The Western Empire

Africa

For the first century and a half of late antiquity Africa was protected by its geographical situation from many of the pressures that weighed on the rest of the empire.⁶⁷ The population was a mixture of indigenous peoples, collectively called Berbers (or simply barbaroi, barbarians) and long established immigrants, the majority from Italy, but also others from elsewhere in the Mediterranean.⁶⁸ Since the first century BC the region had enjoyed an enviable stability. The wealth of Africa derived in the first place from agricultural production. The large hinterland of Carthage in Africa Proconsularis and Byzacena was mainly taken up with cereal agriculture. In the early empire two-thirds of the annual harvest is said to have been exported to Rome. In the late second and third centuries oil production became a major feature of the interior of Numidia and Byzacena, and especially of the Tripolitanian hinterland. The territories of Lepcis Magna, Oea, and Sabratha were extensively given over to producing oil, mainly for the Roman market. The pattern continued into the fourth century. Around 350 the anonymous author of the Expositio totius mundi et rerum remarked that Africa was wealthy in all things, including grain, fruit, trees, slaves, and textiles, but “it virtually exceeds all others in the use of the olive.”

Collection of the annona was a priority in the later empire and recent study of the milestones of Africa Proconsularis, one of the main grain producing areas, indicates on-going repairs and the provision of road stations and collecting points to maximize efficiency in the collection of grain and other produce.⁶⁹ Like all tax collection in the late empire these methods were open to abuse and inevitably led to complaints. Aurelius Victor, in the De Caesaribus which he wrote around 360, noted: “the collectors were evil, open to bribery, cunning, rough-necks, grasping and almost naturally adapted to committing and concealing fraud.” He should have known what he was talking about, as he describes his own father as being a poor African farmer (Aur. Victor, Caes. 33). The law codes also contain plenty of evidence for exploitation of this sort, but oppressive tax collection was an unavoidable feature of Roman rule at all periods, and we can no more take this as a reliable index of maladministration than the contrary description of Victor, bishop of Vita, who prefaced his account of the miseries brought by the Vandal invaders with the simple, and equally contestable assertion that the barbarians were unleashed on “a province that was peaceful and quiet and presenting a picture of a whole world in bloom” (Victor of Vita 1.1.3).

Settlement in the African provinces was based on a combination of small towns, municipia, and estates, which were made up from Berber villages. The towns were the focus for the public activities of provincial landowners, many ultimately of immigrant origin, and these preserved their civic institutions until the early fifth century. Local magistracies were a focus for aristocratic ambitions, and not only the secular but also the non-Christian religious institutions of small town life flourished later in North Africa than anywhere else in the empire. The epigraphic and archaeological evidence from the municipalities and colonies of Africa in the late fourth century evoke an age of independent, self-sufficient cities, whose magistracies and priesthoods continued to be filled by members of the local landowning families. Inscriptions imply that these men still spent much surplus income for the benefit of their home towns on buildings, games, and public festivals.⁷⁰ This picture need not correspond wholly to reality, and there is evidence from the law codes, as there is from elsewhere in the empire, that local councilors and others were increasingly reluctant to carry the burdens and expenses of office. However, it is beyond doubt that the African gentry were still anxious to keep up appearances (see Plate 10.4).

Plate 10.4    Mosaic depicting an aristocratic domain in late fourth-century North Africa. The Dominus Iulius Mosaic from Carthage, depicting a lavish, but fortified villa, with outbuildings and a bath house with cupolas, surrounded by hunting and agricultural scenes (© 2013 DeAgostini Picture Library/Photo Scala, Florence)

Changes occurred in the fifth and sixth centuries, which have been summarized by Mattingly and Hitchner, in their invaluable recent survey of archaeological work on Roman Africa. Many fora ceased to function or played a reduced role in urban life. Some of their functions were replaced by buildings attached to churches, which often occupied part of the earlier fora. Temples in particular, and some bath buildings, fell into ruin; large peristyle houses were broken up into smaller units; shop fronts, workshops, and booths appeared along the street frontages. However, it is evident that these changes implied not so much an overall downturn in economic conditions as a change in social priorities.⁷¹

Under the settled conditions of the empire the native populations had a large stake in sustained peace and order. Many, in accordance with the early imperial laws, the lex Manciana and the lex Hadriana, became tenants on huge estates owned by the landed aristocracy, with rights of usufruct and inheritance, but paying a third of their produce as rent. The terms of these leases, which were attractive to long-term tenants, were still in force at the end of the fifth century.⁷² As in Syria, the country villages of the Numidian uplands multiplied and became more densely populated between the second and sixth centuries. Productivity presumably rose as the population increased and Africa established itself as the economic powerhouse of the late Roman western Mediterranean. The diagnostic feature of this success was the penetration achieved by African ceramic production in western markets. The most widely distributed fine pottery ever produced in the ancient world was African red-slip. From the late first century until the end of antiquity, African red-slip table wares were found throughout the empire from Scotland to Ethiopia, and came close to monopolizing the market in the wealthier parts of the western Mediterranean, in Italy, Gaul, and Spain. This was matched by the equally spectacular success of African oil exports. The production of the first amphoras designed specifically for the oil trade, the so-called “piccolo Africano” and “grande Africano,” which appear in the last second and early third centuries, corresponds to the huge development and growth of Tunisian and Tripolitanian olive production. Combined with the enormous export trade in grain to Rome, these oil and ceramic industries made Africa the equivalent of a modern economic superpower. The favorable terms of the leases regulated by the lex Manciana and the lex Hadriana encouraged tenants to produce on a commercial basis for the market. Africa had good products to sell: fine table wares which wiped the floor with locally produced competitors, and abundant olive oil for the growing needs of urban markets. An equally crucial factor was that roads, harbors, transport ships, and the human organization needed to operate large-scale exports had been developed to deliver grain and oil to Rome. Fine pottery and many other African products were sent to overseas markets on the back of the annona system. A third point not to be underestimated in the African economic success story was that the region was almost completely shielded from the burdens of warfare.⁷³

The invasion of the Vandals had a major impact on political and religious conditions in Africa, but seems not to have inflicted heavy damage on the regional economy. The newcomers were now the main political authority, with the ability to impose their will by the threat of military force. The royal palace was sited on the Byrsa, the acropolis of Carthage, and Vandal confidence in their regional supremacy seems to be reflected in the fact that Carthage's city walls, which had been built under Theodosius I, were allowed to fall into disrepair in the mid-fifth century. Vandal warrior families predictably claimed some of the best land in Africa for themselves. There are remarkably few archaeological traces of their presence, but the handful of inscribed epitaphs for Vandals and the precise locations indicated by literary sources can almost all be plotted within the boundaries of Africa Proconsularis, forming Carthage's enlarged hinterland.⁷⁴ According to Procopius, Gaiseric took property from many African landowners and distributed it to his own people. These estates were still called kleroi Bandilon in Procopius' time. The Vandals effectively stepped into the shoes of the Roman authorities that had ruled Africa previously, and took over the existing administrative regime, including the tax system.⁷⁵

Archaeological evidence suggests that the export of goods and produce from Africa was not seriously impaired by the Vandal conquest, but very little Roman bronze coinage seems to have circulated in Vandal Africa, and the Vandals produced their own issues or re-circulated earlier types. The overall economic effects of the Vandal kingdom both within and outside Africa have yet to be fully assessed. There is some indication that goods that had previously been exported to Rome and Italy were now redirected westwards towards Visigothic Spain and southern Gaul. Rome was now presumably obliged to purchase grain that had previously been sent from Africa in the form of tax, and may have turned increasingly to Sicily as an alternative source of supply.

During the fifth and sixth centuries the Moors (Mauri), based in the Aures mountains, emerged as a prominent factor in ethnic and political conflicts. After the reconquest of Africa by Justinian, successive governors encountered much more significant resistance from Moors under tribal leaders, some of whom had served in the Roman army. The Roman reconquest of Africa seems to have damaged the local economy, above all due to the imposition of high levels of taxation (see p. 471). Justinian's forces and successive east Roman commanders assumed not so much the role of liberators as of occupiers. There were aspirations to revive city life. Procopius has an eloquent passage extolling the rebuilding of the previously insignificant and waterless town of Caputvada, where Belisarus' forces had first set foot on African soil:

Justinian…conceived the desire to transform this place forthwith into a city which should be made strong by a wall and distinguished by other constructions as worthy to be counted a prosperous and impressive city; and the purpose of the emperor has been realised. For the wall and the city has been brought to completion, and the condition of the territory is being suddenly changed. The country-dwellers have thrown aside the plough and lead the existence of a community, no longer going the round of country tasks but living a city life. They pass their days in the market place and hold assemblies to deliberate on questions which concern them; and they traffic with one another, and conduct all the other affairs which pertain to the dignity of a city. (Procopius, Buildings 6.6.13–16, adapted from trans. of Dewing)

However, little other evidence supports this optimistic picture of civic renewal outside Carthage, where a broader-based program of urban restoration seems to have been attempted. Fortifications were erected at many of the major cities, including Lepcis Magna, Sabratha, Theveste, and Sitifis, but the new defended circuits were much smaller than those of their fourth-century precursors, and sometimes cut directly across the central areas of the cities.⁷⁶ The aim of most of Justinian's building program was to maintain military control of the African provinces (Procopius, Buildings 6.5–7). In striking contrast with sixth-century Syria, where almost all city fortification was undertaken at local initiative, most of the building inscriptions relating to fortifications in Byzantine Africa indicate that they were constructed by the occupying force, including twenty-four erected by the praetorian prefect Solomon on behalf of Justinian.⁷⁷ The perceived insecurity of the country is in marked contrast to the Vandal period, when the fortifications of Hadrumetum had been deliberately removed (Procopius, Buildings 6.6.1). Indeed Procopius' account of the later years of the Byzantine reconquest of Africa, after the initial defeat of the Vandals, does little to hide the high level of conflict and disruption (see pp. 419–20).⁷⁸

The final eclipse of Romanized civic life in Africa came swiftly. Carthage was abandoned after the Arab conquest and replaced in 670 by the new foundation of Qayrawan. The Berber tribes, which had grown in strength during the period of east Roman rule, at first resisted but later offered support to the Arabs and provided much of the manpower that extended the Muslim conquests to the Atlantic and Spain.⁷⁹ Despite the extraordinary manifestation of small-town culture and civic organization that had prevailed in African history between the second and fifth centuries, post-Roman Africa appears to owe less to its classical antecedents than any other part of the ancient world where city life had flourished. It is important, nevertheless, to observe that the collapse of civic life did not spell the end of North African Christianity. The extensive conversion of Berber tribes to Christianity, which had begun in the fifth and continued through the sixth century, and the arrival of Arab conquerors in the seventh century were followed by a long period of coexistence between Islamic and Christian worshippers.⁸⁰

Gaul

The Gallic provinces were at the heart of the western Roman Empire in late antiquity. The transformation of their political and social structures during the fifth century helps to define our understanding of the collapse of the empire itself. These changes were extremely complex. They involved the abandonment of the frontier defenses of the empire; the encroachment and emergence of new ethnic groups, speaking different languages; wholesale economic restructuring; and major religious realignments.

The third century had already presented an alarming preview of future events when the Rhine frontier had been exposed to major invasions by Germanic groups, which had been able to range and plunder extensively in northern Gaul. Their impact is dramatically illustrated by the discovery at the Rhine crossing of Rheinzabern of three boatloads of loot, which had been plundered from the villas of northeastern Gaul by a raiding party. The hoard contains numerous virtual replica sets of the tools and household equipment of these farms. The region of the Marne and the Somme in northern France was covered with small villas, evenly distributed across the countryside.⁸¹ The armies of Germania Inferior and Germania Superior were heavily dependent on produce from the northern regions of Gaul, and the pattern of well-equipped rural estate centers implies that the organization of rural production to supply the garrisons was systematically standardized.⁸² The settled but undefended population of this region was acutely vulnerable when the frontier defenses collapsed. The raiders went for these exposed rural targets. “They avoid the actual towns as if they were tombs surrounded by nets,” wrote Ammianus of the Alemanni in the 350s (Ammianus 16.2.12).

Roman authority in Gaul was reasserted by the tetrarchs and Constantine. Their overriding aim was to re-establish the old frontier along the Rhine and across southern Germany to the upper Danube. As always Roman military power was made effective by a combination of good communications and effective garrisoned strong points. The frontier roads, ensuring communications between Neumagen, Trier, and Cologne and along the Rhine itself were repaired and garrisoned, and the critical route from Cologne through Tongres and Beauvais to Boulogne, the Channel port, was protected by fortlets. The north–south highway through France from Trier to Lugdunum and then south to Arles and Marseilles in Provence was another key to political control. This was also the route by which supplies moved, by road and river, from the Mediterranean to the northern frontier zone.

Up to the reign of Valentinian I, the emperors invested heavily in maintaining the status quo. Diocletian extended the legionary fort at Kaiseraugst perhaps to accommodate troops from legio I Martia. Constantine rebuilt the bridgehead fort at Cologne-Deutz (ILS 837; see p. 72).⁸³ Julian's campaigns in northern Gaul, and those of Constantius' generals around the Alpine fringes, were aimed at securing the frontier from Alemannic incursions. This policy continued under Valentinian, who, like Julian, sometimes placed forts beyond the river. Within the limes it seems to have been policy to establish a zone 20–40 kilometers broad, where roads were protected by regularly spaced forts and watch towers. Inscriptions attest numerous burgi along the upper reaches of the Rhine (ILS8949, 371). Large numbers of similar fortifications were also built further east in the great Danube bend between Brigetio and Aquincum (ILS 762, 774–5).⁸⁴ During the fourth century fortified silos were built at larger farms near the Rhine.

The Gallic provinces also provided important tax revenues. This emerges clearly in the continuous arguments between Julian and the officials appointed by Constantius to operate in the province. Ammianus points out that oppressive taxation, in areas that already suffered both from the continuous presence of Roman troops and from constant barbarian threats, severely debilitated the provinces close to the northern frontiers.

(Julian) knew that the irremediable damage inflicted by settlements or, to give them a truer name, unsettlements of this kind had often brought provinces to utter poverty; the very thing, as will be shown later, completely ruined Illyricum. (Ammianus 17.3.3, trans. Hamilton)

The same theme is picked up when Ammianus discusses the activities of the Praetorian Prefect of Illyricum, Anatolius, during the campaigns of Constantius II against the Limigantes on the frontiers of Pannonia.

Under the rule of Anatolius, who was then praetorian prefect in Illyricum, supplies of all kinds had been collected even before they were needed and were pouring in without expense to anyone. It is an established fact that the northern provinces have never to this day been so prosperous under any other prefect. Anatolius…relieved the provincials from the enormous expense of the imperial post service, which had ruined countless homes, and from the hardships of taxation. In fact the inhabitants of those parts might well have lived thereafter happy and untroubled lives with all their grievances settled, but for the abominable refinements of taxation to which they were later exposed. (Ammianus 19.11.2–3, trans. Hamilton)

The imperial palace and administrative center at Trier, Augusta Treverorum, which had been the capital of the breakaway Gallic empire under Postumus in the 260s, was at the hub of Roman power in northwest Europe. The city covered an area of c.285 hectares with an estimated population of 80,000, the largest urban settlement north of the Alps. Construction in the city was on a comparably grand scale. The fortifications of the third century were completed by the Black Gate, whose four-storied towers with a great frontal facade of arcades and colonnades were built not for functional defense but to make a deliberate statement about imperial power. The central palace, adjoining a hippodrome, housed the tetrarch Maximianus and his Caesar Constantius. In 326 work began on a new religious core for the city: two large basilicas, the smaller already complete by 330, and a baptistery. This formed the north end of an enlarged palatial complex, including the finest imperial audience chamber to survive to modern times (Plate 10.5). Trier was a residence for Constans in the 340s, for Valentinian from 365–75, and for Gratian and the usurper Magnus Maximus in the 380s. It has been argued that Julian, who spent little time in Trier, deliberately avoided the city between 358 and 360 as he was building up his Gallic power base further south in Gallia Belgica, in preparation for his own bid for power in 360.⁸⁵ The city was also the headquarters of the Praetorian Prefect for Gaul. Trier was already notable for a very large bath house built in the second century AD. This was to be eclipsed by the so-called Kaiserthermen, the imperial baths (Plate 3.2). Work on these began before the middle of the fourth century, but they seem to have been converted for use as another palatial or administrative center under Valentinian. Large warehouses were also built near the river Mosel at this period, presumably to supply the imperial comitatus.

Plate 10.5    Trier, Basilica of Constantine (Photo by Berthold Werner / Wikimedia Commons)

Trier was eclipsed as the empire's western capital at the end of the fourth century. Valentinian II was briefly based there in 389–90, but Honorius, who was raised to the rank of Augustus to succeed him in January 393, never ventured beyond the protective ring of the Alps. He was resident in Milan for almost a decade, but moved for greater security to Ravenna at the end of 402, marking a definitive change in the regime style of the western empire (see pp. 95–101). At the same period the residence of the regional praetorian prefect was moved to Arles in Provence, at the southern end of the central Gallic highway. Arles had become a ruler's residence in the early fourth century, both for Maximianus between 307 and 310, and for Constantine between 314 and 316. A hippodrome built at this period incorporated an obelisk of Egyptian granite, a clear mark of a major imperial center.⁸⁶ Arles was memorably described as Gallula Roma Arelas, a small Gallic version of Rome (Ausonius, Ordo Urbium nobilium 10). The success of its riverine harbor depended partly on its commercial role, but much of the activity which took place there has been related to the Roman annona system. This is shown in the brief account of Gaul by the mid-fourth-century author of the Expositio Totius Mundi, whose description focused on its two imperial cities:

It is said that [the province of Gaul] has a very great city called Trier, where, it is said, the emperor lives; this is in the middle of the land. It also has another city which comes to the aid of Trier in all things, it is situated by the sea and is called Arles. It receives the merchandise of the whole world and sends it on to the previously mentioned city. (Exp. Tot. Mundi 58)

The fortunes of Arles were therefore intimately linked to those of the western empire.⁸⁷ The fortifications were rebuilt in late antiquity, probably during the fifth century as a defense against the Visigoths. The walls were evidently the result of local efforts at self-help, and the size of the defended area of the city, around seventeen hectares, placed Arles among the smaller towns of Gaul, much reduced from the settlement in the first three centuries.⁸⁸

This was part of a general pattern in fifth-century Gaul, a region whose economic resilience lay in the countryside. By the beginning of the sixth century the region between the Rhine and the Loire was controlled by the Frankish kingdom of the Merovingians. Franks had been a presence in the Rhineland during the third and fourth centuries, and clearly began to settle in numbers south and west of the Rhine as the frontier disintegrated at the beginning of the fifth century. Through the early fifth century their settlements spread until they became political masters of the region. The tomb of the first king of the dynasty, Childeric, who died around 481, was discovered at Tournai in the seventeenth century, and the grave goods, including imported coins from the east Roman kingdom, establish a benchmark for our grasp of the material culture of the Frankish nobility (see pp. 226–30).

A recent survey of the evidence on the ground suggests that the evolution of a new Frankish polity in northern France was gradual not dramatic. Archaeology has conclusively refuted the traditional belief that the regional economy in late antiquity depended on large estates, owned by the wealthy few and worked by an increasingly depressed tenantry.⁸⁹ In contrast to the eastern empire, only one large landholding, comprising several rural settlements enclosed by a land-wall, has been identified in northern Gaul, near Trier. The larger houses of the Roman elite, with evidence for reception rooms, galleries, bath-houses, and heating systems, are no longer found after 300, but peasant life continued uninterrupted. Dispersed villages, often in forested country and built from wood rather than stone, were the most important settlement type. It is difficult to distinguish settlements of the incoming Franks from those of the existing inhabitants, although the appearance of a new settlement type extending from the Rhine to Normandy and comprising groups of large rectilinear wooden longhouses, resembling cattle sheds in design, may be associated with the Germanic peoples.⁹⁰ During the later fourth and through the fifth centuries the number of identifiable sites was reduced, and imports of fine pottery from the Mediterranean production centers of northern Italy, southern Gaul, and Spain were not resumed until the sixth and early seventh centuries, when some African and eastern products occur. However, the local pottery manufacturing centers of L'Argonne and Mayon continued until the end of the sixth or seventh centuries, implying that the trading economy did not collapse but was re-oriented around local consumers, who had replaced the Roman army as the main market. Trade and exchange at a local level certainly facilitated the mixture and fusion of former provincials and the incoming Franks. By 500 the decline in population had bottomed out and most of northern France established a sustainable, largely self-sufficient rural economy.⁹¹

This was the social and economic underpinning of the Frankish kingdom. The Frankish kings dispersed their power by maintaining royal residences in some of the old centers, notably Metz, Trier, Soissons, and Paris, and these were to emerge again as significant centers in the later Merovingian period.⁹² The evolution of Metz between the fourth and seventh century is the best illustration of this process. The walls of the Roman town enclosed a defended area of about seventy hectares. Bath houses and churches were being constructed up to the late fourth century, and Metz's prosperity is probably due to its links via the river Mosel and by road with Trier. However, excavations at different sites through the city all point to a break at the beginning of the fifth century, sometimes marked by fire damage, and this may be linked to a similar collapse of the rural villa system. The downturn is surely connected with the barbarian incursions after 407 and the imperial retreat from Trier. A low point in this process came in 451, when Metz was sacked by Attila's Huns, with only the cathedral of St Stephen being spared (Hydatius, Chronicle 100–1; Greg. Tur., Hist. 2.6). Halsall, in a recent study, argues that many of the inhabitants of the rump of Roman Metz lived outside the walls south of the former civic center, but the main church remained protected. The city did not revive until after 560. It was chosen as a royal capital by the Frankish kings, and their impact can be seen in energetic church building (some directly due to royal foundations), aristocratic burials, and the introduction of important ritual processions which were the occasions for competitive aristocratic display. By this period traders were again active (Greg. Tur., Mir. 4.29) and traffic along the Mosel with Trier had been resumed. The political evolution of the Frankish kingdom and the behavior of the Frankish aristocracy appear responsible for this economic upturn which continued into the seventh century and the later Merovingian period.⁹³

To the north and beyond the English channel, the picture is less rosy. The Ardennes region of Belgium and the Hunsrück between Cologne and Trier, for instance, appear largely deserted at this period. In fifth- and sixth-century England agricultural settlement was significantly poorer, and the absence of wheel-made pottery suggests that a market economy hardly survived in the post-Roman period.⁹⁴ The complex and detailed evidence for the break-up of Roman society and the emergence of Anglo-Saxon settlements in Britain does not mask the fact that the “fifth century saw a dramatic collapse in the sophistication of material culture on the island,” and the “extreme weakening of its economic, political and tenurial structures.”⁹⁵

The pattern in southern Gaul was different, particularly in the areas with ready access to Mediterranean ports and in regular contact with Italy. The countryside had been dominated by large landowners, closely linked by tradition, family ties, and shared culture to the senatorial aristocracy of Italy. By far the best-known representative of this class was Sidonius Apollinaris. The southern Gallic landowners were the first provincial group outside the imperial government that had to work out a modus vivendi with the barbarian kingdoms during the mid-fifth century. They ceded land, tax revenue, and important strong points, especially at Toulouse and in Aquitaine, to the Visigoths, and in the Rhone valley to the Burgundians, while attempting to retain their regional influence, economic purchase, and social position. Such men were not lacking in experience and a measure of military steel. After the virtual disappearance of Roman troops, they looked to their own defense, organized local militias, and manned the walls of the southern Gallic cities.

As the political structures of the western empire disintegrated, the major families built their hopes for survival on the church. Sidonius' career is emblematic, as he transformed himself from being a seigneurial landowner, and one of the last prominent representatives of western classical culture, to become bishop of the embattled town of Clermont-Ferrand, holding out against the pressures from the Burgundian and Visigothic courts.⁹⁶ A generation later Caesarius of Arles, now schooled in the Christian ascetic tradition rather than in classical literature, came from a similar landed Roman background to become bishop of Arles, the most powerful ecclesiastical position in southern Gaul.⁹⁷ Gradually in this transition, great private estates evolved into church properties and monastic foundations. Merovingian Gaul, however, was not a peaceful world, as the constant feuding and warfare among its rulers makes abundantly clear. The fortunes of the class of the formerly senatorial landowners can be traced through the sixth and seventh centuries, and now two distinct careers and lifestyles presented themselves to members of the aristocracy, ecclesiastical and military.⁹⁸

Growing insecurity was certainly a factor which led to the decline of the larger villas, especially during the fifth century, but it seems likely that many of their inhabitants regrouped in the cities. The main towns remained viable centers both of local administration and of economic activity. Although their wall circuits were reduced and their populations were significantly smaller than in the early empire, Geneva, Arles, Bordeaux, Vienne, and Lyon contained numerous churches and monastic establishments. The ruling elite, now in ecclesiastical garb, were able to commandeer the resources of the countryside and transform them into public buildings. Previously inhabited areas, however, were abandoned, while larger urban houses were overlain by smaller structures and sometimes replaced by intra-mural burial sites around the churches (see p. 483).⁹⁹

The conspicuous exception to this pattern was Gaul's main Mediterranean harbor at Marseilles, which displays signs of economic vitality up till the seventh century.¹⁰⁰ Marseilles came into its own after the end of the fourth century. The occupied area of the city, unlike those of the other Gallic cities, seems to have grown substantially, and Christianization was very marked. The baptistery of the Episcopal church, built around 400, was over twice the size of any other baptistery in southern France and larger even than that of the imperial capital of Milan. Marseilles illustrates the rule that harbor cities had a much greater capacity for population growth than towns in the interior. Trade, commerce, and manufacture all flourished. Imports from the eastern Mediterranean and from Africa passed through the port and into the interior of France along the traditional riverine routes. As African trade with Italy may have declined during the Vandal period, the economic axis between Carthage and the north Mediterranean probably shifted west to Marseilles. Conversely, the port became the Frankish kingdom's outlet to the Mediterranean world.

The fall of the Roman Empire had a major economic impact on Gaul. After the end of the fourth century, as territory in the western provinces was surrendered to the Germanic kingdoms, the Roman tax base, deriving largely from landed income, shrunk (see p. 471). The northwestern provinces were thrown back on their own resources. The most vivid single illustration of this underlying trend relates not to Gaul but to Britain. In 410 the British civitates at first chose to govern themselves independently of the legitimate western emperor Honorius and the usurper Constantine and then, when they appealed to Honorius for troops to protect them from the Saxons, received the reply that they should look to their own defense (Zosimus 6.5.3, 10.2). The writing was also on the wall for their continental neighbors.

In Gaul, the transition to a post-Roman, early medieval world differed between the north and the south. In the north the withdrawal of permanent Roman garrisons and garrison towns along and near the Rhine frontier removed the raison d'être for the complex and sophisticated system of communications which had been created in the early empire, employing both river and road transport, and which had been the basis for most economic activity. The Franks, who soon fused with the existing settled population, put in place their own networks of exchange which were smaller scale, more self-sufficient, and based on local markets. In the south civic traditions were stronger and the towns survived, although they were now transformed into smaller fortified centers, dominated by a new ecclesiastical hierarchy. Contacts and commercial exchange with the Mediterranean were still maintained, above all through the port of Marseilles. However, the driving forces of the economy were no longer the needs of the Roman state but the demands and aspirations of the Frankish kings and of the organized church, which through its monastic foundations, growing landholdings, and expanding network of churches, was the main stakeholder in the regional economy.

Italy

Italy's relative importance within the western empire increased during late antiquity. This was due both to the practical division of the empire itself after 395, and to the loss of territory in the west and in Africa to the barbarian kingdoms. A moment of major significance came when Honorius transferred the imperial capital of the western empire from Milan to Ravenna in 402. Ravenna had been important under the empire as the headquarters of the Roman fleet in the Adriatic, and as such had been a center of shipbuilding and some commercial activity. The main harbor was at Classis, about four kilometers from the city, and the site had been chosen for defensive reasons. The marshes and canals of this coastal region provided a buffer against attack from the land, while the harbor guaranteed communications and enabled supplies to be brought in by sea. After 476 Ravenna continued to serve as the administrative center first of Odoacar's kingdom and then of Ostrogothic Italy, and in the 560s was transformed into a dependency of Byzantium, the Exarchate. The city is famous above all for its extraordinary monuments associated with the rulers who made their homes there: the Mausoleum of Galla Placidia, constructed around 450; the baptisteries respectively for the orthodox Catholic and Arian Gothic communities; the basilica churches of the first half of the sixth century, S. Apollinare Nuovo, S. Apollinare in Classe, and S. Vitale (Plate 5.6), which contain the finest surviving ecclesiastical mosaics of late antiquity; and the circular Mausoleum of Theoderic (Plate 6.1). The lavishness and quality of these buildings provides a corrective to the impression of political weakness that is implicit in all narratives of the final generations of the western empire. They bear witness to the vigor of the Ostrogothic court. Whereas the written sources, above all the Variae of Cassiodorus, portray Theoderic and his successors as rulers of Italy in the style of western emperors, operating in close concert with the Italian senatorial aristocracy, the art of Ravenna shows close links to the cultural world of the eastern empire and the court at Constantinople.

Ravenna depended on a northern Italian hinterland, which comprised the Po Valley to the west, and Istria, across the Adriatic, to the east. Cassiodorus described the latter as Ravenna's Campania, “the storeroom of the royal city, covered with olives, glorious for its corn, rich in vines, where all crops flow in desirable fertility, as though from three udders generous in their milk.” The sea produced fish and shellfish, but also substantial tax revenues which were collected both in gold and in kind (Cassiodorus, Var. 12.22, trans. Barnish). Most of the produce which was consumed in Ravenna must have been brought in by sea through the port at Classis, and excavations here are an important index of Ravenna's rise and decline. There is no evidence for commercial activity after 700.¹⁰¹

The overall picture of Italian settlement in late antiquity suggests that rural settlements became fewer and more impoverished between the third and seventh centuries.¹⁰² The evidence of several recent rural survey projects in central Italy suggests that the numbers of sites occupied around 500 were only 20 percent of the figures for the first century. Smaller sites in particular virtually vanish from the record.¹⁰³ By the seventh century the lack of coinage in circulation, especially in northern regions, suggests that the market economy had in large part collapsed.

On the other hand the population and economy of southern Italy appears to have held up. No comparable decline in settlement numbers is apparent in Apulia, Basilicata, and Lucania, and the cities of Lucania and Bruttium were still going concerns in the early sixth century.¹⁰⁴ This was due to the growing contribution made by south Italy to the needs of Rome itself. Rome's economic horizon had begun to shrink during the fourth century, as Constantinople acquired the status of an imperial capital and received imports, especially from Egypt, that had previously been shipped to Italy. Under the Vandals the city no longer received tax grain from Africa. It drew instead on local resources and there was a huge traffic in food staples from the Italian countryside to Rome, whose population still approached half a million in the mid-fifth century. The implications of Rome's continuing food requirements and their impact on the economy especially of southern Italy has been explored in a remarkable study by Barnish. Legal sources, literature, and archaeological evidence can be combined to produce a picture of complex and vibrant economic activity with wide social ramifications. These factors help to explain why southern Italy remained prosperous when areas to the north were in sharp decline. The recent excavations of villa sites in Samnium and Lucania demonstrate that they were heavily involved in pig-rearing, unquestionably to supply the Roman market, where pork had been added to the grain ration of the early imperial period. During late antiquity wine production in southern Italy appears to have outstripped the output from Etruria, Latium, and Campania, which had traditionally supplied early imperial Rome. Amphoras of the type Keay LII, which occur in considerable quantities between the late fourth and late fifth centuries at Rome and Marseilles, as well as at other coastal sites of the western Mediterranean and the Adriatic, were produced in Calabria. They suggest large-scale wine exports.¹⁰⁵

Between the fourth and sixth centuries these developments particularly benefited senatorial families, which had consolidated their wealth in great estates, particularly in the south and in Sicily. The influence of such families, and their importance to Italy's rulers, emperors and kings alike, is best illustrated by a famous letter, penned by Cassiodorus in the name of Theoderic, which asked the Senate to confirm the rank of patricius on his own father after he had served as governor of Sicily, the province that virtually adjoined the family's main estates in Bruttium. The letter refers to the achievements of the honorand's father and grandfather as well. His grandfather had defended Bruttium and Sicily against the Vandals around 440. His father had been a member of a delegation to Attila acting on behalf of Aetius, but had subsequently retired from public service to the family estates in Bruttium. Another branch of the family had distinguished itself in Constantinople in the later fifth century (Cassiodorus, Var. 1.4). Cassiodorus himself retired from service in 537, and transferred from Ravenna to Constantinople at the end of Belisarius' campaign in 540. He returned to Italy after the peace of 554 to his family estate at Squillace in Bruttium, now to create a monastic retreat amid a community of monks and a library of Christian literature. His retreat from the challenges of secular politics mirrors that of the emperor Justinian in his later years.

A serious down-turn in the villa economies only seems to set in after the first quarter of the sixth century. This was due to a complex interplay of factors, including the climatic disasters of the 530s and 540s, the declining population and market in Rome, the impact of the Roman invasion of 536 and the Gothic wars, which devastated Italy for nearly twenty years up to the mid-550s, and perhaps above all the ravages of the plague (see pp. 412–6 and 481).¹⁰⁶

Italy now became a frontier province of the east Roman Empire. The Longobards (Lombards) invaded Italy in 568 and began a lengthy struggle for control with the east Roman exarchate. Their impact may be compared with that of the Berbers in Africa at this period. Unlike the Ostrogoths, who had taken control of Italy with Roman blessing and exploited the Roman administrative framework, the Longobards had no stake in the Roman Empire or interest in perpetuating it in a modified form. Italy was divided into a patchwork of protectorates between the newcomers and the Romans. There are many indications of discontinuity on city sites, even when these continued to be occupied. Wood replaced stone as the main building material in many areas; imported pottery was entirely supplanted by local coarse wares; population levels certainly dropped. It is a matter for debate how far these indications should be seen as symptoms of absolute economic and social decline, or of a radically altered social system. It is hard to dispute that most identifiable traces of the political and economic structures of the ancient classical world had now disappeared.¹⁰⁷ But this is not quite the end of the story. The Latin language was not supplanted by a Germanic dialect, the Lombard aristocracy in the eighth century had tombstones carved that transparently mimicked Roman gravestones of the early imperial period, and Lombard leaders were buried in Roman-style sarcophagi.¹⁰⁸ They, like the Burgundians, continued to assert a claim to a form of Roman identity after Roman political power had collapsed.

The Danube Region and the Balkans

The Balkans under late Roman rule was the least developed and least classical part of the Roman Empire. The mark left by Rome was nevertheless clear and took a familiar military form. Along the Danubian frontier itself there were legionary fortresses, smaller military installations, and a frontier road. There were bridges and other major engineering works along the Danube and its tributaries. The Black Sea fleet patrolled the Danube's lower reaches, and freight was transported along the river. The main east–west trunk road from Constantinople ran through Adrianople, Serdica, Naissus, Viminacium, Singidunum, Sirmium, Mursa, and Poetovia to Aquileia in northeast Italy. This route witnessed all the major troop movements and the engagements of the civil wars between Diocletian's victory over Carinus in 285 and Theodosius' defeat of Eugenius in 394. These wars were on a larger scale than any of the confrontations between Romans and barbarian groups in the region, and their regional impact was correspondingly very large.¹⁰⁹The via Egnatia, from Constantinople to Thessalonica and across northern Greece to Dyrrhachium, which had been built in the second century BC, played a subsidiary role in imperial communications in the third and fourth centuries, but became more important in the fifth and sixth centuries, as the central Balkan region became insecure. These roads and military installations were the direct product of imperial policies, decisions, and initiatives.

Civic life, however, remained seriously underdeveloped away from the eastern Aegean, Propontic, and western Black Sea coasts. Between the first and third centuries there were no more than sixty-nine municipalities and colonies in all the Rhine and Danube frontier provinces between the North Sea and the Black Sea.¹¹⁰ Local aristocracies only sporadically emerged to dominate provincial society through their continued control of landed wealth. There is little evidence for large landed estates, and the Balkans, with the marginal exception of Dalmatia on the Adriatic, produced few Roman senators.¹¹¹ Military officers and soldiers were the most powerful element in society, and it was not unusual for such men to serve as magistrates in local towns, many of which had grown up in the immediate neighborhood of military camps.¹¹²

The archaeological study of settlements in the Danube region is less developed than in Gaul or Italy. Most work has been done in the provinces of Valeria and Pannonia Prima (modern Hungary), where the late Roman remains, notably a series of rectangular, fortress-like enclosures of the later fourth century, attest a general climate of insecurity. Their walls, strengthened by numerous round towers, protected stone-built granaries, churches, and slighter structures. They have been variously interpreted as fortified refuges for the local populations, an interior frontier line behind the Danube limes, or secure collection points for the imperial annona.¹¹³

In the later Roman Empire the two most important urban centers were Sirmium in the West and Thessalonica in the East, both of which were used as headquarters by the praetorian prefects of Illyricum. The defenses of Sirmium were improved by Petronius Probus during Sarmatian and Quadic attacks on Pannonia in 373. Ammianus' description indicates that the civic building of a theater was sacrificed to allow the town defenses to be improved:

He cleared out the moats, which were choked with rubbish, and indulged his native taste for building by raising the greater part of the walls, which had been neglected and allowed to decay owing to the long peace. Even high towers with battlements were now erected, and the works were quickly finished, because he found that materials collected some time before to build a theater would suffice for his purpose. (Ammianus 29.6.11, trans. Hamilton)

Thessalonica, which was accessible by sea from Constantinople, proved a more secure base than Sirmium. Under the tetrarchy it became Galerius' capital, whose buildings included the surviving triumphal arch built to celebrate his victories over the Sassanians in 297, a rotunda intended as his mausoleum, and a palace with a hippodrome next to it. Thessalonica was the imperial base for Constantine's naval campaign against Licinius in 324 and for his campaigns on the Danube in the 330s, and became Theodosius' headquarters from 380 to 382 during his attempt to redress the disastrous situation after the battle of Adrianople. The city was a secure haven for Valentinian II and Justina when they fled to Theodosius from Magnus Maximus. As well as high-level administrative personnel attached to the praetorian prefect of Illyricum there was always a substantial garrison. In the later 440s Thessalonica definitively replaced Sirmium, which had been sacked by the Huns, as the administrative center of Illyricum, and was equipped with extensive new fortifications, accompanied by new government buildings and major churches.¹¹⁴ According to the seventh-century Life of St Demetrius, when the emperor Zeno showed signs of relinquishing imperial control over the city to Theoderic the Amal in the 470s, the inhabitants threw down imperial statues and besieged the palace of the prefect of Illyricum, seized the keys of the city gate from the prefect, and handed them over to the bishop of the city. St Demetrius' life also shows that in the early seventh century Thessalonica was the only safe haven against the barbarians, as the Balkans were lost to Roman control.¹¹⁵ A large pilgrimage church was built around the hexagonal martyr's shrine of Demetrius (reconstruction and plan in CAH 14, 959, fig. 56).

The populations of the western Danubian provinces used Latin for official and administrative purposes, but this was replaced by Greek in the east. The dividing line ran diagonally northwest from the Adriatic to the Danube, with the cities of Dyrrhachium, Scupi, and Serdica, and the coastal settlements of the western Black Sea on the Greek side of the frontier. The Danube basin itself, even in the lower stretch of the river, remained Latin-speaking throughout, thanks to the influence of the army camps and veteran settlements.¹¹⁶Military careers provided many with access to a distinctive and militarized version of Roman culture. Here were the origins of the Romanitas of Diocletian and his colleagues in the tetrarchy, and of Justin and Justinian in the sixth century. It was the Latin linguistic experience of the Balkan regions, which ensured that Latin remained the language of government at Constantinople up to the time of Justinian.

These Balkan emperors did not lose touch with their roots. Diocletian, on retirement, returned to his place of origin on the Dalmatian Adriatic coast at Split; Galerius constructed a stately home for himself at Romuliana (Gamzigrad). That was named after his redoubtable (and pagan) mother Romula, whose own name, a female version of Romulus, also precisely advertised the family's Roman culture. Both Split and Gamzigrad were designed to be simultaneously fortresses and palaces.¹¹⁷ It is significant that retirement in these cases did not mean moving to a Mediterranean city. Justin and Justinian showed similar local loyalty in creating the city of Justiniana Prima, equipped with massive fortifications and conspicuous churches, in the rural country south of Naissus where their families had grown up.¹¹⁸

Another fundamental feature both of the culture and of the settlement and economic pattern of the Danubian regions was the continual stream of immigration from north of the Danube. Larger movements and population displacements are mentioned in the historical narratives and occasionally in inscriptions, including the 100,000 Transdanuviani who were brought across the frontier into Moesia by Ti. Plautius Silvanus Aelianus to become tribute-paying subjects of Rome in the mid-first century AD (ILS 986). Many more large groups were to follow over the following centuries.¹¹⁹ Barbarians were also continually being absorbed piecemeal, some finding their way as recruits to auxiliary military units, others simply crossing the river frontier as individuals or in family groups, to find land, work, and livelihoods. The pattern of immigration surely implies that there was land enough for all. The appearance of dark burnished wares in late fourth-century sites on the middle Danube has been linked to the growing predominance of barbarian foederati in the limes forts.¹²⁰

The presence of well-armed and organized barbarian groups under strong leadership had a huge impact on the Balkan regions. The fear that such groups aroused in the cities of the region is especially brought home by Zosimus' account, derived from Eunapius, of Alaric's attack on Greece in 396.

They immediately fell to plundering the countryside and the utter destruction of the cities, killing the men of all ages and carrying off women and children in droves as well as all the wealth as booty. The whole of Boeotia and the other parts of Greece the barbarians passed through after they entered Greece at Thermopylae were so ravaged that they exhibit from that day to this the signs of their overthrow. Only Thebes escaped, partly because of the city's strength, and partly because Alaric was too impatient to lay siege in his haste to capture Athens. (Zosimus 5.5.5–7, trans. Ridley)

Athens was saved, according to the pagan historians, by dramatic apparitions of the goddess Athena on the city wall, and of Achilles at the head of the local militia, but other evidence suggests that the city was overrun.

Half a century later the most vivid literary testimony to the condition of the central Balkans is to be found in the historian Priscus' compelling eye-witness account of the Constantinopolitan embassy of 449 to Attila the Hun (see pp. 216–7). The delegation reached Naissus, deserted by its inhabitants except for the sick who sought refuge in the ruins of its churches. The party made camp a little way outside the city some distance above the river valley, which was full of the bones of men who had fallen in the recent war with the Huns. The following day they reached the camp of dux Illyrici, Agintheus, who was obliged to hand over five escapees from the Huns, to be returned by the party to Attila and certain death. After spending a night there they reached the Danube, where they were ferried across the river on boats made from hollowed logs, built for the convenience of Attila when he chose to go hunting on Roman territory. These paragraphs offer a haunting picture of Roman impotence and the utter collapse of the former frontier. What city or what fort, Attila was later to claim in an incensed altercation with the embassy, had been able to hold out against him when he was bent on its capture? The narrative of Priscus documents the fall to the Huns of Sirmium, the greatest fortress city of the western Balkans, of Viminacium, the Moesian frontier city on the Danube, and Ratiaria. Almost all the major fortified cities between Aquileia at the head of the Adriatic and Constantia on the Black Sea were overrun during the period of Attila's hegemony.¹²¹

Another written source, the life of St Severinus, written by Eugippius around 511, throws light on the upper Danubian frontier region at the end of the fifth century, as the provinces of Pannonia and Noricum were being abandoned by the Romans. Noricum covered the northeast access to Italy from the Danube. By the early fifth century the neighboring regions of Raetia and Pannonia had been given up to barbarian control. Ammianus noted that even in the early 370s the legionary fortress of Carnuntum of the early empire was a filthy and abandoned town (Ammianus 30.5.2). Noricum is one of the few areas of the Danube basin where archaeological survey can provide a range of data to match the literary record. This suggests that the local populations began to build fortified hilltop refuges, corresponding to the castella mentioned by the saint's life. Roman frontier positions along the upper Danube on either side of Vienna were maintained until the 480s. Severinus himself had moved into this region in the period shortly after Attila's death in 453, and combined the roles of a secular and ecclesiastical leader until the end of his life in 482. He founded a monastery at Favianis, on Vienna's western outskirts, and himself lived in a fortified tower, a burgus, where he was eventually buried.¹²² His life reveals the precarious security of the region. The frontier was still held by small isolated groups of limitanei, including a group of barbarian foederati at the fort of Comagenis (Life of Severinus 1–2). Their situation is reminiscent of the pockets of Roman military populations that had survived along the Rhine, as northern Gaul was gradually engulfed by Franks and Alemanni. Barbarian raiders regularly crossed the river to round up livestock or raid for food and other booty; Roman troops mounted sorties to make reprisals (Life 4.1–5). Ammianus had observed similar conditions in Pannonia a century before (Ammianus 29.6.5). The region was largely self-sufficient in agricultural produce – grain, wine, livestock – but there was some capacity for local trade. Severinus himself distributed olive oil, a notable luxury, brought from Italy. Foodstuffs were brought down the river Inn, on the western frontier with Raetia, to the Danube and thence to the Vienna region, when it was suffering from local shortage (Life 3.1–3).

Severinus maintained relations with the Rugians, the main barbarian group settled north of the Danube, but his death left the region more exposed to their attacks and his monastery was pillaged. Odoacar, now king of Italy, intervened and sent his brother Hunwulf (Onoulph) to reclaim the frontier, but this attempt ended in 488 with the evacuation of the remaining Romans and the body of Severinus himself from Noricum to Naples. The withdrawal claimed victims. A group of soldiers from the frontier fort at Batavis undertook to march to Italy to collect pay owing to them, but were ambushed and cut to pieces by barbarians on the journey (Life 20). Noricum was occupied in the early sixth century by the Rugi, followed in the middle of the century by the Baiovari, the Bavarians. Like the encroachment of the Franks in northern Gaul, this brought a definite end to Roman control.

Despite the pressures and the spread of barbarian settlements, the Romans never definitively abandoned the attempt to maintain control of the middle and eastern stretches of the Danube.¹²³ According to the Notitia Dignitatum there were eight legions in these provinces around the beginning of the fifth century, that is 8,000 men, who may have been supported by more than 40,000 limitanei and comparable numbers from the mobile field armies of Illyricum and Thrace.¹²⁴ These hypothetical figures represent troop numbers on paper, which may have greatly exceeded the true strength on the ground. It is clear, however, that the east Roman emperors, and above all Justinian, made major efforts to secure the closest frontier region to Constantinople itself. Garrisons on the lower Danube were supplied by sea, as is shown by significant finds of late Roman amphora types 1 and 2, which were used between the fourth and sixth centuries to transport wine and olive oil as part of the military annona to this area.¹²⁵ Procopius provides an impressive account of the construction of fortifications throughout Illyricum (Buildings 4). This involved strengthening the key linear defenses at Thermopylae and the Isthmus of Corinth, to protect Greece, and the wall across the Thracian Chersonese and the Anastasian wall west of Constantinople, which served a similar strategic function in Thrace. Procopius claimed that Justinian also improved the city walls of Ulpiana (modern Ljubljana), Serdica, Naissus, and Pautalia, and refurbished the forts of the lower Danube. Events were to show that these measures were insufficient to restore security to the Balkans as a whole (see pp. 442–7).