10

Society and Economy in the Mediterranean and the Near East

1.  The divergence between the eastern and western empire

2.  The Near East and the “fertile crescent”

3.  The limestone massif of northern Syria

4.  The Amuq plain and the Hauran

5.  Syrian cities

6.  Settlements of the Negev and Gaza

7.  Asia Minor and the development of civic life

8.  The economic dynamism of southern Asia Minor

9.  The militarized zones of the eastern frontier

10. Egypt

11. Africa's prosperous agricultural economy under Rome and the Vandals

12. Africa in decline after Justinian's reconquest

13. Gaul in the fourth century on the empire's front line

14. The social and economic transformation of fifth-century Gaul and the origins of Medieval European society

15. The Italian economy in the late empire

16. The Danube region and the Balkans under barbarian pressure

17. The impact of the Roman state on the economy in the East and the West

There was a profound divergence between the economic and social history of the eastern and western parts of the empire. The fundamental reason for this can be expressed in a single word – security. Until the middle of the sixth century most of the Roman East was free from war and invasion. During long periods of peacetime, urban culture thrived, the population rose both in the cities and in the countryside, and the economy benefited from long-term stability. The disastrous effects of war were only felt in the Thracian hinterland of Constantinople, and in North Syria and Mesopotamia, where the empire was exposed to Sassanian attacks in the sixth century. By contrast, the western provinces were exposed to continuous threats from Germanic barbarian groups and later from the Huns. Former provinces were successively relinquished to the invaders. This led not only to the collapse of the western empire in a political sense, but to a transformation of its economic and social structures. As the infrastructure created by Roman administration and taxation was removed, cities shrunk in size and lost their significance, and populations were dispersed into villages with narrow economic horizons. Only the institutions of the church were resilient and adaptable enough to carry forward the legacy of the empire into the Middle Ages.

The Eastern Empire

The Near East

In the late empire the Roman diocese of Oriens, the modern Near East, comprised fourteen Roman provinces. The zone as a whole was the western half of the famous “fertile crescent,” and in social and economic terms it was one of the most vibrant and flourishing areas of the Roman Empire. Its relative importance naturally increased after the empire was divided and the center of economic gravity shifted from the western to the eastern Mediterranean.

In the south the settled provinces of the Near East abutted the edge of the south Syrian and Arabian deserts. Between the permanent settlements and the true desert there is a broad zone of steppe, where the rainfall is too low to allow agriculture without the use of irrigation, but where livestock can be raised. This region was mainly exploited by small mobile populations of nomadic or semi-nomadic tribes, and settled agriculture was not possible outside the oases. Rome attempted to control this marginal region by coming to agreements and treaties with the tribal peoples, and sometimes maintained a small military presence at key locations. During the late fifth and sixth centuries their main allies were the Ghassanid Arabs, under the leadership of the Roman client Aretas (al-Harith b. Jabala).¹

The most distinctive development in the Near East between the third and sixth centuries was the spread of settlements to parts of the region which had hardly been exploited previously. Marginal areas were now brought under cultivation and the population grew. The limestone massif of north central Syria has been intensively studied. This hilly country, divided between the city territories of Antioch, Apamea, and their smaller, eastern neighbors, Chalcis, Beroea, and Cyrrhus, attracted enough rainfall to support mixed agriculture and was particularly well suited to olive cultivation. Between 700 and 800 village settlements have been archaeologically identified, with stone houses, churches, monasteries, and a small quota of public buildings. The smallest villages were occupied by only a few dozen inhabitants, but the larger ones often had several churches, and some even had public bath houses (Plate 10.1). They were occupied by well-off peasant families who drew a livelihood from a mixed agricultural economy, and especially from olive cultivation. The main market for surplus olive oil seems to have been the urban markets of northern Syria. Antioch and Apamea probably consumed most of the crop. It was no doubt the abundant supply of cheap olive oil which made it possible for Antioch to install public street lighting, so bright that “it made night as clear as day” (Ammianus 14.1.9). There is no evidence for crop specialization or monoculture, and the technology of oil production does not show the standardization found in the great oil producing regions of Tripolitania and Tunisia, but suggests simply a multiplication of individual production units.²

Plate 10.1    Serjilla. A large village in the territory of Syrian Apamea, with a bath house and an andron with two-story façade amid the ruins of substantial stone houses (Photo by Bernard Gagnon / Wikimedia Commons)

No other part of the Roman Empire provides a better archaeological perspective on the spread of Christianity to a rural area. By the middle of the fourth century single-nave churches appear in the smaller settlements, while three-aisled basilicas on a simple pattern can be found in the larger villages.³ Construction accelerated through the fifth and early sixth centuries with higher technical standards and more elaborate ornamentation.

The chronological pattern of finds suggests an initial spread of settlements through the second and first half of the third centuries, followed by a pause, and then rapid and continuous development from c.330 to 550. The steepest growth in the number and density of the settlements seems to occur during the peaceful fifth century between 410 and 480, with the graph leveling off thereafter. Calculations of the ratio between the number of rooms in the villages and the size of village territories suggest that Jebel A'la and Jebel Barisha between Antioch and Chalcis may have had populations of between 50 and 100 persons per square kilometer in the fifth and sixth centuries, reaching a total population of around 300,000 persons, a density higher than most of the rural areas of contemporary western Europe or western Turkey, and in any case presenting an extraordinary contrast with today's conditions, when these hills are almost deserted.

By the mid-sixth century the archaeological evidence suggests that the size and wealth of the villages had reached a plateau. The plague in 542, by reducing the level of population, may even have contributed to the viability of the communities, by reducing the demands on their agricultural resources. The discovery of a treasure of fifty-six silver pieces from the Church of St Sergius at the village of Kaper Koraon, all datable to the century from 540–640, with some of the best pieces clearly later than 577, indicates that some villages still possessed surplus wealth.⁴ The suggestion that life continued broadly as before into the seventh century is partly confirmed by the excavation of the village site at Déhès, which shows that the village continued to be occupied, although new building was rare after 550.⁵ It is open to question how far the excavation results indicate continued regional prosperity until the eighth century (see pp. 486–8).

The economic resilience of late sixth century Syria is also clear in the region further east, in the steppic territory east of Chalcis, and north of Jebel Balas. There had been few settlers here before late antiquity, but underground canals (qanats) were constructed to tap the water courses of Jebel Balat, and made irrigation possible. Two striking settlements emerged in the mid- or later sixth century, at Androna, where the structures include a barracks building dated to 558, and at Qasr Ibn Wardan, a complex which combines a church, rooms and stabling for soldiers, and a palatial residence dated to 564. The design and construction technique of the church show a clear awareness of developments at Constantinople. This could be the residence of a Ghassanid tribal chieftain who had responsibilities for the defense of this frontier zone, and was also a focus for cult among the recently Christianized population.⁶

As population levels increased in the limestone massif east of Antioch, it is certain that something similar occurred in the fertile and more accessible areas close to the city. A combination of satellite imagery and field survey in the Amuq plain north of Antioch has identified evidence for intense exploitation of these lowlands, as well as for irrigation channels and canals by which goods could be transported to the Orontes and into the city.⁷ The finds substantially confirm the enthusiastic encomium of Antioch's territory to be found in Libanius' speech In Praise of Antioch (Or. 11).

The settlement evidence for the region of the Hauran, the basaltic region around Jebel Druze between Damascus and Bostra in southern Syria, has a broadly similar profile to that of the north Syrian limestone country. Villages, with associated field systems and traces of hydraulic engineering to collect and conserve rainfall, spread across the region from the first century, especially in the late third and fourth and in the late fifth and early sixth centuries. There is evidence in the early imperial period that there were extensive imperial properties in the region, and the estate centers were designated metrokomiai.

While Syria as a whole shows no evidence for economic decline before the end of the sixth century, there is a significant exception in the larger cities of northern Syria, above all Antioch and Apamea. Excavations at Apamea, a huge city “where all the nobility of Syria” had lived in the fifth century,⁸ suggest that the formerly thriving city went into recession in the late sixth century. In 573 a Sassanian army led by Khusro I's general Adarmahan,

captured Apamea…which of old was prosperous and populous but which had been largely ruined by time. After taking the city on certain conditions, since they had been unable to resist as the wall was lying on the ground through age, he burnt it completely and pillaged everything contrary to the agreements; he departed and went away, after enslaving the city and the adjacent districts. (Evagrius, HE 5.10, trans. Whitby)

Procopius clearly shows that the north Syrian cities were seriously damaged by the Sassanian invasions in the sixth century (see pp. 141 and 429–32). Archaeology appears to confirm that they never fully recovered from this.

The cities of southern Syria, which were not the target of Sassanian attacks, were thriving centers until the middle of the sixth century. Inscriptions suggest that Justinian undertook to build the walls, an aqueduct, and a church at Bostra, but none of this work is dated later than 540 (Inscriptions grecques et latines de la Syrie XIII.1, 540, 541). Damascus and Hama were both major urban centers through the sixth century. The cities of the Decapolis to the south probably reached their highest population levels in the first half of the sixth century. A recent survey shows that in all cases the pattern of urban living was not seriously disturbed either by the conflict between Rome and Persia at the beginning of the seventh century or by the Arab conquest and the coming of Islam. In this period the population of Beth Shean (Scythopolis) has been estimated at between 30,000 and 40,000 inhabitants, a level that could only be supported by a very complex and active economic infrastructure. The classical urban character of the city did not begin to change substantially until the late seventh and first half of the eighth centuries, a process which was finally sealed by a major earthquake in 749 (Plate 10.2).⁹ Christian communities continued to build new churches into the seventh century. Dated mosaic inscriptions testify to new church construction at Gerasa in 559, 565, 570, and 611, and there is a final building inscription from 629. However, archaeological evidence suggests that some parts of the site were abandoned in the sixth century, and the rash of seven new churches built in the sixth or early seventh century may be more a testament to local piety, or even to anxiety in the face of threatening times, than to prosperity (see p. 488).¹⁰ A small early mosque has been identified at Gerasa and a large one at Amman (Philadelphia), but in general the impact of the Islamic conquests on the physical structures of the cities was slight.¹¹

Plate 10.2    Scythopolis in Palestine. The end of the ancient city was caused by the earthquake of 749 (S. Mitchell)

A series of small market towns, including Elusa, Nessana, and Mampsis, all equipped with substantial churches and fortifications, developed in the northern fringe of the Negev region, previously a marginal region. Irrigation systems enabled the semi-desert environment to be cultivated, producing cereals, fruit crops, and vines. Nessana is remarkable for a find of papyrus documents, written in Greek and Arabic, which date to the seventh century, immediately following the Islamic conquest of Palestine. These include marriage documents, land titles, contracts of sale and other merchant records, evidence for the handling of church property, as well as documents relating to the poll tax and military requisitioning (see p. 474).¹² An archaeological counterpart to the papyrus finds from Nessana is the nearby site of Shivta, perhaps to be identified as the ancient village of Soubaita, where the state of preservation of the ancient buildings is remarkable even for this region (see Map 10.1). They date from the fourth to the eighth centuries, and the effect of the Arab conquest is marked not by destruction but by the building of a small mosque beside the southernmost of the site's three large churches in the eighth or ninth centuries. Two large pools next to this church and its central placement suggest that this was the focal point of the community. The settlement was a dense honeycomb of houses clustered in irregularly shaped but roughly even sized blocks, which were separated by streets three or four meters wide. Livestock pens encircle the village. About 2,000 inhabitants are estimated to have lived in the settlement, where 170 houses have been identified, mostly substantial stone structures with about seven rooms grouped round a central courtyard. The houses were mostly evenly sized between 250 and 550 square meters, implying a generally homogeneous population whose livelihood depended on livestock, cereal agriculture, vine and olive cultivation.¹³ The chronological evidence makes it hard to determine whether the quality of life was sustained through late antiquity beyond the Arab conquest or whether decline set in during the later sixth century (see p. 487).

Map 10.1    Plan of the village of Shivta (Negev) (Y. Hirschfeld)

These small towns and large villages lay on the overland route, popularly known as the “spice road” and followed by Roman roads which are still well preserved today, that linked Petra and the Gulf of Aqaba with the Mediterranean port of Gaza. Gaza was one of the most thriving commercial and cultural communities of the Levant.¹⁴ There is no reason to doubt that the prosperity of the communities in the northern Negev was dependent on their ability to sell surplus produce in this major trading port. To judge from finds of amphoras, wine exported from Gaza was a major commodity in eastern and even some western Mediterranean markets in the fifth and sixth centuries.¹⁵ In general the communities of southern Syria appear to maintain their levels of prosperity through to the Arab conquest and the first century of Islam, although population levels may have been reduced by the plague.

The cities of the Orontes valley, especially Antioch and Apamea, suffered heavily from natural disasters and the impact of the Sassanian invasions of the second quarter of the sixth century, from which they never fully recovered. Perhaps too they were politically and socially more closely tied to Constantinople and more dependent on imperial support than other cities of the region. By contrast, most of the other major cities of the central and southern Levant from Aleppo to Gaza, including Damascus, Bostra, Jerusalem, and the cities of the Syrian Decapolis, provide significant evidence of settlement continuity between the late Roman and early Islamic period. There is considerable scholarly debate as to whether the cities declined after the mid-sixth century (see pp. 484–90). Certainly there was a change in urban functionality. The wealthy members of society continued to make these cities their homes, but the public spaces characteristic of Roman civic planning are replaced by a denser pattern of houses, smaller streets, shops, and workshops. Large mosques were built in Aleppo, Hama, and Damascus by the mid-eighth century, but the most striking characteristic of these cities in their post-Roman form was the intensification of artisanal activity in their central areas, and this illustrates, to use the title of a famous study of the period, the transformation of the urban centers from Greek polis to Arab madina.¹⁶

Asia Minor

Asia Minor, divided between the dioceses of Asiana in the west, Pontica in the center and northeast, and Oriens in the southeast, was part of the secure heartland of the eastern Roman Empire. By the early imperial period, the old provinces of Asia, Bithynia, and Lycia were among the most urbanized and economically successful parts of the Roman world. Like Africa, most of the region remained unaffected by war, although an exception must be made for the southern mountainous region of Isauria in the later fifth century. Social and economic conditions had favored the emergence of stable provincial elites, families which owned much of the land and expended their wealth in a sophisticated and ostentatious urban culture. Even modest cities contained an impressive range of public buildings. Athletic and cultural festivals were a particularly notable feature of life in the third-century cities.¹⁷

During late antiquity city culture can be studied more closely in Asia Minor than in any other part of the Roman Empire, thanks to an abundance of inscriptions and the archaeological remains of buildings. Several trends can be identified in late antiquity which point to the differences between the great flowering of cities in the second and third centuries and under the Christian empire that followed. The profile of public building was naturally affected by religious change. No new temples or sanctuaries were founded after the end of the third century and major centers of pagan worship fell into disrepair. In functional terms they were replaced by churches (and to a lesser extent by synagogues for the numerous Jewish communities), but the transition was a gradual one. The earliest church that is epigraphically attested in Asia Minor was built by a survivor of the Diocletianic persecution, the Novatian bishop and a member of a prominent local family, M. Iulius Eugenius of Laodicea Catacecaumene, in a part of central Anatolia which became largely Christian at an early date:

Having held the office of bishop with much honour for twenty-five full years, and built up from its foundations the entire church and all the adornment around it – that is the colonnades, and the quadrangles with colonnades, and the paintings and the decorated screens and the water supply and the gateway with marble decoration – and having in a word constructed everything, and renouncing the life of men, I made for myself a tomb building and a sarcophagus on which I arranged for all these things to be inscribed. (MAMA I, no. 171)

This is a remarkable text. Apart from providing one of the most valuable documents of the persecution, it is a rare example of a text that shows a church leader taking credit for building a church in the manner of an old-style city benefactor. Subsequent church building inscriptions, which are preserved in large numbers from Asia Minor and Syria, suggest that the costs of construction were usually the result of a collective effort by clergy and laymen. Bishops frequently took the initiative for these buildings but did not present themselves as seigneurial benefactors to their communities. Eugenius of Laodicea, however, lived in the age of Constantine and still retained much of the outlook that had sustained civic life in the early empire. He was as ready as the emperor was to claim personal political credit for his actions, although these were now directed towards the institutions of the church. A more explicit Christian ideology, based on personal humility and acts of charitable giving, did not become well rooted until later in the fourth century.¹⁸

It was not until the last decade of the fourth century that existing buildings in the city centers began to be taken over and converted into churches. The evidence seems to be concentrated in two periods in particular, at the end of the reign of Theodosius I in the early 390s, when imperial legislation against pagans and heretics was particularly strident, and in the last decade of the reign of Theodosius II, reflecting the heightened religiosity of the court and the strong influence of the emperor's sister Pulcheria.¹⁹ At Ephesus the major cathedral church of St Mary and the large baptistery next to it were built at the end of the fourth century, adapting the foundations of a large market basilica which had fallen into disrepair.²⁰ At Sagalassus, the metropolis of Pisidia, two large civic temples were literally turned inside out in the 440s to become the city's largest churches. The walls of the cella were re-erected along the line of the external colonnades, and the columns were placed inside the enclosed structure to divide the naves and support the roof. Additional building material was brought in from other Roman buildings, including the imperial temple of Antoninus Pius, and these spolia were systematically re-used for the new constructions.²¹ A similar large-scale conversion occurred at Aphrodisias, also probably in the 440s, when the columns along the short east and west sides of the temple of Aphrodite were relocated to extend the long colonnades.²² It is important to emphasize that such reconstructions represented major, well organized civic undertakings, and should not be seen as the random pillaging of available building stone.²³

The rising force in late Roman society were those who made careers in the imperial government service. Such men avoided local liturgies and tax demands, and provided access to cash salaries and other opportunities for enrichment. It is overwhelmingly men in this category who became civic benefactors. The cities of Asia Minor and Greece in the fourth and fifth centuries produced numerous verse inscriptions composed in honor of imperial officials, usually provincial governors, who organized and funded public building or provided other forms of benefit for the cities where they served.²⁴ One consequence of this trend was that urban prosperity and visible evidence for organized public life in the later empire was increasingly focused in the more important centers, provincial capitals such as Ephesus, Sardis, Ancyra, Aphrodisias, and Side.

In addition to church building there was much other new construction of a secular nature. Most cities had put up fortifications either in response to the Gothic and Sassanian threats of the mid-third or during the fourth century. There is also ample evidence for new aqueducts, bath houses, and public fountains. Roman bathing culture flourished throughout the period, although it had distanced itself from previous associations with the pagan Greek culture of the gymnasium. Theaters and stadia were now used for new types of entertainment. Horse racing and pantomimes remained major attractions not only in the great cities such as Constantinople and Antioch, but also in provincial capitals such as Aphrodisias in Caria.²⁵ There was a notable boom in the large cities of the lower Maeander valley in the hinterland of Ephesus and Miletus, including Magnesia on the Maeander, Tralles, Nysa, and extending upstream to Laodicea on the Lycus and Hierapolis. The colossal bath building at Tralles, built from spolia in late antiquity, is one of the largest of the ancient world (Plate 4.3). The stadium at Magnesia is decorated with reliefs illustrating the events that it hosted and with inscriptions on the banks of seating which show that the contests and horse races attracted fans from all the neighboring cities as part of an entertainment culture that was as lively as that to be found in Constantinople. Laodicea and Hierapolis also witnessed large-scale church building and reconstruction of the cities' colonnaded streets. The boom in urban building and adornment during late antiquity can be seen on the grandest scale at Ephesus, the provincial capital of Asia.²⁶ It was rarely necessary to erect buildings a novo, but existing structures were renovated, adapted, and enlarged. The theater at Pisidian Antioch, for instance, was more than doubled in size at the beginning of the fourth century when the city became metropolis of the new province of Pisidia, and an arch, dedicated to the imperial college of 311, was built at the entrance to a tunnel which took one of the city's main streets underneath the wing of the enlarged auditorium.²⁷ Smaller structures such as odeia were also maintained. Many aspects of public cultural activity were transformed in the fourth and fifth centuries, mainly due to the religious change from paganism to Christian monotheism, but communal life continued, and urban populations had to be entertained as well as fed.

The major regional economic shift to be seen in late antiquity is towards the coastal regions of south Asia Minor. Large new urban settlements of stone-built houses and substantial churches were created along the southern Lycian shoreline from Finike to Kaş, notably on the island of Kekova and at the harbor towns of Tristomon (Üçağiz) and Andriake. Access to these settlements was largely from the sea and they owed much of their prosperity to their position on the shipping routes from Alexandria, Syria, and Cyprus to the Aegean and Constantinople.²⁸ A remarkable example of the impact of the empire on a specialized local economy has been observed at the Lycian coastal settlement of Aperlae. This small city's livelihood depended on harvesting the mollusks which produced the royal purple dye, an imperial monopoly and status symbol. The shellfish were bred in off-shore fish tanks, and the processed dye-stuffs were then transported by a short paved road to an anchorage about a mile away. A milestone of the tetrarchic period marked the route, further testimony that imperial attention was focused on this small place.²⁹

The growth of coastal settlements also stimulated economic activity in the interior. The sixth-century Life of St Nicholas of Sion, who lived in the time of Justinian, provides a detailed picture which illustrates the interdependence of coastal and upland Lycia.³⁰The villages, which were administratively dependent on the coastal cities, supplied them with products of the mountain country, including timber, cheese, and honey, as well as providing upland pasture and relief from the summer heat.

The most dramatic evidence for economic development in the late Roman period comes from the hilly limestone area of Rugged Cilicia, north and east of Seleuceia on the Calycadnus. The most important settlement here in late antiquity was the harbor city of Corycus. In the first three centuries Corycus remained only of local importance, but from the fourth to the sixth century the settlement boomed. It contained numerous large stone houses, churches, and bath houses, all enclosed within substantial fortifications. The cemeteries around the site have produced nearly 600 funerary inscriptions, carved on sarcophagi, which identify not only the persons who were buried there but their professions (MAMA III 200–788). There is an extraordinary diversity of urban trades, which indicate that Corycus was no ordinary harbor town but a major commercial entrepôt.³¹ The texts mention substantial numbers of olive and wine merchants, ship owners, shipbuilders, sail-makers, and potters. They were often organized into guilds. This new city did not appear in a vacuum. There is a dense concentration of late Roman building along the coastline from Seleuceia eastwards, including Corycus' immediate neighbor Elaiussa-Sebaste, which it now overshadowed. The pattern of stone-built settlements, many with large churches and private houses containing fine mosaics, extends into the hinterland.³² The economic key to the area's rapid development in late antiquity was trade, and the products which were exported were timber, wine, cheese, and above all olive oil. Olive presses are to be found throughout the hinterland of Rugged Cilicia, as they were in the Syrian limestone massif. The Cilician production centers, however, had the advantage of being near to the sea, and thus the olive oil could be transported to the eastern empire's largest market, Constantinople, where it formed part of the annona ration for the population. A late-fifth-century inscription found at Abydus in the Dardanelles names the Cilicians as one of the major categories of ship owners who were involved in transporting the annona to Constantinople (see p. 128).³³

While many cities of western and central Asia Minor exemplify the burgeoning urban culture of the later empire,³⁴ other parts of Anatolia defy this characterization. In particular the Isaurians behaved in a way which was closer to the disturbances and systemic insecurity caused by barbarians in the western empire. The Isaurians, attested in our sources from the fourth century BC until the later Byzantine centuries, occupied the mid segment of the Taurus mountain range, roughly from the river Melas above Side in the west to the river Calycadnus in the east. They encroached both on the central Anatolian plateau around Konya to the north, ancient Lycaonia, and on the coastal cities of Pamphylia and Cilicia on the south. The name Isaurian is almost synonymous in modern historical writing with the terms “brigand” or “bandit,” and a large part of the bibliography devoted to them flows from this identification. Thus the Isaurians, especially in the early Roman Empire, are discussed under the rubric “brigandage,” and much of the debate about their historical significance is also a discussion of banditry.³⁵ However, in the perspective of the late Roman world, the Isaurians look less like regional highland bandits and more like non-Roman barbarians in the mold of Goths or Vandals. This is especially evident in the last quarter of the fifth century when an Isaurian, Zeno, became emperor from 474 to 489, a highly contentious reign in which his tenure of power was determined by military support either from his Isaurian kinsmen (who also provided his main challengers) or from the rival Gothic groups led by Theoderic Strabo and Theoderic the Amal who dominated the power struggles around Constantinople in the 460s and 470s (see pp. 125–8).

During the 350s and 360s the Isaurians staged a series of attacks on the settled areas both in the region to the north of their mountainous strongholds and on the cities of the south coast, including Side and Seleuceia on the Calycadnus, which were garrisoned by Roman troops. One Roman response was to build fortifications, including the sea walls of Anemurium in 382 and defensive fortifications to protect the sanctuary of St Thekla outside Seleuceia, a major destination for pilgrims. Additionally military strongpoints were built in the mountains themselves in the late 350s.³⁶ However, these attempts at containment failed, and in the first decade of the fifth century the church historian Philostorgius, in a passage supported by other sources,³⁷ records that Isaurian raids extended across Asia Minor and the Levant as far south as Palestine:

Besides these calamities, the tribe of the Isaurians inflicted several disasters on the Romans. For in the East they overran Cilicia and the neighbouring parts of Syria, and not only what is called Coele-Syria, but all that tract which stretches on till it joins Persia. But towards the north and north-west they invaded Pamphylia and laid waste Lycia. They also devastated the island of Cyprus, and likewise carried off the Lycaonians and Pisidians into slavery; and having driven the Cappadocians out of their settlements, and taken them captive, they pushed on as far as Pontus, and treated their captives far more savagely than was customary among the other barbarians. (Philostorgius, HE XI.8).

At an immediate level the counter to the Isaurian threat was to build fortifications and keep the plunderers from the gates. The point was simply made by an unexpected source, Jerome, who attests this response as far afield as Jerusalem:

There has been a sudden raid of the Isaurians; Phœnicia and Galilee have been laid waste; Palestine has been panic-stricken, and particularly Jerusalem; we have all been engaged in making not books but walls. (Jerome, ep. 114.1)

Many cities of Asia Minor, including Aphrodisias and Sagalassus, built walls between the mid-fourth and early fifth century, which probably offered adequate protection, although rural settlements were exposed.³⁸

The specific threat posed by the Isaurians raises the wider issues. Their strongholds were far from major imperial centers such as Constantinople and Syrian Antioch, their lands and resources were scarcely taxable, and local cities and landowners relied on self-defense, in the form not only of fortifications but of private militias to protect themselves. A law of Leo of AD 474 specifically prohibited estate-holders from maintaining bands of armed slaves, bucellarii, or indeed Isaurians to defend cities and landed property (CJust. IX.12.10). Along with other evidence this law acknowledges the incapacity of official state forces to maintain security in the recesses of Asia Minor. Roman military power remained formidable when deployed against a major adversary, whether the Huns or the Sassanians, but was impotent against a diffused threat. It seems likely that even the vaunted road system was no longer maintained at the level of the third and fourth centuries. Imperial milestones are not found along the major routes after the beginning of the fifth century, and Procopius accused Justinian of neglecting the empire's interests when he reduced the provision of post-stations, and the numbers of mounts available on all routes except the road to the Persian war front (Procopius, Secret History 30.10).

Even though Isauria was brought back under Roman control by a laborious seven-year campaign after the death of the emperor Zeno in 491, during the sixth century the remoter parts of Asia Minor began to slip from their grasp. Justinian introduced a wholesale reorganization of the provincial commands in sweeping acts of legislation dated to AD 535 drafted by John the Cappadocian, the praetorian prefect, which deal with the Pontic provinces, Cappadocia, Pisidia, Lycaonia, Thrace, and Isauria. With the curious exception of the rule for Isauria, all the regulations say that the duties of the new governors were to improve security in the face of latrones, and to enforce tax compliance (Justinian, Novellae VIII). Already in these areas much of the taxation, when levied at all, was probably collected by the landowners. The threat to imperial authority came not so much from recalcitrant local groups, as from landowners, who preferred to take the law into their own hands rather than pay taxes to a central authority which offered inadequate protection. A striking sixth-century inscription from Paphlagonia, rural northwest Turkey, documented repeated attempts to outlaw the armed mounted retainers of large senatorial landowners, who were a major cause themselves of local violence and lawlessness:

Flavius Soterios Mariano Michaelios Gabrielios Ioannes Theodors Nicetas Theodoros Bonos Eutropios Olympios Ioannes, thanks to God scribo of the sacred great palace, to the most god-loving, illustrious and most splendid land-owners, both current and future, of this city of the Hadrianopolitans in the province of Honoria. Since the master of the whole universe, protected by God and enobled by victory, has instructed my moderation through his divine commonitorium, which was also declared through my presence in the secretum of Ioannes, the most holy bishop of this city, giving me instructions on the basis of which I am to place under arrest the so called “horsemen with clubs” and those living in a disorderly fashion, and to make these liable to lawful penalties, and to take away from them their horses and weaponry and send these to the ever-fortunate ruling city, and bring it about that these people in future conduct themselves on their own properties…from now on I order you through this public declaration that all the current and future land-owners and phrontistai should have not more than five men for their service, unarmed and living on freehold property, knowing that if anyone wishes to transgress any of these measures at all and to set aside the divine orders, he shall be liable to a penalty of ten gold litrai, and not only this but he will put his life itself at risk.

The only civic authority mentioned in the document was the local bishop.³⁹ Imperial weakness in rural Asia Minor in the sixth century bears more than a passing resemblance to the similar impotence of the Roman state in mid-fifth-century Gaul.

The Eastern Frontier Zones

The nature, location, and economic impact of the Roman eastern frontier changed during the later empire. At the beginning of our period, Diocletian and the tetrarchs created an elaborate defensive system in the Near East which was typical of the early empire. Garrisons were stationed in Mesopotamia at Singara and Nisibis, on the middle Euphrates at Sura and at Circesium, along a frontier road, known as the Strata Diocletiana, which ran southwest from Sura to Damascus,⁴⁰ and from the city of Bostra through the oasis of Azraq to Dumatha in the Wadi Sirhan. The forts and milestones of the frontier were marked by grandiose Latin inscriptions dedicated by soldiers and army officers to Diocletian and his colleagues.⁴¹

Writing in the 520s, Malalas attributed to Diocletian the essential elements of the organization of Roman defense in Syria that was familiar in his own day:

Diocletian also built fortresses on the limes from Egypt to the Persian borders, and stationed limitanei and duces for each province, to be stationed further back from the fortresses with a large force to ensure their security. (Malalas 308, trans. Jeffreys and Scott)

Soon after this date Procopius records warfare between the two Saracen confederations of the Ghassanids and the Lakhmids over an area called Strata, which was surely the former frontier road built by Diocletian (Procopius, Bell. Pers. 2.1.6–7). Nevertheless, the organization of the defense of the region changed significantly between the end of the third and the sixth century. Responsibility was shared between a settled militia of limitanei and friendly tribes acting as Roman clients. By the end of the fourth century archaeology suggests that the forts along the steppic frontier southwest of the Euphrates were hardly manned by regular troops, although some reinforcement occurred during the early sixth century as open warfare was resumed against the Sassanian Empire.⁴²

Justinian entrusted responsibility for the defense of the northeastern frontier to a magister militum per Armeniam. Procopius records the measures which he applied in the mountainous hinterland of Trabzon, occupied by the Tzani; an account which has been counted as “the best description we have of what the arrival of Roman imperialism looked like on the ground for any moment in Roman history.”⁴³

Tzanica was a very inaccessible country and altogether impossible for horses, being shut in on all sides by cliffs and for the most part by forests, as I have said…Accordingly he cut down all the trees by which the routes chanced to be obstructed, and transforming the rough places and making them smooth and passable for horses he brought it about that they mingled with other peoples…After this he built a church for them in a place called Schmalinichon, and caused them to conduct services and to partake of the sacraments and propitiate God with prayers and perform other acts of worship, so that they should know that they were human beings. And he built forts in all parts of the land, assigned to them very strong garrisons of Roman soldiers and gave the Tzani untrammelled intercourse with other peoples. (Procopius, Buildings 3.6.9–13)

Procopius follows this general account with specific details of the deployment of troops under duces and fort building at Horonon, Charton, Barchon, Sisilisson, Bourgousnoes, Schamalichon, and Tzanzacon. All of these are unfortunately unlocalized, although Procopius' description makes it clear that some of them had already served as locations for earlier Roman garrison positions.

The main northern stronghold of western Armenia, and the headquarters of the magister militum, was Theodosiopolis (modern Erzurum), originally a hilltop fortress built by Theodosius II, but expanded by Anastasius into a city with a lengthy circuit wall.⁴⁴This, like the construction of Dara (see pp. 371–2) was a response to the invasion of Kavad in 502, when the place had fallen to the Persians without resistance. Justinian is said to have built churches and monasteries (Procopius, Buildings 3.4.12–14), strengthened the fortifications, and made the city the headquarters of his general of the two Armenias (Procopius, Buildings 3.5.1–12). The fortresses of Pheison and Citharizon respectively guarded the routes between Persian Armenia and the districts of Sophanene and Arzanene (Procopius, Buildings 3.3.1–8). Citharizon, in a strategic position southeast of modern Bingöl, was the more important site and the headquarters of one of the two Roman field commanders in the region.⁴⁵

Substantially more can be reconstructed from the history of Martyropolis–Maifaraqqin (modern Silvan). The city occupied a liminal position between the Roman Empire and the Persians, and controlled the southern access to the Bitlis pass, the only practicable route from Mesopotamia into highland Armenia around Lake Van.⁴⁶ The activities of Marutha, bishop of Martyropolis between 380 and 408, illustrate the cultural and religious interchange which became possible in the atmosphere of détente between Rome and Persia in the early fifth century.⁴⁷ When Martyropolis surrendered without resistance to the Persian invasion led by Kavad in 502, its citizens were able to preserve themselves from reprisals or deportation by returning to the king a gold chalice which had been presented to Marutha by Kavad's grandfather, Yazdgird I, and presenting him with two years' worth of taxation. Anastasius is said to have acquiesced when the Roman satrap of Sophanene, Theodorus, based at Martyropolis, was issued with his tokens of office by the Persians, since he thought that the city was indefensible. Justinian took a more robust approach and rebuilt the city walls (Procopius, Buildings 3.2.2–14).

Dara is the best studied of the eastern frontier positions (Plate 10.3). Anastasius took the decision to convert this frontier village, a day's journey northwest of Nisibis, into a fortified city in 505 and it took his name, Anastasiupolis (see p. 132). The sources for the foundation of Dara include two invaluable Syriac accounts, the near-contemporary Chronicle of Ps-Joshua, and an extended chapter of Ps-Zachariah's history. The latter, clearly drawing on a very well-informed source, shows that a discussion between Anastasius and his generals after the loss to the Sassanians and recovery of Amida (modern Diyarbakir) in 504 led to the decision to fortify this advanced position, barely ten kilometers from the frontier. The specific objective was to enable the recapture of Nisibis. The generals argued that:

It was not easy…to capture Nisibis, because they had no siege works ready or a place of refuge for rest, because the fortresses were remotely located and were too small to receive the army, and the water and food supplies that were in them were not adequate. (Ps-Zachariah, Chron. 7.6a, trans. Phenix and Horn)

Plate 10.3    The fortifications of Dara in Mesopotamia (Nihat Erdoğan, Mardin Museum)

The task of converting the village of Dara into a new fortified garrison town was entrusted to Thomas, the bishop of Amida. Imperial funds underwrote generous pay rates for skilled builders and craftsmen from the entire region, supervised by a team of the bishop's dependents drawn from the clergy of Amida. Indeed, as much attention was given to building churches as secular buildings which included fortifications, an aqueduct, a large bath house, cisterns, and granaries. The Sassanians objected to the project, which breached the terms of the treaty of 363, but their king Kavad, preoccupied by threats from the Caucasus and from the Huns on his steppic northern frontier, was powerless to intervene. Dara-Anastasiupolis henceforward became the key Roman frontier position in upper Mesopotamia.⁴⁸ The invaluable contemporary witness of Ps-Joshua meanwhile shows that the fortification of Dara occurred in the context of other defensive building work along the Euphrates, largely organized and funded by local civic leaders:

The emperor gave orders that a wall should be built for the village of Dara, which is situated on the frontier. They selected workmen from Syria, and they went thither and were building it; and the Persians were sallying forth from Nisibis and forcing them to stop. (91) The excellent Sergius, bishop of Bîrtâ-Kastra, which is situated beside us on the river Euphrates, began likewise to build a wall at his town; and the emperor gave him no small sum of money for his expenses. The Magister also gave orders that a wall should be built at Europus, which is situated to the west of the river in the prefecture of Mabbug; and the people of the place worked at it as best they could. (Ps-Joshua, Chron. 90, trans. Wright)

Procopius' lengthy account of Dara, which he knew from his service under Belisarius, exaggerates Justinian's contribution at the expense of Anastasius'.⁴⁹

This sixth-century frontier from Mesopotamia to the Black Sea differed radically from the early imperial limes along the upper Euphrates. The new strongholds of the late empire were strategically sited castles, designed to secure major routes (including access to the Black Sea) and to provide centers from which the region could be controlled. They could only be bypassed at the enemy's peril, for they cut off an attacker's lines of reinforcement, supply, and retreat. The warfare of the period between Romans and Persians did not involve pitched battles in the open plains, but assaults, sieges, and subversion aimed at securing these positions.

At a local level the late Roman frontier was porous and economic life in the frontier regions was typified by exchange and commerce, which often conflicted with imperial attempts to control cross-border traffic. This is well described by Procopius for the people of Chorzanene in the highlands of northeast Anatolia, south of Erzerum:

So the inhabitants of this region, whether subjects of the Romans or the Persians, have no fear of each other, nor do they give one another any occasion to apprehend and attack, but they even intermarry and hold a common market for their produce and together share the labours of farming. And if the commanders on either side ever make an expedition against the others, when they are ordered to do so by their rulers, they always find their neighbours unprotected. Their very populous towns are close to each other, yet from ancient times no stronghold existed on either side. (Procopius, Buildings 3.3.10–12, trans. Dewing)

This description shows local conditions in a zone where the imperial boundary cut through the region on an arbitrary basis, dividing peoples who had innumerable social, economic, and religious ties among themselves. In spite of this, the Romans and the Sassanians made a series of agreements that trade (see p. 141) between the two empires should be restricted to a small number of trading posts. In 408/9 a regulation was introduced, that all cross-border trade was to be restricted to the three cities of Nisibis in Mesopotamia, Artaxata in Armenia, and Callinicum on the Euphrates (CJust. 4.63.4; the regulation was confirmed in 422, CJust. 4.63.6). These constraints on free trade remained in force through the fifth and sixth centuries, since the fully preserved treaty of 562 between Khusro I and Justinian states simply that Roman and Persian merchants of any sort should conduct their business in accordance with the old regulations at the established customs stations (Menander Protector fr. 6; see p. 433). Procopius indicates that in the mid-sixth century the northern emporium for this long distance trade between the Roman Empire and the Far East was no longer Artaxata, but the nearby bishopric of Dubios (Dwin), eight days journey east of Theodosiopolis:

In that region there are plains suitable for riding, and many very populous villages are situated in close proximity to one another, and numerous merchants conduct their business in them. For from India and the neighbouring regions of Iberia and some of those under the Roman sway they bring in merchandise and carry on their dealings with each other there. (Procopius, Bell. Pers. 2.25.2–3)

Tolls of 25 percent were imposed on imports to the Roman Empire from the East. It is clear that the merchandise involved in these transactions were not everyday goods but the luxuries of the East – precious metals, gems, spices, and silk. It had never been an economic proposition to transport low-value goods across the vast overland distances of these Asian routes. Conversely neither Rome nor Persia had the ability to prevent localized exchange in everyday goods across the imperial boundary. There is thus no contradiction between a state policy of restricted trade and free commerce throughout the border region.

Roman interest in access to eastern luxuries is most dramatically illustrated by the history of the silk trade. Silk was produced in a region called Serinda, to be identified with southern China, and traded both overland and by sea, through numerous intermediaries, to the West. All routes passed through the Sassanian Empire and were accordingly liable to heavy levies which were charged before they reached the Roman West. Justinian's interest in southern Arabia in 529 was aimed at restricting the Sassanian stranglehold on this trade (see above, chapter 4). Eventually the secret of silk production was cracked. In 551 missionary monks, returning from India, are said to have discovered and explained to Justinian how silk was produced, and were sent back to collect silk worms so that they could be bred in Constantinople, thus avoiding the need to obtain silk, at huge price, which had passed through the hands of the Sassanians (Procopius, Bell. 8.17.1–8).