A History of Warfare

by John Keegan

  It would be odd if Finley were wrong. Court art may triumphalise; it may also perpetuate as symbolic the purely antique; it must not, of its nature, ridicule. Thus it was just possible, at a time when the idea and trappings of chivalry had come back into fashion, to depict the Prince Consort in armour without provoking Victorians to laughter; the representation of Hitler, mounted and armoured, was an absurdity.30 Pharaohs, Assyrian kings and Persian emperors clearly did not think it absurd that they should be shown shooting the composite bow from a chariot. Their court artists may have exaggerated their masters’ prominence in the battle line; but if it was as chariot archers that these great men wished to have themselves depicted, we must therefore infer that chariot archery was a dominant means of winning battle over a considerable period of time, from the first appearance of the chariot about 1700 BC to its supersession by horsed cavalry about a thousand years later.

  It has already been suggested that the initial advantage of the charioteer lay in the suddenly and very greatly increased speed of movement he enjoyed on the battlefield, on the long-range lethality of his composite bow and on a cultural readiness to kill. All these advantages would have eroded with time. Familiarity with a novel weapons system ought not to breed contempt, but it does stimulate counter-measures. Those attacked by charioteers acquired chariots; non-charioteers learnt to target the horses of the enemy’s chariots, to form chariot-proof ranks, to carry arrow-proof shields, to make use of broken ground on which charioteers could not manoeuvre. Nevertheless, while the great men of opposed armies reckoned charioteering glamorous, there must have been a complicity between friend and foe to see that battles were so fought that chariots got their chance. Ritualism or ceremonialism are, as we have seen, deep-rooted in man’s conception of how combat should be conducted and are suppressed only by the necessities of battle to the death — not something that war is always about.

  The first chariot battle of which we have an account, the battle of Megiddo in northern Palestine, fought in 1469 BC between the pharaoh Tuthmosis III and a confederation of Egypt’s enemies under Hyskos leadership, was concluded with almost no bloodshed on either side. Megiddo is also generally counted as the first battle of history, in that we can date it, locate its site, identify the contestants and follow its course. Tuthmosis, who had just come to the throne, was pursuing the new Egyptian strategy of vigorous offensive against the outsiders who had violated the immunity of the river kingdom. Collecting an army, he marched in stages of ten or fifteen miles a day — an impressive speed of advance — along the Mediterranean coast, through Gaza and then up to the mountains on the Syrian border. The enemy seems to have counted on the difficult terrain forming a barrier against his attack. There were three routes through the mountains to the town of Megiddo; the pharaoh chose the most difficult, against advice, on the ground that he might thereby surprise them. The approach march took three days, the last spent negotiating a pass of less than two chariots’ width. Late in the evening he camped on the plain in front of Megiddo and next morning deployed his army for battle. The enemy had also come forward, but, when they saw the extent of the Egyptian line, with one wing on each flank of the valley and the pharaoh commanding from his chariot in the centre, their morale collapsed and they fled in panic to the protection of the walls of Megiddo in their rear. Tuthmosis ordered a pursuit, but his soldiers stopped to plunder the enemy’s abandoned camp on the way and two of the principals in the opposing army managed to get inside Megiddo. Since the city had an ample water supply within its massive walls, it managed to hold out against the Egyptians — who constructed a line of circumvallation around it against any relief operation — for seven months. Only eighty-three of the enemy had been killed in the battle and 340 taken prisoner; the fugitives, however, did not rally and the besieged kings eventually surrendered, sending their children out as hostages and begging the pharaoh that ‘the Breath of Life be given to their nostrils’.31

  The most valuable booty of the victory came in the form of horses, of which the Egyptians captured 2041; as they were perhaps still importers of bloodstock, such numbers would have been an important addition to their chariot arm. We have no indication of how many chariots were committed by either side at Megiddo. When, however, 200 years later in 1294 BC Rameses II defeated a Hittite army at Qadesh on the River Orontes in southern Syria — sustaining the New Kingdom’s policy of aggressive warmaking at the far limit of strategic outreach from the Nile delta — the Egyptian army appears to have had fifty chariots and 5000 soldiers. It is said that the much larger Hittite army had 2500 chariots, which must be an exaggeration — its front of attack would have been 8000 yards wide — but an Egyptian bas-relief of the battle depicting fifty-two chariots indicates that the numbers committed were very considerable.32

  There is some doubt whether the Hittites used the composite bow. Their chariot crews are usually represented as spearmen, which may explain why the Egyptians were able at Qadesh to extricate themselves from potential defeat. In any case, both at Megiddo and Qadesh chariot fighting had not yet reached the developed form it did at the height of Assyria’s imperial power in the eighth century BC. Weapons systems take long periods to be assimilated, the more complex the longer, and the chariot system, which comprised not only the chariot itself, but also the composite bow, the horse, and all its trappings — all foreign to the lands where the chariot kings ruled — was a very complex system indeed. It is not surprising if both the Egyptians and the Hittites were as yet clumsy charioteers and that the system had to wait until comparatively late in the development of Assyrian battlecraft before it achieved its full potential. By then it may well have become, as the scribes of Sargon and Sennacherib describe, a weapon of shock and terror, manipulated by the driver to charge at breakneck speed behind a team of perfectly schooled horses and used by the archer as a platform from which to launch a hail of arrows; squadrons of chariots, their drivers trained to act in mutual support, might have clashed much as armoured vehicles have done in our time, success going to the side that could disable the larger opposing number, while the footmen unlucky or foolhardy enough to stand in their way would have been scattered like chaff.

  THE WARHORSE

  At the apogee of its effectiveness, the chariot was overtaken in importance by a single element in the chariot system, the horse itself. It has been suggested that the Assyrians themselves were responsible for this ironic revolution, which brought the downfall of their empire.

  Horses had been ridden in the civilised world since the second millennium. Riding is represented in Egyptian art as early as 1350 BC and reliefs from the twelfth century show mounted soldiers, one of whom is taking part in the battle of Qadesh.33 None, however, is a cavalryman. All ride bareback, without stirrups, and straddle the horse toward its rump, not a control position. That indicates, indeed, that the horses were not yet strong enough in the back to be ridden in the modern style. By the eighth century BC, however, selective breeding had produced a horse that Assyrians could ride from the forward seat, with their weight over the shoulders, and a sufficient mutuality had developed between steed and rider for the man to use a bow while in motion. Mutuality, or perhaps horsemanship, was not so far advanced, all the same, that riders were ready to release the reins: an Assyrian bas-relief shows cavalrymen working in pairs, one shooting his composite bow, the other holding the reins of both horses. This, as William McNeill observes, is really charioteering without the chariot.34

  Out on the steppe, however, man may have been riding even earlier than in the civilised lands, and it is possible that the use of the bow from horseback bled back from the Assyrians across the steppe frontier and was taken up by peoples who were better advanced in horsemanship. We know that as late as the reign of Sargon II the supply of horses still ran from the steppe, where unbroken foals were caught yearly for training and then sale, to Assyria; it is not improbable that the skills of mounted archery went in the opposite direction.35

  At any rate, the fall of the Assyr
ian empire was due to the irruption, at the end of the seventh century BC, of a horse people known to us as the Scythians, an Iranian race whose place of origin may have been as far away as the Altai mountains in eastern Central Asia. They appear to have ridden on the heels of another Iranian horse people called the Cimmerians, who raided into Asia Minor about 690 BC, leaving a shaken world behind them; the Assyrians themselves, at the moment of the Scythians’ appearance, were hard pressed at the borders of their empire — to the north in Palestine, to the south by the allegedly vassal state of Babylon and to the east by the Medes of Iran. All these pressures might have been resisted, for Assyria had recovered from troubles before. In 612 BC, however, the Scythians joined the Medes and Babylonians in a siege of the great city of Nineveh, which they succeeded in taking. Two years later, despite help from Egypt, the last Assyrian king was again defeated by an alliance of Scythians and Babylonians at Harran and in 605 the power of Assyria passed to Babylon.

  Babylonia’s power shortly passed to Persia, the last of the great empires to arise in the heartland of civilisation, but Persian power was not rooted in any advanced military technique. It ultimately rested on the chariot: despite the recruitment of mercenary infantrymen and the training of Persian noblemen to fight as cavalry, it was as charioteers that the Persian emperors chose to go to battle and when the Persian emperor Darius met an enemy with revolutionary military means at his disposal, he was overcome. His empire passed to Alexander’s successors, and a fragile form of the Alexandrian military system defended it for a century and more after his death. Along the 1500 miles of borderland that separates the steppe from settled land between the Himalayas and the Caucasus, however, neither charioteering nor Alexander’s European tactics were appropriate once the horse peoples had learned that civilisation was vulnerable to their attack. Thus the first Scythians who made their raid into Mesopotamia at the end of the seventh century BC were harbingers of what was to be a repetitive cycle of raiding, despoliation, slave-taking, killing and, sometimes, conquest that was to afflict the outer edge of civilisation — in the Middle East, in India, in China and in Europe — for 2000 years. These persistent attacks on the outer edge of civilisation of course had profoundly transforming effects on its inner nature, to such an extent that we may regard the steppe nomads as one of the most significant — and baleful — forces in military history. The innocent agents of the harm they were to do were the descendants of the little, rough-coated ponies which man had been breeding and eating on the Volga only a few dozen generations before the Scythians made their first ominous appearance.

  THE HORSE PEOPLES OF THE STEPPE

  What is the steppe? To those who live in settled and temperate lands, the steppe means the enormous expanse of empty space that fills the map between the Arctic Ocean to the north and the Himalayas to the south, between the irrigated river valleys of China to the east and the barriers of the Pripet marshes and Carpathian mountains in the west. On the civilised man’s mental map, it appears as not only featureless but climatically undifferentiated, a zone of sparse and uniform vegetation, without mountains, rivers, lakes or forests, a sort of waterless ocean without known voyagers.

  This impression is quite incorrect. In modern times its western reaches have been settled by Russian and Ukrainian city-dwellers in the millions; yet even before they began to populate the banks of the western steppe’s great rivers — Volga, Don, Donetz, Dnieper — travellers who ventured into the wilderness recognised that climate and topography marked it out into several distinct regions. Geographers generally denote three: the taiga or sub-Arctic forest, which runs from the northern Pacific to the Atlantic’s North Cape; a wide band of desert, touching the Great Wall in the east and the salt marshes of Iran in the west; in between the two, the steppe proper.

  The taiga is forbidding territory; the climate is extreme — near Yakutsk the ground has been found to be permanently frozen to a depth of 446 feet — and the fishermen and hunters who survive on the banks of the rivers that drain the plateau to the Arctic Ocean — Ob, Yenisei, Lena and Amur — are shy forest-dwellers; among them only the Tungu, who live in eastern Siberia and the basin of the Amur River, are known to history, principally as the Manchu who captured the Chinese throne in the seventeenth century AD.

  In the desert belt,

  none of the rivers reaches the sea; they lose themselves in the sand or flow into salt marshes. The Gobi Desert is a cheerless solitude of sand, rock or gravel that extends 1,200 miles, inhabited, in popular belief, only by demons, whose thundering wailings are more plausibly attributed to the noise of shifting dunes, dislodged by strong winds.

  Vegetation is confined to scrub and grassy reeds; the climate is extreme; icy sandstorms blow furiously in winter and spring; rain falls but seldom, though after a brief shower the desert floor suddenly blooms with small green plants. The Takla-makan is a smaller Gobi, so swept in summer by choking dust storms that travel across it is tolerable only in winter. The Dasht i-Kavir, or Persian desert, 800 miles wide, consists less of sand than of salt swamps, but [is] dotted with oases.

  These oases were the nodal points, in William McNeill’s theory, by which the Indo-European charioteers made their way to China.

  The true steppe is an elongated belt of grassland, 3000 miles long and averaging 500 miles in depth, bounded to the north by the sub-Arctic and to the south by desert and mountains; it gives at its eastern end on to the river valleys of China and at its western on to the approach routes to the fertile lands of the Middle East and Europe. It forms

  a treeless pasture, a grassy plain between the mountains, unsuited to agriculture unless expensively irrigated, but admirably suited to the breeding of cattle, sheep and goats, the sub-alpine valleys of the Altai providing exceptionally fine grazing-grounds. The vegetation consists principally of rich grass; the surface of the soil varies from gravel to salt and loam; the climate, though severe, and in the high steppe appallingly cold in winter [below freezing in the Altai mountains for 200 days in the year], is dry and therefore bearable, and the shepherds of these regions often live to extreme old age.36

  Geographers distinguish between a high and low steppe, respectively east and west of the Pamirs, which outcrop from the Himalayas. The ‘gradient’ therefore runs westward, and the grazing improves in that direction, thus encouraging migration towards Europe and the Middle East. Historically, however, there has been much movement in the opposite direction; the Dzungarian gap to the south of the Altai mountains, the steppe heartland, offers a natural gateway toward the Chinese plains. This is easier of passage than the western gateways — at each end of the Caucasus mountains, in the gap between the Caspian and Aral seas, and around the top of the Black Sea into the Adrianople corridor — which are narrower and easier to defend.

  The Scythians, the first steppe people of whom we know, probably originated in the Altai and followed the steppe gradient westward to attack Assyria. Of those who came later, the Turks seem certainly to have come from the Altai, and their tongue (to which those of the Kazaks, Uzbeks, Uighurs and Kirghiz among others are allied) was, and remains, the principal speech of Central Asia; the Huns who appeared outside Rome in the fifth century AD spoke a language that belonged to the Turkic group. Mongolian, by contrast, spoken by comparatively few steppe peoples, apparently originated in the forest lands north of Lake Baikal and east of the Altai; Manchu, also Tungusic, derives from eastern Siberia. Some of the first horse peoples were, however, like the original charioteers, Indo-Europeans, and spoke what has become Persian; allied languages, now forgotten but spoken by warriors in their time, were Sogdian and Tocharian, and another was that of the people known to the Romans as Sarmatians.37

  What drew the horse nomads out of the steppe? We cannot easily fit their warmaking behaviour into any of the patterns discerned in other societies by social anthropologists. They were certainly not ‘primitive warriors’: from the start they fought to win, so that explanations couched in terms of kinship quarrels or ceremonialis
m do not fit them. Territoriality also seems an inappropriate concept; though nomadic tribes undoubtedly were attached to certain grazing-lands and conceded the claims of others to theirs, it was also a salient characteristic of nomadism that tribal composition was fluid; chiefship was precarious and followings split or coalesced unpredictably. Perhaps the most useful idea is the ecological one of ‘carrying capacity’. William McNeill has persuasively argued that life on the steppe was subject to sudden and highly disruptive climatic change: warm, moist seasons, making for good grazing and a higher survival rate of animal — and human — offspring, were commonly followed by harsh times, which left larger flocks and families stranded for sustenance. Migration within the steppe did not help, since neighbours suffered similarly and resisted incursions. The obvious means of escape was therefore outward, into gentler climes where cultivated land offered emergency rations.38

  The perceptible flaw in this explanation — perceived and allowed for by McNeill himself — is that nomads would in time have learnt to anticipate the intermission of bad with good times and made their homes elsewhere than on the steppe, which ought in consequence, after their acquisition of the riding-horse, to have reverted to emptiness. In some sense it did so: the widest-ranging aggressors among the steppe peoples — the Mongols and Turks — did establish tribute-yielding empires over settled peoples that liberated them from the cycle of famine on the sea of grass. Nevertheless, the nomads had a weakness: they liked the nomadic way of life and despised the weary cultivator, bound to his furrows and his plough-ox. What the nomads wanted was the best of both worlds: the comfort and luxury that settled ways yielded but also the freedom of the horseman’s life, of the tented camp, of the hunt and of the seasonal shift of quarters.