15
Being a tradesman in Constantinople around 900 was by no means a straightforward process. According to the Book of the Eparch (or the Prefect), a set of official regulations from this period, merchants, shopkeepers and many artisans had to be members of a guild (systma) to operate, and had to sell their wares in specific places, the gold- and silver-dealers in the Mese, the merchants of Arab silk in the Embole, the perfumers in the Milion beside Hagia Sophia, the pork butchers in the Tauros. Ambulant sellers were banned; they would be flogged, stripped of guild membership, and expelled from the city. Sellers of silk could not make up clothes as well; leather sellers could not be tanners. Some guilds, such as the merchants of Arab silk or the linen merchants, had to do their buying collectively, with the goods then distributed among guild members according to how much money they had put in, to keep down competitive buying. Sheep butchers had to go a long way into Anatolia to buy their sheep, to keep prices down; pork butchers, by contrast, had to buy pigs in the city, and were prohibited from going out to meet the vendors; so also were fishmongers, who had to buy on shore, not on the sea. The eparch, the city governor, had to be informed if silk merchants (divided into five separate guilds) sold to foreigners, who were prohibited from buying certain grades of silk. He determined all bread prices, by which bakers had to sell, and the price of wine the innkeepers sold; and he also determined the profits that many vendors made - grocers were allowed a 16 per cent profit, but bakers only 4 per cent (with another 16 per cent for the pay of their workmen), over and above the price they paid in the state grain warehouse.
Later medieval western towns often had quite elaborate guild regulations like these, aimed at maintaining monopolies and internal hierarchies in trades. The Book of the Eparch stands out, however (apart from in its early date), in the degree of state control it assumes. The regulation of profit was particularly important here, and also the regulation of the ways sellers were allowed to buy their goods. Silk was controlled because its production and distribution reflected directly on imperial prestige (the regulations for linen merchants were looser). Above all, however, it was vital that the food market was controlled, for Constantinople had to be fed reliably, at prices the inhabitants could afford. Bread was no longer free, as in the late Roman empire; that had stopped abruptly by imperial decree when the Persians took Egypt in 618 (above, Chapter 11). Constantinople was much smaller now; it did not need Egyptian grain any more, and could provision itself from its Aegean and southern Black Sea hinterland. All the same, as we have seen, it was still very substantial in size; it was the largest city in Europe even at its low point in the seventh and eighth centuries, and was now growing again, reaching maybe 100,000 inhabitants in 900. (Cordoba may well have surpassed it in size in the tenth century, but it shrank in the eleventh, leaving the top spot to the Byzantine capital again.) Emperors and eparchs could not afford the trouble from its inhabitants that would inevitably appear if there were food shortages - and these, indeed, were seen by the urban population as the fault of public authorities. Trade was independent in Constantinople, but the terms of trade were closely linked to the state. We can of course doubt how effective all the rules in the Book of the Eparch were, but they are very striking as an aspiration, and it is at least true that narrative sources regularly ascribe this sort of power to officials. Liutprand of Cremona did buy prohibited silk in 968, but it was discovered and confiscated, to his fury. The Byzantine government had the infrastructure to make its laws obeyed, at least sometimes.
This introduces us to a standard feature of both Byzantine and Arab exchange, its close link to the state. This varied, certainly. It was probably greater in Constantinople than in the Byzantine provinces; it seems to have been greater in Egypt than in al-Andalus; and state control was always more likely to be enforced in the arena of urban provisioning than in that of the international luxury trade (silks and other state-interest goods apart), for that trade relied so much on private mercantile risk-taking. Arab port authorities in the tenth and early eleventh centuries even then regularly assigned official prices to imported goods, but these were only guides to market prices, which varied by supply and demand. But grain in Constantinople was only one out of several commodities which were bought from government warehouses; in Egypt, too, flax (for linen), one of that region’s principal productions, was also sold to merchants (whether for internal sale or for export) by state offices, and some of the major linen-weaving centres, such as Tinnis and Damietta, were largely publicly owned. Egypt, as already implied, had in every period a rather more dominant state sector than existed in some other regions, but the existence of operations on this scale is striking. Commerce itself might be in the hands of independent merchants, but they operated in a framework in which the public power had a considerable say. And, above all, states were huge sources of demand. Egyptian documents from the decades around 1000 show merchants regularly (and sometimes unwillingly) selling to the government itself; and, even when this did not take place, the focusing by merchants and artisans on great political centres such as Constantinople, Baghdad, Fustat-Cairo and Cordoba was because these cities had so many rich buyers who were paid by the state, bureaucrats or soldiers and their own dependants.
As we have seen, and as we shall see again, after the end of the Roman empire in the West, which was a strong and centralized state and which moved large quantities of goods around on its own behalf, exchange in the post-Roman kingdoms depended for its intensity on the wealth of landowners - aristocrats, churches and kings. The richer landowners were, the more exchange there was, and the more complex its patterns. This was broadly true in the eastern Mediterranean as well; but state power, based on tax-raising, continued here, and state buying-power was normally on a somewhat larger scale than that of private landowners. Furthermore, private wealth allowed people access to state office, and thus access to the greater emoluments made possible by taxation. This was so even in the Islamic world, where private landowners were usually less automatically linked to political power, and so could be seen as a rival source of demand to that of officials and soldiers. Taken as a whole, it is the changing wealth of the state sector that is the best guide to the changing scale of demand, and thus exchange, in the Byzantine and Arab East. Where private landed wealth had a different trajectory to the wealth of the state, it must have affected demand as well, and its local variation adds a further level of complexity to our analyses. But broadly the two moved in tandem in most of the East, and the state system is also rather better documented. I shall be saying more about the latter in this chapter as a result.
The gap in our evidence for the landed aristocracy matches the very serious gap in our seventh- to tenth-century evidence for the peasant majority in the East. The millions of documents regularly produced for governments and private individuals in Byzantium and the caliphate have almost all been lost. Only for Egypt do we have the sort of local land documentation that we can find in Francia and Italy, thus allowing in a few cases the reconstruction of peasant societies, as in the case of the eighth-century Coptic village of Jeme, in western Thebes in Upper Egypt; and the uneven publication of Egyptian documents in Arabic means that we cannot as yet easily do this for the period after 800. Rural archaeology is currently poorer for the period after 650 or so than for before, too, in nearly every region. We looked at Byzantine and Andalusi aristocracies in Chapters 13 and 14, and I shall of course be referring to some aspects of peasant economy and society in this chapter, for they will inevitably impinge on issues of wealth-creation, taken as a whole: put simply, the richer élites were (whether from tax or rents), and the higher aggregate demand was, the more the peasantry was exploited - an equation which must be understood to underlie this whole chapter. But we shall have to wait for future research before we can confront the detail of most eastern peasant social realities after 600-650, so as to compare them with those of the West. Urban society is better attested, as we have also seen in the last four chapters. One urban society is particularly clearly documented, the Jewish sector of the city of Fustat in Egypt, whose genza or storehouse of waste paper (kept because Jews would not destroy the word of God, and thus any paper with writing on), founded in 1025, preserves thousands of texts, which begin to be numerous around 980. Most of these are eleventh-century or later, rather than from the tenth, but I shall use some early eleventh-century genza texts here as well, as they transform our understanding of how urban societies could function at the very end of our period. Despite the wealth of the eastern Mediterranean, then, our surviving information about the socio-economic history of the period 600-1000 is even bittier than it is for the West. I shall focus here, necessarily briefly, on three regions in turn: Byzantium, with its seventh-century crisis and ninth-century revival; Syria and Iraq, rivals throughout, where economic protagonism moved decisively from the first to the second in 750; and Egypt, the region with the most continuity. We shall then look at the international commerce which linked them.
As we saw in Chapter 11, the military disasters of the 610s and 640s caused the Byzantine state to change markedly. It adopted a localized and mostly demonetized tax structure, matching a localized military structure, focused on defence. Never again would the state transport its own goods long distances on any scale, even if Constantinople maintained itself as a fiscally supported focus for commercial demand. It is also likely that the landed aristocracy, never as rich as in the West, lost some ground, given its invisibility in the sources before 850 or so, and given the constant raiding that will have reduced agricultural productivity in much of Anatolia until the frontier stabilized in the eighth century; as noted in Chapter 13, even in the tenth century, when our sources all agree that a process of local affirmation of aristocratic power was firmly under way, it is hard to argue that they were as dominant across the whole empire as was normal in the West. The tiny amount we know about peasant society at least shows that there were indeed some areas of the empire where aristocrats did not have full control in the seventh and eighth centuries. The lands west of Ankara described in the early seventh-century Life of the ascetic Theodore of Sykeon had largely independent peasant communities already in the years leading up to the Persian invasions, indicating that aristocrats never had been wholly hegemonic in parts of the Anatolian plateau. If the Farmer’s Law, a private handbook of agrarian law from the period 650-850, can also be located in Anatolia (as the absence of reference to olive-cultivation in it may imply), then such peasant communities continued to exist there after the invasion period as well. In both texts, the state remains present, unquestioned, as a tax-raising and judicial power. There were also considerable wealth differences in each, with richer peasants dominating the community and leasing land to poorer peasants. But external landowners are relatively unimportant in the earlier text, and absent in the later. This is not a guide to the empire as a whole, or even to the whole of Anatolia (aristocrats were rather strong in Cappadocia, further east, in both the fourth to sixth centuries and the ninth to eleventh, so plausibly in between as well); but the patchiness of local aristocratic dominance is made clear by these texts, and this almost certainly increased in the crisis centuries.
Corresponding to the difficulties experienced by the Byzantine state and aristocracy, the seventh and eighth centuries show, particularly clearly in fact, a crisis in urbanism. Archaeologists and historians argue about whether there was already a dip in urban vitality in the Byzantine lands after 550; but no one any longer seriously argues that there was not a systemic crisis in the early seventh century. Urban archaeology makes this too clear. Building cannot be shown to have continued after 650 in most of the dozen or more cities with decent excavation; most show areas of systematic abandonment in the same period, as with the particularly well-excavated street of shops in Sardis, in the Anatolian lowlands close to the Aegean, which were abruptly deserted in the 610s, or the gymnasion in Ankara whose burning can be precisely dated to the Persian sack of 622, for a Persian ring-stone was excavated in the burnt level. I am normally cautious about drawing too catastrophist conclusions from anecdotal examples like these (prosperous cities have abandoned areas too, and can also recover from being sacked), but the accumulation of evidence in the Byzantine lands is too great to be gainsaid. It is significant that the best counter-example, Gortyn on Crete, was on an island, and thus safer from Persian/Arab or Avar/Sclavenian raids: here Heraclius (610-41) reconstructed the city after an earthquake, and a late seventh-century artisanal quarter, probably extending later as well, has recently been excavated. Elsewhere, all we get is new walls, sometimes enclosing only portions of the ancient city, and sometimes on hills above the old town.
The Byzantine state continued, as we have seen. Even small hill-top cities (now often called kastra) could still have a political-military role, and also still had bishops (although these, as we have also seen, often preferred to live in the capital). There is some evidence, furthermore, that some hill-top fortifications were citadels for islands of surviving settlement in the ancient cities below, as at Euchaita and Amorion, both on the Anatolian plateau, or at Corinth in central Greece, or at Myra on the south coast of modern Turkey. Whether this scattered occupation was sufficiently dense and economically diversified to be called ‘urban’ cannot yet be said: of these, Amorion and Corinth are perhaps the most likely. Overall, however, we have to recognize a new urban typology. Some ancient cities were wholly abandoned or reduced to small strongholds. Some developed this scattered pattern, with greater or lesser levels of organization or urbanization. A few continued to be active as urban centres, though on a considerably reduced scale, like Ephesos, Miletos and Athens on the Aegean coast - Ephesos’s new walls left much of the old city centre outside them, but still enclosed a square kilometre of land; the city is recorded by Theophanes as having a major fair, yielding a large sum in taxes, in 795-6. And a handful of cities may well have seen rather less change, though excavation is less good in them precisely because of the urban continuities there: Thessaloniki, Iznik (ancient Nicaea), Izmir (ancient Smyrna), Trabzon, major political centres in each case. This is not total urban collapse, but even on an optimistic reading of the evidence we might propose that four-fifths of Byzantine cities lost all or most of their urban characteristics.
The significant feature in common to most ‘successful’ early Byzantine towns is that they were thematic centres. (Ephesos, long a commercial entrepôt, is the main exception.) It looks as if the state focused on its main local military and administrative centres; if landed aristocrats joined the army and civil bureaucracy, they may well have gone to such towns too. These towns thus remained sufficiently potent centres of demand to retain their urban characteristics: markets, perhaps some artisanal specialization. But there were far fewer of them than in 600. When Byzantium achieved greater military and political stability again, slowly after 750, more visibly after 850, the number of active cities did not greatly expand, either. They increased their own sizes again, although it is as yet hard to be sure exactly when from the archaeology; the eleventh century shows it better than the tenth, although in Sardis, and also in Hierapolis on the western edge of the Anatolian plateau, it is already visible before 1000. But the Byzantine empire never again re-created the density of late Roman urbanism in its territory.
Our evidence for commerce outside the capital, also largely archaeological, both mirrors this picture and nuances it. The seventh century saw the abrupt end of the Aegean’s main industrial tableware production, Phocaean Red Slip ware, and its more local imitations; painted wares of reasonable quality sometimes replaced it (for example in Crete), but their distribution was very localized, and in some places (notably in inland Greece) all we find is handmade pottery, indicating the end of professional production. Amphora production, for oil and wine, also localized and simplified; the standardized Aegean globular amphora, LRA 2, was replaced by a variety of related but more local types. These developments, into the eighth century, imply a breakdown in demand for goods, and thus the weakening of concentrations of wealth, whether public or private. But this is not the whole picture. Constantinople itself had an industrial ceramic production, of Glazed White ware (GWW), which began around 600 and continued for many centuries. In the next two centuries there are sporadic finds of this pottery type in a wide range of places across the Aegean, down to Crete, and even Cyprus (which had its own productions). These show that the Aegean did not lose a certain level of medium-distance exchange. This is supported by the (probably) eighth-century Rhodian Sea Law, another private legal manual, which discusses the relationship between ships’ captains and merchants on ships, and which presumes as standard cargoes an array of goods that are hard for archaeologists to find: slaves, linen, silk, grain, as well as wine and oil in (presumably) post-LRA 2 amphorae. Seventh-to ninth-century saints’ lives also regularly feature shipping, often but not only for grain. The Aegean was by now, as we have seen, Constantinople’s agrarian hinterland; the demand of the capital, even if nothing else, kept ships on the sea. GWW tableware was probably one of the things the capital sold in return.
The Byzantine empire at its low point thus never entirely lost a network of exchange that covered its heartland, the Aegean and Marmara seas and the coasts around them. This was so even if most local production had simplified, sometimes radically. This seems to reflect what else we know about the empire: that the state had localized its own structures, but that it was still dominated by a powerful capital. Arguably, the local differences in productive professionalism around 700 reflect areas of greater or lesser aristocratic power on the ground, although the evidence is not yet good enough for this to be developed further. The Aegean-wide exchange we do see was not run by the state; our written sources stress independent merchants in the period before 800, just as theBook of the Eparch, for all its regulatory interest, does in 900. But state-fuelled demand was the most solid agent of buying power all the same; and this commerce focused on the capital first, although secondarily, in other surviving centres as well, Thessaloniki, Ephesos or Smyrna.
As we move into the ninth century, one visible change is an increase in the numbers of coins found on sites. It is normal in excavations to find coins up to Constans II in around 660, and then nothing, or almost nothing, for a hundred and fifty years; even though every emperor still minted coins, they vanished from circulation, and we could not conclude that they were at all commonly available outside the capital. This changed from the 820s onwards. At Corinth, nearly four times as many coins are known for Theophilos (829-42) as for all his predecessors put together after Constans; those for Leo VI (886-912) are six times as numerous as for Theophilos, those for Leo’s son Constantine VII double again, and the figures go on up from there. This can most plausibly be linked to a revival in taxation and army pay in money, which is most often ascribed to Nikephoros I (802-11: above, Chapter 11); such a shift depended on a more reliable supply of metals, but also presumed (and furthered) market exchange, sufficient to move the coins around. In the ninth century, too, we come upon larger-scale finds of GWW outside the capital, for example at Mesembria, a Byzantine port in modern Bulgaria, and even in field survey, in the countryside outside Sparta; in the tenth, this extends to Thebes. Local imitations of Constantinople pottery begin to be found at Athens, and, significantly, at Preslav in independent Bulgaria. Large-scale ceramic production at Corinth also began by the tenth century, and so did the amphorae of the Ganos area, in the Sea of Marmara, destined for the newly systematic export of local wine. The wine trade could already extend far afield, indeed, if the large consignment of wine-amphorae, marked with their shippers’ names, found in a wreck off south-west Turkey dating to around 880, really was from the Crimea, as the excavators think. Linen was exported from Bulgaria and the southern Black Sea (as also from Egypt) to the capital as well, and both Constantinople and Thessaloniki made glass. We are beginning to move into the complex Byzantine productions of the central Middle Ages.
In the ninth century, and still more in the tenth, the state was getting stronger and richer in Byzantium. In the tenth, so was the aristocracy, in some areas - often away from the Aegean focus of the archaeology, but including in southern Greece, where already in the 880s the wealthy Danelis (see Chapter 13) had access to elaborate linens and silks, and the textile workers themselves, whom she gave to Basil I and Leo VI. A century later, Basil II, in his complaints about ‘the powerful’, was worried that they would monopolize rural markets, too. What we see in this whole list of examples is an increasingly elaborate and diversified set of agrarian and artisanal productions, with an increasingly wide and complex distribution, to and from the capital, certainly, but between provinces as well: Thessaloniki was a particularly important entrepôt. This was made possible by élite demand, which was clearly increasing again, and was also furthered by direct élite involvement in artisanal production and exchange. If there was ever a natural location for medium-distance exchange, of course, it was the Aegean, largely land-locked and protected, and studded with islands, as it is. The years around 900 merely saw a return to normality in this respect; they point up the abnormality, the crisis, of the two centuries after the Persian and Arab invasions. But the growing power of the Byzantine state would push that exchange still further in the two centuries to come. After 1000, a demographic expansion, which is quite likely to have already started in our period, begins to be more visible in our documentation, as does a trend to reclaim uncultivated land; the agrarian base of the empire was clearly expanding. The eleventh century shows some agricultural specializations as well, not least in mulberry trees for silk in various parts of the empire: these too must have existed already before 1000, for Byzantium was certainly producing its own silk in our period. The old view that the empire saw economic stagnation in the eleventh and twelfth centuries is now decisively rejected; the roots of the generalized economic expansion of that period lay in ours, even though we can as yet only see occasional signs of it. And that expansion affected areas outside the empire as well: by the early eleventh century the Byzantines were exporting silk to Egypt. This is a point we shall come back to.
Syria did not for the most part see the seventh-century crisis of the Byzantine empire. After 661 it was the political centre of the Umayyad caliphate, and that period saw major monumental building in the capital, Damascus, as also in the regional religious centre, Jerusalem. Damascus was never a huge city, which partly reflects problems of water supply, but is partly also due to the fact that the Umayyads had difficulties getting taxes from the provinces of the caliphate. But enough came into Syria to ensure the wealth of the caliphs themselves, and their urban and rural palaces still survive in the landscape of Syria and Palestine. The Arab conquest was anyway quick enough for Syria not to suffer in its basic infrastructures. Most of the numerous excavations in both Syria and Palestine, both urban and rural, show continuities that extend to 750 at least, particularly in inland areas. In and around the city of Madaba, for example, in what is now northern Jordan, Christian churches were founded into the late eighth century, with impressively decorated mosaic floors which show both wealthy patrons and skilled artisans: in the city, in rural monasteries, and in villages around.
Cities changed in structure. Their Roman monumental centres tended to fall out of use, as the Arabs had a different ceremonial style, with fewer religious or political processions and a focus on the enclosed public space of mosque courtyards. But they continued to be active demographic and productive centres; Roman public buildings were replaced by artisan workshops, colonnaded streets were replaced by rows of shops, often monumentally built (particularly, as we saw in Chapter 12, by Caliph Hisham, 724-43). So at Gerasa (modern Jerash) north of Madaba kiln complexes were built in a Roman theatre and a temple, part of a network which made Gerasa ceramics a major feature of the economy of the Galilee area until 800 or so; at nearby Scythopolis (modern Bet She’an) there were by 700 or so kilns in the theatre and amphitheatre, linen workshops in a bath complex (Scythopolis linen was well known already in the Roman empire), and one of Hisham’s shop complexes on the site of a sixth-century hall. These patterns are repeated, in greater and lesser detail, in twenty other cities; the production of glass, dyeing (and thus textiles), iron, copper are all attested in recent archaeological work. Substantial élite town houses have been found in some cities, too; and of course the Arab period had its own monumental buildings, mosques and governors’ palaces.
This picture was clearly very different from that in the Byzantine heartland, although the sources - almost all archaeological - are the same. There are almost no usable written sources on these issues for Syria and Palestine, in fact, although the Syriac chronicles for Edessa also paint a glowing picture of the commercial activity of that city and of the wealth of its Christian élites: Athanasios bar Gumoye, a great landowner and a tax official for ‘Abd al-Malik in Egypt around 700, reputedly owned 300 shops and nine inns in Edessa. Two changes nuance this picture of continuing élite and rural prosperity, however. The first is that the coast of Syria and Palestine, a major oil and wine export area under the Roman empire, saw stagnation under the Umayyads, the weakening of major coastal cities such as Antioch, and the abandonment of marginal lands. Umayyad Syria was not closely linked to the Mediterranean; it hardly even had any economic links with Egypt, although some Egyptian products still came in through the major surviving coastal entrepôt, Caesarea in what is now Israel. But actually - this is the second change - Syria and Palestine were no longer a single economic unit. The productions that can be best traced by archaeologists, once again mostly ceramics, remain of very high quality in the Umayyad period, and show industries that were large-scale and many-levelled, aimed at elites and non-élites alike; but they were much more localized than in the Roman period. Gerasa pottery rarely reached the Mediterranean coast, or ‘Aqaba on the Red Sea, or northern Syria, for example; even Jerusalem, only 100 kilometres away, largely had its own - again, high-quality - ceramic tradition. So the Syro-Palestinian economy remained prosperous and complex under the Umayyads, but it was much more internally fragmented, and cut off from its neighbours. It was, in fact, even more internally fragmented than the crisis-bound Byzantine empire, as it seems on the basis of the archaeology of the moment.
This economic fragmentation further underscores the difficulty the Umayyads had in centralizing the fiscal system of the state, even in their own political heartland, although they were certainly more successful here than elsewhere. But the complexity of (almost all) the different sections of Syria and Palestine also points at the continuing force of local demand, and thus of the continuing wealth of urban élites, that is to say the local landed aristocracy. It is often said that the Arabs gave more respect to merchants than the Romans had, which is true; Muhammad had been a merchant, and there was never in the Islamic world any stigma attached to wealth ‘from trade’, unlike in much of the West, or even Byzantium. It is often also said that this ideological shift is already visible in the changing forms of cities, with more artisanal and commercial activity in old public centres; this seems less likely, however. These changes are better explained as the normal result of shifts in the focus of monumental building, from colonnaded streets and theatres, etc., to mosques (above, Chapter 10); if a city remains economically active, unused buildings will get taken over for private uses, and so it was here. But we should also not overstate the mercantile element in élite activity. Athanasios bar Gumoye, notwithstanding all his shops, was a great landowner first and foremost; it is likely indeed that most urban patricians in this period (who were anyway mostly still Christians) were above all landowners, and at most used landed capital to get into commerce, if they wanted. This would be so later, too, in ‘Abbasid Iraq, where such élites would usually be Muslims, and in post-‘Abbasid Iran, where ‘ulam’ biographies show land as much as mercantile activity as the basis for elite wealth. Even the Jewish mercantile elites of Fustat in Egypt, who may well have gained their initial wealth entirely in the commercial sector, bought land or tax-farming concessions with their profits, for land remained overwhelmingly the chief source of wealth overall. Exchange was, and remained, only a spin-off of agricultural wealth, even around the great cities of the second half of our period, and still more in Umayyad Syria.
The year 750 marks a change in the economy of Syria and Palestine. The ‘Abbasid takeover marginalized the region politically, and, with the fiscal centralization of the caliphate from the 780s onwards, Syrian taxes were firmly directed to Iraq. Cities which stayed as prosperous as before into the ninth and tenth centuries were rather fewer, Ramla near Jerusalem, Tiberias on Lake Galilee, Caesarea, ‘Aqaba, Aleppo, Damascus, entrepôts or major local governmental centres. The devastating earth- quake which hit the Galilee area in 749 left cities in ruins, which, significantly, were often not rebuilt and can thus be excavated; Bet She’an is a particularly impressive sight, with white limestone columns (including those of Hisham’s shops) even now lying across black basalt roads. Syria would henceforth be mostly governed from elsewhere, from Baghdad, Cairo, or (for the North in the late tenth century) Constantinople; only Aleppo was sometimes independent at the end of our period. This, plus the wars fought over it in the tenth century, sapped its prosperity. But it was by no means in economic crisis even then, and ‘Abbasid centralization brought with it a widening of economic horizons, with more evidence of exchange with Iraq: new polychrome glazed ware spread from Iraq into Syria/Palestine from 800 onwards, the beginning of a new international taste in fine pottery which would by 1100 dominate the whole Mediterranean, Muslim and Christian regions alike. It is for this reason that entrepôts flourished under the ‘Abbasids; inter- regional networks were beginning to develop again, west to Egypt (via Caesarea), south down the Red Sea (via ‘Aqaba), east to Iraq (via Aleppo). This network would continue even after the ‘Abbasid caliphate collapsed, as we shall see in a moment.
The ‘Abbasids, of course, invested in Iraq. Iraq had been a major political and economic centre for millennia; the Tigris and Euphrates created a fertile and irrigable basin matched only by the Nile for its agricultural wealth. The Sassanians were only the most recent rulers to develop its irrigation, with the great Nahrawan canal, probably built in the sixth century, which brought Tigris water to a network of smaller canals north and east of the capital, Ctesiphon, situated just south of what would become Baghdad. An early and influential 1950s field survey of the Nahrawan area by Robert Adams indeed saw the Sassanian period as the economic height for Iraq, with the pre-tenth-century caliphate, however prosperous, failing to match Sassanian levels after the political crises of the 620s-630s, in which canal dykes were not maintained. The dating of sites in Adams’s work, and thus his assumptions about the number of settlements that were actually occupied in each period, were however more influenced by his over-literal readings of narrative sources than a field survey would be today (if one were possible in Iraq in 2007). The land north of Raqqa in modern eastern Syria, a more short-lived ‘Abbasid capital on the Euphrates, showed a clear ‘Abbasid-period settlement peak in a more recent field survey. The Umayyads, anyway, and even more the ‘Abbasids, were committed canal-builders and land reclaimers, and the ‘Abbasids were particularly active in southern Iraq, as we saw in Chapter 14; it was to build dykes and to desalinate land in the marsh areas of the south that they imported the large-scale African slave gangs of the Zanj. The ‘Abbasid construction of the huge metropolis of Baghdad after 762 required systematic provisioning, and it was in the interests of every public official who bought Iraqi land with his tax profits to develop that land with an eye to the urban market. Samarra, at the northern end of the Nahrawan canal, only added to that market in the mid-ninth century. The sharecropping contracts discussed in legal sources from ‘Abbasid Baghdad, which presumably best reflect the Iraq the legists lived in, show landlordly investment; state investment in the irrigation network is assumed as well, largely through wage-labour; the legists say less about the Zanj. Wage-labourers were also used in agriculture, which shows that some landowners were cultivating estates directly, a sure sign of a market-orientated approach. One result was the expansion of Iraqi rice cultivation, which was a ninth-century phenomenon.
Tax revenues only went to the capitals, but their resultant vast size itself created a stimulus to Iraqi agriculture, and the Iraqi commercial economy as a whole. Baghdad (and to a lesser degree other Iraqi cities) was also an artisanal hub which was for a century unmatched anywhere in the world. Silk, cottons, glass, paper (the Baghdad paper-mills were founded in 795, using technology brought from Samarkand and, before that, from China) were all made in the city. Baghdad was a focus for internal Iraqi exchange, and also an entrepôt for interregional commerce between the provinces of the caliphate, which was by now moving ceramics or cloth across the whole terrain from Iran to Egypt. Indeed, this commerce went further; the 1960s-1970s excavations of the Iranian port of Siraf (as yet only partly published) show that the caliphate had opened up to Indian Ocean and Chinese trade on a large scale by the late eighth century. The Seven Voyages of Sinbad in the Thousand and One Nights symbolizes this for most of us, but that is perhaps matched by the remarkable collection of plausible and implausible stories (some of them first-person experiences) made by the Iranian ship captain Buzurg ibn Shahriyar in the 950s, who discusses wonders, strange customs, storms and remarkable animals right across to the South China Sea. The trade thus established continued for the rest of the Middle Ages.
Baghdad’s wealth, and also Iraq’s, faltered in the tenth century. The region had lost its political and fiscal dominance by now. The cutting of the Nahrawan canal in 937 for short-term military reasons was soon reversed, but the precedent was a bad one; the city and the canals were refurbished several times (most committedly by the Buyid ‘Adud al-Dawla in 981-3), but Iraq’s prosperity did not again match that of the ninth century. All the same, that prosperity had been so great that Baghdad remained one of the principal cities of Eurasia, larger than any western city, and a major entrepôt into the twelfth century at least.
None of these regions matched the stability of Egypt. Egypt was the Roman empire’s richest province by far, with the most complex economy, and it remained so in the post-Roman world into the fourteenth century. In the caliphate, too, if it was surpassed by Iraq, that was only in the ‘Abbasid century, and it had regained its primacy by 950 or so. The power-house of the tenth- to fourteenth-century Mediterranean exchange system, which was not driven by fiscal factors as was that of Rome or the caliphs, was Egypt. The basic reason for this was the relative reliability of the Nile flood, which allowed continuous cropping of agricultural land and produced wheat yields of around ten to one (three or four to one, with fallow periods, being the best that dry farming could produce in the Middle Ages). Egypt’s canal system has also almost always been regularly maintained; the country has almost always been governed by a single political authority, which helps, and it certainly was so throughout our period and beyond. The large yields of Egypt’s agricultural land, not only in wheat, but also wine and flax, allowed a whole hierarchy of non-cultivators to be fed from the labour of the peasantry, including landowners, tax officials and soldiers, of course, but also complex networks of artisans, shopkeepers and merchants. It can be plausibly argued that in the later Roman empire a third of the population of Egypt lived in cities, a figure that is unparalleled in the ancient or early medieval world, and there is not much reason to think that it dropped later; if it had, the drop had certainly been reversed by 1000. Certainly the rather restricted archaeology in Arab-period cities shows dense private housing, in apartment buildings, from the seventh to the tenth century: in Alexandria, Fustat, nearby Saqqara, and Akhmim in Middle Egypt.
Egyptian agriculture was carried out through a hierarchy of substantial villages, whose head-men also handled tax-raising, subordinate in this respect to provincial capitals. The records of taxation, which are good for Arab Egypt, show its systematic nature, inherited from the Roman period, and not relaxed later (as eighth- and ninth-century tax revolts show). Landowning was fragmented in Egypt, however; there were always peasant landowners, and the élites which ran villages were usually rich peasants, and little more. Post-conquest documents imply that great landowners were notably fewer and less rich in the early Arab period than under the later Roman empire, and this did not change until the late ninth century. After 850 three developments led to larger landholdings again: more Christians converted to Islam, thus gaining access to state patronage, which was by now sometimes expressed in terms of grants or leases of state land; more Arabs began to acquire land as well (for a long time Arab immigrants had stayed in Fustat and lived off state salaries, as we saw in Chapter 12); and, from 800 or so, the financial administration began to farm out the rights to collect local taxes, rights which could under certain circumstances be turned into effective landholding over wider areas. Tax-farming turned into full ownership less often in Egypt than it did elsewhere in the Islamic world, for the state never relaxed its grip on the mechanisms of taxation, but it certainly helped the establishment of local control. For the first time in many centuries in Egypt, a late ninth-century estate (day‘a) could consist of a whole village (indeed, by the eleventh century day‘a . could simply mean ‘village’). This was not universal, and fragmented ownership survived past 1000 in Egypt, as did direct tax-paying, but a clear change is visible here at the end of our period.
This weakening and renewed strengthening of a landowning aristocracy, which is paralleled elsewhere (for example, in Byzantium) as we have seen, had less effect on the rest of the Egyptian economy, however, than it did in other regions, precisely because of the continuing strength of the tax system, which independently brought wealth into the cities, and, above all, Fustat. This was the basis for an active exchange network which, throughout our period, unified Egypt into a single economic whole. The Nile helped here, as an easy and cheap routeway which ran by or close to nearly all the population of the region. As a result, we can trace artisanal productions which were available from north to south. The fine pottery of Aswan in the far south can be found up to the Mediterranean, 1,000 kilometres away, throughout the early Middle Ages, a unique achievement in scale and continuity in our period. The Aswan kilns continued to produce Red Slip ware in a Roman style until the end of our period and beyond, too, centuries after tastes had changed elsewhere, although increasingly alongside other ceramic types, white-slipped and painted wares, and, after 800, polychrome glaze, following Iraqi fashion. And, although archaeology cannot track it, we can tentatively say the same for cloth; linen and wool production had always been substantial in Egypt since Roman times, and there is never a period in which its sale is not attested in documents. A cache of late ninth-century papyri from the Fayyum, a large agricultural basin to the west of the Nile 150 kilometres south of Fustat, shows a set of Arabic-speaking cloth merchants and related officials buying and selling up and down the Nile from Qus in the south to Alexandria in the far north. The main figure of this papyrus set, Abu Hurayra, lived in Madinat al-Fayyum, the main city of the basin, in the 860s-870s, although others were based in Fustat, which was clearly a major node in the whole exchange process.
These wide exchange networks were not all that Egypt had, either. We can see an exchange hierarchy in ceramics, with local productions (based on local clays) fitting into the Aswan hegemony, and cloth production was certainly associated with many local centres too (based on local flax and sheep), as well as well-known major artisanal cities like Tinnis and Qus for linen, and Bahnasa in Middle Egypt for wool. There were differences here in status, price, taste and convenience, as in all elaborate commercial systems. And the Egyptian system, in the whole period 650-1000, was by far the most elaborate anywhere in Europe and the Mediterranean. Continuous urban demand saw to that. The demand was also, of course, for food, and also certainly for more diversified artisanal goods than cloth and pottery, too; we can say little about them between the sixth century and the late tenth, for our documents are about other matters, but, given the rest, there is no reason to doubt it. One of these goods was still papyrus, an industrial production based in the Delta; it was only in the late ninth and tenth centuries that it was supplanted by paper, a linen by-product.
The genza documents of the late tenth century and onwards thus illuminate a world that had been economically complex for centuries, not to say millennia. But there were also changes at the end of our period. Already in the late ninth century, we can see signs of a larger-scale investment in artisanal production that seems to be new. The governor Ahmad ibn Tulun (868-84), who ruled Egypt more or less autonomously, invested privately in linen according to early tenth-century narratives, and so did lesser officials. The largely state-run Tinnis linen industry appears in these narratives, as it also did in the Fayyum letters, as a major textile centre. It is hard to trace it earlier than 850 with any certainty, but Ibn Tulun upgraded its infrastructure with public money, and dated Tinnis textiles survive from the 880s. These are luxury items, and the state factories were substantially devoted to the production of court fabrics; but the Delta linen towns also sold on the open market, and by the tenth century exported cloth too, to the Mediterranean (Tinnis is on an island, and is also a port) and to Iraq. The word ‘export’ is the main novelty here. Since the Arab conquests, Egyptian production and consumption had mostly been internal. Even with ‘Abbasid fiscal centralization, it is hard to find very much reference to exports and imports in our evidence. Demand inside the region was evidently steady enough to make interregional exchange less necessary, except for the luxury trade, which always existed. But in the tenth century our evidence for it increases, and by the end of the century Alexandria and other ports were full of ships, moving goods from Egypt to Palestine, Tunisia and Sicily; from the latter two, other ships went westwards to al-Andalus. Egypt exported not only made linen cloth but also flax, to be made up in Tunisia and Sicily; sugar, another industrial product, was also an Egyptian speciality. But the range of goods exported from Egypt, and also imported, was by the end of our period very substantial indeed. The Fatimid conquest in 969 meant that Egypt, Tunisia and Sicily were for a while under the same government, which facilitated this; but Egypt was the major motor of this commerce thanks to the continuing strength of its internal market, as the Fatimids recognized and promoted.
Joseph (Yusuf, or in Hebrew, Yosef) ibn Ya‘qub ibn ‘Awkal (fl. c. 970- 1040) is the first really large-scale merchant in the genza documents. His family may have come from Iran initially, but were settled in Fustat by his father’s time; he spent his life at Fustat and in the new Fatimid capital of Cairo just outside it. He and his sons ran an import-export business, employing numerous secretaries in their headquarters, and agents in both Egypt and abroad, above all in Tunisia and Sicily. They exported flax from Egypt, buying it from small towns in the hinterland of Bahnasa and in the Fayyum and sending it down the Nile from Fustat to Alexandria (thus bypassing the linen factories on the other side of the Delta) and then to the west. They also exported dyestuffs, madder (Egyptian-made), indigo and brazil-wood (both imported); imported pepper and spices, and Egyptian-made sugar; and more expensive luxuries, in particular pearls; 83 different commodities in all. The imports were largely from the Indian Ocean trade; Fustat-Cairo was becoming the principal commercial node between the Indian Ocean and the Mediterranean, which it remained for centuries, although that latter trade was not Ibn ‘Awkal’s speciality. The business bought in return, from its Mediterranean partners, gold (North Africa was the contact point for the Sahara gold trade), copper, lead, olive oil (still an important Tunisian product), its by-product soap, wax, animal-hides, and silk. This sounds solid enough, but Ibn ‘Awkal’s business was in reality rather more delicate than that. Thegenza letters are full of descriptions of the difficulty agents had in selling at exactly the right moment to get a decent profit; and Ibn ‘Awkal, like every other merchant, had to make informal deals with friends, clients and even rivals, who were on the spot, trusting them to act in his interests. This did not always work. We have a long indignant letter from Samhun ibn Da’ud ibn al-Siqilli (‘son of the Sicilian’) from around 1000 in which a by now probably ex-friend, or client, complains among other things that he had made a loss on Ibn ‘Awkal’s brazil-wood; that he has had to sell Ibn ‘Awkal’s pearls without taking any profit; worst of all, that the latter had not paid Samhun’s creditors despite promises, and despite all that Samhun was doing for him to the detriment of the latter’s reputation; and overall, that Ibn ‘Awkal had been critical with no reason and high-handed into the bargain. There is no reason to think that the Fustat merchant was an especially sympathetic character, in fact. But most letters to him were highly courteous, and explained how the sender had protected his interests, often in adverse situations (war, water damage, low prices), but usually with success.
Ibn ‘Awkal did not trade with Iraq or further east, or with Byzantium, and little even with Syria/Palestine, but he can in other respects stand for an entire network of (usually smaller) Fustat merchants, above all in the diversification of his activities. He was also, it may be added, a pillar of the Fustat Jewish community, and a local representative of the important yeshivas (religious academies) of Baghdad and Jerusalem; had he been Muslim, he would have been a leading member of the ‘ulam’. He was socially central, that is to say, not just economically representative. The only misleading aspect of the entire Ibn ‘Awkal dossier is that it deals with external trade at all. Most Egyptian commerce remained internal to the country. However active the Mediterranean network was, or any other external exchange network, it was Nile traffic, between the major cities and towns, that dominated Egyptian exchange, in 1000 as much as in 700. The real-life feel of the world of the genza letters leaves such an effect on the reader that one can forget this basic economic fact; but it was important, all the same, and would remain so.
The economic history of each of these regions was different between the seventh and the tenth centuries, but it had structural elements in common for all that. The continuing strength of the state in both Byzantium and Egypt compensated, as a motor of exchange, for the temporary weakening of local aristocratic wealth, though this compensation was rather less pronounced in Byzantium, where the state had its own difficulties in the seventh and eighth centuries. In Syria, aristocracies stayed prosperous until 750, but were less integrated into a single regional market by the Umayyad state than Umayyad governors managed in Egypt; after 750, the reverse occurred, with local foci of prosperity slipping, but a fiscal-led integration of regional commerce developing. In Iraq, finally, both aristocracies and (overwhelmingly) the state increased their force in the late eighth century, and set the region up as a major agrarian, artisanal and commercial focus for a century and a half, after which the region slipped back again. We could add al-Andalus, over in the West, to this gallery of examples too, where a set of localized aristocracies of varying wealth existed throughout, but the state became notably stronger in the tenth century (above, Chapter 14), allowing the integration of the economy of the whole peninsula and the creation of some export specializations, silk, saffron and qirmis (crimson dye) among them. Much the same could be said of the Tunisian heartland of Ifriqiya, though there we can see an effective state already in the ninth. The ninth century in many places (except perhaps Syria) saw more internal exchange than the eighth, the tenth century everywhere (except Iraq) saw more than the ninth.
These broadly drawn trends occurred in the internal economies of these regions; but they had an effect on interregional exchange, too, especially in the Mediterranean. The first great Mediterranean trade network was that of the Roman empire. As the empire fragmented, Mediterranean exchange lessened: slowly in the West from 450 onwards, reaching low levels by 600, and snuffing out by 700, as we saw in detail in Chapter 9; rapidly in the East in the seventh century, in the context of the great wars of the 610s-640s, and the fiscal decentralization of both Byzantium and the caliphate thereafter. In the eighth century there was less Mediterranean-wide trade than there had been for over a millennium. Not none; there was always a small-scale network of boats nosing from port to port. The Aegean, as we have seen, maintained a certain enclosed identity as the focus for one level of Byzantine exchange. So did the Tyrrhenian Sea, in the triangle between Rome, Calabria and Sicily, fortified by the continuing force of the city of Rome as a market, as we saw in Chapter 9. As we saw in that chapter too, Michael McCormick has pinpointed the route from Rome to Constantinople as the most important sea route still open in the eighth century. It is not chance that it is the route which linked these two more localized maritime networks; it must have been further reinforced by the fact that Sicily was still a Byzantine province in that century, and probably one of the richest ones. We must recognize, too, that a luxury trade always existed in the Mediterranean, as also in the Indian Ocean, bringing silk and spices to Italy and Francia in return for timber and slaves. But, as we have also seen, luxuries are marginal items to the economy as a whole. In the eighth century, outside restricted areas, the bulk trade in food and artisanal goods had gone, even in the Arab-ruled provinces of the southern Mediterranean, which were always in our period the richest. The seas must have been relatively quiet.
In the ninth century this was slowly reversed. The rise of Venice and the Adriatic route after 750 or so is one small sign of it: small, because Venice focused on the luxury trade mentioned earlier, although this must have been expanding for Venetian wealth to increase as fast as it did in the ninth century (below, Chapter 22); Venice traded with Byzantium and also with Alexandria, from where it stole the body of St Mark, henceforth the city’s patron saint, in the 820s. The ninth-century Tunisian conquest of Sicily allowed for more movement, for Sicily was a great deal closer to Tunis than it was to Constantinople, and there was much exchange between the two regions henceforth; we have seen them operate as a pair in their links with Egypt two centuries later, and that pairing began here, at the latest. South Italian ports like Amalfi and Naples benefited from Arab connections which were now nearer (they indeed colluded in Arab attacks on the Italian mainland), and Amalfitans were regularly to be found in Egypt and the Aegean a century later too. Inside the Arab world, we find more casual references to movement along the African coast, using Tunisia and Sicily as halfway points in the route from Egypt to Spain; and ‘Abbasid centralization, even if focused on Iraq, helped to link Egypt closer to Syria, a link which remained, for autonomous Egyptian rulers after the 860s tended to control Syria as well. All this movement was doubtless still largely in the luxury trade, but there was more of it, in ever more complex patterns; and not all of it was luxury, as with the Arab merchant ships carrying large quantities of olive oil, captured off Sicily by a Byzantine fleet in the 880s, oil that probably came from Tunisia.
In the tenth century there were two further developments. One was that sections of the Mediterranean which had hitherto been relatively cut out of these developing systems, like southern France, were brought in as well; several Arab wrecks from the mid-tenth century have been found off the French coast, apparently from Spain, containing amphorae (for oil?), tableware, copper or bronze, and glass. Byzantium, too, less of a protagonist as yet in the ninth century, is much more visibly so in the tenth, selling quality silks and timber in the Egyptian market, and, later, cheese, a major source of protein for Egyptians; on the south Turkish coast, Antalya became an important entrepôt for trade with Syria and Palestine, and south to Alexandria. The development of the port of Almería in 955 by the Andalusi caliph ‘Abd al-Rahman III was intended to focus and expand the Spanish contribution to this exchange network, and as far as we can see it did just that; Almería makes frequent appearance in the genza documents around 1000 and later. Though certain routes (such as from Alexandria to Tunis) were doubtless more prominent than others, one gains the impression that by the late tenth century one could sail from almost anywhere in the Mediterranean to almost anywhere else - not always directly, but without very much difficulty.
The second development, already indicated by these references to oil-amphorae and cheese, is that it became more normal to transport bulk goods again, for a relatively large-scale market. Tunisian olive oil reached both Egypt and Italy by 1000, just as it had done in 400, although grain was never again a major item of international exchange; that had depended on the fiscal needs of the Roman empire rather than any natural interchange, since it was produced everywhere. Probably on the back of oil, we also, as in 400, find Tunisian glazed pottery in Italy by the end of the tenth century. And, above all, the astonishing choice by a sector of Egyptian merchants, by 1000 at the latest, to send flax to be made into linen cloth in Tunisia and Sicily rather than in the great Egyptian linen factory towns, testifies to a set of commercial relationships that had become large-scale and symbiotic, as well as complex and competitive. Bulk trade did not dominate everywhere yet, or ever; all the same, it is here that we can speak of real interregional/international exchange systems, rather than the thin luxury-based links of two centuries earlier. By the tenth century, the second great Mediterranean trade cycle had properly begun, and would continue to the late Middle Ages. In the eleventh century, newly active Italian ports, Genoa and Pisa, would begin to take over the western part of these Mediterranean networks by force and direct them northwards; the Crusades had similar results in the East; but the trade cycle remained, and even expanded, thereafter.
The tenth century thus saw Mediterranean trade reach the complexity that North Sea trade already had in the eighth and ninth (see Chapter 9), and indeed surpass it. Egypt’s agricultural wealth and productive complexity lay at the heart of it. Even after Italian fleets had partially taken over the role of middlemen, including for the Arab world, by 1100, Egypt was still the hub of this exchange, as well as being the nodal point for luxury goods coming in from the Indian Ocean; it was arguably the motor that ran the entire medieval trade cycle. What happened in the tenth century was that the economies of other Mediterranean regions began to be, in some sectors at least, as complex as that of Egypt, so that relations of mutual economic dependence became more reliable, less risky, solid enough to be built on. This was the basis of the exchange of bulk goods in every period of history.
All the same, we must end this account by repeating a point already made earlier: in every part of the Mediterranean, the most important exchange systems were inside, not between, regions. City-country exchange, and micro-regional agricultural and artisanal specializations, lay at the heart of this, not the wharves of Venice or Almería, Tunis or Antalya, Palermo or Alexandria. Nor are we looking at self-sustaining exchange processes here; however active the merchants of Fustat and Venice were, these would not develop for many centuries. Internal economic development essentially depended on the force of internal demand, and thus on the wealth of élites, and thus on the extraction of surplus from the peasantry. These increased in the ninth and tenth centuries, in the Mediterranean as in northern Europe, creating a more complex and colourful environment, and some artisanal products (like cloth) that could be cheap enough to be bought in villages; but they are nonetheless signs of exploitation as well as dynamism. We shall come back to this issue in the north European context in Chapter 22, where there is more evidence for its effect on the peasant majority.