The Mediterranean World in Late Antiquity AD 395

Though on the whole in the Roman empire the maintenance of culture, government and administration depended on cities, the proportion of the population working on the land was extremely high, and the proportion of overall revenues that derived from the land was even higher. Only a very few ancient cities – Rome, Constantinople, Antioch and Alexandria – were large by modern standards, and most were extremely small. The population of Constantinople at its height in the sixth century may on a very generous estimate have approached half a million; that of Rome, perhaps more than a million under Augustus, had declined considerably by the later Roman period and was further reduced during the Gothic wars of Justinian;² in the east, only Antioch and Alexandria came anywhere near these two. The countryside accounted for by far the greatest mass of the population, and, through agricultural production, contributed the basis of most of the empire’s wealth; this remained broadly the case despite the great expansion in urban settlements starting in the fourth century, and despite an increased emphasis in modern scholarship on commercial exchange. Even if trade, or rather, production, was more important in the global economic equation than has sometimes been thought (Chapter 4), the land continued to provide the economic base; and cities on the whole, rather than being primary centres of production themselves, continued to depend on their rural hinterland.³ In recent archaeological and economic studies increasing attention has been given to country as well as city, and a growing amount of attention is being given to villages and the village economy.⁴ The change has also come about under the influence of survey archaeology, in which excavation is not undertaken but all surface finds are picked up and recorded over a given geographical area; more modern techniques such as GIS are also now becoming common.⁵ Major surveys have been conducted in widely separated regions, which focus on a given area and include all surface remains, thus taking a broad chronological sweep which can allow insights into diachronic change not possible on the basis of other evidence. The evidence thus produced may of course be much more informative for one period than for another, and there are some basic methodological problems inherent in all such surveys;⁶ nevertheless, some have produced important evidence for late antiquity. Starting in the 1950s, pioneering surveys have covered sites in Italy (the South Etruria surveys),⁷ Spain (Guadalcuivir), North Africa (Libyan valleys and Caesarea and its hinterland in modern Algeria),⁸ Cyprus and Greece (Boeotia, Melos and Methana). Differences in settlement density in the later part of the period soon revealed themselves between west (South Etruria) and east (Boeotia), but the dangers inherent in such generalization have also been pointed out.⁹ In northern Syria the pioneering work of G. Tchalenko long dominated the field, with his attribution of the prosperity observable in the substantial architecture of the 800 or so villages on the limestone massif to an olive monoculture; subsequent work points to greater diversification and the level of prosperity is still clearly visible in the dense network of remains including standing structures.¹⁰ In Cyprus surveys indicate increased density of settlement in late antiquity through to the mid-seventh century. Modern Jordan and Israel have also been well studied, and it is clear that in the now arid Negev, settlement reached a dramatic peak in the late antique period. Studies of this kind focusing on Palestine and Syria are particularly important in assessing population movement and for judging the state of these areas on the eve of the Arab conquests; they will be discussed in more detail in Chapter 8.

Studies of individual towns in conjunction with their rural hinterland mark a valuable first step, but the sites for which an integrated treatment of texts and material evidence is possible remain relatively few. One important site where this kind of work is being undertaken is Sagalassos in south-west Anatolia.¹¹ But in general where urban sites are concerned, all sorts of practical constraints dictate the course of archaeological work, especially the extent of subsequent settlement. Many major late antique cities will never be excavated, simply because they have been the site of continuous settlement ever since the ancient period; on such a site, the traces of the late antique and medieval city may now be barely visible. For similar reasons, in many other cases only small areas can be excavated. This is largely the case with the city of Constantinople.¹² Excavations took place in the Great Palace area as early as the 1930s, but the layout of the palace in its different phases still has to be largely reconstructed from difficult textual evidence; however, a fresh look at the records of the earlier excavations has led to modification of the conclusions drawn by the earlier excavators, and has contributed to the debate about the dating of the Great Palace mosaics.¹³ Further excavation has taken place in and near the Hippodrome and on the site of the great church of St Polyeuktos built by Anicia Juliana in the early years of Justinian’s reign (Chapter 3); the city walls have received attention and recent publications focus on specific neighbourhoods. The spectacular excavation at Yenikapi of the late antique harbour attributed to Theodosius, in the course of rescue archaeology connected with building a new metro system for Istanbul, has so far revealed the well-preserved remains of some forty Byzantine boats from the seventh century and later, and added an extraordinary amount of evidence, some of which is already on display in the Istanbul Archaeological Museum. An important study of the elaborate installations which secured the water supply of Constantinople, using the evidence of physical remains, literary sources, later historical and geographical accounts and the study of Byzantine masons’ marks, has also added greatly to our understanding of how the city could sustain such a large population in the late antique and early Byzantine period.¹⁴ Nevertheless, the fact that so much work in the city has concentrated on individual churches indicates another important factor operative in determining the nature of archaeological research; namely, the motivation for selection of sites. This has often been dictated by an intense interest shown in churches, their architecture and their mosaic decoration. But like the recent work on the city’s water supply, an important study of the dating evidence provided by brickstamps allows a different view of its urbanism to emerge.¹⁵ Carthage, on the other hand, provides an example of an important late antique city where major excavation was prompted in the 1970s by the threat of development and undertaken on an international scale with the support of UNESCO. During the Islamic period, the centre of settlement moved to nearby Tunis, and ancient Carthage became part of a residential suburb. Systematic excavation over a large area was therefore impossible, but teams from several different countries were assigned specific areas within the ancient urban complex. Their interests and priorities differed, and some of the sites chosen also yielded material rich in one particular period and less so in others. But taken together, it would be hard to overestimate the importance of these results (Chapter 4), not least in providing a systematic and large body of well-recorded evidence which would act as a benchmark for methodology and interpretation in other parts of the late antique Mediterranean.

Since the development of late antique archaeology as a serious study, effectively only from the 1970s, archaeological and epigraphic evidence has been a fundamental aspect of all assessments of the period. A spectacular example of what can be shown by such evidence is provided by the case of Aphrodisias in Caria (south-west Turkey), a city only sparsely attested in literary sources, which has yielded an astonishing amount of evidence from its abundant inscriptions and its excavated remains about urban development and city life in late antiquity.¹⁶ Since it was a major centre of sculpture production, drawing on famous marble quarries, it has also turned up a mass of splendid finished and half-finished late antique sculpture which is extremely important not only in the context of Aphrodisias itself but also for wider issues of iconography and style. Some of this evidence, like the literary evidence for the families of Paralius and Asclepiodotus already noted (Chapter 6), tells us much about the survival of pagan and classical culture in a provincial town; this is especially true of the striking series of sculptured heads of late antique philosophers.¹⁷ Finally, many Greek inscriptions also survive from Aphrodisias, through which we can trace the efflorescence of Greek verse inscriptions and thus the availability of training for this specialized literary accomplishment in the fifth-century east. These are only some of the results of the excavations conducted at Aphrodisias over a thirty-year period to date. In particular, the Aphrodisias inscriptions give us a virtually unbroken record of urban history from the city’s acquisition of free and federate status during the Triumviral period to its change of name in the early seventh century from Aphrodisias (city of Aphrodite) to Stauropolis (city of the Cross) and its survival as a shadow of its former self through the eighth and ninth centuries, when sources are almost absent, only to undergo some rebuilding like other Byzantine sites in the tenth and eleventh centuries. As we have seen, Aphrodisias is an important centre for our knowledge of late paganism, but here too the prominent temple of Aphrodite was converted into a church, probably in the late fifth century.¹⁸

Aphrodisias provides an example of a site with an extraordinarily rich and spectacular amount of archaeological remains, including sculptural and epigraphic material of breathtaking quality and importance. A good many of the later inscriptions are undatable, because their conventional language and style remained so constant over the period, but it is possible here, as it rarely is elsewhere, to piece together a real, if incomplete, view of changing patterns in city life in the late antique period. Some other sites offer this possibility too, each in its own way, among them Ephesus in Asia Minor and Apamea in Syria.¹⁹ But even apparently clear archaeological evidence, coins or inscriptions, and or pottery dating may be unreliable. Archaeological evidence can only tell us what happened, not why it happened, and it is only as good as the methodology adopted by the archaeologists in question permits.²⁰ It is of course tempting, in the absence of specific indications, to link certain sorts of archaeological evidence to historical factors or events known from other sources. Procopius’s Buildings, a detailed account of the building activity of the Emperor Justinian, provides a particularly good example of a text frequently used in this way. However, it omits Italy altogether, for reasons on which we can only speculate, and the ample and literary treatment given to Constantinople in book I is not carried through in the rest of the work, which in places consists only of lists of names of fortifications.²¹ Previous emperors, such as Theodosius II and Anastasius, had engaged in the building of major urban fortifications, with famous surviving examples at Constantinople, Thessalonica and Dyrrachium. However, city walls were commonly repaired and rebuilt over long periods and many of the late Roman fortifications in the Balkans and elsewhere cannot be dated from the material evidence alone. It is therefore always tempting to suppose a given site or fortification to be Justinianic work; yet closer study of the Buildings reveals that Procopius often exaggerates or misrepresents the nature and extent of Justinian’s building programme, and recent studies have suggested that some of these structures may in fact, like the great walls of Amida (Diyarbakir), have been earlier constructions, subsequently much rebuilt, or merely refurbished under Justinian. On the other hand, many of Procopius’ statements are confirmed by other evidence, so that we cannot be uniformly sceptical. A case in point is his account of Carthage and North Africa, where omissions and exaggerations make his evidence infuriating to use, but which is nevertheless clearly based on firsthand experience.²² A final illustration of the range of problems encountered is provided by the evidence (both archaeological and literary) for earthquakes and plague in the late antique and early Byzantine period. Non-specific damage to material remains is frequently attributed to a convenient earthquake; however, this may fail to take into account the fact that unless the literary sources give precise details, there is usually no way of knowing its scale – it may well have been a mere tremor. Sometimes indeed the cause of the damage is very clear, as at Scythopolis,²³ but even when major earthquakes are known to have taken place, they have in most historical periods proved a stimulus to rebuilding, often on a large scale, as can be seen in the case of Antioch in our period. As for plague, despite what seem like detailed and authentic literary accounts and references to plague in the sixth century (and later, especially in the Near East), it has proved notoriously difficult to trace its effect on the ground, and this remains a serious puzzle for historians.²⁴

With this growth in archaeological investigation, the question whether late antique cities were in decline has typically been rephrased in terms of urban change or transformation. If we put together the evidence from archaeological investigation of sites very widely scattered round the Mediterranean, a general picture seems to emerge of contraction, and of shifts in urban topography, and there is evidence from widely different regions to suggest that significant urban change was already taking place before the end of the sixth century.²⁵ A contrast has been drawn between the west, where urban life seems to have been in decline already by the late fifth century, and the east, where many, if not all, scholars hold that late antique urbanism continued to flourish well into the Islamic period.²⁶ However, the picture is not uniform, and new evidence and new interpretations are emerging all the time. There is unlikely to be a single or simple cause for these changes, even if in individual cases particular local factors may be plausibly adduced. But by the end of the period now covered, deep-seated social and economic change seems to have been taking place all round the Mediterranean, if at varying pace and for local as well as macro reasons. The rest of this chapter will investigate the process in more detail and we will have cause to return to it in the case of the eastern provinces in Chapters 8 and 9.

The rural people have cast aside their ploughshares and live like city-dwellers, exchanging their rural lifestyle for civilization.

(Buildings VI.6.15, cf. Vandal Wars I.15.31ff.)

Elsewhere Procopius lists among the standard attributes of a city, stoas, a bath, an aqueduct and lodgings for magistrates.²⁸ This model was already coming under strain when Procopius wrote. It was an urban style which had required public and private investment, both to build and to maintain its public buildings. It also implied a life of cultured leisure, or at least civic involvement, if only for the richer citizens, with a range of public activity, in the forum, at the baths, at the circus, while in the Roman period its temples characteristically looking out over the forum implied the survival of paganism.

Just such a city, named Justiniana Prima, and generally identified with Čaričin Grad (45 km south of Niš), is attributed to the initiative of the Emperor Justinian and commemorated his own birthplace in Illyricum – or so Procopius claims, for his account of this new foundation is even vaguer than it is for the new city on the African coast, simply listing some of the standard elements noted above.²⁹ Yet generalizations are dangerous, and close examination of every site and literary reference is needed. ‘Classical’ cities were far from being the norm everywhere, if they ever were. The large sixth-century site of Androna in Syria, for instance, is identified as a kome, a large village, like others in the east in this period, though it had such urban features as two sets of circuit walls, large extra-mural reservoirs, a kastron, a Byzantine bath, and nearly a dozen churches. The ‘dead cities’ of the limestone massif were also in fact large villages, though with urban features.

In the Balkans, urban life had been sharply interrupted by the Hunnic and Ostrogothic invasions, and Anastasius’ and Justinian’s programmes of restoration and fortification were mainly palliative. Despite Procopius’ claims, it seems that there was little secular urban life in these settlements by the sixth century, and building and signs of culture begin to dry up together. This is the picture at late antique Nicopolis and Philippopolis in Bulgaria.³² Military needs and provision for an increased local role for bishops now took precedence over the spacious civic structures of earlier times.³³ Circumstances were also difficult in Greece, where late sources suggest that in the late sixth century some cities, including ancient Sparta, Argos and Corinth, were abandoned by their inhabitants in favour of safer places. Archaeological and other evidence does not always confirm this oversimplified picture, but it does seem that the pattern of urban settlement was changing significantly during the later sixth and seventh centuries. At Corinth, the remaining population retreated to the fortified height of Acrocorinth, and this became typical of Byzantine settlements in Greece. This impression of a search for places of refuge is reinforced in many sites in the Balkans, where inhabited centres contracted and regrouped around a defensible acropolis, or were abandoned in favour of such positions elsewhere. The early Byzantine walls at Sparta enclosed only the ancient acropolis and not the civic centre; it was presumably hoped that they would provide a place of refuge for the population in time of attack.³⁴ The late Chronicle of Monemvasia connects the move of population in Greece explicitly with the Slav invasions of the 580s, but the extent of Slav movement and settlement is a contentious subject, and both the chronology and the archaeological record are hard to trace.³⁵ According to the same source, the population of Lakedaimon, ancient Sparta, settled at Monemvasia, a rocky crag on the east coast of the Peloponnese, very hard of access; however, the actual date of the foundation of Monemvasia is extremely obscure, and Sparta remained inhabited in the Byzantine period. The general phenomenon of population movement, if it happened in this way, was probably more gradual, and a number of different factors may have been operative in bringing it about, including possibly a shift in economic activity. Whatever the reasons for the new pressures from Avars and Slavs (Sclaveni) in the late sixth century, the empire found the threat difficult to deal with, and resorted to a mixture of diplomacy, subsidies and warfare to try to control it. Athens itself was not occupied, but both Athens and Corinth suffered attack in the 580s, which shows clearly in the coin evidence, and which caused considerable destruction by fire, and Thessalonica was besieged; the Long Walls of Constantinople were attacked more than once by Sclaveni and Avars in the same decade. These Slavs were described by the Emperor Maurice in his Strategikon:

[The Slavs and the Antes] are both independent, absolutely refusing to be enslaved or governed, least of all in their own land. They are populous and hardy, bearing readily heat, cold, rain, nakedness and scarcity of provisions … Owing to their lack of government and their ill feeling toward one another, they are not acquainted with an order of battle.

(Maurice, Strategikon, 11.4, trans. Dennis)

Although Athens was used as a base against the Slavs nearly a century later by the Emperor Constans II (662–3), new building, if any, involved subdivision into smaller rooms and the use of former fine buildings as sites for olive-presses. Similar phenomena are also encountered frequently in North Africa, and at the village of Olympos in Attica, for instance, such re-use involved a former baptistery.³⁶

The last two features, subdivision and ‘encroachment’ on the sites of former grand buildings in urban centres, can be seen in different forms in many other regions. Typically, the large houses, maintained in many areas into the sixth century or even later, are divided into smaller rooms for multiple dwelling, often with mudbrick floors over or instead of the splendid mosaics which are so characteristic of the fine houses. This can be vividly seen at Carthage, where a large peristyle house (built round a courtyard in classical style) in the ‘Michigan sector’ was subdivided into much poorer accommodation by the seventh century, and where the same smaller divisions appear elsewhere in the city. At Apamea, peristyle houses were restored after the capture of the city by the Persians in 573, and apparently maintained until the Arab conquest. But elsewhere ‘encroachment’, either by poorer dwellings or, commonly, by small traders and artisans, frequently occurs over existing public spaces, such as the forum or, as at Anemurium in southern Turkey, on the site of the palaestra. In the latter case this change of use had started early, after the disruption to the city caused by Persian invasion in the third century, and the artisanal activity in the area apparently flourished; but by the late sixth and seventh centuries the other civic amenities such as baths and aqueducts were no longer functioning. A particularly striking example is found at Sbeitla in modern Tunisia, where an olive-press, perhaps seventh-century, sits right on top of the former main street.

Figure 7.2 An olive-press astride a former main street, Sbeitla, Tunisia, probably seventh century

Burials within churches, which are common in the great basilicas of the period, indicate changed religious attitudes rather than economic pressures (Chapter 6); however, the presence of burials within central areas of the town, and even on the sites of earlier fine housing or public buildings is another common feature, seen vividly at Carthage and other North African sites in this period and indicating a major shift in the use of urban space. It has been tempting in the past to think in terms of economic necessity, ‘squatting’, and, where there is some textual evidence to support the idea, as at Carthage, of an influx of refugees from invasion in other areas. Local factors will also have been important; for instance, at Luni near La Spezia on the west coast of Italy, where the decline of the marble trade from nearby Carrara must have affected the town, and where, though it survived into the seventh century, a clear decline in material wealth can be seen from at least the sixth century. Local conditions differed: some of the major cities of Asia Minor, Ephesus and Sardis, for example, which had enjoyed a period of prosperity and expansion in late antiquity, seem to have maintained late antique civic life until the Persian invasions of the early seventh century.³⁷ Very little serious archaeological work has taken place on the Byzantine period in any of these sites, and generalizations carry certain dangers, but the phenomena are so widespread, even if the pace varies in different places, that it seems clear that a general process of urban change was going on, and that this must be connected not simply with causes such as plague or invasion, but with overall administrative and economic factors, including the relation of provincial cities to the central administrative organization.

Recent and ongoing archaeological work at sites in the western Mediterranean including Classe, the port of Ravenna, the islands, and North Africa, indicates a changing economy in these urban environments, with centres of production (kilns, olive-presses and metal workshops, for instance) now established within the former public areas, but also with new developments such as warehouses, indicating new patterns in sea-borne trade. By the seventh century new developments can be seen in coastal sites; for instance at Naples.³⁸ A coastal city such as Marselles remained dynamic in the seventh century, and benefited from the trade of Frankish Gaul as well as the Mediterranean. Western shipwrecks suggest that while trade was on a reduced scale and differently configured than previously, it was still lively. Mediterranean trade certainly continued, and eastern amphorae continue to be found at western sites in the seventh century, if not in the same numbers, while new types start to appear. Although interpreting such evidence is heavily dependent on the limited number of excavated sites, such excavations are increasing in number and some of this work, still in its early stages, seems to be pointing in the direction of a break in the eighth, not the seventh century; it thus serves as a corrective to the previously dominant resort to explanations of change in terms of ‘decline’. It is important also to stress that seventh-century Byzantium had not lost its western role, as can be seen in the 660s in the intervention of the Emperor Constans II (641–68) in Italy against the Lombards, launched from Syracuse in Sicily, and his visits to Rome, Calabria and Sardinia. Evidence from the reign of the same emperor and later also shows that Byzantium continued to hope for continued control of parts of North Africa. But there are signs that a new western Mediterranean system was beginning to emerge.

As argued above, archaeological evidence is often difficult to interpret and, in particular, difficult to link directly with historical events. But in some cases, as in that of street building at Caesarea in Palestine, there was still considerable activity going on in the later sixth century, and Justinian’s building programme included some spectacular achievements, such as the great Nea church at Jerusalem, which is shown on the sixth-century mosaic map of the city from Madaba in Jordan. In this case, excavation dramatically and unexpectedly confirmed the accuracy of Procopius’ description.⁴⁴ Major building work also took place in a number of Near Eastern cities after the middle of the sixth century, among them Gerasa (Jerash) in Transjordan, and some magnificent floor mosaics from churches in the area date from the seventh and even the eighth centuries.⁴⁵ Striking and lavish mosaics continue to be revealed from synagogues in the Near East, most recently and spectacularly from Sepphoris. Many of these synagogue mosaics contain motifs and iconography clearly recognizable from pagan contexts and drew without self-consciousness on a common artistic repertoire; while various explanations have been put forward for Jewish use of pagan imagery, the sheer magnificence and lavishness of these mosaics cannot fail to impress.⁴⁶

Arguments about prosperity versus decline are not easy to balance. It is partly a question of what indicators one uses. Whittow, for instance, argues for the prosperity of Edessa (Urfa, south-east Turkey) in the sixth century from the large sums of gold paid to Chosroes I in 540 and 544, and the quantities of silver in the city when it was captured by the Persians in 609.⁴⁷ In contrast it has been deduced from a study of settlement patterns in the region that while settlement density reached an unprecedented peak from the fourth to the sixth centuries, from the seventh century there was a dramatic fall in occupation.⁴⁸ Edessa continued as an urban centre through the Islamic period until the Byzantine recovery in the tenth century, but the silver it possessed in the sixth century does not tell us very much about the general distribution of wealth or about urbanism as such. Complex readjustments seem to have been taking place in many areas, which involved both rural and urban sites and their mutual relationships. There are, moreover, serious gaps in our knowledge due to the uneven degree of excavation and the lack of certain sorts of evidence. For reasons of local settlement, little may survive now of a place known to have been a prosperous city, while casual information from textual sources such as the Life of Symeon the Fool (for Emesa/Homs, seventh century, but referring to the sixth century), the Miracles of St Demetrius (early seventh century, Thessalonica) or the Life of Theodore of Sykeon (for late sixth-century Anatolia),⁴⁹ sometimes belies any general theory of urban decline. Even more important to remember is the fact that the picture is literally changing all the time as new evidence comes to light and existing theories are revisited. Many excavations on major sites are still continuing, and one season’s work can and does frequently modify previous results – the important site of Pella, one of the Decapolis cities in Jordan, a city extensively studied in the late antique and early Islamic periods since the 1980s by Alan Walmsley, is a case in point. Finally, a reliable ceramic typology for the Near East is only now beginning to be agreed, and it also seems likely that there was much more regional variation than previously supposed.

How far it is possible to generalize, even within these limitations, let alone across the Mediterranean, is obviously very questionable. It is worth rehearsing again some of the main factors that have been adduced by historians as agents of urban change, starting with the plague which hit Constantinople and Asia Minor in the mid-sixth century and continued to strike Syria in successive waves throughout the seventh century. Though the effects may be hard to quantify (see Chapter 5), it is hard not to think that plague must have been a factor in undermining the generally thriving state of cities in the Near East in the early part of the sixth century. But since neither epigraphic nor papyrological sources offer clear evidence of the scale of mortality, and one can make only a general connection between urban and settlement decline and the factor of plague, it is dangerous to use the plague of 541 as a dating reference in the absence of other evidence. On the other hand, arguments which seek to downplay its effects in relation to individual sites must logically be equally suspect.⁵⁰ One can find extreme variation among historians in the amount of weight that they are willing to attach to the sixth- and seventh-century plague. Yet this seems to have been the first appearance of bubonic plague in Europe and its impact should have been far greater than that of the regular diseases which ravaged ancient cities as a matter of course. To take just one example from the literary sources outside the three main descriptions of the epidemic by Procopius, John of Ephesus and Evagrius, when the plague struck their monastery, the monks of the Judaean monastery of Chariton went en masse to the ancient holy man Cyriacus, who was living as a hermit at Sousakim, to ask for his help against the disease, and brought him to live in a cave nearby.⁵¹ At present it does not seem possible to do more than leave the matter open. Second, according to the literary sources, the sixth century also experienced a high incidence of earthquakes, which in some cases can, as we have seen, be plausibly connected with the material record. But some of the literary evidence for earthquakes in the period may be attributable to increased recording of earthquakes by Christian chroniclers interested in pointing out the signs of God’s wrath rather than to a quantitative rise in their actual incidence. Third, other external factors can also be adduced for reduced prosperity in certain areas, such as a possible withdrawal of military resources, which would imply a lower level of economic demand in the region in the future, and poorer roads and communications.⁵²

How far did Christianization play a role in the move away from the civic life of classical antiquity, with its baths, temples and public entertainments? Liebeschuetz has argued that the process contributed to a decline in the old civic values, but Michael Whitby points to the role of bishops and the new Christian framework in the continuing success of many cities.⁵³ The munificence of public benefactors did not give way in any simple sense to Christian charity; the two more frequently existed side by side. Bishops fulminated against the games and the theatre, and some objected to public baths on moral grounds, but often to little effect. The great temples slowly and gradually went out of use and were often converted into churches – though not everywhere, not always without protest and sometimes only at a late date; as we saw, Christians themselves tended to make grandiose claims which were not always justified.⁵⁴ But church building on a large scale certainly changed the appearance and feel of towns, and even average-sized towns in the sixth century might contain far more and far larger churches than their population would seem to warrant; furthermore, they often went on being extended and altered after other forms of public building seem to have stopped, a feature which is strikingly exemplified by the large churches of Sbeitla in North Africa. The church and individual bishops gradually assumed more and more responsibility not only for civic leadership but also for social welfare in their communities, in the distribution of alms and maintenance of hospices and by storing food and distributing it in the times of famine which were a regular feature of ancient urban life.⁵⁵ Eutychius, the sixth-century patriarch of Constantinople, performed this service for the people of Amasea during his years of exile beginning in 565, and the early seventh-century patriarch of Alexandria known as John the Almsgiver acquired his epithet from his reputation for urban philanthropy. Other holy men and monks also performed similar roles: a story told about St Nicholas of Sion, near Myra in Lycia, tells how when the plague struck the metropolis of Myra in the sixth century, Nicholas was suspected of warning neighbouring farmers not to go to the city to sell their provisions for fear of infection. The governor and the city magistrates sent for the saint from his monastery, and Nicholas visited several settlements, where he slaughtered oxen and brought wine and bread with him to feed the people.⁵⁶ Bishops and clergy had long been involved in the affairs of cities and countryside and in negotiations with the provincial governor, as is clearly apparent in the letters of Theodoret of Cyrrhus in Euphratensis (northern Syria) in the fifth century, where villages rather than towns seem to have been the norm, while in early seventh-century Alexandria the patriarch John the Almsgiver was dealing with matters of trade and taxation alike. The presence of holy men and, in the east, of stylites, such as Symeon the Elder at Qalaat Semaan in the fifth century and Symeon the Younger near Antioch in the sixth, around both of whom substantial monasteries grew up, impacted on the rural economy (Chapter 3). Pilgrims required services, and bought local goods including pilgrim eulogiai, essentially religious souvenirs, and any important shrine or location of a notable holy man was soon surrounded by substantial and varied buildings which required technical building skills. To some extent these, and other large monastic complexes contributed to a ruralization of the economy.

While the fabric of life in both town and country had thus changed significantly with Christianization, Christianity did not itself directly bring about urban change. Rather, by stimulating church building, by diverting wealth from secular causes and by influencing social practices, it was one among a range of other factors which together converged to change and undermine the urban topography and economic organization inherited from the high empire. The shifting economic relation between the civic authorities and the church, which came to represent, let us say by the later sixth century, an actual shift of resources in favour of the latter, was an important feature of the period.⁵⁷ As a result, the church’s agents, especially bishops, took on the role of providers and distributors of wealth which formerly lay with the civil authorities. Since the role of cities within the empire had always been closely identified with finance – exchange, monetary circulation, collection of taxes – this shift inevitably had profound consequences.⁵⁸

As we saw, since the fourth century, the curiales, the better-off citizens on whom the government depended for the running of cities, had been complaining loudly about their increased burdens (Chapter 4). Both the chorus of complaint and the theme itself were of long standing, and those who, such as the Emperor Julian, the rhetor Libanius or the historians Ammianus Marcellinus and, later, Procopius, saw themselves as champions of traditional values invariably also took up the cause of the cities whose future they perceived to be under threat.⁵⁹ There were some grounds for their fears: government pressure on the curial class, who provided a convenient target for ways of increasing revenue or at least trying to ensure its collection, certainly increased as time went on. The city councils themselves faced financial difficulties, especially those with splendid buildings to keep up. Many found it difficult to keep their councils up to strength with enough curiales of adequate income. The wealth of curials also varied greatly; many of them were village landowners of quite modest means while richer ones might hold widely dispersed estates. Cities showed an obstinate tendency to survive, and two hundred years after Constantine most were still in a reasonable state, while some were more densely populated and more prosperous than they had ever been, but we now also hear of leading citizens under different terms – honorati (ex-officials), or ‘notables’ or ‘grandees’.⁶⁰ The urban elites of ‘late late antiquity’ were not identical with the old curial class, though there was no doubt over-lap. They now included churchmen, and especially in the Balkans, defence needs now loomed large on their horizon. They also included some of the successful landowners whose influence was clearly felt in the economy of the sixth century.⁶¹ The ‘notables’ were recognized in imperial legislation under Anastasius (491–518), and co-existed awkwardly with councils; by the fifth century the governor and other civic officials (defensor, curator, pater) gained increasing importance and councils gradually dropped out of participation in appointments to the latter posts in favour of the notables. Both John the Lydian and Evagrius, in the middle and late sixth century respectively, suggest that councils no longer functioned. As for ordinary people, they had no official role, but they could and often did demonstrate and express their views, as we shall see below.

The style of life which these cities had supported for so long had already begun to change in many places, and urban life was certainly drastically curtailed in the seventh century (and long before that in most places in the west). The pressures of invasion, insecurity and increased military expenditure by the central government were felt in differing degrees in the west and east and the effects of the Persian and Arab invasions in the east will be discussed later. A.H.M. Jones famously saw the increased numbers of monks, bishops and clergy as a component in the excessive number of ‘idle mouths’ who constituted a drain on the late Roman state, and it is true that these men might otherwise have been producers, or indeed soldiers, or even, as Arnaldo Momigliano thought, have lent their abilities to running the cities and the empire better. Yet such a formulation ignores the structural change that was taking place in the balance of town and country. The number and size of gifts and legacies made to churches, strikingly demonstrated in the rich silver treasures owned by small eastern churches, are not so much a sign of the prosperity of the region as of an economic situation in which local donors chose to direct their wealth towards otherwise obscure village churches;⁶² this silverware contrasts with the impressive late Roman silver found in late fourth-century Britain, which demonstrate the still surviving but soon-to-end wealth of rich Christian households.⁶³ John of Ephesus’ Lives of the Eastern Saints gives a vivid picture with much circumstantial detail of village life around Amida (Dyarbakir) in Mesopotamia.⁶⁴ Rural monasteries were also a very important element affecting the relation between town and country; they attracted donations, attracted recruits, sometimes in very large numbers, and were themselves economic units impacting on their surroundings.

We should not therefore be surprised if the rich epigraphic sources of the earlier period give way in favour of Christian funerary inscriptions, which are by comparison disappointingly brief, or hagiographic sources in the form of local saints’ lives or miracle collections. While the latter material is often rightly viewed with suspicion by historians because of its obvious bias and its tendency to conventional exaggeration, it clearly reflects the changed point of view. Several cities in the crucial period are well provided with evidence of this sort, which shows not only that they continued as vital centres but, even more importantly, how their urban life was now articulated. One or two have already been mentioned, such as Thessalonica in the early seventh century, known from the Miracles of St. Demetrius, composed by the archbishop of the city soon after AD 610 and vividly reflecting the dangers of invasion then facing the city; others include Seleucia in Cilicia, and late sixth-century Anastasioupolis in Galatia, known from the Life of St Theodore of Sykeon. Even having made all the necessary allowances, the picture that emerges from these and many comparable texts is of an urban life no less vital but quite different in kind and flavour from what we associate with the late antique city in the first part of the period, still with its municipal pride, its public spaces, its great buildings and its civic autonomy. Times had changed. A comparison has been made between cities at the end of our period and the decaying industrial towns of modern Britain. Plausible or not, at least this suggests some of the complexity of the changes that were underway.

Apart from the prejudice shown by Procopius and others, including Malalas, Menander Protector and Agathias, and the concern voiced in the anonymous sixth-century dialogue On Political Science,⁶⁷ there is no reason to think that the better-off or middling parts of the urban population were any less given to rioting than the really poor; many episodes were sparked off by hostility to individuals or passionate enthusiasm for chariot racing on the part of all classes, and, as at Constantinople and Antioch, especially by members of the ‘factions’ of Blues and Greens, the organized groups, effectively guilds, of charioteers, performers, musicians and supporters who staffed the public entertainments of late antique cities, and the wider constituency of their followers. Graffiti on seats at Aphrodisias and Alexandria vividly testify to their widespread following. Association of the Blues and Greens with urban violence began in the fifth century, and the level of instability, whether or not associated with the ‘parties’, evidently rose all over the east in the sixth century, reaching a peak with demonstrations in many cities towards the end of the reign of Phocas in the early seventh century. Historians have often supposed that this could only be explained on the assumption that the Blues and Greens were associated with particular religious or ideological standpoints, but a strong rebuttal of this theory was mounted by Alan Cameron in the 1970s.⁶⁸ More recently Liebeschuetz has reviewed the evidence again and argued that the position was not quite so straightforward; the Blues and Greens were not merely sporting hooligans, and while many of the recorded disturbances seem to have started in contexts of crowd and sporting excitement, when there were serious issues, as in the overthrow of the Emperor Maurice (602) and the turbulent reign of Phocas (602–10), the factions might well be drawn in to take sides (Chapter 9). As Liebeschuetz points out, the factions could hardly fail to be drawn in, since ‘after the army, the imperial administration and the church, they were easily the largest organizations in the Empire’, with branches in every major city, large staffs and substantial patronage.⁶⁹ Emperors also favoured particular colours and the potential of the factions could be exploited to their own advantage by individuals, including members of the elite and the ‘notables’ who were now prominent in urban affairs. By the end of the sixth century, there were faction groupings even in small towns in Egypt, as we learn from the Chronicle of John of Nikiu and other sources.⁷⁰ Finally, the factions also acquired a military capacity and came to have a role in urban defence. The Blues and Greens were not political parties, and while their members might well adopt particular positions in specific cases, the factions as such followed no consistent policy. But in general, and despite the fact that our evidence is patchy, they seem to have taken on wider roles by the seventh century, and urban violence to have increased in level and frequency, a trend associated in general terms by Liebschuetz with the phenomenon of ‘failing curial government’.⁷¹

It is true, as Whitby points out, that the level and frequency of factional disorders must be read chronologically, and that factions and their supporters were not infrequently involved in incidents of religious violence. But the ceremony and the public theatre that were the hallmarks of urban life in late antiquity, and which had their roots very far back in the Principate, had always been conducive to public manifestations that could easily turn into disorder. In the late antique period, not only did the emperor confront the people (and vice versa) in the Hippodrome at Constantinople, but provincial governors also behaved similarly in their local setting. Great churches were the scene of similar manifestations; here, too, large crowds often gathered in emotional circumstances, and passions could be easily inflamed. When rioting broke out, symbols of authority such as imperial or official statues, or the portraits of patriarchs and bishops, were frequently torn down or damaged. The people, or rather, some among them, acquired a real opportunity to express their views on public occasions, which they often did by chanting acclamations of the authorities, mixed in with political messages.

Chariot racing, and the context of the hippodrome, offered an obvious physical setting for such outbursts, and many riots began in the circus, but theatres were also frequently the scene of violent episodes. In both places a contributory factor was provided by the highly structured festivals and performances in late antique Greek cities, in which each social and professional group had its own designated place, as was the case in the theatre at Aphrodisias.⁷² The so-called ‘Brytae’ had been abolished by Anastasius in 501 after episodes of violence; this involved aquatic displays or contests and dancing, and a similar festival held at Edessa was also abolished. However, festivals continued to be held in Constantinople and elsewhere in the sixth century, such as the Brumalia, celebrated in winter and known from several literary and documentary accounts. Choricius of Gaza describes such a festival at Gaza, probably from the 530s, and such events provided a ready focus for crowd excitement; pantomimes always spelt trouble.⁷³ A remarkable example of organized crowd control is provided by the theatre claque known at Antioch in the late fourth century, but surely not unique to Antioch: these were professional cheerleaders who could manipulate audiences to powerful effect; since local governors were also expected to attend the theatre, they were often at the claque’s mercy.⁷⁴ On the other hand, enthusiasm for star performers, especially charioteers, was also a major factor which could destabilise urban life: many contemporary epigrams celebrate famous charioteers, among them one of the most famous of all, Porphyrius, in whose honour as many as thirty-two are known. Two great statue bases survive, inscribed with these epigrams, which once bore his statues; they were erected, side by side with many other monuments, on the spina of the Hippodrome in Constantinople, round which the chariots raced.⁷⁵ To mark special feats, Porphyrius and his rivals might be commemorated in statues made of silver, gold, silver and bronze, or gold and bronze, the gifts of their loyal fans, the Blues and Greens,⁷⁶ and the peak of such commemoration, to judge from the surviving evidence, was reached under Justinian. The importance and prestige of the Hippodrome is indicated by the fact that it was also a showcase for some of the most famous of the many ancient statues collected in late antique Constantinople,⁷⁷ as well as the scene of several imperial accessions. The theatre at Aphrodisias remained in use in the sixth century, as is clear from the factional inscriptions, but early in the seventh century the stage building collapsed and was not repaired, and a wall-painting of the archangel Michael shows that at least part of the building was already being differently used.⁷⁸ Alan Cameron suggests that increasing financial difficulties are likely thereafter to have made the continued maintenance of chariot racing difficult. Procopius characteristically complains that Justinian closed down theatres, hippodromes and circuses so as to save money,⁷⁹ but while this may fit other indications, such as the fact that the circus at Carthage seems to have gone out of use during the sixth century, it is hardly the whole explanation. In fact factional disturbances continued in Egyptian towns and in Jerusalem, and into the seventh century, and even in Alexandria at the time of the Arab invasions.⁸⁰

A final factor which needs to be emphasized is the vital role played by the government in the food-supply of major cities, especially Rome and Constantinople (Chapter 4), which reinforced the dependency of the population on the authorities, while at the same time encouraging and maintaining numbers of citizens at a large and potentially dangerous level. The system was highly organized, and left little scope for private contractors. The cost to the government of maintaining it was very great. In addition, it ensured that political factors continued to play a major role in the stability of the larger cities; it placed the government and the population of the capital in an artificial position of alternate confrontation and dependence, which could lead all too often to public disturbance. Furthermore, it rendered the capital highly vulnerable to any breakdown in supply. It was eventually external circumstances which brought about this breakdown, and with it a great reduction in the population at both Rome and Constantinople. Chris Wickham in particular has emphasized the huge impact on the general economy of the eastern Mediterranean as well as on Constantinople of the cessation of the annona in the early seventh century, and this was compounded for the capital by the damage done to the Aqueduct of Valens, not repaired until the next century, during the great siege of 626. At Rome, the distribution was continued after 476 by the church, though on a smaller scale, while at Constantinople, the fatal change came with the loss of Egypt, the main supplier of grain, to the Persians in the early seventh century, after which population contraction was rapid.⁸³ It is a useful corrective, when considering urban change in the period, to remember the high level of public investment in certain aspects of urban life, whether through the annona or through building programmes, which should warn us in turn that urban prosperity is not in itself a good indicator of the general prosperity or otherwise of the empire.