Why was the Byzantine empire, which survived the rise of Islam and continued to rule relatively extensive territories in Anatolia and the Balkans, apparently so much poorer than its late antique predecessor, or indeed than its Muslim neighbour? Some believe the answer lies in plague or climate change; others, as we have seen, do not. A more straightforward answer would be to cite Persian and then Arab devastation. But Anatolia, in particular, is a large place.Would it not have required remarkably Stakhanovite raiders to have had this sort of global economic impact? The poorer and more primitive world revealed by the Sagalassos survey seems to require a more complex explanation than direct enemy action. One approach, casting back to what has been said above about the late antique economy, is to look to the economic role of the state; but, not surprisingly in view of the disagreements about that role before the seventh century, there is no more unanimity for the period that follows. Was the Byzantine empire of the seventh and eighth centuries a weak state? If the late antique economy was dominated by the demands of the annona, or even if the state’s role was more a matter of organising taxation and a stable currency, did the weakening of the state in the seventh century lead to economic collapse? The persistence of a monetary economy in Constantinople and its apparent withering elsewhere could be seen as fitting this model rather well. The Byzantine state clearly survived and where it had real control, a more complex economy survived too, but its reach did not extend far into its nominal territories and so neither did a monetary economy. On the other hand, one might think about Byzantium during these centuries in terms of a powerful state, which survived the early middle ages by becoming a highly militarised society, based upon efficient and ruthless exploitation of its resources. This was not so much a poor world, as one where resources were focused on a very particular end.47 Comparison with the caliphate does not answer these questions, but it underlines their significance. Recent research has made the wealth of the Islamic world during these centuries steadily more obvious. This was a society with a prolific silver coinage, with a standard of living certainly no lower than that of late antiquity and capable of undertaking at Baghdad and Samarra building projects on a scale that makes the Byzantinist, let alone the historian of the early medieval west, boggle. Was the caliphate fulfilling the economic role once played by the late antique empire and no longer provided by its Byzantine successor? Or was the wealth of this world due to the fact that the caliphate could afford to be less intrusive? To find high-quality seventh- and eighth-century church architecture, it is interesting that one needs to look outside Byzantium: to Syria, Palestine and, above all, to Armenia (see above, pp. 340–1, 344–5 and fig. 22). Is that because of the paradoxical survival of the late antique state in the form of the caliphate, or is it because the church-building classes outside the empire were able to spend their resources on fine architecture rather than footing tax bills?48 Behind these uncertainties is the fact that written evidence for the seventh- and eighth century economy is virtually non-existent and the archaeological evidence, at first sight so secure, is not. Current interpretations rest, as explained in the first section of this chapter, on treating stray coin finds, pottery and buildings as proxy indices of economic activity. But in each case there are grounds for caution. Copper coins of the seventh and eighth century collected as stray finds from the excavations of ancient city sites in western Turkey and Greece are certainly very rare, but the same coins seem not to be so rare when Turkish villagers show what they have picked up from their fields. Is this because cities were no longer so important as central places, and transactions were taking place at other sites, possibly rural fairs?49 The very fact that the empire continued to mint gold throughout this period is worth noting. To do so required taxation in gold and that in turn depended upon the fact that tax-payers could obtain gold coins through trade. Even if most gold coin passed rather as tokens from the emperor to his servants and back again, it must have been supplemented by gold reaching tax-payers by other means. In the west, minting in gold had to be abandoned in the eighth century. Its continued minting in Byzantium may have been largely due to the survival of the tax system, but it presumably also indicates a more than minimal monetary economy. The same inference may be drawn from the survival in Byzantium of a multi-denominational coinage. From the eighth century to the thirteenth, western mints produced only single-denomination silver coins. The emergence of a more complex system in the west was a response to economic growth and the consequent demand for something more practicable and flexible. A two-tier coinage in Byzantium might be explained as a simple consequence of the need to keep minting the ideologically critical gold solidi, while at the same time providing some other lesser-value coin; but the emergence of a successful three-tier system from the 720s, with the silver miliar¯esion (worth twelve to the solidus) supplementing the existing gold and copper coins, likewise suggests a more than minimal monetary economy. It equipped Byzantium with a highly flexible system of coinage.50 The pottery, too, tells an ambiguous story. Certainly an ancient tradition of manufacturing red-slip pottery and distributing it across the Mediterranean world came to an end in the seventh century. The standard types of late antique amphorae similarly disappear during these years. Constantinopolitan white ware, which certainly dates from this period, is neither common nor widely distributed. But the problem is as much an inability to identify seventh- and eighth-century pottery as it is one of absence. The excavations of the Sarac¸hane site in the middle of Istanbul produced a continuous sequence of coins through the dark age; John Hayes’ study of the ceramics has produced a similarly continuous sequence of tableware, cooking pots and carrying vessels. Large water jars commonly found on sites in western Anatolia were certainly in production by the eighth century. Hannelore Vanhaverbeke’s impression from the Sagalassos survey coincides with that of others: that much of the unidentified pottery found on Anatolian sites is locally or regionally produced, some likely dating to the seventh and eighth centuries. In the Levant, the study of local wares has helped transform the picture of the early Islamic centuries. The same may be about to happen for dark age Byzantium.51 Ceramics are manufactured objects; they are also trace elements for trade. If we cannot recognise seventh- and eighth-century pottery we lose this means of plotting movement and have tended to assume the minimal worst. But there are other trace elements.MichaelMcCormick has assembled the evidence for journeys of all sorts in the dark ageMediterranean and what he has found suggests much more activity during this period than has generally been imagined. Peregrine Horden and Nicholas Purcell note the evidence for dark age piracy and point out that piracy will only occur when there is something to steal. These are straws, but they need to be taken into account.52 With buildings too, there is a danger that we currently underestimate the evidence already to hand. Seventh- and eighth-century Ephesos, for example, was not on the same scale as its late antique predecessor, yet the huge church of St John was kept standing and the cut-down church of St Mary remained in use. In any case the largest building project of these years was not a church but the circuit of walls. Although they enclose only a small part of the ancient city, at something over 2.5 kilometres long and 3.4 metres thick they were a major undertaking in themselves, and more so when seen as part of programme of fortification which covered the empire’s territories with castles.53 When in 766 Constantine V (741–75) wanted to restore the aqueduct of Valens in Constantinople, ‘he collected artisans from different places and brought from Asia and Pontos 1,000 masons and 200 plasterers, from Hellas and the islands 500 clay-workers and from Thrace itself 5,000 labourers and 200 brickmakers.’54 The building industry was not on the late antique scale, but it did exist. Constantine’s activities are a reminder of the continuing importance of Constantinople itself. No doubt much smaller than it had been in the sixth century and far smaller than Abbasid Baghdad, the Byzantine capital was still a large city with, for example, over eighty churches built before 600 and kept in repair through the following centuries. One of CyrilMango’s many contributions to Byzantine studies has been to pour cold water on overoptimistic interpretations of all sorts. His vision of Constantinople is of a city whose population plummeted in the seventh century from perhaps 400,000 to 40,000. Others would be more optimistic. Paul Magdalino suggests a population of not less than 70,000; Michel Kaplan talks of not less than 150,000. For any of these figures the supporting evidence is slight. Until recently the same could be said of early medieval Rome and an equally depressing picture was drawn. But medieval archaeology has come of age in Italy and the remains of early medievalRome are beginning to receive proper attention. A new picture of the city is emerging, certainly far smaller-scale even than its late antique predecessor, but by the criteria of the early middle ages a large and prosperous place. For the moment one can only speculate, but it is at least quite likely that when comparable archaeological work is done in Istanbul, the same sort of positive re-evaluation will follow.55 If seventh- and eighth-century Byzantium was incontestably a poorer society than in late antiquity, or than its contemporaryMuslim neighbour, we need to be careful before coming to bleakly minimalist conclusions (see above, pp. 469–70, 472). This was a world that increasingly looked to the steppes and Turkic central Asia for cultural models. Constantine V married a Khazar bride and by such routes was a new material culture introduced to Byzantium, a culture that displayed wealth in textiles and in metalwork. Wealthy Byzantines of the new age wore turbans and kaftans, prayed on carpets, reclined on cushions, dined off silver. Not-so-wealthy Byzantines likewise wanted to emulate the east. The disappearance of redslip pottery is a fact about production; it is also a matter of cultural choice, as consumers took to green- and brown-glazed wares, a fashion that ultimately had its origins in China. The wealth of late antiquity is obvious to us because it was still spent in the ancient manner on buildings and marble. By contrast much of the wealth of Byzantium was spent on perishable textiles and recyclable silver. We have hints in the texts, but little survives.56 Although ‘dark age’ Byzantium is a term that some find jarring, it has its uses. The term does not necessarily imply that seventh- and eighth-century Byzantium lapsed into prehistoric poverty – which would certainly be overdrawn – but it does carry a message that this period is obscure, with much resting on assumption and more research needed. The currently available evidence strongly suggests a poorer society with fewer resources than its late antique predecessor; beyond that much remains obscure. There is a case, sketched out above, that dark age Byzantium was less poor and more economically active than we have come to believe. The thesis will have to be tested in the years ahead, but at the least it opens interesting perspectives. If the economy of dark age Byzantium was more active than tends to be supposed, this would have implications, not just for the seventh and eighth centuries, but for what followed in the ninth to twelfth and more generally for the grand narrative of middle Byzantine economic history. The mounting prosperity of the Byzantine world from the ninth century is a phenomenon common to the contemporary west; this encourages one to look for explanations valid for both. Two potentially global explanations are the end of outbreaks of the Justinianic plague in the eighth century or a shift to a more favourable climate, in either case triggering population growth. Although both theses attract support, most scholars consider the evidence too slight and inconclusive. As with the arguments that invoke the impact of plague in late antiquity, comparison with the Black Death tends to be discouraging. In the sixteenth-century west and in theOttoman empire the economy and also population levels recovered long before the outbreaks of the plague abated; in fact they coincided with a period of worsening climate.57 Improved security was certainly a factor. In Byzantium it is obvious that economic growth and the end of Arab raiding roughly coincide, and that will have encouraged investment in provincial property. Cappadocia is a case in point. The excavation and decoration of new cave complexes begins just as the region ceases to be an exposed frontier zone. In the west the argument works less well. The end of Viking, Hungarian and Arab raiding probably has a bearing, but in general violence and economic growth seem to go hand in hand. Historians have therefore tended to look elsewhere. Since the 1950s and Georges Duby’s seminal work on the Mˆaconnais, an explanation that envisages aristocratic demand pushing reluctant peasants onto the market has been widely favoured, and for Byzantium this ‘feudal revolution’ model now underlies the prevailing narrative. The great estates of late antiquity, according to this thesis, broke up in the seventh century, leaving an economically static peasant economy, as pictured in the Farmer’s law of the seventh or eighth century. The ending of Arab raids encouraged a new generation of aristocratic landowners to emerge from the later ninth century onwards; this is reflected in the body of land legislation from the tenth century, through which successive emperors tried and failed to halt the trend. The eleventh century was a new era of great estates and these survived the crises of the later eleventh century to flourish in the twelfth. They provided the motor of Byzantine economic revival. The land legislation was not only a failure, it was misconceived, being set against the very process that was enriching Byzantine society as a whole. Emperors might regret the loss of independent peasant farmers and the potential political unreliability of this new aristocratic class; but the richer material culture attested by field surveys and excavation-work, especially in Greece, suggests an economy where a broad section of the population was benefiting.58 What limited the success of the Byzantine economy was not the rampancy of its landlords, but the immobility of its peasants. When Michael Choniates, the metropolitan of Athens and brother of the historian, complained in the late twelfth century about the injustice of already privileged landowners gaining further exemptions while the poor were ruthlessly fleeced, he may have been morally right, but his economic diagnosis was wrong. The Byzantine economy needed not more protection for peasants, but less.59 Up to a point this looks a convincing model, but its assumption that peasants are merely a drag on economic growth is now somewhat dated. It also fails to engage with the effects of the late eleventh-century crisis. Recent work by anthropologists and development economists has emphasised the extent to which peasant agriculture can generate growth, not least because self-sufficiency in an uncertain world requires substantial margins of extra production and the most effective way to insure against shortage is engagement with the market. Far from peasants being averse to the market, producing cash crops for the market is essential for their survival. In an even partially monetary economy this allows the building-up of surpluses and hence the creation of an insurance against the inevitable bad years. The chief brake on Byzantine economic activity may not have been peasant immobility, but aristocratic self-sufficiency. Only the rich could afford to be self-sufficient.60 Such insights in turn fit with the evidence emerging in various parts of western Europe that suggests economic growth started in some areas as early as the seventh or eighth centuries, rather than the tenth or eleventh, and that the roots of such growth lay in peasant enterprise. Rather than thinking of peasants needing to be forced onto the market and so launching an economic revolution, we should perhaps be thinking of a world where landowning aristocrats hijacked the fruits of pioneering peasant enterprise.61 To Byzantinists this will sound curiously like a return to the ideas of an earlier generation. Before the current consensus that the seventh century ushered in a poorer, more primitive world, the seventh to tenth centuries were painted as a golden age for Byzantium, or at least for Byzantine peasants. A contrast was drawn between the evidence from the late antique law codes for a semi-free peasantry tied to the land and a picture derived from the Farmer’s law of independent peasants free from landlordly exploitation, who rebuilt the Byzantine economy on new and healthier foundations. For good reasons it is an idea that has largely fallen out of fashion. It was the product of an intellectual current, widespread in the late nineteenth and early twentieth centuries, that saw an independent peasantry as a moral and economic good in itself, and it was innocent of archaeology. But there is a case for taking its basic thesis more seriously.62 The tenth-century land legislation offers clues. It is now usually taken as evidence for the rise of a landed aristocracy, but it is also undeniable evidence for the existence of a free peasantry whom the emperors wanted to protect. The term peasant does not inevitably imply ‘very poor’. In fact, as Rosemary Morris pointed out over thirty years ago, the terminology of the land legislation makes plain that the lawmakers were specifically not concerned with a destitute rural underclass, but with the ‘powerless’.63 It has become customary to dwell upon the failure of the land legislation, which makes it a natural progression to link eleventh-century growth with the rise of an aristocracy and to see one as a principal cause of the other. But when one turns to the collection of eleventh-century legal judgements known as the Peira, or looks at the contrast between the place of free peasants in Byzantine southern Italy and their position in the same areas following the eleventh-centuryNorman conquest, or considers the contrast between the status of peasants before and after the Frankish conquest of Greece, what stands out is the effectiveness of the land legislation.64 The legislation did protect the rights of free peasants and it did act as a brake on the expansion of aristocratic estates. The growing economy of the eleventh century took place in a world where a substantial proportion of the empire’s output was generated by free peasant farmers responding to their own agenda. Archaeological data from southern and central Greece has been cited already. The many new churches are an obvious feature, and some such as Skripou andHosios Loukas are undoubtedly evidence of aristocratic expenditure, but what is most striking overall is the unaristocratic culture of eleventh-century Greece. The new building revealed at Corinth, Athens or Thebes seems very ordinary. The associated material culture appears to be one that would fit well with a prospering peasant world.65 If such a pattern provides a key to the development of the Byzantine economy, a number of implications follow. They would fit with a reinterpretation of the seventh and eighth century putting a more positive gloss on the cessation of the great aristocratic enterprises of late antiquity; they would highlight the importance of the survival in Byzantium of a flexible monetary system whereby peasant farmers could effectively produce for the market; they might also suggest a less optimistic view of the twelfth century. Since 1970 most historians have followed Hendy’s line that the twelfth century saw a continuation of the growing prosperity of the tenth and eleventh centuries. In terms of its value for the economy, the loss of Anatolia to the Turks did not matter because the most valuable parts of Asia Minor either remained in Byzantine hands, or were soon recovered. Anatolian stock-raisers may no longer have lived under imperial rule, but they would have had little option but to supply Byzantine demand. In other words business as usual; the indices of coins, ceramics and buildings show all was well and, in some areas, better than ever.Michael Choniates was no doubt telling the truth as he saw it when he describes rapacious tax-collectors destroying the poor; but he wrote in the 1180s and 1190s, a time when the empire was ringed by enemies, much territory had been lost and what remained was bound to be taxed hard.66 The crucial issue for the twelfth century as a whole is not taxation, but the role of the great landed estates. A number of documentary texts make it plain that vast areas of the Komnenian empire were run as the estates of various Constantinopolitan landlords – a category that includes the emperor, his kinsmen, the church and those monasteries and other pious institutions known as the ‘pious houses’.67 If the economic revival of the tenth and eleventh centuries was driven by aristocratic pressure, this phenomenon does not matter, or can be seen in a positive light, a return to the conditions underlying late antique prosperity. Rather than taking demand out of the system, the bankruptcy of the previous system of remunerating the state’s servants in cash would have encouraged aristocratic landlords to put the produce of their estates on the market and indeed run their estates with the market in mind.68 But if the economic revival of the middle Byzantine period was in fact driven from below, these vast estates might look more like a stifling of Byzantine economic enterprise. Did the great estates necessarily promote local and regional economic enterprise, or did they dampen such activity in favour of self-sufficiency and the provision of goods in kind to feed their dependents in the capital? Is the growth of Latin commerce a sign of the economic vitality of Byzantium, or rather a symptom of an economy where Byzantine traders were disadvantaged in favour of outsiders servicing the great estates? An economy, in other words, shifting, like that of later medieval eastern Europe, to become one of great estates producing for a profitable export market, but in so doing fundamentally damaging its social base. Good cases can be made on both sides of the argument. For want of sufficient documentary evidence to shed light on how estate and society interacted at local and regional levels, any judgement must ultimately rest on a more general view of the state of the twelfth-century economy. How rich was Byzantium?