Pax Romana

by Adrian Goldsworthy

  One thing all types of potential raider had in common was a desire to stay alive. If there were ever bands for whom casualties were less important than killing the enemy, then they were exceptionally rare. Even when a raid was a means of waging war, it was not expected to win that war on its own, but help to wear the enemy down. The full-time warriors in an aristocrat’s household were drawn from a small pool within the tribes, and such skilled men were hard to replace. Other men who volunteered for a raid did so for adventure, glory or profit and quite naturally wished to survive to enjoy these things, while full-time bandits acted for profit rather than in the hope of finding an heroic death. Raiders were highly motivated, since most had chosen to take part in the expedition, and many were skilful and ferocious fighters, but they were reluctant to take heavy casualties. In later eras much the same could be said of the Vikings or Apaches. The aim was to take the enemy by surprise, striking with overwhelming force, looting what they wanted and then escaping. It was never the intention to let the enemy gather in sufficient force and risk a battle on anything like even terms, for a costly victory – let alone a defeat – could cripple a band and discredit a leader.

  Surprise was vital, which meant that time was always the key factor. Successful raiders needed to strike and escape before the defenders could gather in sufficient force to deal with them – something especially dangerous when the area was garrisoned by the Roman army. Once it was decided to mount an expedition, the first important decision was the choice of targets. If the raiders wanted livestock then they needed to know where these were likely to be, something often dependent on the season. Over winter, animals tended to be gathered in one small place rather than wandering to graze. They were likely to be at their fittest when brought down from summer pastures at the start of autumn, before they became thinner on winter feed. If the raiders wanted slaves then they needed to know where houses, farms, villages – and towns for very large raids – were located. The same was true if they wanted to steal things of value, adding temples and shrines to the list, although sometimes there might be a taboo against touching sacred sites.

  Everything said about the Romans’ lack of maps and precise geographical knowledge applies equally to peoples outside the empire. Fiction and cinema have fed a false image of raiders relying on chance, attacking whatever they happened to find – the Viking long ships gliding out of the mist, the Comanches appearing from nowhere out of the desert. The truth is very different. Raiders carried only enough food for a short time, and wandering around hoping to spot likely targets and then attack them was bound to take too long. More importantly, it increased the chance that they would be discovered and so gave the defenders the opportunity to hide, flee to fortified places of safety, and gather in strength to deal with the raiders. Unless the defenders were considered so weak that they were incapable of posing any threat, the raiders needed to have a good idea of where they were going and how to get there before they attacked, especially if they were coming by sea.

  Settlements which had been raided in the past were known about, but over time these would be abandoned, yield less and less plunder or be fortified and become harder to take. In the fourth century AD we read of a Roman commander sending a tribune fluent in the native languages on a diplomatic mission to the tribes, but secretly tasked with discovering their intentions and also spying out the land. On another occasion the same general ordered some officers to capture one of the Alamanni and force the man to guide a Roman column raiding the tribal lands. This was a common method, but relied on finding someone willing and able to co-operate. Hannibal’s army was once led astray by confusion over the pronunciation of a place name. In 53 BC a band of Germanic warriors responded to Julius Caesar’s announcement that anyone who wished might plunder the territory of the Eburones. Two thousand of them crossed the Rhine and caught large numbers of Gauls fleeing from other attacks. Taking many cattle and captives, the Germans were quite content to act as freelance allies of the Romans in this way, until:

  One of their prisoners said: ‘Why do you pursue this miserable and slender booty, when you now have the chance of the utmost fortune? In three hours you may come to Atuatuca; in that spot the Roman army has concentrated all its stores; the garrison is so small that it cannot even man the walls, and no one dares step outside the entrenchments.’ With this hope offered them, the Germans left in a secret place the plunder they had got and made for Atuatuca, using as guide the very man by whose information they had learnt the news.30

  In this case the information was not entirely accurate. Although the Germans surprised the Romans and managed to cut up a detachment out foraging, as well as merchants plying their wares outside the rampart, the camp itself was too strongly defended for them to take. Abandoning the attack, they retrieved the plunder and captives they had already taken and retired across the Rhine to their homeland.

  Caesar employed Roman merchants operating among the tribes as sources of information, and in the same way traders from outside the empire who worked in the provinces were a valuable source for tribal leaders. Tacitus notes that of all the Germanic tribes, the Hermunduri were uniquely ‘loyal to Rome, and with them alone of Germans business is transacted not only on the riverbank, but far within the frontier in the most thriving colony of the province of Raetia [Augusta Vindelicum, modern Augsburg]. They cross the river everywhere without supervision.’ In contrast other Germanic peoples were only permitted to cross at set places and use specific, closely monitored markets, making it harder for them to learn about the layout of the wider province. In the second century AD one tribe seen as more reliable was still banned from trading outside designated markets, because the Romans were afraid that others, similar in appearance but from hostile peoples, might infiltrate with them.31

  Although difficult to date with any precision, several passages in the Talmud clearly assume that raiding was a common risk, especially to communities in Idumaea, bordering the Nabataean kingdom. For instance, ‘It is ordained that if gentiles come on the Sabbath upon villages that are close to the border, the inhabitants issue forth fully armed to meet them and return fully armed. [This law applies] even in the case that they come only to loot straw and wood. If they fall upon villages further in the interior, one may draw out armed only if they attack with the intent to kill.’ Other provisions concern watch-towers and watchmen, who looked out for natural disasters such as storms, but also wild animals and raiders. Jewish law required slaves to convert to their master’s faith, but ‘if somebody bought a slave from a gentile and he does not agree to circumcision, the new master may attend to his persuasion for 12 months and if he remain adamant he should be sold to a non-Jew. Rabbi Simon ben Elezar said . . . in a town on the border he [the slave] should not be kept at all, since he could pass on to his kinsmen matter he had heard.’32

  Runaway slaves and army deserters carried information with them which was very useful to potential attackers – one reason why peace treaties at the end of wars often stipulated that these people be returned along with captives taken by force. As with intelligence gathered by the Romans, not everything told by traders, runaways or spies was accurate. Several Germanic tribes came to believe that Claudius had bound his provincial legates by strict instructions not to cross the Rhine or mount any aggressive operations. As a result they began to settle on land that was east of the river but previously kept clear by the Roman army, and were forcibly evicted. In the fourth century AD a major raid was prompted by the tales told by auxiliaries recruited from beyond the frontier who had gone back on leave. They repeated a rumour that the bulk of the Roman army was about to march away to the east. Presented with such an opportunity, the tribes attacked, only to discover that it was wrong. In AD 68 the Sarmatian Rhoxolani defeated two auxiliary cohorts. This victory, and the realisation that much of the provincial army had gone to fight in Rome’s civil war, encouraged them to raid across the Danube into Moesia. In this case their information was correct, but the chance arrival of a Roman for
ce coming from Syria, marching through the region on its way to take part in the struggle for the throne, led to their defeat.33

  Once raiders had decided upon their targets, they needed to reach them quickly. If there was a river or sea in the way it had to be crossed, and the boats or rafts employed hidden and protected for use on the return trip. Maritime raiders tended to be reluctant to stray too far from their boats in case these were destroyed or the route back to them cut. Roman fortlets on the coast often overlooked landing places. The garrisons were too small to stop a major force from getting ashore, but were well placed to threaten the men left to guard the boats. Man-made boundaries like Hadrian’s Wall or the German limes presented a different problem. Crossing these was possible for men on foot, especially with the aid of rope or ladders, but it would take time, particularly when it involved getting through obstacles like the stakes and other defences in front of Hadrian’s Wall. It was much harder to get horses or other mounts across them without seizing control of a gateway, and these were guarded. The garrisons of towers, milecastles or other fortlets were small, so a strong band was likely to be able to overpower them. Yet this inevitably took time, risked casualties, and meant that an alarm might be raised. Frontier areas without rivers or other barriers were easier to cross. Towers, fortlets and patrolling garrisons increased the chance of detection, and careful use of ground or the cover of darkness was essential to avoid being seen.

  None of the Roman frontiers were able to prevent every determined attacker from passing through them. They stopped some, and made it difficult for the rest while increasing the chance of spotting them and raising the alarm. In our sources, raiding bands were intercepted while they were plundering or as they returned home, but almost never on their way into a province. Attackers were able to choose when and where they crossed, while the Roman troops were spread widely, covering every possible approach. Thus even when bands did raid into the province, most of the garrisons of towers and fortlets would have been in the wrong place, watching and waiting for something that did not happen in their area. Overall numbers mattered far less in such circumstances than the forces able to react to an incursion. The quicker the alarm was raised, the better chance the defenders had to respond.

  Advancing armies routinely burned settlements as they passed. Caesar once achieved surprise by not doing this, catching many of the tribe he was attacking while they were working the fields – they had not had warning to flee to fortified places. Larger raiding bands often acted in a similarly destructive way, aware that their sheer numbers made them visible so stealth was impractical. They also moved more slowly than smaller forces and, if composed of numerous disparate groups and individuals, were harder for their leaders to control. The alarm could be raised by the civilian communities as they approached. Cicero mentions the guards and watchmen at a temple in Sicily raising the townsfolk with their shouts when robbers attacked the precinct – in this case sent by Verres. In another incident at a different town, temple guards sounded a cow-horn to raise their neighbours, who grabbed arms and gathered to resist the attack. The Talmud mentions a villager put on guard and instructed to blow on a ram’s horn trumpet to raise the alarm. There were also guard dogs, and the rabbis decreed that ‘none shall raise a dog, unless he is kept on a chain except in a town adjoining the frontier, in which he is permitted to keep the dog unchained, but only at night time’.34

  Watch-towers – whether in a village, as part of a grand villa, along a road or part of a system of military outposts – were all defensive in the sense that they offered an elevated platform from which to observe and spot potential attackers. Some towers built by the army were clearly intended to signal, with a beacon or raised flag to warn of danger or in some cases a semaphore system allowing a more detailed message. Even a simple alarm alerted anyone able to see the sign, whether civilians or the largest garrison post nearby. The most detailed means of passing on information was in a written message carried by a despatch rider. The presence of horses at fortlets is highly likely, and it may also have been common to have a cavalryman and his mount at some towers. Otherwise, this would mean the main garrison sending a rider to find out what the tower’s men had seen and then riding back with the information. This is possible, given that, except at times of highest emergency, a garrison would take time to prepare a patrol or relief force to deal with the incursion. Response times were more likely to be measured in hours than minutes.

  Part of a report of this nature survives from one of the outposts watching the routes down to Egypt’s ports on the Red Sea. What we have is a copy, written on the fairly flat side of a used amphora. Such pieces of pottery or ostraka were a common and cheap alternative to papyrus, but in this case part of the amphora has broken off, taking with it a section of the text. The real report was no doubt written on more easily portable papyrus or a wooden writing tablet. It was written and sent by Antonius Celer, a cavalryman from the century of Proculus in cohors II Ituraeorum equitata, to Cassius Victor, centurion in the same auxiliary cohort, and describes an attack by ‘sixty barbarians’ on the outpost/way station at Patkoua. Its location is unknown, but it was one of a series of small outposts along the road, all of them provided with wells and protected by a square of high stone walls. Celer does not appear to have held any formal rank, and was an ordinary soldier appointed as acting commander of a garrison consisting of a dozen or so men. Civilian travellers sheltering in and around the fort were also involved, and were probably the main target of the attackers. A woman and two children were abducted, and one of the latter or another child was later found dead. The ‘attack began at the tenth hour of the day and went on until the second hour of the night, and again around dawn the next day. Hermogenes, an infantryman of the century of Serenus, was killed.’ At least two other soldiers were wounded, including ‘Damanais, cavalryman from the century of Victor & Valerius Firm . . . and his horse . . . of the century of Proculus’. The report was sent along the road to the main garrison at the port of Myos Homos, and then circulated to all the outposts, commanded by prefects, centurions, decurions, duplicarii, sesquiplicarii and post commanders (presumably, like Celer without other formal rank).35

  Other documents record the theft of animals from a quarry site run by the army in the Egyptian desert, while there are mentions in ostraka from North Africa of the sighting of groups of nomads. Information was recorded and circulated. In the case of the attack on Patkoua, all military stations in the area were warned of the presence of the ‘barbarian’ raiders. They were placed on their guard, able to warn travellers of the danger, while the larger garrisons prepared to chase the raiders. Forewarned, civilians hid or gathered together for mutual protection, ideally in buildings. Even a simple Iron Age farmstead with houses surrounded by animal pens gave its occupants a better chance than if they were caught in the open. Seeing them warned and ready, a handful of raiders might well decide to look elsewhere rather than take the risk of fighting. Surrounding ditches and animal pens made it much harder to take livestock or captives by stealth. The sturdily built house and work-buildings of a villa, defended by owners, families, workers and slaves armed with whatever was available, were capable of strong resistance, something shown well in the passage from Apuleius where the travellers were mistaken for bandits and set upon by dogs and bombarded by missiles. Sturdy doors and locks presented barriers to small groups of attackers even if they would not normally be considered as ‘military’ defences. Such things could be overcome, but risked casualties and delay. Large raiding bands were unlikely to be deterred, knowing that the bigger villas also yielded plenty worth stealing. Excavations at a villa site near Regensburg-Harting in the Rhineland showed that it was sacked in the early third century AD, and thirteen skeletons were dumped down a well. Several were dismembered and partly scalped – one of the rare instances of scalping in the European Iron Age.36

  If it came to a fight then the raiders were likely to win, for they were more skilled in the use of violence and better eq
uipped than most civilians. Small detachments of soldiers were also vulnerable, if able to put up much more of a struggle. The handful of men stationed in a watch-tower were incapable of prolonged resistance against a serious attack. Their protection relied on the reluctance of raiders to take the time and suffer the losses involved in killing them, balanced against the minimal plunder to be gained, apart from the soldiers’ weapons. The fortlet at Patkoua held out against a long, if probably sporadic attack, even though its garrison was small.

  Once the fighting was over, plundering the victims took time. When Caesar wanted to punish the Eburones in 53 BC, he was reluctant to give the task to his legions. The tribesmen lived in scattered houses and farms, hard to find for anyone who did not know the area well. ‘These localities were known to the dwellers round about, and thus the matter required great care, not for protection of the army as a whole . . . but for the preservation of individual soldiers . . . For the passion for plunder was apt to draw many men too far afield.’ Reluctant to risk his own men, Caesar ‘sent messengers round to the neighbouring states and invited them all, in the hope of booty, to join him in pillaging the Eburones, so that he might hazard the lives of the Gauls among the woods rather than the soldiers of the legions’.37

  Cattle, sheep or other livestock had to be gathered and then watched to prevent them from straying. Human captives had to be restrained and prevented from escaping. Hunting out hidden valuables required time and ingenuity. A few of the coin hoards found in the modern era were buried as offerings, but the majority were surely put in the ground to keep them safe by owners who never returned to retrieve them. The mutilation of the inhabitants of the villa at Regensburg-Harting may have been savagery for its own sake or torture intended to force them to reveal secrets. Warfare in the ancient world was extremely brutal, and war captives were routinely subjected to violence or murder and women to rape, with almost their only protection being their value for ransom or sale as slaves.