War: What is it good for?

by Ian Morris

  In these forts the remnants of Agricola’s army stayed for the next forty years. They wrote home, they fought deadly little skirmishes with Caledonians (“there are lots of cavalry,” another urine-soaked memo from Vindolanda observes), and—above all—they waited. Only in the 120s A.D. did they move on, but not to new triumphs. Rather, the emperor Hadrian set them to building the great wall across Britain that bears his name. Rome had abandoned the conquest of the North (Figure 3.1).

  Figure 3.1. The limits of empire in the West: sites in western Eurasia mentioned in this chapter

  As Tacitus saw it, all this came about because the emperor Domitian was jealous of Agricola’s triumphs. Perhaps he was right, but it was the ruler’s job to see the big picture, and in the 80s A.D. that picture was turning distinctly dark. Even before the battle at the Graupian Mountain, Domitian had been withdrawing contingents from Agricola’s legions to bolster defenses along the Rhine, and when the emperor pulled the best troops out of Britain in A.D. 85, it was to plug gaps in the crumbling Danube frontier. This strategic pivot worked, and the river frontiers held. But Domitian drew a radical conclusion from it: that Rome no longer had much to gain from productive war.

  Romans had been drifting toward this conclusion for nearly a century. Between 11 B.C. and A.D. 9, the emperor Augustus had methodically pursued what would—had it succeeded—have been the most productive war Rome ever fought, pushing the frontier northeast to the Elbe River to swallow up what is now the Netherlands, a slice of the Czech Republic, and almost all of Germany. But it ended in disaster: stretched out along ten miles of winding paths through dark forests, their bowstrings and armor soaked by torrential rains, the Romans were betrayed by their guides and ambushed. In the three-day running battle that followed, about twenty thousand Romans were killed, and—even more horrifying to Rome’s warrior class—three legionary standards were captured. Roman armies took revenge with a decade of rape, pillage, and killing, but in the end the disaster prompted them to rethink the empire’s grand strategy. Conquest seemed to be more trouble than it was worth. When Augustus died in A.D. 14, his will contained just one piece of strategic advice: “The empire should be kept within its boundaries.”

  Most of the men who followed him onto the throne did what he said. Claudius broke the rule by invading Britain in A.D. 43, only for Domitian to close the campaigns down in the 80s. Trajan broke it more flagrantly after 101, overrunning much of modern Romania and Iraq, but when Hadrian succeeded him in 117, abandoning many of these gains was almost his first act.

  Rome’s emperors were groping their way toward a profound strategic insight, which would be formalized seventeen centuries later as one of the basic maxims of war-making by Carl von Clausewitz, arguably the greatest of all military thinkers. “Even victory has a culminating point,” Clausewitz observed. “Beyond that point the scale turns and the reaction follows with a force that is usually much stronger than that of the original attack.” Whether Clausewitz learned this from his own career (he witnessed Napoleon’s disastrous experience with culminating points in 1812 at first hand, fighting for Russia because his native Prussia had dropped out of the war) or from his deep study of Rome’s wars remains unclear. It is perhaps no coincidence, though, that Edward Luttwak, the modern strategist who has looked hardest at the paradoxical nature of culminating points, has also written the best book on Roman grand strategy. “In the entire realm of strategy,” Luttwak notes, “a course of action cannot persist indefinitely. It will instead tend to evolve into its opposite.”

  For centuries, wars of conquest had (over the long run) been productive, creating larger empires that gradually made people safer and richer. But as ancient imperialism neared its culminating point, the back-to-front logic of war threw everything into reverse. War did not just stop being productive; it turned downright counterproductive, breaking down large societies, impoverishing people, and making their lives more dangerous.

  The first sign that the ancient empires were approaching their culminating points was the onset of diminishing returns to conquest. So long as the Romans stayed near the Mediterranean Sea, size was no great issue, because water transport was relatively cheap and fast. But in a world where armies moved at the pace of an oxcart, pushing inland—into Germany, Romania, and Iraq—drove costs upward. It cost almost as much to load a ton of grain onto carts and drag it ten miles overland as it did to ship it from Egypt to Italy, and despite the Romans’ famous roads, by the first century A.D. the gains from war—whether measured in gold or glory—rarely seemed to justify the costs.

  At the other end of Eurasia, China’s rulers were wrestling with the same calculus (Figure 3.2). Between about 130 and 100 B.C., Han armies had gone on a rampage, bringing into the empire what are now the Chinese provinces of Gansu, Fujian, Zhejiang, Yunnan, and Guangdong, as well as a great chunk of central Asia, most of Korea, and a piece of Vietnam (not to mention punitive campaigns deep into Mongolia). After 100 B.C., though, the feeling grew in the court at Chang’an that the cost in blood and treasure was just not worth it. The farther the armies got from the Yellow and Yangzi Rivers, the higher the costs rose and the lower the benefits fell. There were renewed pushes into central Asia and toward Burma in the 80s and 70s B.C., then another lull, and in the aftermath of a terrible civil war in A.D. 23–25 expansion more or less ended.

  Figure 3.2. The limits of empire in Asia: sites mentioned in this chapter, and the greatest extents of the Sasanid (around A.D. 550), Kushan (around A.D. 150), and Tang (around A.D. 700) Empires

  By the first century A.D., the Roman and Han Empires had conquered similar areas (around two million square miles each) and ruled similar populations (fifty to sixty million people each). The problems their emperors faced were similar too, and both sets of overlords reached the same conclusions. They recalled their ambitious generals, built walls along their increasingly rigid frontiers, and settled hundreds of thousands of soldiers in forts much like Vindolanda. Some sites on China’s arid northwest frontier in fact outdo Vindolanda; since the 1990s, excavators at Xuanquan, a Han military post office, have found twenty-three thousand undelivered letters, painted on bamboo strips between 111 B.C. and A.D. 107 (many of them complaints about how unreliable the mail was).

  First-century-A.D. emperors could see perfectly well that war did not pay the way it used to, but they could not see that the very success of productive war had transformed the larger environment in which it operated. To be fair to them, it is always hard to know when to stop. “If we remember how many factors contribute to the balance of forces,” Clausewitz mused, “we will understand how difficult it is in some cases to determine which side has the upper hand.” Over the next few centuries, however, it would become all too clear who had it.

  War Horse

  The ancient empires reached—and passed—their culminating points because by the first century A.D., productive war had entangled them with the horsemen of the steppes. This was a long, drawn-out process, which made it all the harder for emperors to identify what was going on. We saw in Chapter 2 that the entanglements began as early as 850 B.C., when the Assyrian Empire began buying the big new horses—strong enough to carry riders on their backs—that herders on the grasslands had succeeded in breeding. Over the centuries that followed, the empires kept expanding. Their farmers plowed up the edges of the steppes to grow grains, and their traders pushed deeper into central Asia to buy animals; and as they did so, the nomads along the ecological frontier where arid grasslands blurred into cultivated fields learned that they had new options. Often, they found, they could do better by selling horses to imperial agents than by rushing from oasis to oasis to fight other horsemen for a few mouthfuls of muddy water. Better still, they learned, when the imperialists would not pay the price they demanded, they could shoot their way into the empires and take what they wanted from the unarmed, peaceful peasants.

  We first hear about an empire having trouble with steppe nomads in Assyrian sources before 700 B.C. Assyria had e
xpanded into the Caucasus Mountains, right at the edge of the steppes (Figure 3.3). When Scythian riders began terrorizing the borderlands, Assyrian kings simply hired some nomads to fight the other nomads for them. They quickly found, though, that the skills that made Scythians attractive employees—mobility and ferocity—also made them uncontrollable. The seeds of disaster were being planted.

  Figure 3.3. Storms on the steppes: a millennium of asymmetric wars, ca. 700 B.C.–A.D. 300

  In the seventh century B.C., gangs of Scythians went into business for themselves, robbing anyone who came along and effectively taking over much of what we now call northern Iraq, Syria, and eastern Turkey. “Life was thrown into chaos by their aggression and violence,” the Greek historian Herodotus wrote, “because they rode all over, carrying off everything.” In the 610s, anti-Assyrian rebels hired Scythians of their own, and before the decade was out, the empire was in ruins. This, though, left the victorious rebels with the problem of what to do with the Scythians. They eventually solved it in the 590s (according to Herodotus, by getting the Scythian leaders drunk and murdering them).

  The more Eurasia’s empires grew, the more they found themselves facing a peculiarly modern problem: how to fight asymmetric wars around the edges of central Asia. In the late 1990s, when Osama bin Laden perpetrated his first massacres, the United States found no way to “neutralize” (the preferred term) him in his Afghan lair except by firing million-dollar cruise missiles at terrorists’ ten-dollar tents. Ancient empires, with their vast, ponderous infantry armies, found chasing bands of horsemen around the wilderness similarly difficult.

  This was a matter not of Western versus non-Western ways of war but of agrarian versus nomadic ways. From Europe to China, the rulers of wealthy empires all faced more or less the same challenges in dealing with steppe horsemen, and by the age of Agricola they had all worked through every possible permutation for waging asymmetric war. Then as now, the obvious approach was preemptive war, and Persian kings sent a series of armies onto the steppes to chase down the Scythians. But pursuing the nomads into their hiding places, the Persians learned, could be almost as problematic as doing nothing, because infantry could not force nomad cavalry to fight if they did not want to. Sometimes preemptive wars paid off quickly, as in 519 B.C., when the Persians crushed a confederation that they called the Pointy-Hatted Scythians, but often they did not. In 530 B.C., nomads had killed King Cyrus, the founder of the Persian Empire, and annihilated his army. In 514, King Darius of Persia—after chasing the Scythians around the steppes for months without being able to catch them—only avoided the same fate by scrambling back across the Danube under cover of darkness.

  Assyria and Persia were the first empires to become entangled with the steppes, but by the third century B.C. China was going the same way. In 213 B.C., the Qin First Emperor launched a preemptive war, annexing a great swath of the steppes in an effort to push the Xiongnu nomads away from his frontiers. But it brought the Middle Kingdom little joy: in 200 B.C. the Xiongnu lured a Chinese army deep into the steppes and destroyed it completely.

  In 134 B.C., the emperor Wudi tried preemptive war again and half a dozen times over the next fifteen years sent armies hundreds of thousands strong into the steppes. Few of his men returned, and the costs wiped out the budget surpluses his cautious predecessors had accumulated, driving the government deep into debt. But despite spending so much, Wudi—like Darius—never did get his decisive battle with the nomads.

  From Athens to Chang’an, intellectuals denounced preemptive war as disastrous. But in another peculiar parallel with modern experiences, in the long run it proved surprisingly difficult to tell who had won the preemptive wars, or even when they were over. The costs in blood and treasure had been terrible, but the Scythians never again threatened Persia after 513, and Xiongnu raiding had declined sharply by 100 B.C.

  The conclusion emperors eventually reached was that the hard power of expensive expeditions onto the steppes worked best when combined with softer, albeit still pricey, techniques. The most popular was containment, which usually meant building walls to keep nomads out. The most famous, the Great Wall of China, goes back to the 210s B.C.; the wall that Hadrian built in the 120s A.D., mentioned at the beginning of this chapter, was its distant cousin. Walls could not keep nomads out altogether, but they did at least channel where the horsemen came in.

  The most successful (or, perhaps, least unsuccessful) strategy was bribery. Nomad raids killed a lot of people and lowered the empires’ tax take, so why not just pay nomads not to raid? As long as bribes cost less than preemptive war, protection money was a win-win-win proposition—the emperors saved cash, the peasants in the borderlands saved their lives, and the nomads saved themselves a lot of trouble. Two thousand years on, bribery retains its appeal in asymmetric warfare: by handing out $70 million in cash to Afghan warlords in 2001, the CIA also saved a lot of money, lives, and trouble.

  There is a saying in Chicago that an honest politician is one who, when you buy him, stays bought, but expectations are lower in asymmetric wars. The Afghan commander who took $10,000 in December 2001 to guard escape routes from the Tora Bora Mountains, only to let al-Qaeda fighters through when they offered him more, would have fit right in on the ancient steppes. Scythians and Xiongnu regularly took payoffs and then raided anyway. Bribery, it turned out, was the worst way to deal with nomads—except for all the other ways. Persian and Chinese strategists found that handouts worked best as part of a package of carrots and sticks. A stream of sweeteners combined with the occasional massive, violent, preemptive war could more or less keep the peace.

  Combining all these tricks, rulers in the last few centuries B.C. learned to manage their frontiers. They turned their relationships with the steppe nomads into something resembling bad marriages, in which the partners could live neither with nor without each other. When an empire was strong, it could impose a settlement on part of the steppes and keep violence within tolerable limits; when it was weak, it had to pay more and suffer more.

  The only way to keep the upper hand, every empire discovered at some point in the millennium between roughly 500 B.C. and A.D. 500, was to beat the nomads at their own game. This meant that emperors had to leaven their huge infantry armies with more and more cavalry. Historians who think there is a Western way of war rooted in ancient Greek culture often see fighting from horseback as typical of Eastern evasiveness, while fighting on foot is the hallmark of Western values. In reality, though, the great shift toward cavalry between 500 B.C. and A.D. 500 was driven by geography, not culture. Empires whose frontiers ran right up to the steppes shifted toward cavalry relatively soon after 500 B.C.; those that were shielded by mountains and forests shifted later, and less completely. But, willingly or not, all the empires in the Old World’s lucky latitudes moved in the same direction.

  Not surprisingly, the shift began in Persia, the empire most exposed to nomadic raids. When Darius chased the Scythians around Ukraine in 514 B.C., almost all his men were walking, but by 479, when the Persians fought the Greeks at Plataea, they relied almost as much on cavalry as on infantry. And in 334, when Alexander the Great invaded Persia, the empire looked almost entirely to riders for victory. China, the next-most-exposed empire, was also the next to move down the riding trail. Emperor Wudi raised huge mounted forces before launching his preemptive wars. In 110 B.C. he had 180,000 horsemen on his payroll, making up one-third of the army and costing twice as much to feed each year as the entire empire paid in taxes. India, largely shielded from the steppes by the Himalayas and the Hindu Kush, was less exposed, and between the fifth and the second centuries B.C. its kings felt safe sticking to what they knew. Head-on clashes of armored elephants still won battles, with horsemen doing little more than covering the elephants’ flanks—until another oddly modern development shifted the ground under their feet.

  In 1954, faced with mounting demands to do something about the spread of communism in Southeast Asia, President Dwight Eisenhower warn
ed America about “what you could call the ‘falling domino’ principle. You have a row of dominoes set up,” he explained, “you knock over the first one, and what will happen to the last one is the certainty that it will go over very quickly. So you could have a beginning of disintegration that would have the most profound influences.”

  Whatever its strengths or weaknesses as an analysis of 1950s Indochina, this is an excellent description of the steppes in the first century B.C. As Han China’s huge cavalry armies began getting the better of the Xiongnu, many of the nomads migrated westward into lands where the Yuezhi peoples had traditionally grazed their flocks. The terrified Yuezhi then moved still farther west, which took them into Scythian territory. As the next domino fell, the Scythians (called Shakas in India) moved south through what is now Afghanistan, crossed the Khyber Pass, and descended into the Indus Valley. By 50 B.C. the Shakas had overrun much of northwest India.

  A century later—after more half-forgotten cavalry wars on the steppes—the Yuezhi followed the Shakas over the Hindu Kush. Pushing the Shakas deeper into India, the Yuezhi conquered a huge domain stretching from modern Turkmenistan to the middle Ganges, known to historians as the Kushan Empire. The Kushans prospered mightily, becoming one of the great cavalry powers of the day. By the second century A.D., their fearsome mounted archers, commemorated in countless sculptures in what are now Afghanistan, Pakistan, and northern India, controlled the Silk Roads linking Rome and China. The Kushans even fought their own preemptive wars, including one against a Han expedition to Afghanistan.