Why geography — unfortunately — is destiny for South Asia's troubled heartland.
BY ROBERT D. KAPLAN
Perversity characterizes Pakistan. Only the worst African hellholes, Afghanistan, Haiti, Yemen, and Iraq rank higher on this year's Failed States Index. The country is run by a military obsessed with — and, for decades, invested in — the conflict with India, and by a civilian elite that steals all it can and pays almost no taxes. But despite an overbearing military, tribes "defined by a near-universal male participation in organized violence," as the late European anthropologist Ernest Gellner put it, dominate massive swaths of territory. The absence of the state makes for 20-hour daily electricity blackouts and an almost nonexistent education system in many areas.
The root cause of these manifold failures, in many minds, is the very artificiality of Pakistan itself: a cartographic puzzle piece sandwiched between India and Central Asia that splits apart what the British Empire ruled as one indivisible subcontinent. Pakistan claims to represent the Indian subcontinent's Muslims, but more Muslims live in India and Bangladesh put together than in Pakistan. In the absence of any geographical reason for its existence, Pakistan, so the assumption goes, can fall back only on Islamic extremism as an organizing principle of the state.
But this core assumption about what ails Pakistan is false. Pakistan, which presents more nightmare scenarios for American policymakers than perhaps any other country, does have geographical logic. The vision of Pakistan's founder, Muhammad Ali Jinnah, in the 1940s did not constitute a mere power grab at the expense of India's Hindu-dominated Congress party. There was much history and geography behind his drive to create a separate Muslim state anchored in the subcontinent's northwest, abutting southern Central Asia. Understanding this legacy properly leads to a very troubling scenario about where Pakistan — and by extension, Afghanistan and India — may now be headed. Pakistan's present and future, for better or worse, are still best understood through its geography.
THE MUSLIM EXPERIENCE in South Asia begins with the concept of al-Hind, the Arabic word for India. Al-Hind invokes the vast tracts of the northern and northwestern parts of the Indian subcontinent that came under mainly Turko-Islamic rule in the Middle Ages and were protected from the horse-borne Mongols by lack of sufficient pastureland. The process of Muslim conquest began in Sindh, the desert tract south and east of Iran and Afghanistan, adjacent to the Arabian Sea, easily accessible to the Middle East by land and maritime routes.
The Umayyad Arabs conquered and Islamicized Sindh in the early eighth century. Then came the Turkic Ghaznavids (based out of Ghazni, in eastern Afghanistan), who conquered parts of northern India in the 11th century. The Ghaznavids were followed by the Delhi Sultanate, a military oligarchy between the early 13th and early 16th centuries, which preceded the splendorous rule of the Persianized Mughal dynasty on the subcontinent. All these Muslim warriors governed immense inkblots of territory that were extensions of the Arab-Persian world that lay to the west, even as they interacted and traded with China to the north and east. It was a land without fixed borders that, according to University of Wisconsin historian André Wink, represented a rich confection of Arab, Persian, and Turkic culture, bustling with trade routes to Muslim Central Asia.
To the extent that one area was the ganglion of this Muslim civilization, it was today's Pakistan. Fertile Punjab, which straddles the Pakistan-India frontier, "linked the Mughal empire, through commercial, cultural and ethnic intercourse, with Persia and Central Asia," writes University of Chicago historian Muzaffar Alam. This area of Pakistan has been for centuries the civilizational intermediary connecting the cool and sparsely populated tableland of Central Asia with the hot and teeming panel of cultivation in the Indian subcontinent. Pakistan's many mountain passes, especially those of Khyber and Bolan, join Kabul and Kandahar in Afghanistan with the wheat- and rice-baskets thousands of feet below. The descent from Afghanistan to the Indus River, which runs lengthwise through the middle of Pakistan, is exceedingly gradual, so for millennia various cultures occupied both the high plateaus and the lowland riverine plains. This entire middle region — not quite the subcontinent, not quite Central Asia — was more than a frontier zone or a bold line on a map: It was a fluid cultural organism and the center of many civilizations in their own right.
What we know as modern-day Pakistan is far from an artificial entity; it is just the latest of the many spatial arrangements for states on the subcontinent. The map of the Harappan civilization, a complex network of centrally controlled chieftaincies in the late fourth to mid-second millennium B.C., was one of its earliest predecessors. The Harappan world stretched from Baluchistan northeast up to Kashmir and southeast down almost to both Delhi and Mumbai, nearly touching present-day Iran and Afghanistan and extending into both northwestern and western India. It was a complex geography of settlement that adhered to landscapes capable of supporting irrigation, and whose heartland was today's Pakistan.
The Mauryan Empire, which existed from the fourth to the second centuries B.C., came to envelop much of the subcontinent and thus, for the first time in history, encouraged the idea of India as a political entity. But whereas the area of Afghanistan, Pakistan, and northern India all fell under Mauryan rule, India's deep south did not. Next came the Kushan Empire, whose Indo-European rulers conquered territory from the Ferghana Valley, in the demographic heart of Central Asia, to Bihar in northeastern India. Once again, the heart of the empire that linked Central Asia and India was in Pakistan; one of the Kushan capitals was Peshawar, Pakistan's frontier city today.
Later on, throughout the Middle Ages and the early modern era, Muslim invaders from the west grafted India to the greater Middle East, with the Indus River valley functioning as the core of all these interactions, as close to the Middle East and Central Asia as it is to the Ganges River valley. Under the Delhi-based Mughal dynasty, which ruled from the early 1500s to 1720, central Afghanistan to northern India was all part of one polity, with Pakistan occupying the territorial heartland.
Rather than a fake modern creation, Pakistan is the very geographical and national embodiment of all the Muslim invasions that have swept down into India throughout its history, even as Pakistan's southwest is the subcontinental region first occupied by Muslim Arabs invading from the Middle East. The Indus, much more than the Ganges, has always had an organic relationship with the Arab, Persian, and Turkic worlds. It is historically and geographically appropriate that the Indus Valley civilization, long ago a satrapy of Achaemenid Persia and the forward bastion of Alexander the Great's Near Eastern empire, today is deeply enmeshed with political currents swirling through the Middle East, of which Islamic extremism forms a major element. This is not determinism but merely the recognition of an obvious pattern.
The more one reads this history, the more it becomes apparent that the Indian subcontinent has two principal geographical regions: the Indus Valley with its tributaries, and the Ganges Valley with its tributaries. Pakistani scholar Aitzaz Ahsan identifies the actual geographical fissure within the subcontinent as the "Gurdaspur-Kathiawar salient," a line running from eastern Punjab southwest to the Arabian Sea in Gujarat. This is the watershed, and it matches up almost perfectly with the Pakistan-India border. Nearly all the Indus tributaries fall to the west of this line, and all the Ganges tributaries fall to the east. Only the Mauryas, Mughals, and British bonded these two regions into single states. For those three empires, the Indus formed the frontier zone and required many more troops there facing restive Central Asia than along the Ganges, which was under no comparable threat.
Likewise, the medieval Delhi Sultanate faced so much trouble in Central Asia that it temporarily moved its capital westward to Lahore (from India to Pakistan, in today's terms) to deal with the military threats emanating from what is today Afghanistan. Yet, for the overwhelming majority of history, when one empire did not rule both the entire Indus and the entire Ganges, the southern and eastern parts of Afghanistan, most of Pakistan, and northwestern India were nevertheless all governed as one political unit. And the rich and populous Indus Valley, as close to the wild and woolly Central Asian frontier as it was, formed the pulsating imperial center of that unit.
Here, alas, is the conundrum. During the relatively brief periods when the areas of India and Pakistan were united — the Mauryan, Mughal, and British — there was obviously no issue about who dominated the trade routes into Central Asia. During the rest of history, there was no problem either, because while empires like the Kushan, Ghaznavid, and Delhi Sultanate did not control the eastern Ganges, they did control both the Indus and the western Ganges, so that Delhi and Lahore were under the rule of one polity, even as Central Asia was also under their control. Today's political geography is historically unique, however: an Indus Valley state, Pakistan, and a powerful Ganges Valley state, India, both fighting for control of an independent and semi-chaotic Central Asian near abroad — Afghanistan.
Despite its geographical and historical logic, this Indus state is far more unstable than the Gangetic state. Here, too, geography provides an answer. Pakistan encompasses the frontier of the subcontinent, a region that even the British were unable to incorporate into their bureaucracy, running it instead as a military fiefdom, making deals with the tribes. Thus, Pakistan did not inherit the stabilizing civilian institutions that India did. Winston Churchill's first book as a young man, The Story of the Malakand Field Force, wonderfully captures the challenges facing colonial border troops in British India. As the young author then concluded, the only way to function in this part of the world is through "a system of gradual advance, of political intrigue among the tribes, of subsidies and small expeditions."
PAKISTAN'S GEOGRAPHICAL COHERENCE, albeit subtle and problematic, is mirrored in its subtle and problematic linguistic coherence. Just as Hindi is associated with Hindus in northern India, Urdu is associated with Muslims in Pakistan. Urdu — from "horde," the Turkic-Persian word for a military camp — is the ultimate frontier language. Reflecting its geographical links to the Middle East, Urdu is written in a Persianized Arabic script, even though its grammar is identical to Hindi and other Sanskritic languages. It is often believed that Urdu came into existence through the interaction of Turkic, Persian, and indigenous Indian soldiers in Mughal army encampments, not just on the Indus frontier but in the medieval Gangetic cities of Agra, Delhi, and Lucknow. Thus, it is truly the language of al-Hind.
Urdu is Pakistan's lingua franca, even as Punjabi, with links to the non-Islamic Sikhs and Hindus, enjoys a plurality of native speakers in Pakistan. Under Pakistan's military dictator Muhammad Zia ul-Haq in the 1980s, the combination of Urdu literacy programs in religious institutions and the teaching of Arabic in state schools gave Urdu more of a Middle Eastern and Islamic edge, writes Alyssa Ayres, now U.S. deputy assistant secretary of state for South and Central Asia, in Speaking Like a State: Language and Nationalism in Pakistan.
The linguistic, demographic, and cultural organizing principle of the Indus Valley is Punjab, whose name means "five rivers": the Beas, Chenab, Jhelum, Ravi, and Sutlej, all tributaries of the Indus. Punjab represents the northwesternmost concentration of population and agriculture before the ground starts to climb toward the wilds of Central Asia. As such, it is coveted because of its special access to Central Asian trade routes, though it was a frontier battleground in its own right relative to the rest of British India.
Because of Sikh uprisings, the Mughals had a difficult time securing Punjab. The British fought two wars to wrest the region from the Sikhs in the 1840s, after the rest of India had already been subdued. Once Punjab was conquered, however, the Pashtun northwest frontier, the gateway to Afghanistan and Central Asia, beckoned for the British. Because Punjab abutted the northwest frontier zone, which in turn abutted southern Central Asia, its soldiers became known for their military prowess — the "sword arm of India," contributing 28 of the 131 infantry units in the Indian Army by 1862.
But with the re-creation of an Indus state and a Gangetic state upon the demise of the British Raj in 1947, Punjab, rather than a frontier province of greater India, became the urban hub of the new Indus Valley frontier state: Pakistan. Although eastern Punjab fell within India, western Punjab still contains more than half of Pakistan's population. With close to 90 million people, western Punjab would be the world's 15th-largest country, putting it ahead of Egypt, Germany, Turkey, and Iran. Punjabis have accounted for as much as 80 percent of the Pakistan Army and 55 percent of the federal bureaucracy.
Punjab is like an internal imperial power ruling Pakistan, in the way that Serbia and the Serbian army ran Yugoslavia prior to that country's civil war and breakup. "Punjab is perceived to have 'captured' Pakistan's national institutions through nepotism and other patronage networks," writes Ayres. Its rural female literacy rate is nearly twice that of Sindh province and the province on the northwest frontier with Afghanistan, and it's more than triple Baluchistan's. Punjabis, she adds, "are better off than everyone else [in Pakistan], with more productive land, cleaner water, better technology, and better educated families."
Pakistani historian and anthropologist Muhammad Azam Chaudhary writes, "If the motherland of the five rivers [Punjab] had not been obtained, then in terms of geography, it would have been impossible to establish Pakistan." Yet Punjab itself is not indivisible, for the southern part of the province is made up of speakers of Saraiki — a linguistic mixture of Punjabi and Sindhi — with their own separate identity. And while the rest of Pakistan sees Punjab as hegemonic, Punjabis themselves harbor an inferiority complex (again, like the Serbs), claiming that they have sacrificed much for a state that doesn't work and, as a result, get insufficient respect from other Pakistanis.
The tension between Punjabis and other Pakistanis overlaps with the tension that exists among the other ethnic groups. Chronic urban conflict in Karachi, Pakistan's largest city, pits local Sindhis against Baluchis and Pashtuns, just as in Baluchistan there are tensions between Baluchis and Pashtuns. Islamic ideology, like communism in Yugoslavia, has proved an insufficient glue to form a prideful national identity. Instead, this frontier region between the Middle East and Hindu India has become an explosive amalgamation of often warring ethnic identities.
This is not, of course, how Jinnah envisioned Pakistan. He imagined a federalized state in which the various ethnically based provinces retained a high degree of autonomy. With such freedom, the angst of domination by Punjabis — and by each other — would not have existed, allowing for a civil society to emerge and, with that, a state with vibrant institutional capacity. Indeed, history shows that central authority can only be effective if it is strictly delimited. Regrettably, Pakistan has been what 20th-century European scholars Ernest Gellner and Robert Montagne call a "segmentary" society. Hovering between centralization and anarchy, such a society, in Montagne's words, is typified by a regime that "drains the life from a region," even though, "because of its own fragility," it fails to establish lasting institutions. This is the byproduct of a landscape riven by mountains and desert, a place where tribes are strong and the central government is comparatively weak. Put another way, Pakistan, as King's College London scholar Anatol Lieven notes, is a weak state with strong societies.
India is the counterfactual to Pakistan's dilemma. India's individual states are linguistically based and thus have confident identities: Kannada-speaking Karnataka, Marathi-speaking Maharashtra, Telugu-speaking Andhra Pradesh, Bengali-speaking West Bengal, Hindi-speaking Uttar Pradesh, and so forth. This might, in some scenarios, lead to local nationalism and irredentist movements, as is the case with Pakistan. Because central authority in New Delhi is restricted, however, diversity is celebrated and has become, in turn, a healthy basis for a pan-Indian national identity.
If India were less diverse and consisted of only the "cow belt" of Hindi-speaking northern India, observes Lieven, it might not have become a democracy but rather "some form of impoverished Hindu-nationalist dictatorship." Instead, India is like Indonesia: a geographically sprawling and diverse democracy united by a common language that does not threaten the use of local tongues and dialects.
Kashmir, the contested region over which India and Pakistan have fought for decades, is where the two countries' different personalities are most in evidence. According to Indiana University's Sumit Ganguly, India requires the Muslim-dominated Himalayan territory to substantiate its claim as a multiconfessional democracy, rather than as a Hindu-dominated state, whereas Pakistan requires Kashmir to substantiate its claim as the chief remnant of Muslim al-Hind.
And so we come to the core reason for Pakistan's perversity. The fact that Pakistan is historically and geographically well-rooted is only partially a justification for statehood. Although a Muslim frontier state between mountains and plains has often existed in the subcontinent's history, that past belonged to a world not of fixed borders, but rather of perpetually moving spheres of control as determined by the movements of armies — such was the medieval world. The Ghaznavids, the Delhi Sultanate, and the Mughal dynasty all controlled the subcontinent's northwestern frontier, but their boundaries were all vague and somewhat different from one another — all of which means Pakistan cannot claim its borders are legitimate by history alone. It requires something else: the legitimacy that comes with good governance and strong institutions. Without that, we are back to the medieval map, which is what we have now — known in Washington bureaucratic parlance as "AfPak."
The term AfPak itself, popularized by the late diplomat Richard Holbrooke, indicates two failed states — otherwise, they would share a strong border and would not have to be conjoined in one word. Let me provide the real meaning of AfPak, as defined by geography and history: It is a rump Islamic greater Punjab — the tip of the demographic spear of the Indian subcontinent toward which all trade routes between southern Central Asia and the Indus Valley are drawn — exerting its power over Pashtunistan and Baluchistan, just as Punjab has since time immemorial.
This is a world where ethnic boundaries do not configure with national ones. Pashtunistan and Baluchistan overlap with Afghanistan and less so with Iran. About half of the world's 40-plus million Pashtuns live on the Pakistani side of the border. The majority of the more than 8 million Baluchis live within Pakistan, the rest in neighboring Afghanistan and Iran.
In recent decades, the age-old pathways in this region have been used by Islamic terrorists, as well as by traditional traders. The link between Pakistan's premier spy agency, Inter-Services Intelligence (ISI), and the so-called Haqqani network tied to al Qaeda merely replicates the arteries of commerce emanating from Punjab outward to southern Central Asia. Punjabis dominate the ISI, and the Afghan Pashtun Haqqani network is both an Islamic terrorist outfit and a vast trade and smuggling operation, unto the Amu Darya River to the northwest and unto Iran to the west.
Because al-Hind has historically been so rich in cultural and commercial connections, when modern states do not sink deep roots into the land, the result is a reversion to traditional patterns, albeit with contemporary ideological characteristics. The U.S. State Department and many policy analysts in Washington have proposed a new silk route that could emerge in the event of a peace treaty in Afghanistan. What they fail to recognize is that a silk route is already flourishing outward from Punjab — it is just not oriented to Western purposes.
The longer the fighting goes on in Afghanistan and along the Afghanistan-Pakistan borderland, the weaker Pakistan as a modern state will become. As that occurs, the medieval map will come into even greater focus. Jakub Grygiel, a professor at Johns Hopkins University's School of Advanced International Studies, points out that when states or empires involve themselves in irregular, decentralized warfare, central control weakens. A state only grows strong when it faces a concentrated and conventional ground threat, creating the need to match it in organizational capabilities and thus bolstering central authority. But the opposite kind of threat leads to the opposite result. Pakistan's very obsession with the ground threat posed by India is a sign of how it requires a conventional enemy to hold it together, even as its answer to India in the contested ground of Central Asia — supporting decentralized Islamic terrorism from Afghanistan to Kashmir — is having the ironic effect of pulling Pakistan itself apart. It is unclear whether invigorated civilian control in Pakistan can arrest this long-term process.
This process could even quicken. With the Soviets abandoning Afghanistan in the late 1980s and the Americans on their way out in coming years, India will attempt to fill the void partially by building infrastructure projects and providing support to the Afghan security services. This will mark the beginning of the real battle between the Indus state and the Gangetic state for domination of southern Central Asia.
At the same time, as Pakistan is primarily interested in southern and eastern Afghanistan, the part of Afghanistan north of the Hindu Kush mountains may, if current trends continue, become more peaceful and drift into the economic orbit of the former Soviet Central Asian republics, especially given that Uzbeks and Tajiks live astride northern Afghanistan's border with Uzbekistan and Tajikistan. This new formation would closely approximate the borders of ancient Bactria, with which Alexander the Great was so familiar.
Indeed, the past may hold the key to the future of al-Hind.