Broad economic structures had once spanned the Mediterranean and fostered Byzantine commercial interaction with the west. The sweeping changes of the seventh century naturally affected communications between the two former halves of the empire. The occasional western shipwreck and growing ceramic evidence of imports confirm that, although dwindling, economic links to Byzantine Africa and the easternMediterranean persisted well into the seventh century, perhaps reinforced by supply efforts to the last garrisons perched along the Ligurian coastline.34 But even the trickle of sea communications between Constantinople and the west seems almost to dry up towards 700.35 A high-status ecclesiastical community whose traces have been recovered at Crypta Balbi in Rome was still importing some luxury wares from Byzantine Africa and the IslamicMiddle East in the late seventh century. But the imports practically ceased early in the eighth century, if one can judge from their absence and a turn towards recycling and away from Levantine raw materials in glass manufacture.36 Long-distance shipping routes declined in importance, and regional shipping networks emerged as characteristic.37 The reasons for all this remain controversial; certainly the causes were multiple. Networks of easterners trading with the west may have withered under the cyclical plagues, whose contagion contemporaries linked with shipping. Declining economic fortunes presumably shrank western purchasing power even before the Islamic conquests redistributed eastern wealth and reorganised macro-economic structures, fanning demand in the east. To grant Henri Pirenne his due, warfare around the Mediterranean rim probably played a role. The final fall of Byzantine Carthage in 698 disrupted a crucial pivot for shipping linking the eastern and westernMediterranean. 38 Even before it came to conflict, the rapid build-up of Arab and Byzantine fleets will – initially at least – have competed with merchant ventures for such sailors and ships aswere available.39 Greek and Coptic papyrus archives of 698–711 from the inland town of Aphrodito on the Nile paint an astonishing picture of how the new rulers mobilised local wealth and conscripted landlubbers for sea raids (koursa) launched from Africa, Egypt and the east.40 Land travel too was disrupted: Byzantine loss of control in the Balkans blocked the old Roman overland routes to the west, essentially cutting Italy off from Constantinople during the winter months.41 The structure of exchange within the territories that remained Byzantine took new shapes, as the shrunken fiscal component, formerly dominated by massive northbound shipments from Egypt and Africa, sought new bases. And, it has been argued, very differently administered economic zones emerged in the capital and the provinces.42 Byzantium’s own long-distance trade in the eighth century seems to have been reorientated along a new axis linking the Aegean and the Black Sea.43 Direct documentary evidence of trade between Byzantium and the west is slim, and complicated by the ambiguity of the notions of trade and ‘Byzantine’. Should we classify as ‘Byzantine’ middlemen the Venetian merchants who recognised Byzantine sovereignty and sailed between Italy and Africa, Egypt and Palestine? In any event, over the next two centuries the old infrastructures of travel gradually recovered or were replaced. After the nadir of the earlier eighth century, communications and also commerce rebounded smartly in the final quarter of the century; they continued to climb into the reign of Louis the Pious (814–40) and, after a period in which they leveled out, growth resumed in the later ninth century.44 Although the structures and volume of trade differed greatly from those of the late Roman period, communications and commerce were again significant factors in the relations between western Europe and the Middle East. Practically in the shadowof the Alps, the more or less autonomousByzantine outpost of Venice rose out of the Adriatic mists from insignificance to embody this change. The ancient trunk route linking the Tyrrhenian coast with theMiddle East via the straits ofMessina, around southernGreece and across the Aegean had never ceased completely to function. Indeed, in 746– 7, it transmitted eastwards to the Byzantine capital the last major seaborne outbreak of the bubonic plague (until 1347), even though the epidemic had begun in the Levant.45 But over the next century, branch routes sprouted again on the Adriatic or through the Gulf of Corinth, feeding piracy along the coast. The old Balkan and Danubian overland routes, including the Egnatian Way, returned to activity, even if the Hungarians were to make the Danube corridor short-lived.46 The rare data on Venetian shipping between c. 750 and 850 point mostly to trade between Italy and the Islamic world. Muslim traders show up at Rome around 800. Slave trading ran along the west coast of Italy and involved Rome and the shipment to Africa of enslaved Europeans by Venetian andGreek merchants c. 750–75; and Emperor Leo V (813–20) was eager to block Venetian commerce with the caliphate.47 Further west, it may be more than coincidental that all four of the tenth-century shipwrecks discovered off the French Riviera were carrying goods from the Islamic world.48 At the head of a reinvigorated Adriatic shipping route, Arab gold and silver coins competed with Byzantine and Frankish money, to judge from the way coins found in the earth converge with those mentioned in contemporary records. They testify in their way to intensifying economic links between the Islamic world and eighth-century Italy. That people involved in this sort of contact also frequented at least Byzantine provincial centres is also suggested by finds of Byzantine bronze coins. Although these coins never served as a medium of international exchange, their presence in western Europe nevertheless indicates that the people who lost them had direct or indirect contacts with the Byzantine world.49 Merchants make only sporadic appearances in Carolingian sources, and ‘Greeks’ are rarest among them.Traders who do showup tend to be Frisians, Anglo-Saxons, Jews or Italians, and only rarely can we discern the horizons of their activities.50 Some Jews and Italians may have been subjects of the Byzantine emperor; in any event commercial contacts with Byzantium may have been realised through non-Greek intermediaries. Thus, in 885–6, an astute observer in the caliphate described the trading patterns of Jewish merchants who alternated their export voyages of eastern luxury wares, travelling one year to Constantinople, and the next, to the Frankish court. One of their main western routes probably ran through Venice.51 A few ‘Greek’ merchants do crop up in the eighth-century Tyrrhenian Sea. Towards the tenth century, some Italo-Byzantines imitated their neighbours in Amalfi, Naples and Gaeta, such as the Greek slave trader from Armo near Reggio di Calabria who would not sacrifice his trade’s superior profits for less reprehensible commercial ventures. A near-contemporary life of a Sicilian saint compared his crossing from Africa back to Sicily c. 880 to ‘some huge ship filled with all kinds of merchandise’. Such hints perhaps explain the concession of a landing for Greek merchants to the church of Arles by one of the last Carolingians, Louis the Blind who, as we shall see, had other connections with Constantinople.52 The account of the North African crossing underscores that Greek merchants, like their Venetian peers, might have found more profit linking western Europe with the huge Islamic economy than with Constantinople. In other words, western contacts with Byzantine merchants may have been an indirect result of commercial relations with the Islamic world. But this does not diminish the significance either of the intermediaries or of the overall growth in infrastructures which permitted and channelled relations between Byzantium and the west.