of any ability to use aid as a tool to maintain peace and order. The result, as in Somalia, could be a backlash against the western presence. That in turn could prompt the west to throw up its hands and abandon the region as beyond saving. But we have seen in Afghanistan (and Somalia, if reports of al Qaeda's presence there are true) the dangers for the US which can result from abandoning failed states in this way.

Instead, we need a strategy based on two things. The first is a recognition of present Afghan power relations on the ground. Following the destruction of the Taliban, power in the various areas is held by heavily-armed tribal warlords. In the northern half of the country, these are loosely grouped in the "Northern Alliance"-though this seems unlikely to last absent its raison d'àªtre of resistance to the Taliban In the Pashtun areas, confusion reigns and it will be hard to create any unifying political structures. These groups will not surrender their arms or their local power to any national government.

A central government which could achieve the goal of overcoming the warlords and the ethnic militias would require not only the army and international subsidies of Abdur Rahman, but also his unspeakable ruthlessness-or indeed that of the Taliban, the only force of the past generation to have made a moderately successful attempt to unify the country. No such army exists or could be created, and it is unlikely that the west or anyone else would subsidize a process involving so much inevitable cruelty.

This is an area where analogies between territories under UN and/or NATO supervision today and the "mandated territories" of the League of Nations break down. The mandates were given to a colonial power which ruled by colonial means-in other words, extreme ruthlessness. When the British took over Iraq after the first world war, they suppressed local rebellions by Kurds and Assyrians with the help of poison gas dropped by the RAF and massacres by local Iraqi auxiliaries. The contemporary west has profited too from atrocities by local auxiliaries in Croatia, Kosovo and, most recently, Afghanistan but it can hardly back a campaign of this kind lasting many years.

Consequently, the second necessity on the part of western policymakers is a willingness to devote not years, but decades of steady effort to ameliorating the awful conditions of life of most Afghans, restoring the bases of a productive economy, creating basic state services-especially a minimally stable and useful currency-and preventing major conflicts.

Given Afghan realities, this does not require the creation of a strong Afghan central government with control over large amounts of western aid. Instead, we need to learn how to attain the best possible results in the quasi-medieval circumstances of contemporary Afghanistan: that is to say, a country in which any foreseeable state will lack one of the distinguishing features of the modern state, a monopoly of armed force. Therefore, our immediate aim for the Afghan government should be not the impossible fantasy of an administration governing the country, but a kind of national mediation committee, representing local interests and negotiating between them. The deal in Bonn was a first step.

The aim of this negotiating process should be to create the minimal conditions for medieval civilization: the avoidance of major armed conflict; the security of the main trade routes; and the safety and neutrality of the capital. In the case of Afghanistan, this should be secured not by an Afghan national army-another fantasy at present-but by an international force created by the UN, led first by Britain and then by Turkey, and backed by US and British airpower.

Most of western aid should not be directed through the "Afghan government" but should as far as possible be sent directly to Afghanistan's regions, from across the borders of Afghanistan's neighbors. It should be used in a clear-headed way as an instrument not of state-building but of peacekeeping-crudely, as a way of bribing warlords and armies into not going to war with each other. At the same time, if we can secure Kabul through an international force backed by the US air force, we can do our utmost to turn that city into a model for the rest of the country.

While this may be the most we can attempt, it is also the least that we should attempt. We should never again abandon Afghanistan to its demons. On 11th September, they came back to haunt us.