A History of Warfare

by John Keegan

  The persistence of the nomadic ethos is nowhere better caught than in the Topkapi at Istanbul, palace of the Ottoman Turkish sultans, where, until the beginning of the nineteenth century, the rulers of an empire that stretched from the Danube River to the Indian Ocean spent their days as they might have done on the steppe, seated on cushions on carpeted floors of makeshift pavilions set up in the palace gardens, dressed in the horseman’s kaftan and loose trousers, and having as their principal regalia the mounted warrior’s quivers, bow cases and archer’s thumb rings. Planted though it was in the capital city of the eastern Roman empire, the Topkapi remained a nomadic camp, where the horsetail standards of battle were processed before great men, and stables stood at the door.

  Another explanation has been offered for the nomads’ warmaking: that it was a means of forcing the civilised lands to trade. The steppe peoples certainly learnt to trade at an early time and in their horses, and probably also slaves, had commodities that professional merchants were eager to buy or exchange against manufactured goods; one of the conditions for peace that the Huns requested from the Romans in the middle of the fifth century AD was that a market on the Danube be reopened ‘as in former times’.39 The success of commercial interests at both ends of the Silk Road that linked China to the Middle East, first opened in the second century BC, in maintaining traffic along it for over a thousand years also suggests that the nomads generally perceived the advantage in encouraging rather than plundering a flow of goods through their lands. Nevertheless, it was frequently interrupted, when local greed got the better of commercial sense; and, moreover, forced trade does not work when there is a structural imbalance between what is sought and what can be offered in return. The steppe simply did not produce enough of what civilisation wanted for transactions initiated by military means to become self-sustaining by normal commercial incentives. As the British found when they sought to foist unwanted opium on China in the nineteenth century, a demand to sell, backed up by force of arms, inevitably leads the seller into imposing his political will on the unwilling buyer, and so becoming an imperialist in substance if not name. Such a sophisticated two-step was, in any case, probably beyond the early horse peoples.

  THE HUNS

  The first steppe people of whom we have any detailed knowledge were the Huns, who invaded the Roman empire in the fifth century AD. If they can be identified with the Hsiung-nu, they had seriously destabilised the unified China of the Han dynasty in the second century BC. The Huns, who probably spoke a Turkic language, were without writing; their religion was ‘a simple nature worship’. They may have used shamans — spirit callers who were believed to mediate between god and man whom we also know among the northern forest peoples who migrated to North America — and are certainly known to have practised scapulimancy, the telling of omens from patterns on sheeps’ shoulder-blades. Foretelling the future was important to the Huns; it was apparently for Hunnish mercenaries in his service that Litorius took the omens before the battle of Toulouse in 439, the last Roman general known to have performed the ancient pagan rites.40 The Huns’ social system was simple: they recognised the aristocratic principle — Attila prided himself on being well born — and they kept a limited number of slaves but accepted no other divisions.

  They sold slaves, of course, and in very large numbers after they had made a conquest: their inhumanity in breaking up families for the market appalled Christian writers in the fifth century.41 Slave-selling certainly exceeded in profitability the trade in both horses and furs as soon as the Huns established themselves in the outer provinces of the Roman empire, but they also derived an enormous income in gold by way of ransom for military and civilian captives, as well as in straight bribes from the later emperors: during the period AD 440–50 the eastern provinces paid them 13,000 pounds of gold, about six tons, to buy peace.42 It is these sort of transactions that cast doubt on the interpretation of horse peoples’ excursions from the steppe in terms of ‘escaping climatic change’ or ‘enforcing trade’. The truth seems much simpler: nomads — physically tough, logistically mobile, culturally accustomed to shedding blood, ethically untroubled by religious prohibitions against taking the lives or limiting the freedom of those outside the tribe — learnt that war paid.

  Whether the conquests that successful warmaking brought could be sustained was another matter. Nature seems to impose limits on the depth of penetration that nomads can make into settled land. Nomadic demand on irrigated land for grazing rapidly disrupts the system and returns it to a state where it will support neither beast nor man; if cleared from forest, the land reverts to woodland when the ploughing population is dispersed. (The trend became disastrous in Mesopotamia after the arrival of the Turks in the thirteenth century.43) Nomadic expansion could be consolidated, therefore, only in the borderlands between steppe and agriculture, but such lands support only small populations. In the Far East, where the conquering nomads were halfway to being Chinese already, they were easily assimilated, even if as a ruling class. In the west, where religion and civilised custom imposed a much sharper differentiation between them and the agriculturalists, the borderlands became a permanent battleground, where use of the soil had to be sustained by force of arms.

  To Attila’s Huns, the ploughed fields of Gaul and the gardened floodplain of the Po must have presented bewildering environments. Food they would have found in abundance, but not of the accustomed sort, and not in varieties that regenerated after foraging. Grass does not replace wheat or beans in a single season. Attila is said to have brought his following of families in waggons, but he cannot have brought his sheep or any large number of his horses; his traditional economic base must have been left behind, perhaps as far away as the lower Danube valley. It may have been the pull of his flocks and herds that explains his mysterious departure from Italy in 452, when the peninsula lay undefended before him. A return to the grasslands in such circumstances would have made logistical sense. It was not his retreat which shook the Roman empire, however, but his advance, and before that the thrust of the Huns into eastern Europe, which provoked a massed assault by the Germanic tribes on the Danubian frontier. The sequence of the Hunnish offensive from the steppe gives us a clear example of how disruptive a campaign by horse people could be once they took to the warpath.

  If the Huns were the Hsiung-nu who menaced China in the second century AD (the identification rests on a single piece of Scythian evidence), nothing was heard of them between the first century BC and AD 371, when they defeated the Alans, an Iranian people, at the battle of the Tanais River, between the Volga and the Don; many Alans joined the Huns, others reached the Roman borders and became mercenary cavalrymen.44 In 376 the Huns advanced from the Volga to invade the Gothic lands between the Dnieper and the Roman frontier on the Danube. The Goths were the most aggressive of the German tribes who had been pressing against the border of the Roman empire for at least a century. Their western (Visigoth) branch was established on territory which had been Roman between 106 and 275 — the province of Dacia (modern Hungary) — and, in this time of troubles for the empire, its leaders had been treating on equal terms with the emperors. The advancing Huns, driving the eastern (Ostrogoth) Goths in front of them, turned the Visigoths into supplicants overnight. Reluctantly — there were too many barbarians inside the empire already — the Romans gave them permission to cross the Danube, and their cousins followed in their wake. Local officials badly mistreated them, however, and though they had given up their arms as a condition of entry, they conjured up other weapons and stood to fight at the Willows, near the Danube delta. The Romans might easily have overcome them but, panicked by a rumour, true or false, that the Goths had made an alliance with the Huns, who were camped across the Danube, they retreated into the Balkan mountains.

  Trouble, perhaps fomented by the Goths, now flared up along the whole of the Roman frontier with Germany and while the youthful emperor Gratian tried to contain the Alemanni on the Rhine, Valens, emperor in the east, collected th
e best army he could and moved to restrain the Goths who were pillaging eastern Greece. On 9 August 378, he came up against their fortified camp outside Adrianople, was wounded in the course of a chaotic battle and died in the massacre that followed it. The death of an emperor in battle, all too soon after that of Julian in a Persian war (363), was a grave blow to Rome. The irretrievable consequence of Adrianople, however, was neither the moral nor material damage it caused but the enforced barbarisation of the Roman army, imposed upon the new eastern emperor, Theodosius, by the Visigoths as condition of their good behaviour. In return for being allowed to settle south of the Danube (382), inside the empire and with their weapons, the Visigoths agreed not only to keep the peace but to fight for the emperor as ‘federate’ allies.

  ‘The settlement was … a grave breach with precedent.’45 The Romans, as the Assyrians had done before them, had traditionally incorporated barbarian contingents into the army, but as specialists and in small numbers. As the pressures on the empire had increased, so had the numbers — there may have been 20,000 ‘Roman’ Goths at Adrianople, and some mercenary Huns were serving in the cavalry, along with representatives of other horse peoples — but thitherto the Romans had always retained control of the leadership, either by appointing imperial officials as generals or by elevating barbarians to the still greatly coveted — and well paid — higher ranks of the Roman army. Theodosius’s settlement changed that: thereafter barbarian armies operated autonomously within the empire and, in circumstances where a continuous press of barbarian numbers from outside provoked successive crises of leadership within, the barbarian chiefs threw their weight this way and that between competitors for the imperial title, with catastrophic economic and military results.

  Thus, though Theodosius succeeded in reuniting the empire under a single throne, he allowed more Goths to enter the empire in the course of his pacification campaigns, and this Visigothic contingent under Alaric, after the death of Theodosius in 395, caused irreparable damage to what remained of the imperial structure in the west. In 401, from a base in Greece, Alaric invaded Italy across the Alps, opening a campaign of despoliation which Stilicho, the last but one great Roman general, took three years to bring under control. At its end Stilicho’s army was so depleted in numbers that he lacked the strength to move against the next major threat. During 405 the largest barbarian host yet seen, a collection of Germanic peoples including Vandals, Burgundians, Swabians and Goths under the leadership of Radagaisus, crossed the Danube and then the Alps to winter in the Po valley. They had apparently been displaced from northern Germany by Huns who were pushing upward from the homeland they had established in Dacia — which forms the last extension of the steppe grassland on the edge of forest Europe. Stilicho was eventually able to confine Radagaisus’s horde into an area near Florence, starve it into surrender and drive the remnants back across the Alps into southern Germany. Thence, in the next few years, the separate tribes made their way across the Rhine to initiate the culminating barbarisation of Gaul.

  Roman loss of control over the remaining western provinces proceeded apace, with Alaric playing a malign role. In 410 he captured and sacked Rome and then marched south to cross to Roman Africa but died before he could find ships. Meanwhile the eastern empire also came under threat from the Huns who had briefly invaded Greece in 409. Fortunately some Huns proved willing to change sides for the right inducement and these mercenaries provided Aëtius, ‘the last of the Romans’, with much of the force with which he sustained imperial authority in the second quarter of the fifth century.46 From 424 onward, campaigning largely in Gaul, he succeeded in holding the Teutonic invaders at bay, even while Spain and Roman Africa were crumbling under the attack of the Vandals. Between 433 and 450 Aëtius’s war in Gaul was almost continuous.

  In 450 he was confronted by a new challenge. The Huns of Hungary had been acting as an independent power on the flank of the eastern Roman empire for twenty years, taking tribute from the emperor but also raiding his territory and cooperating with the Teutonic rulers to their mutual advantage. In 441 they had raided into Greece again, under the leadership of their king’s nephew, Attila, who in 447 appeared under the walls of Constantinople. In 450 he transferred his effort to Gaul and in 451 was laying siege to Orléans. Siegecraft was not an art that the Huns had yet mastered, or that any horse people before the Mongols were to do, and while Attila was engaged under the city’s walls, Aëtius, by frantic diplomacy, assembled an army of Franks, Visigoths, Burgundians and Alans and drew him into battle on the open plains of the Champagne between Troyes and Châlons.

  Châlons, fought in June 451, has been called one of ‘the decisive battles of history’. There were Teutons and horse people on both sides and it was Aëtius’s Alans who succeeded in holding Attila’s Huns in a pitched struggle. When Attila detected that Aëtius had profited from this check to make an encircling move toward his rear, he took refuge in his waggon laager and, under cover of Hun archery, succeeded in disengaging and beating a retreat to the Rhine. In the following year he moved from the Rhine to Italy, his appearance driving people from the plain of the Po to take refuge on the islands which would become Venice and, also in popular supposition, causing Pope Leo I to visit his camp and dissuade him from attacking Rome. In the event Attila did not march further south but, after agreeing to ransom his more important captives, turned about and retreated. Within two years ‘the scourge of God’ was dead and the Hunnish empire collapsed.

  There were circumstantial reasons for Attila’s decision to leave Italy. It had just suffered famine, while pestilence had broken out in his army and an eastern Roman force had crossed the Danube to campaign in Hungary. Such circumstances do not, however, explain why the Hunnish empire failed to survive Attila’s death or why, on the death of his sons, the Huns disappeared from history. One suggestion is that, during their sojourn on the borders of the Roman empire, they had abandoned their steppe habits, adopted Teutonic methods of fighting and so become absorbed.47 This is dismissed by Maenchen-Helfen, the most meticulous collater of Hunnish data: ‘Attila’s horsemen were still the same mounted archers who in the 380s had ridden down the Vardar valley into Greece.’ Another explanation is that the Hungarian plain is not large enough to support horse herds of the size the Huns needed to sustain their cavalry organisation. Horse peoples certainly need very large strings. Marco Polo, who crossed Central Asia in the thirteenth century, noted that a single rider might keep as many as eighteen changes of mount. It has been calculated, moreover, that the Hungarian plain can graze only 150,000 horses, too few, even allowing ten horses per rider, to have mounted Attila’s horde. That calculation, however, leaves out of account the much milder climate that prevails there by comparison with the steppe, making for a richer and longer growth of pasture. In 1914 Hungary mounted 29,000 cavalrymen at one horse per man and, though the horses would have been larger than Attila’s and partly grain-fed, such differences are not sufficient to explain a tenfold diminution of requirements.48 Hun horses must have thrived in the seventy years they were there and it is most unlikely that Attila was short of them when he set out for the west in 450.

  On the other hand, it is highly probable that a large proportion of the horses he took were ridden to death and that they could not be replaced down his line of communications. Cavalry campaigns kill horses in huge numbers if they cannot be regularly rested and grazed. During the Boer War of 1899–1902, for example, the British army lost 347,000 out of the 518,000 that took part, though the country abounded in good grazing and has a benign climate. Only a tiny fraction, no more than two per cent, were lost in battle. The rest died of overwork, disease or malnutrition, at a rate of 336 for each day of the campaign.49 Attila, moreover, had no means of moving his horses by waggon or ship, as the British transported theirs to and within South Africa. The likelihood is, therefore, that any remounts he received along the overland route from Hungary arrived in little better shape than those his men were already riding, and that the retreat to the
grasslands finished off many of the survivors. The Scourge of God may have been an even worse enemy of his own army. He seems to have left little substantial force to his sons, and their deaths in battle, one at the hands of the Goths, the other against a general of the eastern Roman empire in 469, is the last news we have of the Huns.50

  THE HORSE PEOPLES’ HORIZON, 453–1258

  Yet despite the abrupt disappearance of the Huns from history, the horse peoples had arrived. They were to remain an ever-present menace to the civilisations of Europe, the Middle East and Asia for the next millennium. Theirs had been an extraordinary rise to power in little more than 1500 years. They were, moreover, truly a new sort of people, previously unknown to the world. Military force was, of course, already established as a principle before their coming but as a resource available only to governments and the settled populations they ruled, and one strictly limited by the yields of the economies they controlled.

  Armies that were fed from an agricultural surplus and limited in range of manoeuvre by their pace and endurance on foot simply could not undertake free-ranging campaigns of conquest. Nor did they need to: enemies similarly constrained could threaten them with defeat in battle but not with Blitzkrieg.

  The horse people were different. Attila had shown an ability to shift his strategic centre of effort — Schwerpunkt, as Prussian general staff doctrine later denoted it — from eastern France to northern Italy in successive campaigning seasons, 500 miles apart as the crow flies and considerably further in practice, since he was operating along exterior lines. No such strategic manoeuvre had been attempted or had been possible before. Freedom of action on this scale lay at the heart of the ‘cavalry revolution’.

  The horse people fought unconstrained in another sense. They did not seek, as the Goths did, to inherit or adapt to the half-understood civilisations they invaded. Nor — despite a suggestion that Attila contemplated marriage with the daughter of the western Roman emperor — did they seek to supplant others’ political authority with their own. They wanted the spoils of war without strings. They were warriors for war’s sake, for the loot it brought, the risks, the thrills, the animal satisfactions of triumph. Eight hundred years after the death of Attila, Genghis Khan, questioning his Mongol comrades-in-arms about life’s sweetest pleasure and being told it lay in falconry, replied, ‘You are mistaken. Man’s greatest good fortune is to chase and defeat his enemy, seize his total possessions, leave his married women weeping and wailing, ride his gelding [and] use the bodies of his women as a nightshirt and support.’51 Attila might have spoken thus; he certainly behaved in that spirit.