Legions tended to be based in the same place for long periods, sometimes for several centuries. Auxiliary units also appear at the same fort generation after generation, although in this case we should be more cautious, as inscriptions attesting their presence but separated by decades need not imply that the unit was based there all the time. Even when units were in place for a long time, this may mean no more than that the site acted as their depot. They were administrative centres, the headquarters building or principia in each base maintaining the unit’s records. Legionary fortresses, and to a lesser extent auxiliary forts, also housed workshops to manufacture and maintain equipment. Some or all bases may also have played a role in training, for we do not really understand how the army trained recruits and whether this was done at unit level or collectively. Much of the time, perhaps especially in spring and summer, the bulk of a garrison was away from its base. Excavation shows substantial sections of the base of Legio XX Valeria Victrix at Deva (modern Chester) in Britain were abandoned in the second century AD, with buildings derelict, demolished or replaced with unusual structures. Some continued military presence is likely at all times, even if this was little more than sufficient manpower to maintain the site, keep the unit’s paperwork in order, carry out minor duties, and care for the sick and injured. Presumably there was a point at which a base was fully abandoned and the unit moved elsewhere if there was no prospect of it ever returning and the site no longer performed any useful function.20
On the well-studied frontiers, especially in Britain and Germany and increasingly in the Danubian provinces as well, we can locate the legionary fortresses and most auxiliary forts as well as smaller installations such as fortlets and watch-towers. This is not the same thing as knowing where the bulk of their garrisons were and what they were doing. At times substantial detachments or vexillations were in another province, and it was even more common for smaller parties and individuals to be away. Service elsewhere in the same province was even more likely. This does not necessarily mean that a fort was empty save for an administrative rump of personnel. A fort built originally to accommodate a particular auxiliary unit could readily house another of the same or a different type. Legionaries could live in barrack blocks designed for auxiliaries and vice versa, although it would be difficult to station a cavalry unit in a fort without adequate stabling facilities, at least for any long period or over winter. There are plenty of hints of men from different units in or passing through other bases. None of this ought to have surprised us, but the great gaps in our evidence for the Roman army have encouraged scholars to piece each fragment together under the assumption that they were part of a simple neat reality. Instead the picture is of a far more dynamic, ever changing, often confusing situation, much like armies in more recent periods. It is worth noting that detachments rarely consisted of formal sub-units such as centuries or turmae, let alone the eight-man contubernium who shared a tent on campaign and a pair of rooms in barracks. Instead as many men as were felt necessary were placed under temporary command to fulfil a particular task.
Unit records tracked the movements of each soldier individually. For instance a first-century AD papyrus from Egypt records the detached service of the legionary Titus Flavius Celer: ‘ . . . left for the granary, 11 February, 80. Returned the same year, (date lost). Left with the river patrol (month lost, 81). Returned the same year, 24 May. Left (destination lost) 3 October, 81. Returned 20 February, 82. Left with grain convoy, 19 June, 83. Returned (date lost).’
Men were sent for as individuals by name. For example: ‘Aponius Didymianus, decurion, to Iulianus the curator, greeting. Please send me quickly Atreides, cavalryman of the turma of Antoninus, when you receive the ostrakon from me, since the prefect has sent for him. Farewell.’ And: ‘Claudius Archibios to Aristoboulos his colleague greeting. I have sent Paprenis of the turma of Antoninus and Iulius Antoninus of the turma of Tullius to Aphis . . . in place of Aponius Petronianus and Iulius Apollinaris. I pray that you are well.’21
It was rare for the entire unit stationed at one base to be there, especially for such large formations as a legion. Smaller sites, such as fortlets and watch-towers, were more likely to have something like the garrison for which they were designed, if only because it was easier to abandon these when they were no longer needed. Yet even with these small installations, there is no need to assume that the men stationed there spent all their time in the tower or behind the ramparts of a fortlet, or even that they were occupied all the year round. As we have seen, systems of towers were common along roads as well as on the frontiers, and there is evidence from Egypt for civilians as well as soldiers manning them. This seems less likely when they formed part of a bigger frontier system, which means that the larger garrisons faced further calls on manpower to station men in these outposts.
The design of Hadrian’s Wall shows that at times it was thought useful to concentrate a large number of troops along one fortified zone. It is not known for certain whether the wall itself had a parapet and walkway, allowing men to patrol along it, or whether it was simply a barrier like the stockades in Upper Germany and Raetia. Even when the design was changed and Hadrian’s Wall made narrower, it was certainly wide enough to have a walkway, so this seems likely but cannot be proven. For every one of the eighty Roman miles of its length there was a milecastle, able to accommodate at most a few dozen men, and between them three turrets. There were fifteen forts, each designed for a single auxiliary unit, on the wall itself and several more, including Vindolanda, close behind or in front of it. A wide, steep-sided ditch, known today as the Vallum, lay to the south, sealing off a military zone. It looks as if this was not carefully maintained throughout the occupation of the wall, but even today it is a formidable obstacle along much of its length. In front of the wall was a narrower ditch – except in a few places where the rock proved too hard to clear or where nature provided a very steep slope or cliff . Between the ditch and the wall were rows of sharpened stakes, some mounted on beams or concealed in pits – the ancient equivalent of barbed wire.22
None of this was intended to be a barrier to the movement of the Roman army. Every milecastle had gates on its north and south sides, allowing passage through the wall, although in some cases these opened out onto slopes impractical for men on foot or horseback, let alone a vehicle. Little was lost when many of them were sealed. More important were the fort gateways, with at least one, and sometimes three, of the four double gateways in each fort lying to the north of the wall. There were also two other main crossing places which may have been partly intended for civilian traffic and which lay beyond the only crossing places of the Vallum. Whenever it wanted to, the army could muster one or more columns and march out beyond Hadrian’s Wall. There were also permanent outposts, several of them forts capable of holding an entire auxiliary unit, maintained well in advance of the wall. In most respects these bases appear typical of the forts built elsewhere, with no hint of exceptional defences to suggest that the troops there faced a permanent and serious threat of attack. Diplomatic activity reached even further north. There is evidence for centurions sitting in on tribal councils on other frontiers. The Vindolanda tablets mention a centurio regionarius at Carlisle, a type of officer who appears elsewhere in the empire and probably mixed political and military responsibilities. Hoards of recently minted Roman silver coins from far in the north of what is now Scotland look very much like subsidies paid to chieftains to secure alliance.23
Negotiation and bribery were backed by the permanent threat and occasional use of major military intervention. The first Roman army to operate in a region may well have had little reliable information about the country or its peoples until it arrived there – just as Julius Caesar was able to discover little about Britain before he launched his first expedition to the island. Much would be discovered during campaigning, and even more if the army settled down in or near the territory. Most of the frontiers of the empire were occupied for a very long time. Some locals welcomed the
Romans from the start, while others came to accept their presence as a reality and did their best to come to some mutual accommodation; both types were sources of information. This was reinforced by direct observation, and peaceful exchange and encounters. Quite a few bases and even more towers and fortlets lay near indigenous settlements. One inscription from Africa records the construction of a ‘scouting outpost (burgus speculatorius)’ which provided ‘new protection for the safety of travellers’, a rare case where a site was named for its function. Roman frontiers were not an ‘Iron Curtain’ intended to keep the two sides apart, so much as networks of bases, garrisons, mobile troops and individual officers and men designed to observe, control and dominate an area. The function was the same, whether the main line lay on a river or some artificial boundary.24
THE ANATOMY OF A RAID
Raiding was the most common form of warfare in the ancient world, and even some large-scale invasions were raids on a grand scale, where permanent conquest and occupation was not the aim and instead the attackers wanted to plunder or take captives. Prestige was also important, for if an army was able to march up and down at will in enemy territory it showed that the attacker was strong and the defender weak. Some would choose to fight a battle to avoid this humiliation, which was often precisely what a stronger attacker wanted them to do. Defeat in battle or the devastation of territory persuaded some peoples to seek peace, offering the attacker hostages, wealth and other prizes to obtain it. Even if this did not occur, and they waited for the invader to withdraw, their perceived weakness invited more attacks from the same enemy and from others. There were also consequences for domestic politics and power. A leader who failed to protect the community from being plundered lost reputation and became more vulnerable to challenges. The Táin Bó Cuailnge, an Irish epic written down by monks in the early medieval period, but with echoes of an earlier Iron Age world of old gods, heroes in chariots and head-hunting, includes this tactless conversation between a king and his queen:
‘It struck me,’ Ailill said, ‘how much better off you are today than the day I married you.’
‘I was well enough off without you,’ Medb said.
‘Then your wealth was something I didn’t know or hear much about,’ Ailill said.
‘Except for your woman’s things, and the neighbouring enemies making off with loot and plunder.’
After this the couple compare their possessions, and Queen Medb discovers that her husband is wealthier, but only because he owns a finer bull. She promptly musters an army and invades the neighbouring kingdom of Ulster to take the even better animal owned by its ruler. The poem, whose title translates as The Cattle Raid of Cooley, is at the heart of the Ulster Cycle and tells the story of this invasion motivated by plunder – here a very specific theft. In this respect it is not too different from the story of Troy, where the great war is prompted by wife-stealing. In the Odyssey, when the hero Odysseus visits the Underworld and is surprised to encounter the spirit of King Agamemnon, he asks, ‘Most glorious son of Atreus, king of men, Agamemnon, what fate of pitiless death overcame you? Did Poseidon overcome you on board your ships, when he had roused a furious blast of cruel winds? Or did hostile men do you harm on land, while you were cutting off their cattle and fine flocks of sheep, or were fighting to win their city and their women?’ A willingness to attack and plunder is taken for granted, which surely explains the ‘hostile’ reception.25
Many wars in the Greek and Roman world focused on plundering attacks, which might sometimes provoke a battle. Many Roman operations, for instance Germanicus’ campaigns beyond the Rhine from AD 14 to AD 16, were raids on a vast scale. This makes it unfortunate that scholars tend to see major wars and raiding as clearly distinct and unrelated to each other. Often dubbed ‘High Intensity’ and ‘Low Intensity’ warfare, the former is the only one seen as important and of concern to emperors. It was also rare for much of the first and second centuries AD on most frontiers. One study of the Rhine frontier in this and later periods denies any large-scale threat, while repeatedly admitting that raids into the empire were common by comparison. These are dismissed as not a threat, which then leads to the claim that emperors created an artificial image of powerful and aggressive tribes beyond the frontier to serve their own needs to win military victories.26
It is true that a raid by a few dozen men – or a few hundred or even a few thousand – would not bring down the empire. It is equally true that the disposition of forts and fortifications in frontier areas were not on their own well designed to stop major invasions. Garrisons spread out along the whole line of a frontier would have permitted a strong enemy to concentrate overwhelming force at a single point and break through. Yet this ignores all the evidence for the mobility of garrisons. Frontier lines were not intended to stop major attacks, which would be met by a field force mustered from the troops in the area and any available reinforcements. Given the diplomatic activity and scouting beyond the frontier, in most circumstances it was expected that there would be sufficient warning to gather enough strength to meet and defeat the enemy in the open, perhaps even in advance of the frontier. Nor was raiding a matter of no concern; an indication of this is the common practice of scattering small detachments of troops in towers and fortlets – something that would have made it more difficult to gather a strong force for a large-scale campaign. Several inscriptions from the Danube record how the Emperor Commodus ‘fortified the whole river bank with fortified posts [burgis] constructed from the ground, and also with garrisons [praesidiis] stationed at advantageous places against the clandestine crossings of robbers’.27
These robbers (latrunculi) came from the lands outside the empire across the river. Dismissing them as bandits and criminals rather than formal enemies helped to vilify them, although no doubt their likely victims did not need any encouragement in this respect. It also preserved the legal distinction between proper enemies in a formally declared war and those whose use of violence was irregular. That an emperor wanted it known that he had taken extensive measures to deal with such bands makes it very clear that the authorities did not ignore low-level raiding. The Roman army dealt with hostile raids and sometimes launched raids of its own, but rarely have scholars given any thought as to the nature of such plundering attacks. Although no single incident is recorded in great detail, it is possible to piece together a composite picture of the factors involved and how both sides acted, and much of this applies equally to bandits within the empire as well as ones coming from beyond the frontiers.
A good deal depended on the purpose of an attack. The most common motive was plunder. Cattle, sheep and other livestock could be consumed or taken home to augment the attackers’ own herds or traded for profit. Some communities relied for their survival on supplementing what they could produce with food stolen from others. Tacitus noted that the Germans took great pride in their cattle, which meant that gaining more would be attractive as well as a blow to the prestige of the victim. Another form of livestock was human captives, especially ones suitable to be kept or sold as slaves. Then there were objects of value, ranging from money, jewellery and statues or vessels made of precious metals to clothing, material, tools, especially metal ones, and other practical items. Such things were useful and had the added prestige of having been won by strength. Another motive was to kill or capture people as proof of courage, perhaps confirming warrior status. Head-hunting was common in many Iron Age societies. The desire to kill was not incompatible with a desire to rob, but the dominant motive would influence the conduct of a raid. Alternatively, both might be secondary if the raid was first and foremost about making war on an enemy, whether that was Rome or its provincial and allied communities. In this case the objective was to inflict injury and humiliation on the target, with plunder a means to this end and to feed the attacking band.28
The leader planning to raid needed to recruit followers. A chieftain’s household warriors were obliged to accompany him, and indeed expected the rewards brought through w
ar. Depending on his importance, this meant a force of a few dozen up to 200 or 300. Other men from the tribe might be persuaded to join, depending on their attitude and obligations to the chieftain as well as an assessment of the chances of success. Accomplished raiders with a past record of winning plunder and glory were better placed to attract other warriors. Another option was alliance with other leaders who brought their own household warriors and anyone else willing to join. Warriors joining the comites of the aristocrats were likely to be less well equipped, probably on foot rather than horseback – except among nomadic or pastoral peoples like the Sarmatians and Numidians. Warfare was not the main occupation for these men, who were mainly farmers or herdsmen, which meant that they were only available for short periods and at certain times of year. Finally there were cattle-rustlers, horse-thieves and bandits, for whom raiding was their main occupation. Some were on the fringes of the community, others preyed on its enemies. Such men might join in a larger raid, but also operated on their own, usually in small groups – larger bands either became chieftains and their followers or were hunted down by established chieftains as threats to the communities they ruled.29
Pax Romana Page 39