War: What is it good for? (22 page)

Rome fared just as badly. With population, agriculture, and trade in free fall, cash-strapped emperors stinted on soldiers' pay or debased the coinage to make their limited stock of silver go further. The result, predictably, was that worthless coinage set off vicious inflation, depressing the economy even more.

Angry soldiers took matters into their own hands. In
A.D.
193 and again in 218 the imperial guard sold the throne to the highest bidder, and between 218 and 222 the empire was ruled—if that is the right word—by the crazed teenager Elagabalus, who stood out even among Roman emperors
for his corruption, cruelty, and incompetence. Between 235 and 284, Rome had, depending on how you count, as many as forty-three emperors. Most were military men, and all died violently except one, who was carried off by the plague. Of the other forty-two emperors, Gothic invaders killed one in battle, and the Sasanid Persians captured another, whom they threw in a cage, mocked, and tortured until they got bored and murdered him. The remaining forty were all killed by fellow Romans.

Forced to face multiple military threats, emperors had no choice but to entrust large armies to subordinate generals, even though these subordinates repeatedly repaid their rulers' confidence by launching coups (this even though hardly anyone survived promotion to the purple by more than a few months). Once a general rebelled, his army would normally abandon its post on the frontier so that it could wage civil war, leaving the empire open to anyone who wanted to enter.

The Goths built ships, sailed over the Black Sea, and looted Greece. The Franks (then based in what we now call Germany) rampaged across Gaul and into Spain. Other Germans raided Italy, while Moors overran North Africa and Sasanid Persians burned Syria's prosperous cities. Realizing that the central government could not or would not protect them, the eastern and western provinces formed their own governments, and in
A.D.
260 the Roman Empire—like the Han—split into three smaller states.

The bloody breakdown of great empires was becoming the norm. India's Kushan Empire, defeated by Sasanid Persian armies and Scythian raiders, split in two in the 230s. The western kingdom was absorbed by Persia after a final defeat in 248, and in the 270s the eastern kingdom shrank to a rump after losing control of the Ganges cities. Farther south, the great second-century trading kingdom of Satavahana also struggled to handle the Scythians, and in 236 it too collapsed.

Mancur Olson, the economist from whom I borrowed the term “stationary bandit” in
Chapter 1
, liked to draw a contrast between these relatively benign thieves and completely malign “roving bandits.” Whereas stationary bandits came, saw, conquered, and administered, roving bandits came, saw, stole, and rode off again. The empires of the first millennium
B.C.
flourished largely because their stationary bandits were usually strong enough to keep roving bandits out, but by the third century
A.D.
this was no longer the case. Almost everywhere in Eurasia, war turned counterproductive, tearing the huge, peaceful, and prosperous ancient empires apart.

Almost everywhere—but not quite everywhere. The big exception to
the rule of third-century imperial collapse was Persia, where, after overthrowing the Parthians in
A.D.
224, the new Sasanid dynasty went from strength to strength. It smashed Kushan and Roman armies, rolled back the steppe nomads, and centralized power. By 270, when the great conqueror Shapur I died, the Sasanid capital at Ctesiphon was one of the world's grandest cities.

But on closer inspection, the Sasanid exception turns out not to have been an exception at all, because the rule in these years was not simply one of imperial collapse. Rather, the twelve hundred years between roughly
A.D.
200 and 1400 were an age of cycles of productive and counterproductive wars. As we saw in
Chapters 1
and
2
, the millennia leading up to
A.D.
200 were an era of expanding Leviathans, rising prosperity, and falling rates of violent death, and as we shall see in
Chapters 4
–
7
, this is even truer of the centuries since 1400. But the long Middle Ages separating these two periods constituted a complicated, messy, and violent interlude.

It is a tangled story. For a while in the late third century it looked as if the Sasanid revival were actually the first example of a new trend toward imperial recovery. After half a century of anarchy, Rome had regained control over the whole Mediterranean Basin by 274, the Western Jin dynasty had reunited China into a single empire by 280, and in the 320s the Gupta dynasty had begun doing the same in India. By then, though, the recovery was already ending in other parts of Eurasia. Xiongnu nomads burned China's ancient cities, executed a string of Western Jin emperors, and massacred millions of refugees. Sixty years of fighting followed, until in 383 it looked as if a new dynasty were about to unite China once again; but its army mysteriously dissolved in panic after a minor defeat, and another cycle of slaughter engulfed East Asia.

Rome too slid back toward chaos in the late fourth century. Goths destroyed the empire's field armies at Adrianople in 378 and the frontiers began dissolving. Westward migrations of Huns (the most terrifying of all the ancient nomads) toppled more dominoes, and on New Year's Eve 406, thousands of Germans flooded across the frozen Rhine River. Western Europe spiraled down into violence and chaos, and in 476—just seventy years after the Rhine frontier failed—a Germanic king announced that the western half of the Roman Empire had ceased to exist.

In 484, it looked as if Sasanid Persia would go the same way when another branch of the Huns wiped out its army and killed its king. But the Sasanids hung on, and by this time China too was moving back toward unity. In the fifth century another new dynasty reunited the Yellow River
region, and in 589 the Sui dynasty finally brought the whole of China back under a single government.

For a few giddy years, the Mediterranean also seemed to be swinging back toward unity. In the 520s, Justinian, ruler of the Byzantine Empire—as the surviving (eastern) portion of the old Roman Empire is often called—won back Italy and parts of Spain and North Africa. By 550, though, expansion had stalled, and in the later sixth century fresh invasions rolled the Byzantines back. India had an equally rough ride: after 467 the Gupta Empire started disintegrating in the face of attacks from another branch of the Huns, and despite a great victory over the nomads in 528, by 550 the empire was to all intents and purposes history. And on it went, century after chaotic century, all across Eurasia's lucky latitudes.

I have not tried to tidy up the fact that this is a confusing narrative, and I think that
Figure 3.6
sums up its messiness nicely. The graph divides the lucky latitudes into four regions (Europe, the Middle East, China, and India) and charts the geographical size of the biggest empire in each across the first fourteen centuries
A.D.
Admittedly, there are all kinds of technical problems in simply using size as a measure of Leviathanness (by which I mean the strength of centralized government). The most obvious is the great spike in the Middle Eastern curve between
A.D.
650 and 850, representing the Umayyad and Abbasid caliphates established by the Arabs. In
theory, the caliphs ruling from Damascus and Baghdad controlled 4.3 million square miles, one of the biggest empires in history, but in practice hardly anyone outside Syria and Iraq took much notice of them. The Indian spike around
A.D.
150, representing the Kushan Empire, raises another problem: the Kushans ruled 2.3 million square miles, but most of these square miles had hardly anyone living in them.

But despite these (and other) issues, this tangled graph does make one big point. Between the second century and the fourteenth, there were few years when every part of the lucky latitudes was moving in the same direction. For every empire that rose, another fell. One society's golden age was another's dark age.

What does this mean? The obvious interpretation, and the one most favored by historians, is—as the brilliant polymath Arnold Toynbee put it back in the 1950s—that the past is simply “a chaos unamenable to [scientific] laws; a meaningless succession of events that a twentieth-century novelist who was also poet laureate called Odtaa, standing for ‘one damned thing after another.'” On the face of it,
Figure 3.6
looks like the poster child for Odtaa. Empires rise and fall, battles are won and lost, but nothing much changes. Everything is an exception to everything else.

Toynbee, however, conjured up the image of Odtaa only to dismiss it. After spending decades studying world history, he knew perfectly well that the story is full of big patterns that go beyond Odtaa, and I think he would have seen several such patterns in this graph. First, he might have observed, there is an obvious trend here, which
Figure 3.7
draws out. Behind all the noise, the size of empires steadily declined across the first fourteen centuries
A.D.
The lucky latitudes had become the graveyard of empires.

Second, Toynbee would surely have seen that the wild swings in the size of states are not simply Odtaa: they came in a repetitive boom-and-bust pattern. Counterproductive wars that drove empire size down were followed by productive wars driving it back up, only for counterproductive war to return and break Leviathan down again. Rather than Odtaa, the lucky latitudes were trapped in a terrible cycle.

The explanation is not hard to find. Because productive war had overshot its culminating point, the steppes and the agrarian empires had become tied together. Every action now had an equal and opposite reaction. At one moment, plagues, rebellions, and invasions would bring an empire crashing down in counterproductive war, leaving millions dead; at the next, local warlords—or perhaps an invader—would wage new productive wars, exploiting the vacuum to bring forth another Leviathan. Its king,
enthroned in pomp and circumstance, would struggle mightily to bring back the rule of law and squeeze taxes out of his subjects, only for the wealth of the new state to draw in more raiders and rebels, setting off a new downward spiral of counterproductive war … and so on.

Each region in the lucky latitudes flipped back and forth between productive and counterproductive war on its own schedule, mostly because one kingdom's success in driving off raiders tended to increase the pressure on neighboring kingdoms. Some migrations off the steppes were so overwhelming that they seemed to hit everywhere at once, as when Huns plundered all the way from India to Italy in the fifth century and Mongols attacked from Japan to Germany in the thirteenth, but even then the accidents of battlefield victory and defeat randomized the results, producing the apparently chaotic outcomes we see in
Figure 3.6
.

There had been counterproductive wars before, but even the worst of these had been breakdowns within a larger pattern of productive war. Some of the collapses had lasted for centuries, but despite the fall of the Akkadian Empire and the Egyptian Old Kingdom around 2200
B.C.
, the Indus Valley cities around 1900
B.C.
, and the kingdoms of the eastern Mediterranean's international age around 1200
B.C.
, Eurasia's lucky latitudes had kept moving toward Rome, Chang'an, and Pataliputra. For every step backward, there were two or three steps forward.