The Fall of the Roman Empire: A New History

by Peter Heather

  We have quite a lot of written source material generated in the heat of the subsequent conflict, but unfortunately it largely denounces rather than describes the activities of the Vandal-Alan supergroup. Amongst other things, we have some of the letters of St Augustine who was caught up in the action and eventually passed away when the Vandals were besieging his episcopal see of Hippo Regius, together with a set of contemporary sermons written for an audience in Carthage. Geiseric had recently declared his allegiance to the so-called Arian form of Christianity as espoused by Ulfilas (see Chapter 3), which may well have spread to the Vandals from the Visigoths in the early 410s. The Vandals not only did all the normal damage that an invading army does, but also targeted Catholic Christian institutions and expelled some Catholic bishops from their sees. What the sources transmit, therefore, is not detailed information, but the outrage of good Catholics in the face of persecuting heretics.

  The big question left unanswered is: how, exactly, did Geiseric get his army across the sea? It used to be argued, for instance, that the Vandals and Alans went a long way east from Tarifa by sea, landing close to Carthage itself. If so, where was the Roman army of North Africa? According to the listings of the Notitia Dignitatum, which records the state of the Roman field forces in or around the year 420, Boniface, the count of Africa, had at his disposal 31 regiments of field army troops (a minimum of 15,000 men), as well as another 22 units of garrison troops (at least 10,000 men) distributed from Tripolitania to Mauretania.29 It is normally reckoned that, for a successful landing, a seaborne force needs five or six times more troops than land-based defenders. So if, as we think, the Vandal-Alans could put into the field at best maybe 20,000 warriors, they shouldn’t have stood a chance, especially since they were also bringing with them large numbers of noncombatants.

  The Constantinopolitan historian of the mid-sixth century, Procopius, tried to explain this conundrum by supposing that, faced with his own possible extinction in the three-way struggle for control of the young Valentinian III, the senior Roman commander in Africa, Count Boniface, invited the Vandals and Alans into his province – although even Procopius supposes that he later repented of the deed.30 But there is no reference to any such treachery on Boniface’s part in contemporary western sources (even after he was defeated by Aetius), and if you think about it, such an invitation would not make any sense: by 429, Boniface had made his peace with the imperial court, so he would have had no reason to invite them into Africa at this point.31

  The real explanation for Geiseric’s success is twofold. First, on simple logistic grounds, it is nigh inconceivable that he could have got together enough shipping to move his followers en masse across the sea. Roman ships were not that large. We know, for example, that in a later invasion of North Africa an east Roman expeditionary force averaged about seventy men (plus horses and supplies) per ship. If Geiseric’s total strength was anywhere near 80,000, he would have needed over 1,000 ships to transport his people in one lift. But in the 460s the whole of the western Empire could raise no more than 300, and it took the combined resources of both Empires to assemble 1,000. In 429, Geiseric had nothing like this catchment area at his disposal, controlling only the coastal province of Baetica. It is overwhelmingly likely, therefore, that he would not have had enough ships to move all his followers in one go.

  To move a hostile force piecemeal into the heart of defended Roman North Africa would have been suicidal, offering the Romans the first contingent on a plate, while the ships went back for the second. So rather than trying to move his force a long distance by sea, Geiseric simply made the shortest hop across the Mediterranean, from modern-day Tarifa across the Straits of Gibraltar to Tangier (map 10): a distance of only 62 kilometres – even a Roman ship could normally make it there and back again inside twenty-four hours. For the next month or so, from May 429 onwards, the Straits of Gibraltar must have seen a motley assortment of vessels shunting Vandal-Alans across the Mediterranean. The intinerary is confirmed by the chronology of the subsequent campaign. It was not until June 430, a good twelve months later, that the Vandals and Alans finally appeared outside the walls of Augustine’s town Hippo Regius, about 2,000 kilometres from Tangier, having travelled there by the main Roman roads (map 10). As the Allies found in late 1942 and early 1943, much of the area is far too rough and rugged for any straying off the beaten track, and the Vandals were presumably hauling a wagon train to boot. French historians have busily calculated that, having finally assembled itself during the summer of 429, the force then moved east towards Hippo at a comfortable average of 5.75 kilometres a day.32

  This also explains why the seaborne landing was successful. By choosing Tangier, Geiseric did not land his men in Roman North Africa proper. Tangier was the capital of Rome’s westernmost possession in North Africa, the province of Mauretania Tingitana (modern Morocco). More than two thousand kilometres to the west and separated by the barren mountains of the Er Rif, it was in fact so far from the heart of Roman North Africa that, administratively, it was part of Spain (map 1). Its defence, correspondingly, was not the responsibility of the Count of Africa, but of the Count of Tingitana. He had under his command five field army regiments, which might be reinforced by a further eight units of garrison troops, giving a grand total of thirteen units, or maybe 5–7,000 men. However, the main job of the garrison troops had always been to police the comings and goings of nomads, and it must be extremely doubtful whether they were really up to set-piece confrontations with Geiseric’s battle-hardened force. It had fought its way from the Rhine to Spain, and, at least from its unification in 418, had shown itself capable of standing up to major Roman field armies. The mismatch was even greater than first appears. As we saw in Chapter 5, Constantius responded to the huge losses suffered by western field armies after 405 by promoting garrison troops into mobile field army units. Of the force available to the Count of Tingitana, only two regiments were real field army units; the other three were promoted garrison troops.33 So he had maybe 1,000, at best 1,500, decent-quality troops with which to frustrate Geiseric’s designs. This despatches any idea of a contest, and with it, any mystery surrounding the Vandal-Alan coalition’s ability to get ashore.

  Once disembarked, the coalition headed slowly east. The one possible cross-check we have on their progress is an inscription from Altava, dated to August 429, which records that one of its leading inhabitants had been wounded by a ‘barbarian’, but whether of the Berber or Vandal-Alan variety we do not know. After Altava, 700 kilometres from Tangier, there was only another thousand or so to go before the invading force hit the richest provinces of North Africa: Numidia, Proconsularis and Byzacena. The sources give us no details of the march, but a great deal of vituperative rhetoric:

  Finding a province which was at peace and enjoying quiet, the whole land beautiful and flowering on all sides, they set to work on it with their wicked forces, laying it waste by devastation and bringing everything to ruin with fire and murders. They did not even spare the fruit-bearing orchards, in case people who had hidden in the caves of the mountains . . . would be able to eat the foods produced by them after they had passed. So it was that no place remained safe from being contaminated by them, as they raged with great cruelty, unchanging and relentless.

  Stirring enough in its own way and probably fair historical colour, but not much help when it comes to historical reconstruction. Finally, on the borders of Numidia, the advancing horde was met by Boniface and his army. Boniface was defeated, and retreated to the city of Hippo Regius, where in June 430 a siege began that would last for fourteen months. While Geiseric’s main army got on with the business of besieging, some of his outlying troops, lacking credible opposition, spread out across the landscape. Leaving devastation in their wake, looting the houses of the rich and torturing the odd Catholic bishop, they moved further west towards Carthage and the surrounding province of Proconsularis.34

  Boniface’s failure to hold the line was the result of the same financial stringe
ncies that had hampered Constantius’ reconstruction of Empire everywhere outside Italy. In the fourth century there had been no field army in North Africa, only garrison troops under a dux (duke), supplemented when necessary by expeditionary forces sent from Italy. By 420, and probably from rather earlier, Africa had acquired a field army commander (a count rather than a duke), and a substantial field force (see p. 268). Of the thirty-one regiments only four – maybe 2,000 men – were top-grade imperial field army units. Augustine, in a letter of 417, reports on the good job that Boniface, then a regimental commander, had done with just a small force of barbarian allies in securing the North African provinces against the new threats.35 I take it that these allies will have been some or all of the four units. But the reinforcement was only small-scale: apart from these four, Carthage had to make do with the same old forces that the Empire had always maintained in Africa.

  When Geiseric finally trudged into Numidia, therefore, it was the Tingitana story all over again, but on a larger scale. Boniface did what he could, but the Vandal-Alan coalition was much more awesome than the Berber nomads that most of his troops had been trained to deal with. The key North African provinces were now under direct threat, and the future of the western Empire lay in the balance. For while the province of Mauretania Tingitana in the far west was in no sense imperial heartland, Numidia and its two eastern neighbours, Proconsularis and Byzacena, clustered around their administrative capital Carthage, were a different matter. These provinces played such a critical role in the Empire’s political economy that it is no exaggeration to state that, once the siege of Hippo had begun, Geiseric’s forces were looming directly over the jugular vein of the western Empire.

  Jewel in the Crown

  TWO PICTORIAL representations of Roman North Africa survive in medieval copies of late Roman originals. Between them, they take us to the heart of the region’s role in the western Empire. The first is the Peutinger Table, a copy of a fourth-century Roman world map made at Colmar in the Rhineland about the year 1200. It shows the inhabited world, stretching from Spain and Britain (all but fragments of which are missing) through the Mediterranean world, and on as far as India. The scroll is 6.82 metres long but only 34 centimetres deep: the world as you’ve never seen it before. Highly elongated, its proportions betray the place of its manufacture: about five-sixths of the total is devoted to the Mediterranean, and about a third just to Italy itself. North Africa appears as a line at the bottom, strung out below the west coast of Italy. Immediately below an elaborate Rome is a hardly less impressive representation of Ostia and Portus, Rome’s great ports. Through Portus passed the tribute of Empire on its way to the capital. Clearly visible are its lighthouse, breakwater, quays and warehouses. Immediately below Portus is a considerably more modest representation of Carthage, capital of Roman North Africa, given just a couple of towers. But for all its geographical oddity, the Table focuses our attention on a triangular relationship of key importance to the western Empire: Rome–Portus–Carthage.

  The nature of this relationship emerges clearly from the other late Roman image of North Africa: the Notitia Dignitatum, as well as its military lists, also gives us an illustrated list of the main civilian office-holders of the Empire, along with their staffs. The upper half of the picture accompanying the post of Proconsular Governor of Africa shows, between an inkstand and a desk (on which is pictured the official’s letter of appointment), a female representation of the province holding out sheaves of corn.36 Below, ships loaded with sacks of grain strike out across the sea. By the fourth century, Carthage was the port from which North African grain tributes flooded into Portus, to be offloaded on to carts and smaller boats for the shorter trip inland and upstream to Rome. Carthage and its agricultural hinterland were responsible for feeding the bloated capital of Empire. But keeping the capital fed was no more than a specific application of a much more general point. By the fourth century AD, North Africa had become the economic powerhouse of the Roman west.

  Given its ancient past, there is an irony here. The city of Carthage was founded in 814 BC, or thereabouts, as a Phoenician colony. And once it had come to dominate its hinterland, it spent a fair part of the next seven centuries vying, often violently, with Rome for mastery of the western Mediterranean. In 146 BC, when Carthage was captured after the three-year siege that ended the Third Punic War, the whole city was destroyed and its site symbolically ploughed with salt to prevent any resurrection of this great enemy of Rome. It seems odd, from the modern perspective, to think of North Africa, now very much on the periphery of the western European economy, as the powerhouse that it was back then. When the European colonial powers moved into the region in the nineteenth century, they were staggered by the wealth of its Roman remains – as tourists often are today – especially by the contrast between those remains and their barren and deserted surroundings.37

  Most of the African continent north of the 15-degree parallel now consists of 10.25 million square kilometres of desert. Beneath it is a sunken water table, which, topped up by as little as millimetres of rain each year, maintains a dispersed network of oases. In the very ancient past, the area was much more humid and the water table higher, and the nineteenth-century Europeans originally supposed that part of the answer to the prosperity of Roman North Africa lay in the fact that agricultural conditions had been considerably better then. But the land had dried up long before the rise of Rome, certainly by 2000 BC. The only holdover by Roman times of this different ecological era was the survival close to the Mediterranean of lions, elephants, giraffes and other animal species now confined to sub-Saharan Africa.38 North African hillsides may also have been wooded, but otherwise, conditions were the same in the Roman period as they are today.

  There are exceptions to the pattern of African aridness north of the 15-degree parallel: Egypt is watered by the Nile, for example. The Maghreb, heartland of Roman North Africa, receives rain from nearby uplands.39 In the modern world, the Maghreb comprises Tunisia, Algeria and Morocco, an extensive landscape caught between the Atlas Mountains and the Mediterranean, varying between 300 and 500 kilometres from south to north over its 2,200-kilometre length from the Atlantic to the Gulf of Gabes. Hilly and interspersed with patches of desert, the region’s agricultural possibilities are defined by the exact distribution of rainfall. Where the average is 400 millimetres per year or more, wheat can be grown straightforwardly. The broad river valleys of Tunisia and the great northern plains of Algeria, together with parts of Morocco in the west, fall into this category. Where precipitation is between 200 and 400 millimetres, some kind of irrigation is required, but Mediterranean dry farming can still be practised. Where rainfall is between 100 and 200 millimetres, olive trees will grow – olives requiring less water, even, than palm trees. The North African climate, then as now, was a constant, and had the potential for a wide range of produce.

  The first Roman Africa, ruled directly from Rome after the destruction of Carthage in 146 BC, comprised only a small part of the Maghreb: about 13,000 square kilometres of northern and central Tunisia, bounded by the ‘royal ditch’ (fossa regia) which stretched from Thabraca to Hadrumetum (modern Souss). This territory was divided into 700 square-metre blocks. Apart from six small towns that had supported Rome in the war against Carthage, many of these blocks were kept as public property, rented out to settlers under long-term leases, while some were sold off to Roman magnates with money to invest. To this great sell-off is attributed the fact that in the fourth century AD many old Roman senatorial families (such as Symmachus’) still possessed large estates in this territory – lands that had been passed down and shuffled around over the ages by inheritance and marriage. The rest of the Maghreb was still in the hands of local dynasts, but over the next century these individuals were increasingly sucked into the Roman orbit, while Roman settlers began to move beyond the confines of the royal ditch. As with so much of the Mediterranean, these developments paved the way for the extension of direct Roman rule: first into Numidia
(modern eastern Algeria) in the time of Julius Caesar (46 BC), on the pretext that the last of the local kings had supported his great rival Pompey; then, under Claudius, when two provinces were created in Mauretania (western Algeria and Morocco). From that point on, all of the Maghreb was Roman, although for geographical reasons it came to be administered in two parts: western Mauretania Tingitana run from Spain, and Mauretania Sitifis, Numidia, Proconsularis and Byzacena run from Carthage.

  From early on, the Roman authorities grasped the potential of the well watered coastal lands to provide grain for Rome. Caesar’s expanded province of Africa – called ‘New Africa’ (Africa Nova) – was already shipping to the capital 50,000 tons of grain a year. One hundred years later, after the expansion of direct rule, the figure was 500,000 tons, and North Africa had replaced Egypt as the city’s granary, supplying two-thirds of her needs. A substantial process of development was required to guarantee and facilitate this flow of grain.40

  The first priority was security. In a view heavily coloured by nineteenth- and twentieth-century experience, French archaeologists, some of them military men, considered that ‘civilized’ Roman life in the region must have been under constant threat from indigenous Berbers pursuing a more pastoral lifestyle. Air surveys in the 1930s revealed two lines of fortification and a series of fortified houses and storage facilities in the pre-desert, which were taken as proof of habitual confrontation. No doubt, as on all Rome’s frontiers, smallscale raiding was a constant irritant. It could also lead to larger-scale trouble. The Lepcis affair (see pp. 100-101) began, we noted, when a neighbouring tribal leader was burned alive in the city for unspecified crimes, then escalated into a more serious confrontation. But this clearly didn’t happen often, and even the larger-scale conflicts were not that large. Throughout the first three centuries of its existence, Roman Africa required no more than one legion and a range of auxiliary forces (a maximum of 25,000 men) to guarantee peace and stability throughout its vast extent. Britain, by contrast, needed four legions.