Even in its later years the Roman state was ideologically and politically resilient. The narrative of events in the sixth century after the setbacks of the 540s demonstrates that its rulers and their military forces had not lost their sense of purpose. They were still focused on Rome's hegemonic role, including the assertion of power in the western Mediterranean by establishing enclaves of territory in Italy, Spain, and Africa, and by facing down the rivalry and challenge of the Sassanians in the east (see pp. 424–38). Nevertheless, the empire, measured in economic and military terms, had become markedly weaker as a political force during this period. Did deeper structural factors, largely beyond the control of the state, contribute to this weakness, or were they indeed its root cause? At the heart of this is the question whether the waves of bubonic plague that swept across Europe, the Mediterranean, and the Near East between 542 and the middle of the eighth century cumulatively played a decisive role in the decline of Rome. Wickham systematically and consistently excludes mass mortality resulting from plague from his explanations of social and economic change. In this he joins the company of many archaeologists specializing in the late Roman period, who regularly omit to discuss or downplay the significance of the plague. Other research, notably several papers assembled in Lester Little's volume on Plague and the End of Antiquity, give plague a central explanatory place in accounts of the period.35 If arguments that maximize the impact of the plague are to be upheld, it is essential to be able to demonstrate that plague mortality really did result in the abandonment, collapse, or degradation of settlements across at least most of the Mediterranean and Near Eastern region from the middle of the sixth century.
It is difficult to demonstrate or refute claims about the effects of plague on the basis of archaeological evidence. Direct evidence remains very sparse. Certain telltale signs, such as the hasty mass burial of bodies, diverging from normal local funerary practice, are a reasonable criterion to apply, but such discoveries by their nature are sporadic and could have other, strictly localized explanations.36 There is some evidence for a more subtle and pervasive change in burial practice, evolving as a response to mass mortality caused by recurrent plague outbreaks. This is the shift towards the frequent re-use of a single burial site, effectively the practice of piling bodies on bodies in graves previously reserved for one person.37 But the question has been little investigated. Modern DNA analysis has the potential to detect the plague bacillus in well-preserved dental remains, and there has been some success in identifying victims of the Black Death by this means, but testable samples for the plague of late antiquity are very rare.38 Thus it is necessary to turn to proxy indicators supplied by archaeology, namely the evidence for the occupation and abandonment of settlements in the relevant periods, the form that these settlements took, and any clues derived from their location or material remains that these should be interpreted as a response to plague conditions.
Information obtained from survey and excavation only becomes fit for purpose when it meets certain conditions: that it is sufficiently comprehensive to be representative of the regional situation, and that it is adequately dated by reliable criteria, in most cases by the chronology of the fine pottery. Further important distinctions need to be established about settlements during the period after 550. Did new construction of domestic housing occur on any significant scale? If not, did earlier structures continue to be occupied, or were they abandoned? Is church building, which is widely attested in the later sixth century, an indication of an active and viable local population, and therefore a sign of economic vitality, or simply evidence for the piety of desperation, as communities attempted to appease God's wrath and avoid his punishment by pious works? Is there any reliable indication that communities successfully moved from place to place, or even region to region, to avoid the plague? From the outset it has to be said that none of these questions is easy to answer. Matters are made harder by the fact that rigorous archaeological study of late antiquity and the early Middle Ages, especially in the Mediterranean regions, is a relatively young discipline, and much possibly relevant information has simply been too casually recorded to be used with any confidence, especially when fine judgments about chronology can critically affect the argument.
Despite the inherent interpretative problems, the changing nature of both urban and rural settlement has been an important focus of modern attention. One of the critical chapters of Wickham's analysis of the period from 400 to 800 is his survey of rural settlement and village societies.39 In the territories of the former western empire, a key question concerns the disappearance of the Roman landed estate centers, the villas. As far as survey and excavation has been able to demonstrate, the villas of the northwest European regions, Britain and northern Gaul, began to disappear from the rural landscape in the late fourth century and had vanished entirely by c.450. In the west Mediterranean region the period of abandonment came around a century later. In southern Gaul most villas ceased to function in the sixth century, and all traces had vanished by 600. In Spain abandonment began in the late fifth century, with very few surviving beyond the sixth century. In Italy villa settlements cease in the early or mid-sixth century. In Africa, the end of the villa settlement pattern is placed around 550. Subsequently, to judge from the sparsity and often minute size of their rural settlements, these west Mediterranean regions remained in a condition of acute recession until at least 700, and sometimes much later in the eighth or ninth centuries. Such seventh century settlements as can be identified are small but scattered, and according to Wickham's careful economic analysis it is difficult to see them as forming part even of a rudimentary site hierarchy, an important criterion for a dynamic and differentiated society with an active trading economy or significant state institutions. After about 550 these regions, evidently, were both depopulated and impoverished.
Regions outside the west Mediterranean zone also show the same features. In Denmark, which had always been outside the Roman Empire, society was based on gradually evolving village settlements but recession is clear in the period c.550–700, before there are signs of an upturn, some urbanization, and evidence for the introduction of imported goods. In Britain, Wickham characterizes the Anglo-Saxon settlement of the late fifth and sixth centuries as being on a small political scale in a landscape dominated by an economic crisis. There was no general upturn until after 700. On the neighboring continent, even in the west Frankish Merovingian heartland between the Seine and the Rhine, there is no sign of any large-scale or complex construction in settlements before 650, and the colorful and often bloody feuds and rivalries of the Frankish kings in the later sixth century, recorded by Gregory of Tours, have left virtually no material trace in the form of buildings. Overall only around 50 percent of Frankish sites were continuously occupied between the late Roman and the early medieval period. Wickham concludes that there had been a population decline of around 50 percent, which he evaluates as a drastic reduction but not a demographic collapse. Tentatively he infers that the evidence from Italy and Spain points the same way.40
The cities of the west were in comparable decline after the mid-sixth century. Urban degeneration is obviously a large and complex subject, but its visible symptoms are characterized in deliberately simple terms by Wolf Liebeschuetz: “the increasing prevalence of more primitive building techniques and the appearance of large empty spaces in what had previously been built-up areas”.41
Many of the cities of Africa, which had flourished and maintained their monumental character through the fifth century both before and after the Vandal conquest, received new fortifications after the reconquest. This development is analogous to the building of walled defensive strongholds in the lower Danube region by Anastasius, Justin, and Justinian, usually on the sites of former towns or legionary fortresses. However, the function of these fortified centers was primarily military, and their construction was obviously initiated by the regime in Constantinople. Moreover, these fortifications enclosed only small areas within the settlements, a likely, but not certain, indication of reduced population levels.42 Public building otherwise was restricted to a few churches, while private houses were often subdivided, or of much simpler construction.43 Crucially the whole region's economy, which had thrived as Africa had exported an agricultural surplus and manufactured products, notably pottery, through the western Mediterranean, shrank to the level of very localized intra-regional exchange. After the Justinianic reconquest the cities of Africa, without entirely disappearing, became impoverished.
Italian cities have been more intensively studied than those of any other region. Many of the urban centers of the late Roman period continued to be occupied and in due course formed the basis of the settlement pattern of early medieval Italy. They had, however, undergone a drastic transformation. Wickham notes that after c.550 “even those where urban activities survived…were very much poorer, with most administrative functions (such as street cleaning) reduced to a minimum, and with often very simple buildings, or Roman buildings fairly crudely re-used.” This decline was a universal phenomenon, coinciding with the general devastation of the Gothic War, but it is hard to see how the war as such could have been the cause of the pervasive impoverishment of Italy's cities. The forces involved were simply too small, and many of the campaigns too localized, to have had such a widespread and lasting impact. High mortality caused by the plague, with catastrophic disruption of all the forms of urban administration, seems a more likely catalyst than warfare for this implosion of civic life, which coincided with a severe decline in the rural economy.44 One of the most detailed accounts of the ravages of the plague was written by Paul the Deacon, the eighth-century chronicler of the Lombard kingdom. His account particularly emphasizes the desertion of settlements:
For as common report had it that those who fled would avoid the plague, the dwellings were left deserted by their inhabitants, and the dogs only kept house. The flocks remained alone in the pastures with no shepherd at hand. You might see villas or fortified places lately filled with crowds of men, and on the next day, all had departed and everything was in utter silence. Sons fled, leaving the corpses of their parents unburied; parents forgetful of their duty abandoned their children in raging fever. (Paul the Deacon, History of the Lombards II.4, trans. Foulke)
Paul's History noted further outbreaks of plague at Rome in 590, at Ravenna and Istria in 593, and at Rome and Ticinum (Pavia) in 680. The last of these accounts specifically described the abandonment of the urban settlement: “this pestilence also depopulated Ticinum so that all citizens fled to the mountain ranges and to other places and grass and bushes grew in the market place and throughout the streets of the city” (History of the Lombards III.24; IV.4; VI.5).
Gaul as a whole in late antiquity was less urbanized than Italy.45 The centers north of the Loire, Paris, and above all Trier and Cologne, which were emphatically Roman imperial creations, shrank in size and lost much of their monumental character (although ancient buildings must have survived) as their territory was ceded to the Franks through the fifth century. Only Metz appears to have re-established itself as a significant urban center around 560, as a direct result of Frankish royal patronage (see pp. 388–9). The cities south of the Loire in the late Roman period – Lyons, Clermont Ferrand, Poitiers, Arles, Aix-en-Provence, Marseilles, and Toulouse – re-emerged as key centers of the medieval period, but in general all these cities were reduced to small, usually fortified strongholds in the seventh and eighth centuries. Very few of them have been sufficiently well investigated for the process of decline to be followed in detail, but two important cases show a marked reduction of settlement size in the sixth century, Arles around 550 and Lyons around 600.46 Marseille is an important exception to this pattern, in that it remained an important focal point of administrative and above all commercial activity in the later sixth and seventh centuries. However, Marseille was the main entrepôt to the Mediterranean for the Frankish kings, and it is clear that they invested considerably in maintaining the urban infrastructure of this key port. Even here the late sixth-century development of an important zone in the east part of the city around the inner harbor appears to be counterbalanced by the abandonment of other intramural occupation areas, and this prefigured a major gap in the material record between the mid-seventh and tenth centuries.47 We know, as a matter of fact, that plague was brought to Marseilles in 588 and flared up repeatedly over the following months and years up to 594:
Meantime a ship from Spain put in at the port with its usual wares and unhappily brought the seed of this disease. And many citizens bought various merchandise from her, and one household in which were eight souls was quickly left vacant, its inmates all dying of this plague. But the fire of the plague did not at once spread through all the houses, but after a definite time like a fire in standing grain it swept the whole city with the flame of disease…The plague passed away in two months, and when the people, now reassured, had returned to the city the disease came on again and they who returned perished. Later on the city was many times attacked by this death. (Gregory of Tours, Hist. 9.22).
Marseille's continuing role as a port city rendered it extremely vulnerable to plague, although Merovingian royal support was available to restore its fortunes.
Summarizing the evidence for sixth-century Gaul as a whole Liebeschuetz concludes that most of the cities were very small, some inhabited by fewer than a thousand persons even on an optimistic reckoning, and that much urban space was occupied by a growing number of churches.48
The evidence for settlement density and occupation in post-Roman Spain is hard to evaluate. In general, as in Gaul, urban centers were small and thinly scattered over a very large land area, with poor inter-regional connectivity. For this reason Wickham emphasizes that late and post-Roman Spain played only a marginal role in the emergence of a new early medieval order in Europe. From another perspective Michael Kulikowski has argued that although Spain was affected by the plague, it was too thinly populated for it to host the disease for extended periods, and thus mortality may have been proportionately less than in other regions.49
The picture from the eastern Mediterranean and the Near East is more complicated and controversial. It is particularly difficult to integrate the information about regional settlement patterns provided by archaeological surveys as projects have adopted different and sometimes conflicting pottery chronologies, especially as far as the Near Eastern evidence is concerned.50 Greece and the southern Balkans were in evident decline through the fifth and sixth centuries, not least due to their basic insecurity against barbarian incursions from the north. The vulnerability of Thessalonica, the most important city of the region, is vividly illustrated by several episodes in the Miracles of St Demetrius, which documented both Slavic attacks and outbreaks of plague in the later sixth century. Plague ravaged Thessalonica and Thrace in 597–8, and doubtless also in earlier years of the sixth century. Amid general evidence for very sparse rural occupation at the end of the sixth century, the appearance of new hand-made pottery types has been identified as an indication of Slav settlements, apparently moving into a largely desolate landscape.51
Asia Minor, which was secure from barbarian invasion, presents a quite different appearance in the fifth and early sixth century. The major cities of the Aegean region and provincial capitals such as Aphrodisias in Caria, Sagalassos in Pisidia, and Ancyra in Galatia were in good shape at the beginning of the sixth century, although the public areas and buildings of some previously substantial cities, such as Aezani in Phrygia, seem to have been neglected or abandoned in the course of the fifth century.52 The south coastal regions of Asia Minor enjoyed an astonishing economic boom through the fifth and into the sixth century, especially demonstrated by the building of houses, churches, and bath houses. In Lycia the settlements on the mainland and on offshore islands presented a dense ribbon of ugly rubble and mortar housing, interspersed with churches and occasionally bath houses, which were functional, unspectacular, but robust enough to have survived until today. The stories found in the very detailed life of St Nicholas of Holy Zion, a monastery complex in the territory of Myra, which was written in the 560s, illustrate the economic interdependence between the coast and the hinterland.53 A very similar pattern, including the excellent preservation of many structures and buildings from late antiquity, can be seen farther east in Cilicia, whose ports shipped the products of the hills, primarily olive oil, to Constantinople (see pp. 365–6). The economic upsurge of the fifth and sixth century also extended to the cities and countryside of Pamphylia where olives were the major cash crop.54 However, there is a major question about when this relatively short-lived boom came to an end. None of the architecture of Lycia and Cilicia is sufficiently distinctive for more than generalized fifth/sixth-century dating to be possible. The exhaustive territorial survey around the Lycian city of Kyaneai has produced almost nothing that can confidently be dated to the seventh century.55 An economic and demographic collapse due to the plague cannot be ruled out. Equally the catalyst for decline may have been the warfare of the early seventh century, including the fall of Rhodes to the Sassanians in 623, although no Persian attacks on Lycia and Cilicia are explicitly attested in the meagre narrative sources.
Most of the evidence from the major cities of the interior of Asia Minor in the sixth century is similarly ambiguous. Clive Foss demonstrated in a fundamental study that the collapse of urban life in Anatolia was not the result of Arab invasions in the eighth and ninth centuries, but had already occurred in the early seventh century, when the destructive effects of Sassanian attacks become clear in the archaeological record.56 However, it remains an open question how much recession and decline had already occurred from the middle of the sixth century. In city sites where recent excavations have paid specific attention to this issue, some important evidence points in this direction. At Pergamum a new Byzantine fortification wall seems to have been undertaken in the early Justinianic period, but was then left incomplete. No new public or domestic construction is recorded at Aphrodisias after 550. At Aezani, although decline is already apparent in the fifth-century record, the city became desolate in the second half of the sixth century, with streets and buildings abandoned.57 The fullest record comes from the excavations at Sagalassus. The city was struck by a major earthquake around 500. The immediate response was positive and the streets and major public buildings such as the nymphaea on the two city agoras, the source of the public water supply, were restored. However, much of this restoration work was incomplete and ceased before the middle of the sixth century. Although the site remained occupied after 550, activity was at a low level, with no new building before the city was finally abandoned in the early seventh century. The repair and paving of the colonnaded streets was abandoned, the aqueducts no longer served the fountain houses, which became locations for rubbish dumps, and the shops and workshops around the city agoras were abandoned. Sagalassos provides a notably well dated and documented example of abrupt civic decline, beginning around the mid-sixth century, and the plague has been seen as the most likely cause for this collapse of civic life.58 The same pattern, moreover, has been identified in all the other major excavated cities of western Asia Minor, although only the more recent excavations have provided reliable chronological data.59 The evidence is compatible with the hypothesis that the plague dealt a mortal blow to the life of these cities after its first outbreak in 542. This provides at least a partial explanation of the ease with which the Sassanians were able to pick off the Anatolian cities and advance to the Bosporus between 609 and 623. One source, the Life of Theodore of Sykeon, demonstrates that the countryside was not depopulated, and that the villages of the northwest contained active communities of peasant landowners. However, their evident resilience may be due to the fact that resources were now controlled by a smaller rural population, and one which was able to stand up in the face of the now weakened urban elites.60
The evidence from the Near East, at first sight, presents a very different pattern, and is highly controversial. It should be noted that the two most recent modern syntheses of the evidence in English, by Liebeschuetz and Wickham, reach opposed conclusions. Liebeschuetz argues that although there was settlement continuity across the region from c.500 to 800, all the cities experienced a major drop in population and economic activity after 550.61 Wickham on the contrary emphasizes the evidence, especially from the southern Levant, for flourishing rural and urban communities through the late sixth and seventh centuries, as the Caliphate based in Damascus replaced the Roman state as the dominant political force.62
Outside the major cities debate has focused on three areas containing the most extensive non-urban ruins that have survived from antiquity: the limestone hills east of Antioch and north of Apamea in northern Syria, the basaltic country of the Hauran astride the Syrian–Jordanian border, and the northern extremity of the Negev, southeast of the port city of Gaza. These areas of dense village or small-town settlements are located between major cities which were also occupied throughout late antiquity into the period of the Arab conquests. The cities for which the best archaeological information is available include Antioch and Apamea in the north, and Caesarea, Scythopolis (Beth Shean), Pella, Gerasa, and Bostra, which form a belt across south Syria. They have been a major focus for scholarly attention in the study especially of “late Late Antiquity” roughly defined as the period from 500 to 800, spanning the Islamic conquest of the Roman Near East.63
The three village and small-town regions have been cited as providing strong counter-indications to the decline observed in the western Mediterranean and Asia Minor. Tchalenko's ground-breaking survey of the northern limestone massif indicated that there had been a building boom through later antiquity until around 550. Thereafter it is suggested that houses continued to be occupied and used until 600, and church building combined with pottery imports as late as 650, thus implying that the region was occupied through the seventh century. These conclusions rely heavily on the only significant excavation in the region, of three village houses at Déhès. This showed that the fourth-century structures were rebuilt in the early sixth century using new construction techniques, but subsequently these buildings appear to have been combined into one. Evidence for later occupation has been found, but on a small scale. Phocaean red-slip ware, which had been one of the main forms of imported fine pottery, is no longer present at the site after 550, and the chronology of the later pottery is extremely uncertain. There are no signs of intense activity on the site in the seventh and eighth centuries, apart from bronze coin finds, which actually increase in the Muslim period, and show that Déhès continued to be occupied at a modest level and was integrated into the economic or fiscal system of the early caliphate.64
In the Hauran, the key excavated site outside the city of Bostra is Umm el-Jimal in north Jordan, a thriving late Roman town with as many as fifteen identified churches dated to the fifth and sixth centuries, and over one hundred houses as well as a military castellum. The site remained an important center in the Ummayad period in the seventh century, but with a diminished population. However, under Muslim occupation the previous domestic structures were literally stripped down to their foundations, and although the new settlement took shape within the shell of the late Roman site, it stretches the definition to regard this as evidence for substantial site continuity. The new Ummayad settlement, which occupied much less space than its predecessor, was evidently founded on the initiative of the new Arab power center at Damascus. As at Déhès the population level had declined significantly.65
In the Negev region along the overland route that ran southeast from Gaza, there are several late Roman villages or towns, with robust stone-built houses and churches, which have survived until today – a counterpart to the survival of similar sites in the North Syrian limestone country. However, they present similar problems of interpretation. The most completely preserved settlement is the village of Shivta (ancient Sobata, see pp. 361–2), but unfortunately the excavation records are now lost and may never have been adequate to answer the crucial questions about continuity of settlement and population density after the mid-sixth century. The settlement plans show a small mosque, probably to be dated around 700, at the center of the settlement alongside its three large churches. However, there is no way of telling whether this now served all the religious needs of a much reduced population, or simply those of a Muslim minority alongside substantial numbers of Christians who continued to worship in the existing churches. There is no means of telling how many of the village's houses and churches were actually occupied after 550 or 600. Potentially critical evidence comes from a group of about twenty now lost inscribed gravestones and building inscriptions relating to the churches, which carry era and indiction dates. However, it is ambiguous whether Shivta used the era of the nearest large city of Gaza, beginning in 60 BC, or that of the province of Arabia, beginning in AD 106. In the former case the inscriptions date between AD 423 and 513, in the latter between 589 and 679.66 At all events the archaeological exploration of the large villages or small towns of the northern Negev region provides a valuable context for the evidence of the papyri from Nessana that around 700 this was a region where the social and legal structures of the late Roman Empire remained fully recognizable after the Arab conquests (see p. 474).
The monumental appearance of the major cities of the Syrian and Palestinian Near East changed in important ways between AD 400 and 800, although significant settlement continuity is undeniable, especially in the south. After 550 Roman public buildings fell out of use, many more churches were built, and domestic housing and streets encroached on former public areas. In North Syria the great cities of Antioch and Apamea were in recession by the mid-sixth century and never recovered their former stature. Well-documented disasters, including the earthquakes of 526 and 528, the Persian attacks of 531, 540, and 573, as well as plague outbreaks between 542 and 594, proved cumulatively fatal. It seems impossible that the exceptional rural prosperity of the villages of the limestone massif, which mostly belonged to the territories of Antioch and Apamea, could outlast the collapse of the cities, which provided the main market for their surplus production of olive oil and other commodities. The evidence for some continued settlement at Déhès is not sufficient to refute the hypothesis that the whole of northern Syria underwent a sharp reduction in population and economic activity well before the Muslim conquests.
Changes during the crucial period of the later sixth century can also be detected in the southerly regions. The evidence points to a transformation in the character of urban settlements, although not to a catastrophic decline. The cities of the Decapolis and the Hauran region did not disappear in the transition from late Roman antiquity to the Ummayad caliphate, although there was no major new public building after 540 at either Scythopolis or at Gerasa, two of the best documented sites. Scythopolis is estimated to have had a population of more than 30,000 in the early sixth century, but this became much smaller during the seventh and eighth centuries, when the main city street was reduced from twelve to three meters in width, and débris rapidly accumulated in courtyards and at street edges. As at Sagalassos in Asia Minor a main street which had been restored in the Justinianic period, as well as all the city's secular public buildings, had been abandoned well before the end of the century.67 There was an attempt to put new life into the settlement when a new soukh (bazaar) was built by caliph Hisham between 724 and 743, but this revival was ended by the devastating earthquake of 748. East of the Jordan river the development of late Roman Gerasa is well documented both archaeologically and epigraphically. Three churches were built at the beginning of Justinian's reign between 529 and 533, and mosaic inscriptions documenting the dedication of new churches are recorded for 559, 565, 570, and 611. There was also new activity under Muslim rule, including the building of a small mosque. Many small workshops for ceramic and metalwork production, also observable at Scythopolis, attest the presence of a significant artisan population. However, this evidence needs to be set alongside signs of population decline, including the abandonment of houses and workshops around the hippodrome before the mid-sixth century. The immediate reason is not far to seek. Archaeologists report that skeletons were found piled up in huge numbers beneath the seats of the hippodrome, clear evidence of emergency burial during the plague, presumably in 542. Plague mortality is also the obvious explanation for the re-use of late Roman graves at Pella so that corpses could be crammed in without ceremony, wherever space could be found.68
Intensive church building in the later sixth and seventh centuries was, however, a widespread phenomenon both in southern Syria and further afield in Asia Minor. Mosaic inscriptions from Jordan as well as the Life of Theodore of Sykeon show many new churches and monasteries being built in this period.69 As in Europe in the years of the Black Death, the drastic reduction of the population, although it led to declining economic activity, also concentrated wealth and resources into fewer hands. Churches and monasteries were evident beneficiaries of this evolution. As institutions they now acquired land that had been abandoned in a period of demographic collapse, and also profited from pious donations by the survivors of the epidemic and other disasters.
In summary, the cities and villages of northern Syria either collapsed or declined steeply after the mid-sixth century. The picture in southern Syria and northern Arabia is less clear cut, with evident continuity of settlement to the eighth century, but here too population declined and communities were materially transformed. There is very little evidence for new construction, apart from churches, after 550. Cumulatively this evidence provides a very strong case for seeing the bubonic plague of 542, and its regular recurrence until the middle of the eighth century, as the major cause of demographic decline, with a catastrophic impact on population levels, economic productivity, and the administrative capacity of communities to maintain organization at its former level.
The communities and regions which defy this pattern are almost all to be found in the southern Levant and Egypt. Since these regions were just as exposed to the plague as the rest of the empire, and mortality will have been comparably high, explanations need to be found for this apparent resilience. Two considerations seem particularly important. The first can be expressed in two words: size matters. The cities and villages of Syria and Egypt had grown spectacularly in late antiquity. The north Syrian limestone massif may have had a population density higher than that of the countryside in contemporary western Europe. The Egyptian village of Aphrodito had grown into a substantial town, with over ten thousand inhabitants, complex social stratification, and diverse economic activity; Scythopolis (and by analogy other major cities) had a sixth-century population far exceeding the average city of the ancient world, and incomparably larger than the small urban conglomerations of the western empire. When they were decimated by the plague, reduced to half their former size or even less, there remained a sufficient population to maintain the community in viable form. Indeed, in cities such as Constantinople or other major centers, where economic structures remained intact, surviving workers could charge more for their labor and thus prosper in adversity. On the land, provided that there was still a large enough work force to maintain agricultural activity, individuals and families could make a living. Although population losses in Syria and Egypt are likely to have been proportionately as great as elsewhere in the empire, the sheer size of their existing cities and villages, and the complex and diverse commercial and administrative infrastructures that they had evolved in the boom of the fifth and early sixth centuries, enabled them to pull through the demographic devastation which laid other regions waste.
The other explanation for the resilience of the southern Levant was surely political. These were the earliest and most important provinces of the Arab conquerors, who had every reason to ensure that the cities and villages of Palestine, Syria, and above all Egypt survived as sources of wealth and tax revenue. Populations could be maintained by bringing in new settlers from the Arab south. Archaeological evidence suggests that the Ummayad caliphs took active measures to revitalize urban life at Umm el-Jimal, Gerasa, Scythopolis, and elsewhere, just as they ensured that the infrastructure necessary for tax collection was kept in place. Another relevant factor was the economic re-orientation of the southern Levant towards the centers of Arab power. The Negev towns, which lay directly on the route leading southeast from Gaza, could now benefit from the new economic opportunities offered by access to the Arab heartland in the Hejaz region. As happened elsewhere in the Mediterranean the power of the state could be harnessed to shore up a city or region in the face of general decline. Rome and, self-evidently, Constantinople overcame the catastrophe. The economic and political interests of the Frankish kings enabled Metz and Marseille to avoid the fate of most other Gallic cities. On a much larger scale, and with more significant long-term consequences, the towns and cities of the southern Levant and Egypt owed their survival until the eighth century to the political and strategic interests of the early Arab caliphate.
There has been a noticeable polarity in recent explanations of the decline and fall of the Roman Empire by archaeologists and historians, and this has been brought into focus by their respective assessments of the effect of the Justinianic plague (see pp. 412–3). Archaeologists, surveying the material record, have tended to be “continuists,” looking for and emphasizing evidence for the survival and evolution of institutions and communities through and beyond the upheavals of war, natural catastrophe, and epidemic disease. The emphasis of archaeological explanation is usually placed on continuity rather than abrupt or drastic breaks in the sequence of settlement activity. Communities are seen to persist and evolve within a framework of rising and falling economic well-being. Historians working from written evidence are more likely to be “catastrophists” and to identify major events as responsible for large social changes. Traditionally, war and conquest are seen as the largest catalysts for historical change.70
The great plague of late antiquity, however, extending over two centuries across a geographical range which can be traced from the Atlantic to the Persian Gulf and may have been much wider than this, caused incomparably higher levels of mortality than any ancient war. Ancient observers from Procopius and John of Ephesus to Gregory the Great and Paul the Deacon, appalled and helpless, recorded the plague's symptoms and the speed and virulence of its attacks, fully aware of the immediate devastation that it caused. No ancient writer was able to adopt a perspective that surveyed the demographic catastrophe over two centuries, a task that has to be left to modern analysis. The written accounts have been the object of excellent and rigorous modern study. A systematic and dispassionate re-evaluation of the archaeological evidence, addressed to the question of the lasting impact of the plague, is overdue. The preceding pages offer a brief preliminary appraisal of the growing body of archaeological literature and the conclusion that the long-term impact of the plague on the population level and the social and political organization of Mediterranean and Near Eastern societies from 550 to 750 was indeed as devastating as the immediate impact recorded by the contemporary observers.