CHAPTER 10

Urban continuity and change

Cities and civic elites

Cities and their amenities were one of the distinctive features of the Roman empire.1 To be sure, the empire’s main eastern neighbour in late antiquity, Persia, was also urbanised, partly as a result of the long tradition of urban settlement in the Middle East, and the fresh injection of city-founding brought by Alexander the Great and his successors, but also through the building activities of the Sasanian kings themselves.² However, this was the exception among the empire’s neighbours during the late Roman period, the remainder of which were still at considerably earlier stages of development. In his accounts of Roman military expeditions beyond the Rhine and Danube in the mid-fourth century, Ammianus refers to barbarian villages of rudimentary wooden structures which Roman troops destroyed with ease (e.g., 18.2.15); according to Priscus (fr. 11.1 [lines 356–72]), Attila’s main settlement north of the lower Danube was ‘a very large village’ which included a palace and a bath house, but the palace was made of timber and the modest bath house had been constructed by a Roman prisoner with stone imported from Roman territory (presumably spolia – plundered building materials – comprising dressed blocks). One would be justified in wondering whether these historians’ descriptions were influenced by prejudiced stereotypes of barbarian capabilities, but in this case archaeological investigation has corroborated the essential lack of any settlement in barbaricum remotely comparable to a Roman city, in either size or the presence of monumental public structures.3

None of this is to say that most of the empire’s own inhabitants were city dwellers, for it remained the case that the bulk of the empire’s population lived in rural contexts. However, since, for administrative purposes at any rate, cities were conceived of as comprising the sometimes extensive rural hinterland of villages, hamlets and dispersed settlement surrounding the built-up urban core, the city was also a focal point for rural inhabitants, even if many rarely had occasion to visit the city proper. Historical circumstances meant that the distribution of cities across the empire was uneven. Broadly speaking, there were heavier concentrations in the east than the west, although Italy and north Africa were exceptions to this. Late Roman emperors did continue the long-established practice of founding new cities which bore their name – most famously, Constantinople. In theory, these expanded the empire’s stock of cities, but in practice many (such as the most famous) simply involved the ‘rebranding’ of already existing cities, albeit sometimes in conjunction with restoration work after major damage arising from war or earthquake. While, then, there can be no denying the broader historical significance of the investment by Constantine and his successors specifically in Constantinople, imperial foundations in late antiquity more generally did little to change the overall number or distribution of cities.⁴

During the late Republican period, the Roman ruling class had absorbed from the Greeks a cultural appreciation of the benefits of polis-centred life, but they were also quick to grasp the practical advantages of the Mediterranean’s network of cities as a convenient infrastructure to assist effective governance of a vast empire. For this purpose, the crucial institution in a city was its council (curia in Latin, boulē in Greek), the members of which constituted the civic elite (curiales or decuriones in Latin, bouleutai or politeuomenoi in Greek). Varying in number depending on the size of the city, these were local landowners whose wealth and connections made them the ideal candidates to oversee the administration of their city and to shoulder the responsibilities and burdens that this entailed. These localised administrative duties included ensuring adequate availability of staple foods and water, maintenance of public buildings, preservation of public order, and the staging of religious festivals and games. City councils could use city revenues to help fund such matters (e.g., rents from civic land), but these were rarely sufficient to cover everything and there was an expectation that councillors would draw on their personal resources to supplement civic funds. Indeed in the early empire there developed a culture in which rivalry between members of the local elite expressed itself in competitive display and generosity to their community – so-called ‘euergetism’ – as attested in the numerous inscriptions which memorialised their gifts, whether it be helping to purchase grain, fund games or construct a new theatre. From the perspective of the emperor, however, the most important function of city councils was to oversee and underwrite the collection of imperial taxes from their city’s rural hinterland. It was their assumption of this role which enabled emperors during the Principate to make do with a tiny total number of administrators (in the low hundreds) – an arrangement aptly characterised as ‘government without bureaucracy’.5

The curial order were prepared to take on this role partly because it was accepted that, in collecting imperial taxes, they could extract more than the emperor required and pocket the difference. They also benefited from minimal interference by central government in the internal affairs of cities, which allowed councillors to run local affairs and to engage in political and social rivalry with their peers. By the fourth century, however, the balance between the benefits of status and the burdens of responsibility involved in performing curial duties had begun to shift inexorably away from the former towards the latter. This was at least partly due to the expansion of central government resulting from the reforms of Diocletian and Constantine. The number of administrators grew significantly, into the tens of thousands, as the central bureaucracy expanded and the number of provinces was increased by the subdivision of existing ones. This facilitated closer oversight of provincial and civic affairs, and above all of the levying of taxes, for which, with a larger army to support, there was even greater demand. The scope for city councillors to enrich themselves through the tax collection process accordingly contracted, while cities lost control of civic lands and local taxes to the imperial treasury. With the material and less tangible incentives for curial service receding, many of those liable for service sought ways to escape; one of the themes of the fourth century was the attempts of successive emperors to maintain numbers on the city councils by, for example, insisting on the hereditary nature of curial obligations.6 Yet, ironically, emperors were also responsible for creating escape routes, through the expansion of the central bureaucracy and the establishment of the new senate at Constantinople, both of which drew on city councillors and gave immunity from curial duties. Service in the army or the church offered further modes of egress.

The correspondence and speeches of the rhetorician Libanius provide some of the best evidence for the phenomenon as it affected the city of Antioch in the second half of the fourth century. At various points he comments on the sharp decline in the number of city councillors, no doubt with an element of rhetorical exaggeration – ‘Instead of the six hundred of days gone by, there aren’t even sixty now’ (Or. 2.33) – while elsewhere he remarks on the difficulties of councillors in finding a wife of any wealth, since no prospective father-in-law wanted to risk their daughter’s dowry being consumed by the demands of curial obligations (Or. 2.36). At the same time, many of Libanius’ students – often the sons of city councillors – were keen to escape their looming obligations and sought entry to the imperial bureaucracy, which Libanius, proud teacher that he was, assisted by writing letters of recommendation.7

While repeated imperial legislation to restrict avenues of escape and require absconding councillors to surrender their property to the council might suggest that emperors were fighting a losing battle, there is other, admittedly patchy, evidence from the fourth and fifth centuries indicating that the laws were not consistently ignored.⁸ An inscription from the reign of Valentinian I, for example, records one man’s restoration of a basilica at Cuicul in Numidia ‘for the completion of his required curial obligation (munus)’ (ILS 5535), while a fifth-century saint’s life reports the conversion of Rabbula, a leading councillor in a Syrian city (probably Chalcis), whose first step towards an ascetic life involved his giving to the council the property on which it had a claim.⁹ Furthermore, curiales were not such an endangered species as to be on the verge of extinction: as already noted (p. 166 above), the Emperor Anastasius’ institution of vindices to oversee tax collection did not, as sometimes assumed, replace the involvement of councillors in this process, and they continue to feature in the extant sources for the eastern half of the empire throughout the sixth century.¹⁰

Nonetheless, the city council as the body responsible for civic administration was increasingly superseded by more informal arrangements, with leadership in cities being assumed by a less clearly defined elite of what modern scholars have come to refer to as ‘notables’, comprising all or some of the following: honorati – that is, locals of senatorial rank (in its expanded and diluted form [p. 64 above]); imperial officials based in the city; some city councillors (probably the wealthiest); and – most strikingly – clergy, above all the local bishop.¹¹ The growing dominance of these notables partly reflected changing patterns of land ownership, as indicated by the collective terms often used in the sources to refer to them – possessores in Latin, and its Greek equivalent, ktētores.12 Some city councillors were prepared to escape their obligations by selling their land and using the proceeds to facilitate their appointment to a post in the imperial bureaucracy, creating opportunities for other notables to expand their local land-holding and influence.¹³ In one sense, bishops could be seen as an exception to this pattern, since even if they did not always relinquish all their property at ordination as the church encouraged them to do – many bishops came from a curial background – they were unlikely to add to what they already owned; in another sense, however, they epitomised this new urban elite, since they held office independently of the curial order, and they controlled (sometimes significant) local church resources which necessarily gave them influence in their community.¹⁴

The first clear indication of the shift of power from council to notables is a law issued in the west in 409 regarding appointments to the civic post of defensor civitatis (‘defender of the city’, who acted as a judge in minor civil cases), previously a decision of the city council, but now to be determined by the bishop, clergy, honorati, possessores and curiales (Cod. Iust. 1.55.8). This law was reiterated by the Emperor Anastasius in the east in 505 (Cod. Iust. 1.55.11), and a number of inscriptions from Corycus in Cilicia at about the same time also record imperial rulings confirming that the defensor and other civic posts were to be appointed by the bishop, clergy and ktētores.¹⁵ In the post-imperial west, the curiales (and indeed the council) retained a presence in some of the law codes and other documentation of early barbarian rulers, but the most important figures in cities were the bishop, the possessores and the comes civitatis (‘count of the city’) – a royal official with responsibility for maintaining public order and overseeing the levying of taxes.¹⁶ This reconfiguration of civic elites represented an important change in the leadership and administration of cities during late antiquity, both symbolically and because the less formal organisation of the notables is likely to have made it more difficult for emperors and kings to hold them to account in the way that had been possible with city councils.17

Urban fabric and amenities

Changes in the fortunes and character of civic elites are ones which can be traced primarily through legal and other documentary sources. Another important dimension of urban continuity and change is the physical character of cityscapes, for which the primary evidence comprises inscriptions and the archaeological remains of buildings and urban spaces. There is now a vast (and ever-increasing) volume of relevant data, to which it is impossible to do justice within the constraints of a study such as this. The intention here is rather to sketch the outlines of the most important developments. In doing so, a fundamental theme is the difficulties of generalisation. These difficulties are twofold. First, the haphazard survival of evidence from individual sites and the finite resources available for their archaeological investigation mean that it is rarely possible to gain a comprehensive overview of the development of a city. Secondly, it is difficult to generalise more broadly because of regional variation, not only between east and west, but also between areas within those broader designations. Patterns of development in Anatolia were not necessarily the same as in the Balkans, just as those in north Africa could be quite different from those in Gaul, while the pace of change was also variable – a reflection of differing geographical and historical circumstances and differing cultural traditions. Nor was the experience of cities even within the same area necessarily uniform.

By way of illustration, it is instructive to consider some of the most intensively studied sites from around the empire. Turning first to the east and more specifically to Palestine, the material remains from Scythopolis indicate that the city reached the peak of its development and prosperity in the late fifth and early sixth century, while excavations at the coastal city of Caesarea show significant expansion of the city’s area in the fifth century, a range of evidence for the upkeep and repair of public buildings and infrastructure throughout the sixth century, and a flourishing economy during this period.18 In western Anatolia, a wealth of epigraphic and archaeological evidence from Aphrodisias points to growing prosperity from the mid-fifth to the mid-sixth century and significant investment in public buildings, while the commercial quarter of Sardis appears to have been thriving in the sixth century.¹⁹ At Sagalassos in central Anatolia, on the other hand, the picture is less uniformly positive. Important public buildings, such as the Neon Library and the baths, were refurbished in the second half of the fourth century, while a large urban villa was also constructed in this period. On the other hand, the library was destroyed c. 400 and not rebuilt, and a defensive wall was also constructed at this time, defining an area a third of the size of that occupied during earlier centuries. Other indications that the fifth century was a less prosperous period in the city’s history include the complete disappearance of inscriptions honouring members of the local elite, and the abandonment of the city’s main pottery workshop. A severe earthquake c. 500 did major damage to the city’s infrastructure, and although there was some rebuilding, it is apparent that the fifth and sixth centuries here were not a period of prosperity akin to that enjoyed by cities in Palestine and western Anatolia.²⁰

In the Balkans, the picture is even bleaker. Corinth, which once prided itself on one of the most spacious civic centres in the empire, experienced significant earthquake damage in 365 and again in 375, and although a certain amount of restoration work was done in their aftermath, the city was sacked by Alaric’s Goths in 395 (p. 110 above); while some new structures were built (from spolia) in the fifth century, most public buildings were not restored, honorific inscriptions disappeared from the agora (the focal point of public life) after the early fifth century, the city wall enclosed a much reduced area of only 1.5 km², and by the sixth century the agora was being used for burials.21 Meanwhile, Nicopolis ad Istrum, in the northern Balkans, was destroyed, presumably by the Huns, and although the city was subsequently rebuilt in the later fifth century, it was located on a new site adjacent to the original city, covering an area only a quarter the size.²²

Moving westwards, inscriptions and archaeology from urban centres across north Africa show local elites continuing to invest in public buildings and infrastructure throughout the fourth and early fifth century.²³ While the advent of the Vandals in the middle of the fifth century did not for the most part entail the destruction of the fabric of cities (cf. pp. 185–6 above), it is apparent that civic centres increasingly ceased to be maintained in traditional fashion. At Belalis Maior (Henchir al-Faouar), to the west of Carthage, rooms around the forum were modified by the addition of partition walls, there was a build-up of soil over the forum pavement, and the area began to be used for burials;²⁴ at Thuburbo Maius, inland to the southwest of Carthage, the forum area was gradually occupied by housing and oil presses;²⁵ and at the Tripolitanian city of Sabratha the forum area ‘was eventually turned into a cemetery … [and] the adjacent temple of the Unknown Divinity was overlain with dense housing and workshops’.²⁶ The return of imperial rule during the reign of Justinian brought fresh investment, but as far as secular structures are concerned, this was mostly spent on fortifications to defend communities against Moorish tribesmen, as described by the contemporary historian Procopius in Book 6 of his Buildings and corroborated by inscriptions and archaeological evidence.27

Continuous occupation of many ancient urban sites in the Iberian peninsula has made it more difficult to trace the evolution of cities in this region during late antiquity. At one of the more accessible, Emerita (Mérida), to the west in Lusitania, there is a range of material evidence for substantial investment in private housing throughout the fourth century and into the fifth, even if some of these structures increasingly encroached on public space; on the other hand, some of these residential quarters experienced major damage during the middle decades of the fifth century, a period when the city is known, from literary sources, to have been fought over by imperial and barbarian forces. At Tarraco (Tarragona) on the Mediterranean coast, the civic centre was maintained into the fifth century until, in the 440s, the paving was removed from a section of the forum, which became a pit for domestic rubbish, again implying the encroachment of private housing into public space. Elsewhere, at Caesaraugusta (Zaragoza), in the north, there is evidence for the deterioration of the city’s infrastructure rather earlier, in the later fourth century, when, for example, all the city’s sewers became silted up, and it is thought that this pattern was in fact more typical of cities in the region.²⁸

In Gaul, the southern half of the region had been exposed to Roman influence for longer than the north, and so there was a stronger tradition of urbanism in the former. Nonetheless, a number of southern cities saw the construction of walled enclosures during the fourth century which encompassed much reduced portions of the urban settlements of earlier centuries and, significantly, excluded the established civic centres. In the case of Bordeaux, this shift in the city’s centre of gravity was due to the pull of the river port, while in the case of Périgueux, it can be attributed to the desire to save money and labour by incorporating the city’s amphitheatre into the wall-circuit.²⁹ While the influence of such practical considerations is understandable, shifts of this sort are nonetheless symbolic of changing priorities. Even at Arles, which particularly benefited from imperial patronage during the fourth century and became an administrative centre at the end of that century, the forum lost its paving and accompanying portico during the first half of the fifth century as simple commercial structures encroached on this public space; although the circus seems to have remained in sufficiently regular use for Frankish kings to watch chariot races there in the mid-sixth century (Procop. Wars 7.33.5), private housing was erected around its exterior.30 A striking departure from this pattern was Marseille, which remained prosperous throughout the fifth and sixth centuries. There was little, if any, contraction of its city wall or of intramural habitation, and even evidence of extramural expansion; and archaeological and textual evidence confirms its role as a thriving emporium, channelling goods from Africa and the eastern Mediterranean up the Rhone valley. It was, however, the exception, which owed its good fortune primarily to its liminal location between the Mediterranean and the emerging Frankish kingdom, and its excellent harbour.³¹

Turning finally to Italy, the city of Rome itself remained in many respects unique and, together with cities which functioned as imperial capitals in the fourth or fifth centuries (Milan, Ravenna) or as royal capitals in the sixth century (Ravenna, Verona, Pavia), cannot be taken as a guide to more general trends in the physical evolution of Italian cities. In many of those cities which did not benefit from imperial, royal or senatorial patronage, it seems, on the one hand, that the traditional forum and paved street grid were nonetheless preserved at least through to the sixth century, but, on the other hand, that private investment in public buildings ceased at the end of the fourth century or in the early fifth century.³² While Italian cities did not experience the growth which many eastern cities did during the fifth and sixth century, neither did they suffer the contraction which cities did in other regions of the west – at least, not until the imperial campaigns of the 540s against the Goths, followed closely by the Lombard invasion in the late 560s.³³

This brief overview of a selection of sites from different parts of the empire demonstrates the dangers of generalisation about the fate of cities during late antiquity. While cities in the east tended to fare better for longer than those in the west, there were exceptions in both halves of the empire, often depending on the vulnerability of a particular location to military or geophysical dangers, or alternatively its favourable position vis-à-vis trade routes. These examples also hint at some of the problems raised by employing the term ‘decline’ in the context of the urban history of late antiquity. Clearly some cities were expanding and enjoying enhanced levels of prosperity well into the sixth century. However, even for those cities which were not so fortunate, the use of the term ‘decline’ can still be challenged, on a number of grounds. The term is loaded with much unhelpful baggage, and it is such a broad term which can be applied to so many diagnostic features of urban well-being of varying degrees of significance that its value as a descriptive term is questionable.34 Dissatisfaction with the term has in turn prompted scholars increasingly to deploy alternative terminology, such as ‘demonumentalisation’ for changes to civic centres. Moreover, the diagnostic features used to assess urban well-being are themselves often more ambiguous in their implications than is often recognised. Take the case of the construction of city walls, often seen as symptomatic of urban decline because they are assumed to have been erected in response to military threats, because their construction frequently included the use of spolia (which is taken to imply haste in the face of such threats), and because they often enclosed a much more restricted area of the city. However, while military insecurity undoubtedly must have been a factor in many instances, fortifications could also be a statement of civic pride; there are instances where spolia have been deployed in city walls with great care and discrimination; and the walls did not necessarily mark the extent of human habitation, since (as archaeological evidence often shows) in some cases there were unoccupied areas within the walls, while it was perfectly possible for some of a city’s populace to live in extramural suburbs.³⁵

Running in tandem with all these developments was another over-arching trend with important implications for the debate about urban change – the Christianisation of cityscapes. The most visible manifestation of this was obviously the construction of churches, with leading citizens who might in the past have displayed their philanthropy to their community through contributing to new civic structures now redirecting their resources towards the construction of churches and related buildings such as hospices. Church resources were, of course, also directed into such structures, as well as related charitable activities – care for widows, orphans and the poor – which contributed to the growing prominence and influence of bishops in cities.36 The other side of this development was the increasing neglect of temples, or even their destruction and subsequent replacement by churches. Again, the pace of change varied, as also its impact on urban topography. It has been observed that in the west, there was a tendency for churches to be built away from the traditional civic centre, with the result that the classical configuration of cities was lost more quickly than in the east, where churches tended to replace temples in the civic centre. While this marked a significant change to the civic centre, it also had the effect of preserving the broader layout of the city.³⁷

Christianity also had potential implications for buildings related to public entertainment, such as theatres, amphitheatres and circuses/hippodromes, since Christian bishops and clergy usually regarded the events staged at such venues as antipathetic to Christian values and a distraction from church attendance. The plays, mimes and dancing presented in the theatre perpetuated familiarity with the deities of classical myth and exposed audiences to storylines and performances of questionable morality (cf. p. 13 above), while the chariot-racing held in the circus or hippodrome, and the gambling which accompanied it, was evidently an obsessive preoccupation for many (cf. Figs 14–15). Augustine told his congregation that ‘the absurdity of spectacles’ was ‘a fever of the soul’ comparable to avarice, lust and hatred (Serm. 9.8.10 [= PL 38.83]),38 while John Chrysostom, preaching in Antioch, emphasised the distortion of priorities they induced:

Figure 14 A gambling machine, with reliefs of chariot racing (77 × 55 × 57 cm) (marble, Constantinople, c. 500). This was an appropriate device to find at the hippodrome where money was wagered on the outcome of the chariot races; here coloured balls would be rolled down the tracks and through the holes, with gamblers placing bets on which ball would emerge first at the back. The relief at the base shows a four-horse chariot in pursuit of another, with further reliefs on the sides depicting other aspects of the races. © bpk/Skulpturensammlung und Museum für Byzantinische Kunst, SMB/Jürgen Liepe

There are those who are in a state of excited distraction over the spectacle of horse racing and are able to state with complete accuracy the names, the herd, the pedigree, the place of origin, and the rearing of the horses, as well as their age, their performance on the track, which horse drawn up against which other horse will be victorious, which horse will begin best from which starting-gate, and which charioteer will be victorious over the course and outrun the opposition. And no less than these, there are those who devote their time to the theatre and display even greater madness over those who behave in an unseemly manner in the theatre – I mean dancers and those who perform mimes – detailing their ancestry, place of origin, upbringing and all the rest. But if we ask them how many and what letters Paul wrote, they do not know the number; or if some know the number, they are quite at a loss when asked the cities which were the recipients of the letters. (In illud, Salutate Priscillam 1 [= PG 51.188])

Figure 15 A contorniate with inlaid decoration depicting a four-horse chariot (4.3 cm diameter) (later fourth century), found in the amphitheatre at Trier. Contorniates were bronze medallions with raised edges, usually bearing images associated with success in the games, and are thought to have been distributed by aristocrats to the urban populace during such events. The name in the upper label, Pvrfyri (Porphyry), may refer to the charioteer, while that in the lower label, Fontanus, may be the name of the lead horse. © Rheinisches Landesmuseum Trier

Despite ecclesiastical opposition, however, the continued popularity of spectacles meant that the structures in which these events were staged were generally maintained for much of late antiquity, whether from imperial or (in the post-imperial west) royal funds, or by local patrons, so preserving one of the distinguishing features of the classical cityscape.39 In addition to providing a time-honoured distraction from sources of dissatisfaction among the urban masses, these events also presented an opportunity for the populace of a city to voice their complaints to the emperor or his officials through the mass chanting of acclamations, and for a response to be given40 – thereby demonstrating the continuing relevance of cities to the political dynamics of the late Roman world. Such opportunities to voice popular discontent could serve as a valuable safety valve for rising tensions, although as the next section shows, it was not a guaranteed solution.

‘City-destroying civil strife’

Bouts of mass unrest were an inevitable phenomenon in cities of the Roman world. Food shortages were perhaps the most predictable stimulus, with the largest city – Rome – potentially the most vulnerable in this respect.⁴¹ As long as Rome was the emperor’s primary residence during the first two to three centuries AD, measures were in place to minimise the risks of hunger-induced public disorder, but with emperors ceasing to reside in Rome for most of the fourth century and much of the fifth, there was no longer the same imperative to guarantee supplies. It is hardly surprising, then, that food riots appear to have become a much more common phenomenon in the city in this period, with the prefect of the city shouldering the responsibility and bearing the brunt of the anger of hungry inhabitants.⁴² Nor was Constantinople immune from such problems, at least initially, despite becoming a permanent imperial residence from the late fourth century onwards. In 409, rioting over food shortages resulted in the burning of the headquarters of the prefect of the city and the dragging of his carriage through the city streets; the government quickly found 500 lbs of gold to purchase emergency supplies of grain, and subsequently established a permanent reserve of gold to deal with any future food crisis in the capital.⁴³ Nor was grain the only commodity of importance: in 515 shortages of bread and oil in Alexandria led to rioting and the lynching of the senior imperial official in the city (Malalas, Chron. 401–2, Exc. de Insid. fr. 41).

Since the main burden of taxation fell on the inhabitants of rural areas, it was less common for fiscal demands to be a source of urban disturbances. Nonetheless, it did provoke one famous episode, in Antioch, when, in 387, it was announced that taxes were to be levied on the city. Although the episode is comparatively well documented (particularly from a series of sermons by John Chrysostom and speeches by Libanius), the sources are vague about the taxes in question, but the reaction suggests they represented some sort of significant additional imposition. In the ensuing rioting, the wooden portraits and bronze statues of the imperial family outside the residence of the governor were pulled down, mutilated and dragged through the streets, and public buildings were torched. Imperial troops then intervened, arresting and executing rioters identified as ringleaders, while reports were sent to the emperor in Constantinople. Destruction of imperial portraits and statues, which embodied the emperor’s presence in the city, was tantamount to treason (cf. p. 108), so Theodosius’ response was awaited with trepidation. Following the arrival of imperial commissioners, the public baths, theatres and hippodromes were closed, the city was deprived of its metropolitan status, and the city councillors were arrested, tried and sentenced to death, on the grounds that they bore the ultimate responsibility for what had happened. Significantly, it was the bishop of Antioch who led a delegation from the city to the capital to plead for the emperor’s clemency, while local monks interceded with the commissioners; their efforts were successful, with the lives of the councillors spared and the city pardoned.44

There is no surprise in hunger and taxes giving rise to urban unrest in late antiquity, since these had been factors in earlier periods of Roman history. However, there were also a number of novel factors contributing to civic disturbances in late antiquity. In particular, there were many instances of religiously inspired disorder, on the one hand, and, on the other, many instances of violence arising from the activities of the so-called circus ‘factions’ – both factors which had been far less prominent in earlier periods.

While the element of intolerance in polytheistic Roman religion of the pre-Constantinian period should not be underestimated,45 exclusionist religious attitudes intensified during late antiquity and generated much urban unrest. That intensification derived partly from Christianity’s monotheistic theology, partly from competition between different groups within the church, and partly from the polarisation of attitudes provoked by Julian’s attempt to undo the effects of Constantine’s support for the church.⁴⁶ It manifested itself in public confrontations between pagans and Christians, most famously the events resulting in the destruction of the Serapeum in Alexandria in 391/2 (p. 54 above), but also in incidents which are perhaps less well known such as an episode in Sufes in north Africa in 399, when the destruction of a statue of Hercules by Christians resulted in rioting by local pagans during which sixty Christians died (August. Ep. 50),⁴⁷ and the fracas in Alexandria in 486 between pagan and Christian students which resulted in the sacking of an extramural shrine of Isis.⁴⁸ It was also evident in violent confrontations between rival Christian groups.⁴⁹ Episcopal elections provided recurrent flashpoints. In Rome there were disturbances between the supporters of different candidates, most famously in 366 when more than a hundred died in the violence (Amm. Marc. 27.3.13, Coll. Avell. 1), but also in 418 (Coll. Avell. 14–37). There were also public disturbances in Constantinople after the deposition of John Chrysostom in 403 (Sozom. Hist. eccl. 8.22); and the appointment of a pro-Chalcedonian bishop to Alexandria in 451 led to rioting which required the intervention of imperial troops, who nonetheless proved unable to prevent the lynching of the bishop in 457 (Evagrius, Hist. eccl. 2.5, 8). The frequent exiling of heterodox bishops by emperors in this period seems to have been a strategy for removing them from their support base, so reducing the likelihood of their becoming a focus for urban unrest.50

The other new source of urban unrest in late antiquity was the circus ‘factions’. In earlier centuries, there had been four factions, identified by different colours – red, white, blue and green – and they had been responsible for providing horses and chariots for racing in the circus (Latin) or hippodrome (Greek) of major cities. By the fifth century, however, a number of important changes had occurred. First, two of the factions – the red and white – had become much less important, with the blue and green the dominant groups. Secondly, the factions had become responsible not only for the provision of chariot-racing, but also for other forms of public entertainment, such as productions for the theatre (plays, mimes, pantomimes). Thirdly, the role of the factions increasingly extended beyond provision of entertainment to the orchestration of popular opinion in the form of chanting and acclamations at the hippodrome and other venues, as well as during imperial accessions. Fourthly, although there was a notorious hippodrome-related riot in Thessalonica in 390,⁵¹ it was only from the 440s onwards that the Blues and the Greens became involved in regular episodes of violence in major cities in the eastern empire. The following account of an episode during the reign of Anastasius (probably from the year 507) illustrates some of these developments, and the destruction that could result:

During his reign, the supporters of the Greens at Constantinople appealed to the emperor, while the chariot-races were being held, for the release of some people who had been arrested by the city prefect for throwing stones. The emperor did not yield to them but grew angry and ordered troops to attack them, and there was great disorder. The supporters advanced against the palace guard, and approached the imperial box and they threw stones at the emperor Anastasius. Among them was a man named Maurus who threw a stone at the emperor, who stood up and dodged it. The palace guard, having seen the man’s boldness, went for him and dismembered him limb by limb, and so he breathed his last. The crowd, which was hemmed in, set fire to the Chalke, as it is known, of the hippodrome, and the colonnade was burned as far as the imperial box, and also the public colonnade, as far as the Hexahippion and the Forum of Constantine, was completely burned and destroyed, collapsing throughout its length. After many had been arrested and punished, there was quiet, when Plato, who was patron of the Green faction, was appointed as city prefect. (Malalas, Chron. 394–5 [tr. E. Jeffreys et al., with revisions])

Scholars originally tried to account for this sort of violence by seeing the factions as championing a particular Christological stance – Blue Chalcedonians, Green anti-Chalcedonians – or a particular social group – Blue aristocrats, Green non-aristocrats – but it has been recognised for some time now that such identifications do not account for all the evidence.52 However, the proposed alternative – that the violence was simply a case of sporting hooliganism between rival partisans – has also been seen as too simplistic an explanation.⁵³ It has been argued by some that the rise of the factions reflected the decay of traditional civic institutions and the emergence of government by the notables,⁵⁴ but this was irrelevant to Constantinople, where many of the factional disturbances occurred, and the most plausible analysis relates factional violence to the patronage of the factions by prominent figures, above all emperors whose authority needed affirmation, starting with Theodosius II (408– 50).⁵⁵ As he was an emperor with no military credentials who spent nearly all his life in the imperial palace, patronage of the Greens during the 440s offered him an opportunity to harness the potential of the faction to influence the urban populace of Constantinople and other eastern cities.⁵⁶ Needless to say, this could be a dangerous game which sometimes got out of hand, as Justinian in particular was to discover in 532 (pp. 247–50 below).

Cities, education and culture

Cities were synonymous with civilisation in the Roman world not only because of the architecture which adorned them and provided facilities such as bathing complexes and entertainment buildings, but also because they were the focal point for the maintenance of classical literary culture through education.57 Judging by the evidence surviving from Roman Egypt, elementary education could be available in rural communities (though this should not be taken to mean that basic literacy was widespread), but education at a level beyond this could only be acquired in an urban context. This remained the case during late antiquity, as did the assumption that learning was a characteristic of urban society, reflected in the comment of a well-educated inhabitant of fifth-century Gaul: ‘the educated are as far superior to rustics as humans are to beasts’ (Sid. Apoll. Ep. 4.17.2). For those whose families had the resources, education beyond the basic level involved two main stages: first, studying under a grammarian to acquire, on the one hand, a sound grasp of the grammar and pronunciation of literary language and, on the other, close familiarity with the canonical poetic texts; and secondly, studying under a rhetorician to acquire detailed knowledge of rhetorical texts and handbooks, followed by practical instruction and experience in composing and delivering speeches. A prosopography of grammarians in late antiquity confirms that they were to be found in cities and towns, while, judging by what is known of the careers of individual rhetoricians and a law of the Emperor Gratian concerning arrangements in Gaul (Cod. Theod. 13.3.11), teachers of rhetoric usually practised in provincial capitals.⁵⁸ Ideally, this education would equip the student to a high level in both Greek and Latin, but increasingly during late antiquity this could not always be taken for granted.⁵⁹

As has often been observed, the educational curriculum was a very narrow one, but it served to distinguish and unify the empire’s elite through the specialised knowledge acquired, as well as the friendships and contacts made during student days.60 In late antiquity, this education became the usual prerequisite for service in the imperial bureaucracy, and thus was the entrée to some degree of influence and guaranteed remuneration. Gaining such an education for their son(s) therefore became the goal of ambitious parents from more modest backgrounds, of whom the future bishop and theologian Augustine is perhaps the most famous example. Born in the north African town of Thagaste to a father who owned only a little land and a few slaves, Augustine moved first to the regional centre of Madauros, then to Carthage, to acquire the education which eventually became his passport to better things in Rome and Milan. The imperial government and city councils did provide funding for some posts for grammarians and rhetoricians – the Emperor Theodosius II, for example, famously established more than thirty such posts in Constantinople in 425 (Cod. Theod. 14.9.3, Cod. Iust. 11.1.9.1) – but students were also expected to pay fees to their teacher, which limited the pool of potential students significantly. Augustine’s education was delayed for a year while his father saved enough money to send him on from Madauros to Carthage, and when his father died soon after, Augustine was only able to continue his studies there due to the generosity of a wealthy neighbour (August. Conf. 2.3.5, C. acad. 2.2.3).

The most detailed insights into the world of advanced education from the perspective of the teacher emerge from the writings of the fourth-century rhetorician Libanius, the son of an Antiochene decurion, who studied in Antioch and Athens, and briefly held teaching posts in Athens and Constantinople before returning to Antioch, where he established a reputation as one of the pre-eminent teachers in the empire. His speeches and voluminous correspondence illuminate both the educational process and his relationships with students, which endured well after they had completed their studies.⁶¹ The student’s perspective is perhaps best represented by Zachariah’s biography of the anti-Chalcedonian bishop Severus, which recounts student life in the late fifth century in Alexandria and Berytus (Beirut) – the latter a major centre for the study of law – especially clashes between pagan and Christian students and their teachers.62

As these incidents imply, education was not immune from the religious controversies of late antiquity; indeed, education was itself a culturally contested area. The Emperor Julian’s ban on Christian teachers, though short-lived, politicised education as a religious issue and forced Christian leaders, many of whom had themselves been the beneficiaries of a traditional classical education, to review their attitudes. In doing so, they could draw on earlier pronouncements on the subject, but those pronouncements did not speak with a unified voice and anticipated an east/west divergence. In the early third century, the north African lawyer and theologian Tertullian had famously posed the rhetorical question, ‘What has Athens to do with Jerusalem?’ (De praescr. haeret. 7), whereas the Alexandrian theologian Origen did not regard pagan learning as posing an inevitable threat to Christian principles.⁶³ This latter ‘sense of easy and sophisticated superiority’ informed the views of leading Christian figures in the east during the fourth, fifth and sixth centuries, perhaps most famously in Basil of Caesarea’s Address to Young Men: How They Might Profit from Greek Literature. Not an uncritical endorsement of classical literature or education, it nonetheless allowed a place within the formation of Christians for such study, since it could provide an outline of virtue which Christian teaching could then develop in detail.⁶⁴ Although the Emperor Justinian (527–65) was to take a strong line against the teaching of philosophy in Athens (p. 276 below), the prevailing attitude on the part of Christian leaders in the east was to remain one of accommodation with traditional literature and learning.

In the west, on the other hand, there was much greater ambivalence, reflected most famously in Jerome’s report of a dream in which he stood before the heavenly judge and was told, ‘You are a Ciceronian, not a Christian!’ (Ep. 22.30). The fullest discussion of the subject was developed by Augustine in his De Doctrina Christiana (‘On Christian education’). While not as hostile or uncompromising in his attitude to classical learning as Tertullian, he was much more dismissive of its value than eastern commentators. Although western Christians of aristocratic background, such as Sidonius Apollinaris, remained comfortable with their cultural inheritance, the prevailing attitude in Christian circles in the west was much more guarded and sceptical.65 Even so, it was not this attitude which was primarily responsible for the shrinkage of educational opportunities in the west during the fifth and sixth century, nor was it a by-product of the new barbarian elites who gradually came to control the west, since they were for the most part appreciative of classical literary culture (Theoderic the Ostrogoth had, after all, been educated in Constantinople).⁶⁶ It was, rather, a consequence of political and economic uncertainty, the fragmentation of centralised imperial authority, and the deterioration of city and personal finances. Against this background, ecclesiastical institutions – above all, monasteries – increasingly emerged as the focal point for education. The classical heritage was not lost sight of, but it was firmly subordinated to Christian priorities. Perhaps the most explicit articulation of this approach was that of the Roman senator and servant of the Gothic regime in Italy, Cassiodorus. After being prevented by Justinian’s invasion of Italy in 535 (pp. 262–3 below) from establishing the Christian educational institution he had envisaged at Rome, he eventually realised a version of his plan on his estate, Vivarium, in Calabria. Beginning in the early 550s, he founded two monasteries which included a library and a scriptorium for translating works from Greek into Latin. He also wrote his Institutions: the first part focused on guidance in reading and understanding the Bible, while the second part preserved a summary of what he considered to be important aspects of the classical literary heritage. He thereby provided both a practical model and an accompanying handbook which were to prove influential in the medieval west.⁶⁷

1. There is an ever-expanding body of literature on the city in late antiquity. An outstanding overview is S. T. Loseby, ‘Mediterranean cities’ in P. Rousseau (ed.), A Companion to Late Antiquity, Oxford: Wiley-Blackwell, 2009, 139–55, while K. G. Holum, ‘The classical city in the sixth century’ in M. Maas (ed.), The Cambridge Companion to the Age of Justinian, Cambridge: Cambridge University Press, 2005, 87–112, provides another good summary within a more restricted timeframe; C. Wick-ham, Framing the Early Middle Ages: Europe and the Mediterranean, 400–800, Oxford: Oxford University Press, 2005, 591–692, offers a more detailed survey with a particular focus on the economic dimension. Important monographic treatments include J. H. W. G. Liebeschuetz, The Decline and Fall of the Roman City, Oxford: Oxford University Press, 2001, and H. G. Saradi, The Byzantine City in the Sixth Century, Athens: Society of Messenian Archaeological Studies, 2006. There are many conference volumes relating to the subject, including J. Rich (ed.), The City in Late Antiquity, London: Routledge, 1992, N. Christie and S. Loseby (eds), Towns in Transition: Urban Evolution in Late Antiquity and the Early Middle Ages, Aldershot: Scolar Press, 1996, and J.-U. Krause and C. Witschel (eds), Die Stadt in der Spätantike: Niedergang oder Wandel?, Stuttgart: Franz Steiner, 2006 (which includes many papers in English).

2. J. Howard-Johnston, ‘The two great powers in late antiquity: a comparison’ in Averil Cameron (ed.), The Byzantine and Early Islamic Near East III: States, Resources and Armies, Princeton: Darwin Press, 1995, 157–222, 181–2, 198–211.

3. M. Todd, The Early Germans, 2nd edn, Oxford: Blackwell, 2004, 62–75; M. Kulikowski, Rome’s Gothic Wars from the Third Century to Alaric, Cambridge: Cambridge University Press, 2007, 87–94 (who notes the existence of the occasional larger dwelling making some use of stone and terracotta tiles – but still in a village context).

4. Imperial foundations are detailed in Jones, LRE, 718–22, to which can be added the elevation of Didyma, in western Anatolia, from village to city status in 527/33 with the new name of Justinianopolis, as attested in an important inscription discovered in 1991 (AE 2004.1410).

5. P. Garnsey and R. Saller, The Roman Empire: Economy, Society and Culture, London: Duckworth, 1987, 20–40.

6. Valentinian and Valens did also try to ease the strains on city councils by restoring one third of the revenues from former civic estates for the specific purpose of repairing public buildings: N. Lenski, Failure of Empire: Valens and the Roman State in the Fourth Century A.D., Berkeley: University of California Press, 2002, 295–6.

7. J. H. W. G. Liebeschuetz, Antioch: City and Imperial Administration in the Later Roman Empire, Oxford: Clarendon Press, 1972, 174–82.

8. A. Laniado, Recherches sur les notables municipaux dans l’empire protobyzantin, Paris: Collège de France, 2002, 3–26.

9. Life of Alexander Akoimetos 11, 20, with general discussion of this source and a translation in D. Caner, Wandering, Begging Monks: Spiritual Authority and the Promotion of Monasticism in Late Antiquity, Berkeley: University of California Press, 2002, 126–57, 249–80.

10. Laniado, Notables, 63–87.

11. Liebeschuetz, Decline, 104–36; Laniado, Notables, 131–223.

12. The most detailed evidence on changing patterns of land-holding is from Egypt, for which see J. Banaji, Agrarian Change in Late Antiquity: Gold, Labour and Aristocratic Dominance, rev. edn, Oxford: Oxford University Press, 2007, 101–70; terminology is discussed in Laniado, Notables, 171–200.

13. Liebeschuetz, Antioch, 182–6.

14. C. Rapp, Holy Bishops in Late Antiquity, Berkeley: University of California Press, 2005, 211–34.

15. MAMA 3.197 with discussion in Liebeschuetz, Decline, 55–6, 110–11.

16. Liebeschuetz, Decline, 124–36; S. T. Loseby, ‘Decline and change in the cities of late antique Gaul’ in Krause and Witschel, Stadt, 67–104, at 83–97.

17. Liebeschuetz, Decline, 123, 407 (although in the west, the proliferation of counts was designed to counter this). Laniado (Notables, 211–14) has argued for the existence of a ‘conseil des notables’, but the evidence he adduces is extremely thin; cf. J. H. W. G. Liebeschuetz, ‘Transformation and decline: are the two really incompatible?’ in Krause and Witschel, Stadt, 463–83, at 470–1.

18. Y. Tsafrir and G. Foerster, ‘Urbanism in Scythopolis-Bet Shean in the fourth to seventh centuries’, DOP 51 (1997), 85–146, at 99–106; A. Raban and K. G. Holum (eds), Caesarea Maritima: A Retrospective after Two Millennia, Leiden: Brill, 1996; K. G. Holum et al. (eds) Caesarea Papers 2, Portsmouth, RI: JRA Supp. 35, 1999.

19. C. Roueché, Aphrodisias in Late Antiquity: The Late Roman and Byzantine Inscriptions, 2nd edn: http://insaph.kcl.ac.uk/ala2004, Introduction 18; C. Ratté, ‘New research on the urban development of Aphrodisias in late antiquity’ in D. Parrish (ed.), Urbanism in Western Asia Minor, Portsmouth, RI: JRA Supp. 45, 2001, 116–47; J. S. Crawford, The Byzantine Shops at Sardis, Cambridge, MA: Harvard University Press, 1990.

20. M. Waelkens et al., ‘The late antique to early Byzantine city in southwest Anatolia – Sagalassos and its territory: a case study’ in Krause and Witschel, Stadt, 199–255.

21. E. A. Ivison, ‘Burial and urbanism at late antique and early Byzantine Corinth (c. AD 400–700)’ in Christie and Loseby, Towns in Transition, 99–125; Saradi, Byzantine City, 239–42.

22. A. G. Poulter, Nicopolis ad Istrum – The Roman, Late Roman and Early Byzantine City: Excavations 1985–1992, London: JRS Monographs 8, 1995.

23. C. Lepelley, Les cités de l’Afrique romaine au Bas-Empire, Paris: Études augustiniennes, 1979–81, with his recent updating ‘La cité africaine tardive, de l’apogée de IVe siècle à l’effondrement du VIIe siècle’ in Krause and Witschel, Stadt, 13–32; also A. Leone, Changing Townscapes in North Africa from Late Antiquity to the Arab Conquest, Bari: Edipuglia, 2007, 82–96.

24. A. Mahjoubi, Recherches d’histoire et d’archéologie à Henchir el-Faouar (Tunisie), Tunis: Publications de l’Université de Tunis, 1978.

25. L. Maurin, ‘Thuburbo Maius et la paix vandale’, Cahiers de Tunis 15 (1967), 225–54.

26. D. J. Mattingly and R. B. Hitchner, ‘Roman Africa: an archaeological review’, JRS 85 (1995), 165–213, at 212.

27. D. Pringle, The Defence of Byzantine Africa from Justinian to the Arab Conquest, Oxford: British Archaeological Reports, 1981; Leone, Changing Townscapes, 187–98.

28. M. Kulikowski, Late Roman Spain and its Cities, Baltimore: Johns Hopkins University Press, 2004, 85–129, with an overview in his ‘The late Roman city in Spain’ in Krause and Witschel, Stadt, 129–49.

29. Loseby, ‘Cities of late antique Gaul’, 73–5, 80–1.

30. M. Heijmans, Arles durant l’antiquité tardive, Rome: École française de Rome, 2004, 367–71.

31. S. T. Loseby, ‘Marseille: a late antique success-story?’, JRS 82 (1992), 165–85.

32. B. Ward-Perkins, From Classical Antiquity to the Middle Ages: Urban Public Building in Northern and Central Italy, AD 300–850, Oxford: Oxford University Press, 1984, 14–37, 179–86.

33. T. S. Brown, Gentlemen and Officers: Imperial Administration and Aristocratic Power in Byzantine Italy, AD 554–800, Rome: British School at Rome, 1984, 39–45.

34. Cf. Wickham, Framing, 672–3, who lists eight different features which might be characterised as symptomatic of urban decline, though need not indicate ‘urban weakness’. For a vigorous defence of the terminology of ‘decline’, see Liebeschuetz, Decline, and ‘Transformation and decline’.

35. Loseby, ‘Cities of late antique Gaul’, 76–9; Waelkens et al., ‘Sagalassos’, 220; Kulikowski, Late Roman Spain, 210.

36. P. Brown, Poverty and Leadership in the Later Roman Empire, Hanover, NH: University Press of New England, 2002; R. D. Finn, Almsgiving in the Later Roman Empire: Christian Promotion and Practice, 313–450, Oxford: Oxford University Press, 2006.

37. B. Ward-Perkins, ‘Reconfiguring sacred space: from pagan shrines to Christian churches’ in G. Brands and H.-G. Severin (eds), Die spätantike Stadt und ihre Christian-isierung, Wiesbaden: Reichert, 2003, 285–90; Kulikowski, Late Roman Spain, 220–40.

38. On Augustine’s response to the challenge of spectacles, see further R. A. Markus, The End of Ancient Christianity, Cambridge: Cambridge University Press, 1990, 107–23, and D. G. Van Slyke, ‘The devil and his pomps in fifth-century Carthage: renouncing spectacula with spectacular imagery’, DOP 59 (2005), 53–72.

39. Ward-Perkins, From Classical Antiquity, 92–118.

40. Alan Cameron, Circus Factions: Blues and Greens at Rome and Byzantium, Oxford: Clarendon Press, 1976, 157–83; C. Roueché, Performers and Partisans at Aphrodisias in the Roman and Late Roman Periods, London: JRS Monographs 6, 1993.

41. The heading of this section is taken from IAph2007 8.407 (from a fragmentary inscription at Aphrodisias honouring an individual [perhaps a governor] who, probably during the fifth century, ‘drove out city-destroying civil strife’: discussion in Roueché, Aphrodisias, V.37).

42. P. Garnsey and C. Humfress, The Evolution of the Late Antique World, Cambridge: Orchard Press, 2001, 110–14; cf. also pp. 62–3 above.

43. Marcellinus, Chron. Marcell. s.a. 409, Chron. Pasch. s.a. 412, Cod. Theod. 14.16.1 (409), 14.16.3 (434), with discussion in Garnsey and Humfress, Evolution, 112–13. Malalas, Chron. 488, records a shortage of bread in 556 which led to chants against the emperor in the hippodrome, though not, apparently, to violence.

44. D. R. French, ‘Rhetoric and the rebellion of AD 387 in Antioch’, Historia 47 (1998), 468–84.

45. P. Garnsey, ‘Religious toleration in classical antiquity’ in W. J. Sheils (ed.), Persecution and Toleration (= SCH 21), Woodbridge: Boydell, 1984, 1–27.

46. For discussion of some of the issues, see H. A. Drake, ‘Intolerance, religious violence, and political legitimacy in late antiquity’, Journal of the American Academy of Religion 79 (2011), 193–235; also M. Gaddis, ‘There is no crime for those who have Christ’: Religious Violence in the Christian Roman Empire, Berkeley: University of California Press, 2005.

47. Cf. also August. Epp. 90–1, 103–4, for similar violence between pagans and Christians in the city of Calama in 408.

48. E. J. Watts, Riot in Alexandria: Tradition and Group Dynamics in Late Antique Pagan and Christian Communities, Berkeley: University of California Press, 2010.

49. The ongoing controversy between Catholic and Donatist Christians in north Africa generated much violence, particularly associated with the so-called circumcelliones, but their activities focused on rural areas, rather than cities: see B. D. Shaw, Sacred Violence: African Christians and Sectarian Hatred in the Age of Augustine, Cambridge: Cambridge University Press, 2011.

50. Garnsey and Humfress, Evolution, 143–5.

51. Details and references in N. B. McLynn, Ambrose of Milan: Church and Court in a Christian Capital, Berkeley: University of California Press, 1994, 315–30.

52. This was the achievement of Cameron, Circus Factions; see also Roueché, Partisans and Performers.

53. Liebeschuetz, Decline, 253, 259, notes, e.g., that factional violence did not always involve Blues against Greens: sometimes the two groups co-operated, or one remained neutral, and sometimes they did show religious preferences.

54. Liebeschuetz, Decline, 213–18, 248–57.

55. In the episode quoted above, it is an aristocrat, Plato, who is described as patron of the Greens, reflecting the fact that, unusually for an emperor in this period, Anastasius took a deliberate decision early in his reign not to support either the Blues or Greens (Malalas, Chron. 393).

56. M. Whitby, ‘The violence of the circus factions’ in K. Hopwood (ed.), Organised Crime in Antiquity, London: Duckworth, 1999, 229–53; M. Whitby, ‘Factions, bishops, violence and urban decline’ in Krause and Witschel, Stadt, 441–61.

57. For good overviews of the subject in late antiquity, see Averil Cameron, ‘Education and literary culture’ in CAH² 13.665–707, and R. Browning, ‘Education in the Roman empire’ in CAH² 14.855–83.

58. Prosopography: R. A. Kaster, Guardians of Language: The Grammarian and Society in Late Antiquity, Berkeley: University of California Press, 1988, 20–1, 463–78; rhetoricians: Jones, LRE, 998.

59. For the status of Latin in the sixth-century east, see the references at p. 256 n.25 below.

60. P. Brown, Power and Persuasion in Late Antiquity, Madison: University of Wisconsin Press, 1992, ch. 2; E. Watts, City and School in Late Antique Athens and Alexandria, Berkeley: University of California Press, 2006, 7–11.

61. R. Cribiore, The School of Libanius in Late Antique Antioch, Princeton: Princeton University Press, 2007. For a western, Latin perspective, there is much evidence about Bordeaux as an educational centre in the fourth-century: see Kaster, Guardians, 100–6.

62. There is a recent translation of the life (which is preserved only in a Syriac version) by L. Ambjörn (Piscataway, NJ: Gorgias Press, 2008), with discussion of the author and work in G. Greatrex et al., The Chronicle of Pseudo-Zachariah Rhetor: Church and War in Late Antiquity, Liverpool: Liverpool University Press, 2011, 4–8, 15–18.

63. Kaster, Guardians, 74.

64. Text and commentary in N. G. Wilson, St. Basil on the Value of Greek Literature, London: Duckworth, 1975; discussion in Kaster, Guardians, 77–8, N. McLynn, ‘The manna from uncle: Basil of Caesarea’s Address to Young Men’ in C. Kelly et al. (eds), Unclassical Traditions I, Cambridge: Cambridge Philological Society, 2010, 106–18.

65. Kaster, Guardians, 80–95; text and translation of Augustine, De Doctrina Christiana, by R. P. H. Green (Oxford: Oxford University Press, 1996).

66. Cf. Y. Hen, Roman Barbarians: The Royal Court and Culture in the Early Medieval West, London: Palgrave Macmillan, 2007, and pp. 184–6 above.

67. J. J. O’Donnell, Cassiodorus, Berkeley: University of California Press, 1979 (www9.georgetown.edu/faculty/jod/texts/cassbook/toc.html), 177–222; translation and discussion of the Institutions of Divine and Secular Learning by J. W. Halporn and M. Vessey (Liverpool: Liverpool University Press, 2004).