by John Keegan
Nevertheless, no one denies that from the third century AD onward, when population pressure in the west and the strains of war with Persia in the east intensified, the identification of the legions with the fortified frontiers became absolute. There was a rationalisation of borders, particularly on the Danube, where the province of Dacia was abandoned in 270, on the River Rhine, on the lower Nile, where the Romans found the Numidians as implacable as the pharaohs had done, and in Africa, where parts of Mauritania were evacuated in 298. On the shorter lines, however, the legions were to fight for another century, and Roman strategy centred on the protection of the internal territories whose integrity the fortified frontiers defined. That being so, it is not factitious to argue that, even if in diluted strength, the outline of the frontiers, which shifted little between the accession of Augustus in the first century BC and the abandonment of Britain at the beginning of the fifth AD, exerted throughout a determining influence on the Roman military outlook. Historians with a particular knowledge of a period or a province, even of the Roman empire as a whole, may be able to show explicit inconsistencies in the view, perhaps bequeathed to us by Gibbon, that Rome saw itself as the still centre of a world of barbaric disorder. But to do so is to overlook the influence that the psychology of a professional army exerted on the imperial policies of the governments it served. Once frontiers are defined by fortifications which then become the permanent places of garrison of formal and named units, or at least familiar stopping-places through which such units rotate, they take on a symbolic significance for the soldiers who defend them; the emergence of such a symbolism is easily discernible in the history of the Roman army when, for example, we find that the VI Legio Victrix, which arrived in Britain about AD 122 from the Rhine, was still there sixty years later, that III Legio Cyrenaica, raised by Julius Caesar on the Nile, was still based in Egypt in the third century, and that two cavalry regiments, Ala Augusta Gallorum Petriana and Ala I Pannoniorum Sabiniana, raised respectively in Gaul and Pannonia (modern Hungary), served from the second to the third centuries on Hadrian’s Wall, the latter throughout at what is today Stanwix.62 The examples elaborate: between AD 69 and 215, III Legio Gallica was in Syria, from AD 85 to 215, II Legio Adiutrix was in Hungary, and from AD 71 to 215 VII Legio Gemina was on the Rhine.63
Within an army whose backbone was supplied by a body of professional soldiers on whose tongues circulated from generation to generation the litany of the legions’ cantonments and the lore of the life lived there, it is impossible that the soldiers’ consciousness did not eventually come to be circumscribed by the geography of the frontiers. There was, of course, much to distract their attention from imperial defence, notably the recurrent disputes over the imperial succession which, during the third century, called legion into conflict with legion in the service of usurpers and provincial pretenders who laid claim to it. The reorganisation of the garrisons under Constantine (AD 312–37), who succeeded to the imperial title by victory in one of these civil wars, withdrew the legions into several central reserves, reduced them in size and added sizeable formations of cavalry to the army.64 These changes drastically altered the composition of the army, diluting for good the strength of the infantry foundation on which it had rested since republican times. It remained, nevertheless, an imperial army, supported, if with greater difficulty in their collection, by imperial taxes, and still dedicated, though at a greater distance from base, to the defence of the frontiers. The quality of the auxiliaries, left by the Constantian reforms in uncomfortable isolation on the ever more contested borders, declined as a result of their detachment from contact with the legions; increasingly these units of limitanei were formed from locally enlisted peasant militias, who were farmers before they were soldiers. The military worth of the regulars remained, nevertheless, formidable.
After Diocletian (284–305), the empire was divided for administrative purposes into western and eastern halves, with a consequent and progressively separative effect on their military forces. But the next and eventually disabling crisis of the imperial armies was not felt until the fifth century. Despite the disasters of the Persian campaign of 363, in which the emperor Julian the Apostate was killed, and the catastrophe of Adrianople (396), in which Valens died at the hands of the Goths, order within the empire and the defence of its borders was restored by the titanic efforts of Theodosius, who reunited the eastern and western halves and waged a succession of campaigns to repel the outsiders from its territory. Nevertheless, as we have seen, it was Theodosius who took the fatal step of compromising the Romanity of the army by accepting under his command large units of barbarian ‘federates’ who served, not as the auxiliaries of old had done in units raised and officered by imperial officials, but as allies under their own leaders. This step, once taken, could not be withdrawn. Throughout the first half of the fifth century Teutonic soldiers poured into the western empire, and, though the imperial structures there remained nominally in place, and local generals such as Constantius or Aëtius retained sufficient force under command to confine some tribes to limited areas of conquest, and at times even to set barbarian against barbarian, control of the frontiers had to be abandoned altogether, while control within was feeble and erratic. The ‘Roman’ armies of Constantius and Aëtius were Teutonic in composition, carried Teutonic weapons, lost all semblance of legionary drill, and even adopted the German warcry, the baritus.65
In the face of Attila and the Huns, some of these barbarian invaders, who had suffered at Hunnish hands outside the empire, came to Aëtius’s aid; they formed a large proportion of his army at Châlons in 451. While that victory spared Gaul, and perhaps Rome, from devastation by a horse people, Italy and the capital now came under threat from another direction. Gaiseric, leader of the tribe of Vandals who had crossed Gaul and Spain to found a kingdom in North Africa, took to the sea, seized Corsica and Sardinia and from that base captured and sacked Rome in 455; a counter-offensive mounted against him by Leo, the eastern emperor, ended in failure; and the Vandals established a piratical regime that controlled Mediterranean waters from their bases in Sicily and Africa; it was carried on by Saracen and Barbary successors for a thousand years. In Gaul and Italy power passed to three German chieftains, Ricimer, Orestes and Odoacer, who set up a succession of puppet emperors. One of them, Marjorian (457–61), actually reasserted a brief imperial authority in southern Gaul but was then forced off the throne. In 476 Odoacer, who disposed of the largest force in Italy, nominally a Roman army owing obedience to the puppet emperor Romulus, defeated Ricimer in a struggle for power, deposed Romulus and proclaimed himself not emperor but king. The senate, which still survived in shadow form, returned the imperial regalia to the eastern emperor at Constantinople; the Roman army in the west had already long ceased to exist.66
EUROPE AFTER ROME: A CONTINENT WITHOUT ARMIES
The Roman army did not cease to exist in the east; it defended Byzantium, at greatly varying distances from Constantinople — sometimes as far away as the Caucasus or the Nile, sometimes under the foot of its Cyclopean walls — until its remnants were overcome in the great siege of Constantinople by the Ottoman Mehmet the Conqueror in 1453. But from the start of the eastern empire’s autonomy, it was a different army from that of the legions. Under Belisarius and Narses, the generals through whom the great emperor Justinian (527–65) recovered control in Italy and North Africa (destroying Vandal power in the process), it very closely resembles that of Aëtius and Marjorian. At Tricameron (453), where Belisarius overthrew the Vandal Gelimer, and at Taginae (455), where Narses won the victory that returned Ravenna and Rome to imperial rule, the bulk of both generals’ armies was formed of non-Romans, including Huns in Africa and a body of Persian archers in Italy.67 Once the borders of Byzantium had been stabilised, however, roughly on the line of the Danube and Caucasus and a sea frontier that enclosed Cyprus, Crete and the toe of Italy (Egypt, Syria and North Africa were lost to the Arabs between 641 and 685), the empire’s military organisation could be put on a different basi
s. It resembled in structure that of Augustus: the empire was divided into provinces, called themes, under commanders who, with their troops, answered to the emperor direct. The troops were organised into units that derived from those of the Constantian reforms of the fourth century rather than from the heavy marching legions; they were small and independent infantry and cavalry regiments that could be combined as needed to reinforce the frontier militias. In the second century there were thirteen themes, seven in Asia Minor, three in the Balkans and three in the Mediterranean and Aegean; by the tenth century the number had grown to thirty, but the size of the army remained constant at about 150,000 men, half foot, half horse, about as large as that of the legionary army under Augustus. Supported by an efficient bureaucracy and taxing system, and fed and supplied by a prosperous peasantry, the Byzantine army effectively sustained a surviving, if greatly changed and, of course, Christianised Roman empire until the onset of the Turkish assaults in 1071.68
In the west no such army was revived to preserve the remnants of that Roman civilisation for which its destroyers professed such a strong admiration. Indeed revival was impossible, for the basis on which the army had been supported, regular and equitable taxation — very inequitable though it had become in the late empire — had been destroyed. The barbarian kings taxed as best they could, but the returns were insufficient to support disciplined soldiers; in any case the conquerors were deeply antipathetic to discipline, preserving in their hearts a rough Teutonic belief in the freedom of the arms-bearing warrior, and of his equality with his fellows. The Goths, Lombards, and Burgundians had been farmers before pressure from the steppe had pushed them across the Rhine, and they expected to live by farming when they came into their inheritance. In Italy each was allotted a third of the occupant’s plot on which he was settled, an extortionate adaptation of the old imperial system of assigning a third of an occupant’s premises to a billeted soldier; in Burgundy and southern France the allotment was set at two-thirds. Thus the soldiers settled down to plough unwelcome on scattered farms, frittering away the military virtues which had made them so formidable in attack, without yielding to the government the regular surplus by which a civilised, peace-keeping army might have been rebuilt. ‘The barbarian kingdoms combined the characteristic vices of the Roman empire’ — principally the corrupt expropriation of smallholdings to swell the estates of the rich — ‘and of barbarism … To their old abuses were now added the lawless violence of the barbarian tribesman and of [the surviving] Romans who aped their manners.’69
In retrospect, how easy it is to see that Rome’s principal contribution to mankind’s understanding of how life may be made civilised was its institution of a disciplined and professional army. Of course, it had had no such end in view when it embarked on its campaigns of expansion within Italy and then undertook the wars against Carthage; the army was transformed from a citizen militia into a long-range expeditionary force under the demands of the battlefield, not by conscious decision. Its adoption of a system of regular enlistment, offering ‘a career open to talents’ throughout the empire and to citizens and non-citizens alike, had its origins in necessity; the reforms of Augustus merely rationalised a situation that already existed. As if by the workings of an unseen hand, however, the evolution of the Roman army exactly served that of Roman civilisation itself. Rome, unlike classical Greece, was a civilisation of law and of physical achievement, not of speculative ideas and artistic creativity. The imposition of its laws and the relentless extension of its extraordinary physical infrastructure demanded not so much intellectual effort as unstinted energy and moral discipline. It was of these qualities that the army was the ultimate source and often, particularly in the engineering of public works, the direct instrument. Inevitably, therefore, the decline of the army’s powers — even if brought about as much by internal economic and administrative failures as by military crises at the borders — entailed that of the empire’s also, and the army’s collapse that of the western empire itself.
The successor kingdoms in the west did not learn how priceless was the institution they had destroyed and how difficult to replace. Yet moral authority in post-Roman Europe did not altogether lose a home; it migrated to the institutions of the Christian church, firmly established in its Roman rather than Nestorian form thanks to the conversion of the Franks in 496, and in the Church the idea, if not the substance, of the empire found a continuum. Without swords, however, the Christian bishops could not give the Christian covenant force; and though their royal patrons had swords, they used them to make war on each other rather than to establish and keep a Christian peace. The history of western Europe in the late sixth and seventh centuries is a doleful one of constant quarrels among the royal houses of the successor kingdoms, moderated only when, at the beginning of the eighth century, the first of the Carolingians established their primacy in the Frankish lands on both sides of the Rhine. The emergence of the Carolingians was the outcome of an internal struggle, but it may be seen also as a response to the new threats — notably the advance of the Muslims from Spain into southern France and the assaults of pagan Frisians, Saxons and Bavarians on the eastern borders. Charles Martel’s victory over the Muslims at Poitiers in 732 repelled them beyond the Pyrenees for good; the campaigns of his grandson Charles the Great, Charlemagne, consolidated a border as far away as the Elbe and upper Danube in Germany and brought the Italian kingdom of the Lombards, which included the city of Rome, within the new empire established by his coronation by Pope Leo III on Christmas Day, 800.
Charlemagne’s legitimacy derived from his recognition by the Pope as successor to the Roman emperor through fictive descent; his power depended on his armed forces, which resembled the Roman army, even in its final state of decay, not at all. Earlier Frankish kings, like the other barbarian rulers, had maintained as the military core of their retinues groups of chosen warriors who could be depended upon to fight bravely and on demand — the equivalent of Alexander’s Companion cavalry. In the era of the conquests the problem of how they were to be maintained did not arise, and in unsettled times they lived by extempore means. But once a kingdom acquired borders, however ill defined, and sought to maintain stability within them, the ruler’s warriors required a steadier source of support than loot or temporary expropriation. The solution was to accommodate members of the Germanic war band (termed, in the Latin that supplied the new kingdoms with so many of their legal terms, the comitatus) within the old Roman practice of the precarium, effectively the lease by which cultivators tilled plots on a landowner’s estate. In the days of the Roman empire’s prosperity, a precarium had been held for money rent; as the disorders of the fifth and sixth centuries drove money out of circulation, the payment of rent gave way to the performance of services of various sorts. It was not a complex process, though in practice it proved gradual, for a ruler’s followers, who already owed him a personal obligation and in return benefited by his patronage (patrocinium), to translate the relationship into one where military service was returned for patronal favour, but the patrocinium was expressed by the grant of a precarium. The relationship suited both parties: the vassal (from the Celtic word for dependant) received a means of livelihood; ‘the ruler was assured of his military services; and the bond between the two was sealed by the performance of an act of homage which, when Christianised by the intervention of the Church, became known as the oath of faithfulness or “fealty”.’70
The arrangement known to us as feudalism (from the beneficiary feudum, or fief, that the patron granted to the vassal) became the general basis on which kings raised armies and the military class held land in Carolingian Europe from the middle of the ninth century onward; by then it was also established custom that fiefs were heritable within families as long as service continued to be done. The formalisation of these elements is conventionally dated to the year 877, when Charles the Bald, king of the West Franks and grandson of Charlemagne, decreed in the Capitulation of Kiersey that fiefs might pass from father to
son; he had already decreed that every free man, which in effect meant those who had land or bore arms, must have a patron or lord, and that every man who had a horse, or ought to have one, should come mounted to the assembly at which, at least annually, the army was mustered. ‘When every man had to have a lord, when every holder of a benefice had to serve as a mounted soldier, and when offices, benefices and military obligations became hereditary, feudalism was complete.’71
Carolingian feudalism, despite the emphasis it laid on horse-owning, should not be equated with the military system of the nomads. The cultivated lands of western Europe could support a horse population of no large size, and the feudal armies that answered the summons to arms resembled a horse people’s horde in no way at all. The difference derived in great measure from the distinctive military culture of the Teutonic tribes, which encouraged face-to-face fighting with edged weapons, a tradition reinforced by their encounters with the Roman armies before they had lost their legionary training. This culture had been preserved when the Western warriors took to horseback, and it was reinforced by the potentialities of the equipment they wore and the weapons they used from the saddle. The saddle itself had developed into a solid seat, in part because from the early eighth century it became the point of attachment for the newly introduced stirrup.
The origins of the stirrup may have been Indian, but in the fifth century it was adopted by the Chinese and then by the peoples of the steppe, whence its use rapidly migrated to Europe. Its significance is fiercely debated, between those who argue that in giving the horseman a firm seat it transformed him into a mounted lancer, and the sceptics who deny that the stirrupless nomad had united himself with the horse any less well; since contemporary evidence is lacking to validate either view, it is not an argument into which the uncommitted should enter.72 But we do know that in the West, from the eighth century onward, the mounted warrior bestrode his horse from a high saddle, lodged his feet in stirrups, and in consequence could manage weapons and wear equipment hitherto associated exclusively with the foot soldier. True, the Persians and then the Byzantines had fielded squadrons of armoured horsemen and even armoured horses at an earlier date, but we do not know how they were outfitted or how they fought, and to ascribe the origin of heavy cavalry warfare to them is risky.73 By contrast, there is no doubt that by the ninth century the feudal horseman of western Europe wore a coat of iron chain-mail, carried a shield and had sufficiently free use of his hands to manage it and a lance or sword while in motion.