A History of Warfare

by John Keegan

  From the fourth millennium BC, agricultural communities had clustered on the better-watered patches of this plateau; and agriculture probably increased in importance there during the second millennium. On the grasslands around and between these agricultural settlements lived barbarian pastoralists, linguistically akin to the warriors of the western steppe. Through the mediation of agricultural communities in their midst, these pastoralists became increasingly exposed to influences radiating from the distant Mesopotamian culture centre. In this setting, not long before 1700 BC, a critically important fusion of civilised technique with barbarian prowess seems to have occurred.7

  This was the invention, or perfection, of the chariot.

  Why should charioteers, or the pastoralists from whom they directly or indirectly descended, have been more warlike than their hunting ancestors or agricultural neighbours? The answer requires a consideration of factors not for the squeamish, all having to do with how man has killed — or not killed — fellow mammals. It may be taken for granted that the adoption of agriculture reduced the proportion of meat in the human diet; not only is it known that a switch to cereal production always reduces protein intake, as cultivators devote land to crops rather than grazing; it is also a widely observable fact that cultivators seek to prolong the lives of their domesticated animals — so as to maximise their milk yields, carcass weights or muscular power — rather than cull them for food as soon as they reach maturity. As a result, the farmer lacks skills both as a butcher of slaughtered meat and as a killer of young, nimble animals likely to evade his lethal intentions. Primitive hunters, though no doubt excellent butchers, were probably no more skilled in the techniques of the kill; their preoccupations were rather with tracking and cornering their prey than with the precise method by which they struck the fatal blow.

  Pastoralists, on the other hand, learn to kill, and to select for killing, as a matter of course. They must be quite unsentimental about their sheep and goats, which are to them no more than food on the hoof: milk, and its derivatives, including butter, curds, whey, yoghurt, fermented drinks and cheese, but chiefly meat and perhaps blood. It is not certain whether the steppe nomads of antiquity siphoned blood from their animals, as East African cattle-herders do, but it seems possible; they certainly killed the annual crop of young animals and the older breeding-stock, together with injured, deformed or ailing individuals, on a rotational basis. Such a programme of killing demanded an ability to dispatch a living creature with the minimum of damage to the carcass and its valuable contents, and with the minimum of disturbance to others in the flock. Dealing the lethal blow, once, quickly and neatly, was a principal pastoral skill, heightened no doubt by anatomical knowledge gained in regular butchery; the need to castrate most of the males in the flock taught another lesson in cutting flesh, as also did lambing and the rough veterinary surgery of flock management.

  It was flock management, as much as slaughter and butchery, which made the pastoralists so cold-bloodedly adept at confronting the sedentary agriculturalists of the civilised lands in battle. Battle between the two, therefore, may have been not much different from the tentative and procrastinating encounters of the Yanomamö and the Maring, perhaps formalised with ceremonial elements. Even if there were a specialist warrior class, this supposition is not invalidated; lack of armour and truly lethal weaponry speaks for the persistence of ‘primitive’ habits of combat in the Nile kingdom, too, and the Sumerians’ equipment was not much advanced over theirs. In such technological circumstances, battle formations were likely to have been loose, discipline weak and battlefield behaviour crowd- or herdlike. Working a herd, however, was the pastoralists’ stock in trade. They knew how to break a flock up into manageable sections, how to cut off a line of retreat by circling to a flank, how to compress scattered beasts into a compact mass, how to isolate flock-leaders, how to dominate superior numbers by threat and menace, how to kill the chosen few while leaving the mass inert and subject to control.

  All pastoralists’ methods of battle as described at later dates in history disclose just such a pattern. We must make an allowance for the fact that the Huns, Turks and Mongols known to European and Chinese writers had graduated from the chariot to the riding-horse, which made their tactics even more effective; the essentials, nevertheless, must have remained constant. These people, the writers say, did not form lines of battle or commit themselves irrevocably to attack. Instead they approached their enemy in a loose crescent formation, which threatened less mobile opponents with encirclement around the flanks. If strongly resisted at any point, they would stage a withdrawal, the object of which was to draw the enemy into an ill-judged pursuit that would break his ranks. They closed to hand strokes only when the battle was clearly going in their favour, and when they did so they inflicted wounds with superlative edged weapons which often decapitated or dismembered; they thought so little of the quality of enemy steel that they disdained to wear any but the most exiguous armour themselves. To turn the battle in their favour, they harried and intimidated the enemy with volleys of arrows shot at long range with their terrifyingly superior arm of choice, the composite bow. Ammianus Marcellinus wrote of the Huns in the fourth century AD, ‘In battle they swoop upon the enemy, uttering frightful yells. When opposed they disperse, only to return with the same speed, smashing and overturning everything in their path … there is nothing to equal the skill with which — from prodigious distances — they discharge their arrows, which are tipped with sharpened bones as hard and murderous as iron.’8

  Scholars dispute how to date the appearance of the composite bow; it may have been in use as early as the third millennium BC, if a Sumerian stele has been correctly interpreted; it was certainly in existence by the second, since its distinctive ogival or ‘recurved’ shaped — familiar to us as the ‘Cupid’s bow’ whose arrows transfix Watteau’s and Boucher’s lovelorn courtiers — is clearly depicted in a golden bowl of 1400 BC, now in the Louvre.9 It cannot have appeared overnight, for the complexity of its construction, like that of the chariot, speaks of a variety of prototypes, and decades, if not centuries, of experimentation. In its finished form, which did not vary between its perfection in the second millennium BC and its supersession as a weapon of war in the nineteenth century AD (it was last used by Manchu bannermen), it consisted of a slender strip of wood — or a laminate of more than one — to which were glued on the outer side (‘back’) lengths of elastic animal tendon and on the inner side (‘belly’) strips of compressible animal horn, usually that of the bison. The glues, compounded of boiled-down cattle tendons and skin mixed with smaller amounts reduced from the bones and skin of fish, might take ‘more than a year to dry and had to be applied under precisely controlled conditions of temperature and humidity … a great deal of art was involved in their preparation and application, much of it characterised by a mystical, semi-religious approach’.10

  The composite bow began as five pieces of plain or laminated wood — a central grip, two arms and two tips. Once glued together, this timber ‘skeleton’ was then steamed into a curve, opposite to that it would assume when strung, and steamed strips of horn were glued to the ‘belly’. It was then bent into a complete circle, again against its strung shape, and tendons were glued to its ‘back’. It was then left to ‘cure’ and only when all its elements had indissolubly married was it untied and strung for the first time. Stringing a composite bow, against its natural relaxed shape, required both great strength and dexterity; its ‘weight’, conventionally measured in ‘pounds’, might amount to 150, against only a few for a simple or ‘self’ bow made from a length of sapling.

  Similar ‘weights’ characterised the long bow, when toward the end of the Middle Ages west European bowyers learnt to use a billet containing both heart and sap wood to fashion their weapons; it worked by the same principle of opposing the forces of elasticity and compressibility, stored by the archer’s arm when he bent the bow, and released by his fingers, to drive an arrow forward. The disadvant
age of the long bow, however, was precisely its length; it could only be used by an archer on foot. The composite bow was short, reaching only from the top of a man’s head to his waist when strung, and therefore suiting itself perfectly to use from a chariot or horse. It shot a lighter arrow — the best weight was about an ounce — than the long bow would, but could still carry to 300 yards with great accuracy (far longer ranges in free flight have been recorded) and penetrate armour at a hundred yards. The lightness of the arrow was actually an advantage, since it allowed the pastoral warrior to carry a large number — up to fifty in his quiver — into battle, which he counted on winning by subjecting the enemy to a disabling hail of missiles.

  The simple accoutrements belonging to a chariot or mounted archer did not vary for more than 3000 years. The essentials were the bow itself, the arrow and the thumb ring, which protected his skin from flaying at the moment of the arrow’s release; important accessories were the quiver and the bow case, which shielded weapons not in use from variations in temperature and humidity (both degraded the weapon’s range and accuracy). This equipment can be seen in some of the earliest depictions of composite-bow archers; exactly the same objects are displayed as principal parts of the regalia of the Ottoman sultans of the eighteenth century AD in the Topkapi Palace in Istanbul today.11 Many other items also did not vary in the horse world — including tents, floor coverings, cooking-vessels, clothing and the nomads’ simple pieces of furniture. The pastoralists stored their possessions in chests, which could be slung in pairs athwart a pack animal, and used round-bottomed pans and kettles, which could be bundled into nets; the kettle drum, on which the Turks would beat the note of battle, was no more than a nomad’s camp cauldron with skin stretched across its mouth.

  Quite as much as the charioteers’ equipment and their familiarity with animals, their ability to move and readiness to do so fitted them for aggressive warmaking. All war requires movement, but for settled peoples even short-range moves impose difficulties. Their gear is clumsy and heavy; they lack readily mobilisable means of transport, particularly draught-animals — which are needed in the fields; and food both for man and beast comes in an awkward and bulky form. Settled people expect to sleep with a roof over their heads, but do not possess tents; they take shelter when the weather turns inclement, lack weather-proof clothes, and relish regular, cooked meals. The farmer is hardier than the artisan — the Greeks thought it was ponos, agricultural toil, that fitted the tiller of the soil alone to be a warrior — but even he is soft by comparison with the nomad.12 The nomad is constantly on the move, eats and drinks when he can, braves all weathers, is grateful for small mercies. Everything he possesses can be bundled up at a moment’s notice and his food moves with him, as grass and water call his flocks, whenever he shifts camp. Even the nomads most favoured by circumstances, those whom regular availability of summer and winter grazing at fixed locations allows to practise transhumance, are tougher by far than the settled farmer. The ancient nomads of the arid steppe, where tribe had to compete against tribe for what scraps of grazing there were, must have been among the toughest people in creation.

  The American sinologist Owen Lattimore crossed the 1700 miles of arid zone between India and China in 1926–7 — following part of the route which those who brought their chariots to China in the second millennium BC may have taken, oasis by oasis, over several generations. The caravan men among whom he travelled, he recalled,

  become nomads. Many of their propitiatory rites and self-defensive tabus are not only taken over from the Mongols, but from the most primitive instincts of nomadic people. They strive to propitiate the powers and spirits that follow at the heels and lurk about the tents of savage, wandering people at grips day and night with the harsh menace and niggardly resources of a raw, unmastered country. From the moment that the tent is pitched at the first camp … fire and water assume a different importance. Each time that the tent has been set up in a new place, a little of the first water boiled and the first food cooked must be thrown out of the door.

  This was always done even though the food and water available to the caravan men was grossly unpalatable.

  We began the day at dawn by making tea … of the coarsest grade of twigs, leaves and tea-sweepings … In this tea we used to mix either roasted oaten flour or roasted millet — looking like canary seed, which in fact it was — stirring it into a thin slush and drinking it down. About noon we had the one real feed of the day. This would be made of half-cooked dough. We carried the white flour along with us, and would make the same sort of dough every day. We would moisten the flour, roll it and thump it, and then tear it up into little blobs or cut it into a rough kind of spaghetti … The reason we drank so much tea was because of the bad water. Water alone, unboiled, is never drunk … Our water everywhere was from wells, all of them more or less heavily tainted with salt, soda and, I suppose, a number of mineral salts. At times it was almost too salt to drink, at other times very bitter. The worst water … is thick, almost sticky and incredibly bitter and nasty.13

  Lattimore’s nomads probably differed in their habits from those of the second millennium in using tea and flour; in other respects there can have been little to choose between their ways of life, characterised in both cases by its subjection to natural forces, unpredictability and extreme harshness. Anything that alleviated that harshness must have been welcome, and it is in that light that we should consider why — perhaps rather than how — the two extraordinary artefacts, the chariot and the composite bow, appear to have originated in the borderlands where civilisation touched the nomadic world.

  The elements of the chariot — wheels, chassis, draughtpole and their metal fittings — were ‘civilised’ in origin, in that they derived from clumsier prototypes developed for work in farming and building. Archaeologists continue to disagree about who then refined the elements into the light, cross-country chariot itself, but they do not address the question of what the chariot was for.14 That may be clarified if we ask how the chariot was used: for war, of course, but also for hunting. It could be driven across rough ground, and it was used as a platform for huntsmen to shoot game with the composite bow, as much pictorial evidence from many Egyptian and Mesopotamian sources attests; Chinese poetry of the Chou dynasty also makes it clear that the chariot was a hunting-vehicle.15

  That being so, we may perhaps suggest that the chariot and the composite bow appeared together because they served a crucial need of the nomad pastoralist: to give him a means of herding his flocks at a pace faster than his feet could carry him, and also to put him on near, if not equal, terms of mobility with the predators, wolf and perhaps bear and the large cats also, that harried their flanks. It would certainly have made an excellent platform for the composite bowman out after wolf, who would have found accurate shooting at moving targets no more difficult, perhaps easier, than the horseman was later to do from the saddle. Settled people were later to marvel at the horseman’s ability to drop his reins and pick off a victim with an arrow without checking pace. John Guilmartin ascribes it to ‘[the infinity] of time the steppe nomad … spent herding and guarding livestock, which kept him in the saddle but not otherwise occupied … unless by constant archery practice … In view of the number of targets — human and animal, edible and otherwise — which the steppe presented, constant practice made economic sense.’16 If we substitute ‘chariot’ for ‘saddle’ in that passage, the sense remains the same and the argument quite as convincing.

  Toward the middle of the second millennium BC, the peoples who had learnt the skills of making and using chariots and composite bows discovered — by what means we cannot surmise — that the defenders of the settled lands could not stand against the aggressive methods they had initially devised to oppose the predators that attacked their flocks. Charioteers who descended from the highlands to the open and level plains were able to inflict crippling casualties on the Mesopotamians and Egyptians with impunity. Circling at a distance of 100 or 200 yards from the he
rds of unarmoured foot soldiers, a chariot crew — one to drive, one to shoot — might have transfixed six men a minute. Ten minutes’ work by ten chariots would cause 500 casualties or more, a Battle of the Somme-like toll among the small armies of the period. In the face of such an attack by an enemy against which it could not manoeuvre out of trouble, the stricken host had only two choices: to break and run or to surrender. In either case, the outcome for the charioteers would have been a large booty in prisoners, probably rapidly destined to become chattel slaves.

  It is widely suggested that the first interpenetrations of steppe and civilised societies were brought about by long-range traders, who carried cloth, trinkets and worked metal to exchange for the objects of value that the barbarian world yielded, including furs, tin — and slaves. No one knows how slave-trading began. It would have come naturally to pastoralists used to herding four-footed animals, particularly if foreigners adopted the habit of taking their goods to sites at which the herdsmen gathered for seasonal festivals, which, as Lattimore noted, ‘tend to become the scenes of fairs’, and such fairs may have been the first slave markets.17 If the pastoralists had learned to accumulate and convoy slaves for sale on the steppe, one may suppose that when they eventually descended from the highlands on campaigns of conquest, they were attuned to slave-taking and slave management and well prepared to impose their authority over the people they conquered through an intermediate stratum of slaves attached to themselves.

  That would be one explanation of how quite small groups of aggressive intruders not only overthrew but for a time sustained power among peoples who greatly outnumbered them. That the chariot rulers were also slave-masters appears indisputable. Of course, slavery was known in pre-chariot Mesopotamia and Egypt, but its practice, particularly on a trade basis, may have been intensified there by the arrival of the chariot conquerors, while its transmission into Europe may have derived from the migration of the Mycenaeans from Asia Minor, who did not bring the chariot with them but acquired it about the middle of the second millennium BC, at the time when it suddenly came to dominate warmaking in the Middle East.18 Slavery in China is dated to the arrival of the Shang dynasty, while, according to the Rig-Veda, the chariot conquerors of the Indus valley made slavery the basis of what would later become caste.