Centralization/Decentralization in the Dynamics of Afghan History

The analysis of political organization in Afghanistan is clouded by a number of myths (unconquerable, ungovernable and graveyard of empires) that are contradicted by the facts. Historically Afghanistan was peacefully governed by a wide variety of conquerors and native dynasties, but all used combinations of direct and indirect rule to create stable polities. They also relied on theories of political legitimacy that vested authority in ruling elites that, once established, returned to power after periods of disruption to bring order to the country. This pattern of successful governance has been overlooked in rebuilding Afghanistan today to the detriment of political stability.


Afghanistan has never been conquered even by the most powerful of empires
In reality the territory of today's Afghanistan was conquered and successfully ruled by practically every imperial power in the region before the midnineteenth century as the very general summary below indicates: o The Achaemenid (Persian) Empire (sixth-fourth century BCE) o Alexander the Great and Greco-Bactrian kingdoms (fourth-first century BCE) o Kushans (first-third centuries) o Persian Sassanid Empire (third-sixth centuries) o Muslim Caliphate (seventh-eighth centuries) o Ghaznavids and Ghorids (tenth-twelfth centuries) o Chinggis Khan and the Mongols (thirteenth century) o Tamerlane and Timurids (fourteenth-fifteenth century) o Mughal India and Safavid Iran (sixteenth-mid-eighteenth centuries) o Durrani Empire (late eighteenth century) One striking feature of this imperial incorporation was that the ruling dynasties were (with the exceptions of Ghorids and Durranis) all of foreign origin, throwing into question the common assertion that Afghanistan never accommodated itself to rule by outsiders. But perhaps more important is that the territory of today's Afghanistan has served both as an imperial borderland and an imperial center, depending on the power base of the ruling elite. A classic case of its territory as a borderland includes its dismemberment by Mughal India, Safavid Iran, and the Uzbeks of central Asia from 1500 to 1750, or its long history as a fighting frontier between the Muslim and non-Muslim world from the mid-seventh to late twelfth centuries. At other times Afghanistan itself was an imperial center that projected power outward under the Kushans, Ghaznavids, and Ghorids, and again reemerged in this role for a short time under the Durrani Empire in the mid-eighteenth century. Finally, there were also historical periods in which the territories of Afghanistan became core parts of larger empires such as under the Timurids, who moved their capital from Samarkand to Herat, and under the Achaemenid Empire where (according to the ancient Greek historian Herodotus) its satrapies were some of the empire's most politically powerful and economically valuable.
It is true that the pre-modern empires that controlled Afghanistan regularly failed to incorporate its marginal mountainous and desert areas under their direct rule. Thus today's Pashtun and Nuristani inhabitants of Afghanistan's mountainous border region with Pakistan, the Hazaras in the high mountain massif of central Afghanistan, or the Tajik and Pamiri peoples of northeastern Badakhshan province can rightfully claim to have successfully resisted central governments and avoided conquest on many occasions. But since such people and their economically marginal territories rarely repaid the cost of administration, pre-modern empires ignored them or used policies of indirect rule such as access to trade, rule through local proxy leaders, or the occasional punitive campaign to project power. It was a Swiss cheese model of governance in which the state ruled the cheese and ignored the naturally occurring holes. In other words, failure to project state power uniformly across all the territory an empire proclaimed itself to be sovereign over was not viewed as a defect that needed to be corrected, but a political state of nature to which rulers were forced to adapt (Barfield 2010: 67-82).
The parts of Afghanistan that were conquered and successfully ruled included its cities, irrigated plains, and strategic trade routes that constituted the vast bulk of the region's wealth and its only densely populated territories. There were only four of them-each with a major city that dominated its region economically and politically: Herat in the west, Kandahar in the south, Balkh (Mazar-i-sherif) in the north, and Kabul in the west. In addition, Kabul was often tied as a single political unit to its sister city of Peshawar (the old Afghan winter capital) that is now in Pakistan as part of a dyad that linked central Asia to the plains of India via the Khyber Pass. These regional units have great time depth and continuity that spans 2500 years, going back to the time of Achaemenid Empire. It is they, and not the boundaries of modern Afghan state or its modern provinces, that constitute its deep historical political and economic building blocks. Although they are all part of the country of Afghanistan today, this is as more the product of historical accident than any compelling logic. What has been compelling is the pattern of successful incorporation of these regions into larger state structures (usually large empires) that provided for their governance, protected them from attack and secured access to international overland trade networks that was a source of great wealth and cultural sophistication. The acceptance of state authority was the default and viewed as the price of political and economic stability, the alternative to which was not freedom but anarchy.

Afghanistan is a graveyard of empires
No Afghanistan generalization is more popular among journalists than the 'graveyard of empires' trope. Such narratives highlight the destruction of the British expeditionary force ('Army of the Indus') in the First Anglo-Afghan War (1839-42), the massacre of the British diplomatic staff in Kabul, and the Afghan victory at the Battle of Maiwand during the Second Anglo-Afghan War (1878-80). They pass over the fact that such victories did not result in the fall of other Afghan cities nor did stop the British from returning to burn Kabul down in revenge in 1842 or beating back the insurgency in 1880. Similarly, insurgents did not drive the Soviets out of Afghanistan by winning battles but instead by making Moscow recalculate the cost of victory in time, blood, and money. Thus both the British and the Soviets withdrew from Afghanistan not because they were militarily defeated, but because they made the political decision that the benefits of leaving outweighed those for staying. And while each of these invasions ended in withdrawal of foreign forces, none proved fatal to the imperial power involved. For example, in the aftermath of the First and Second Anglo-Afghan Wars, Victorian Britain reached the apogee its imperial power over the whole Indian subcontinent, a rule that would last until 1947. Similarly, while the Soviet misadventure Afghanistan highlighted the fractures that had long undermined its economic, political, and economic power, the failure of the Afghan war was symptomatic of its problems not its cause. Indeed Gorbachev saw ending the Afghanistan war as a positive act that would allow him to focus on those very problems (Kalinovsky 2011).
What does demand some explanation is why the most powerful international actors at the time failed to succeed in Afghanistan after 1840 whereas seemingly weaker imperial polities before that time did. For while Afghanistan was not a graveyard of empires, it was a graveyard for a particular type of colonial policy that attempted not just to rule Afghanistan directly but transform it. Pre-modern conquerors of Afghanistan displaced or co-opted existing local elites, but made no attempt to transform existing social and economic relations in the county. By contrast Western powers coming to Afghanistan had much broader aims. The British set about modernizing Afghanistan's political, military, and economic structures in much the same way they had successfully done through most of South Asia and as the Russians were doing in Central Asia. (American policy in Afghanistan has also followed a similar plan of modernization.) It was opposition to these changes as much as foreign invasions that sparked opposition and generated rural insurgencies in Afghanistan, a pattern of opposition striking absent before the mid-nineteenth century. In south Asia the British had invested the time and resources to overcome such opposition because they deemed it central to their colonial aims in the region. Similarly the Czars, and more particularly the Soviets, transformed the economic and social structures of central Asian territories. While doing the same in Afghanistan was theoretically possible, the cost of doing so always proved to be a bridge too far. For both the British and the Russians Afghanistan was just a piece in a much larger international game they were playing with each other. Its people, territories, and resources were never deemed a vital national interest to either and, thus, staying or going was a policy option and not an existential question. The most powerful pre-modern empires had faced similar frontier policy dilemmas. Rome's failure to incorporate Germany or China's Great Wall boundary that excluded Mongolia were policy decisions made when these empires were at the height of their power, not when their power was in decline. They were policy options employed when the cost of expansion was determined to have exceeded a limit each empire was willing to pay. In part this was because, unlike the pre-modern empires that ruled Afghanistan, Rome and China both assumed that any territory they incorporated needed be directly ruled and its peoples, cultures, and economies transformed to meet imperial standards.

Afghanistan is ungovernable
Failures to stabilize Afghanistan by invading foreign armies have often been explained away by asserting that the Afghans are an ungovernable people who revel in war and anarchy. A recent example was a 2009 comment by the Prime Minister of Canada, Stephen Harper, that "We are not going to ever defeat the insurgency. Afghanistan has probably had-my reading of Afghanistan history-it's probably had an insurgency forever, of some kind." In fact Afghanistan had no insurgencies before 1841 and until 1978 its twentieth century history had been one of peace and stability with only minor disruptions. Portraying Afghanistan and the Afghans as unruly, however, has also been employed by Afghanistan's rulers to deter foreigners contemplating invasions and as a tactic to get outside aid by promising to keep them from causing trouble. Failures of government (the continuous exercise of state authority over the population it governs) have been wrongly conflated with failures of governance (the manner in which communities regulate themselves to preserve social order and maintain their security). Where the majority of the population resides in rural Afghanistan, local communities regularly maintain adequate local governance in the absence of formal government institutions. But communities that regularly resisted state intrusion nevertheless recognized the sovereignty of the Afghan national state and only challenged its legitimacy when provoked by the imposition of too much state authority.

Decentralization as a Tool of State Governance
The Afghan state's physical control of a specific territory has never been a valid reference point in assessing its ability to govern. Instead, the stability of the government was judged by the ability of its leaders to balance their interests against local needs and priorities. Effective leaders leveraged their power by devolving authority to non-state mediators to resolve many diverse local and regional grievances on the government's behalf in the most remote parts of the country. This allowed Kabul to preserve order and enhance its authority even in the absence of state institutions. It was a results-oriented system in which formal government institutions played a decreasingly smaller role the farther one was from the centers of state authority. Only when disputes that threatened the peace grew larger than local communities could handle, or threatened core state interests, did district or provincial government authorities see the need to intervene with state power. The degree that a government could exert direct authority was always variable. Weak governments could extend their writs only to the outskirts of major cities, powerful ones deep into rural areas. The ability to employ coercive power to create a system of direct rule was always the goal of governments in Kabul and to eliminate it the goal of autonomous rural areas. But both sides weighed the significant cost of challenging the other and it was this that produced an equilibrium that provided a surprising level of stability when left undisturbed (Barfield and Nojumi 2010).
This minimalist approach to government suited both Afghanistan's rural residents and Kabul appointed officials, who historically devoted themselves only to collecting taxes, conscripting soldiers and preventing banditry. Those regimes that insisted on imposing greater state authority generated rebellions in opposition to them. In one case, that of Amir Abdur Rahman (1880-1901), the national state eventually prevailed and crushed its rivals. But more recent instances of state expansion under King Amanullah (1919-29) and the PDPA (1978-92) failed and the national government collapsed. It was no accident that the 'do-little' Musahiban dynasty, which fell between these two radical reforming eras, deliberately chose minimalism as its guiding policy. Despite its many weaknesses, this dynasty gave Afghanistan a half-century of peace-the longest in the country's history-and maintained a functioning state structure throughout its existence.
The conclusion that may be drawn from Afghan history is that communities in rural areas that resisted the Afghan government's attempts to interfere in their affairs never rejected the need for governance. They just believed that their own informal institutions better maintained long-term local order than any distant government could. As significantly, all communities in Afghanistan (even those most insistent on preserving their own autonomy) accepted the need for an Afghan government in Kabul that could take on higher-level responsibilities that require a state structure. These include government's role in preserving internal security, protecting the country from hostile neighbors and negotiating on the nation's behalf for benefits from the larger international community. The narrowly constructed Afghan Constitution of 2004 that vests all administrative authority in the Kabul government failed to appreciate this lesson. Its 'one size fits all' approach was a recipe for failure precisely because it did not accommodate the country's historic diversity, a critical variable for any plan to implement models of governance there. Provinces and districts might look the same on a map, but some had much more significance than others. The administrative structure of the government therefore lacked the flexible tools of governance necessary to succeed in areas of low population density with subsistence economies (Barfield 2011).

Cultural Aspects of Political Authority
Debates about Afghan tribes and ethnic groups and tribes are numerous and contradictory. Some assert that such units are fictions that lead to wrongheaded and misleading analyses while others contend that they are the fixed bedrock units of social organization vital to understanding politics there. One problem in this debate is that the unit of analysis for tribe or ethnic group in Afghanistan, glossed as qawm, is context-specific. Qawm affiliations may refer just to a person's immediate kin or home village (hundreds of people), regionally based affinity groups (thousands or tens of thousands of people), or nationally recognized tribes and ethnic groups (hundreds of thousands or millions of people). Another problem is that as the level of generalization expands, the more checkered with exceptions it becomes. Smaller units that display a great deal of uniformity, solidarity, and accepted self-identification dissolve at the higher level. Groups identified simply as Pashtun, Tajik, Uzbek, or Hazara fall into Benedict Anderson's (1983) slot of "imagined communities" that take on salience only in opposition to (or cooperation with) similar such units (for a map of the distribution of main Afghan ethnic groups see Figure 1).
If this is the case, is there any reason to do any analysis or comparison at the large group level at various times in history? One area in which it may well be useful is looking at some comparisons of baseline dynamics of economy, social organization, and models of political authority. This would show that over long periods of time different groups in Afghanistan have displayed distinctly different models of political organization depending on 1) whether they are organized by descent group (tribal) or territory (non-tribal) and 2) the  degree to which they accept hierarchical differences in social and political organization. When applied to Afghan political history such an analysis reveals that the success in maintaining a Durrani royal dynasty for 230 years had its roots in lessons borrowed from dynasties of Turko-Mongolian origin that ruled Afghanistan in the 700 years preceding the establishment of an Afghan state in 1747, not in the culture of egalitarian Pashtun tribal jirgas. Indeed the Pashtuns of eastern Afghanistan, who continually maintained a highly egalitarian social and political structure with no hereditary leaders, remained historically opposed to all organized state power, whether by outsiders or their own ethnic kin.
In an important article comparing the tribal cultures of the Middle East and Inner Asia, Charles Lindholm (1986) identified marked structural differences between the hierarchical Turko-Mongolian cultural tradition of Inner Asia and the egalitarian cultural tradition of tribes indigenous to the Middle East. Kinship terms among Inner Asian peoples made distinctions between elder and younger brothers, junior and senior generations, and noble and common clans. Political leadership was vested in rulers drawn only from distinct 'royal clans' whose claim on political authority, once established, was so strong that a single ruling lineage might rule for many centuries. By contrast more politically egalitarian tribes in the Middle East found it difficult to unite and leaders found it practically impossible to maintain dynastic control for more than a few generations. This was because their large-scale political organizations were undercut by culturally based limitations on a leader's authority held by their own people. Tribes composed of such egalitarian lineages forced leaders to rule by means of consensus or mediation. While they could unite rival groups through use of segmentary opposition, leaders found it difficult to maintain a broad confederation for longer than a single lifetime. Arab Bedouins (and most Pashtuns) characteristically resisted subordinated cooperation at the supra-tribal level because they defined themselves through localized lineages that expected to remain autonomous.
By contrast, in Turko-Mongolian tribal systems, where a hierarchical kinship organization was accepted as culturally legitimate, local lineages, clans, and tribes became the building blocks of political-military coalitions created by hereditary leaders whose authority was rarely challenged from below. Large tribal confederacies (such as those in Iran, Central Asia, and Mongolia) could not have incorporated hundreds of thousands of people without employing such a 'top-down' imposition of order that had no room for independent political agents at the local level.
Beginning with the Ghaznavid rule (975 to 1173), dynasties of Turko-Mongolian origin established almost continual control over the territories of today's Afghanistan for seven centuries. Originally of central Asian nomadic origin, their horse cavalry gave them enough military superiority to conquer the areas they invaded. However, although Turko-Mongolian tribal leaders were skilled at building tribal confederations, they lacked the basic skills of sedentary administration and were initially forced to rely on the assistance of literate Persian sedentary advisers for administration. This led to emergence of a state with a dual organization in which a shah of Turkish origin maintained a Turkish military establishment (men of the sword) and a Persian-speaking administrative structure (men of the pen), a structure successive rulers found maintained long after they had become well-established and sophisticated (Barfield 1991).
This pattern was still the norm in 1500 when northern Afghanistan was ruled by Uzbeks based in Bukhara, western and southern Afghanistan by the Iranian Safavids in Isfahan, and Kabul and eastern Afghanistan by the Mughal dynasty from their Indian capital of Delhi. Pashtun leaders served as client governors for the Safavids and Mughals until the mid-eighteenth century when Ahmad Shah Durrani took control of the eastern half of Nadir Shah Afshar's empire upon his death in 1747. While this marked the first time Pashtuns had ruled over Afghanistan, the Durranis followed a Safavid-like model of administration that relied heavily on Persian speaking administrators and appointed governors who often had ties to the local population. The Durrani kings also managed to instill in themselves the aura of a 'royal clan' that had a monopoly on succession to the throne, something more characteristic of the hierarchical Turks than egalitarian Pashtuns.
The Durranis succeeded in establishing themselves as a ruling dynasty in part because they managed to place themselves above all other Pashtun groups. Non-royal southern Pashtun leaders accepted this state of affairs because they benefited from state subsidies, tax-free land grants, and government favoritism. The eastern Pashtuns, their potential rivals for leadership, also accepted Durrani supremacy, but for a very different reason: they were too egalitarian and too divided by local rivalries to be able to unite under a supreme leader of their own. However, unlike the Turkish dynasties whose unbroken lineages lasted centuries, Afghan rulers regularly faced competition for the throne from rival Durrani clans. Ahmad Shah's Popalzai/Sadozai clan was forced to cede power to Dost Muhammad's Barakzai/Muhammadzai clan in 1825 and another internal dynastic change occurred in the aftermath of the 1929 civil war when Nadir Shah's Musahiban lineage displaced King Amanullah and his heirs. Literate urban Persian speakers ('Tajiks') generally avoided competing for power directly. Instead they followed their time-tested strategy of co-opting the Durranis the same way their ancestors had handled the Turks: through their mastery of government administration, finance, and as purveyors of the region's high culture. (The failed attempt by the Tajik rebel Habibullah Kalakani to secure the amirship in 1929 bore out the difficulties of attempting to rule directly.) The Turks in the north, while politically sidelined, had always been content to accept rule from urban centers, whether by the Uzbeks in Bukhara or the Pashtuns in Kabul.
Such broad scale political strategies played themselves out in post-1978 Afghanistan. The leaders of the Peoples Democratic Part of Afghanistan (PDPA) government, who were eastern Pashtuns, almost completely purged the old Durrani elite from power. However, they proved unable to consolidate their authority among their own people and had a disturbing tendency to murder one another. In the war that followed the Pashtun military leaders on both the PDPA and the mujahideen sides were overwhelmingly eastern Pashtuns and few Durrani military leaders emerged on either side. The Tajiks revived their military tradition and became unified fighting forces under Ahmad Shah Masud in the east and Ismail Khan in the west. The Uzbeks under Abdul Rashid Dostum formed an effective militia, but sought regional autonomy rather than national power. The Taliban leader Mullah Omar (of eastern Pashtun tribal origin) attempted to break the old model of power by declaring a clerical regime, but his Islamic amirate collapsed in the face of the American invasion of 2001. Despite its military victory, the predominantly Tajik Northern Alliance reverted to their traditional template by accepting a Sadozai Pashtun, Hamid Karzai, as leader of the country, while maintaining strong influence over government institutions.
The easy reestablishment of a Durrani ruler once again demonstrated the fractiousness of the eastern Pashtuns and their inability to coalesce politically. (One Ghilzai explanation of this was a curse by an eighteenth century Muslim pir who proclaimed that the Ghilzais would live under Durrani control for seven generations, "Badshahi da Durrani, tura da Ghilzai" [Kingship to the Durrani, but to the Ghilzai the sword] (Ahmed 1982(Ahmed : 1104. While such largescale generalizations may explain no specific political action, the structure of political relationships in Afghanistan does appear to create a bias toward one set of outcomes, as opposed to others.