The Economic History of Byzantium From

Byzantine Specificities

This summary list of the areas useful in a comparison between the Greek and Latin rural situation of the Middle Ages leaves wide open the field of Byzantine particularities. The most striking of these arise from the enduring nature of the Byzantine state and from the continuity of its interventions at the level of the rural economy and society. One may note, in this connection, that through a sort of rebound effect or, one might say, through a process of Byzantinization, western historiography nowadays tends to attach more importance to the tenacity of the influence of public fiscal regulations on the agrarian economy of the high Middle Ages; indeed certain scholars overestimate this influence to an absurd degree.22 It must be admitted that these “fiscalist” concepts do in fact contribute some useful corrections to our understanding of the “barbaric” sixth and seventh centuries. Conversely they cease to be of relevance from the eighth to ninth century on, when the collapse of public structures and the expansion of private lordship led to the fragmentation of power and the establishment, in the tenth to eleventh century, of the rural seigneurie as the basis for the organization and control of all aspects of life in the countryside. It would, therefore, be idle to insist on the original traits conferred in so many ways on the Byzantine rural community, by the presence of a public system of taxation, whether direct or assigned to private beneficiaries. The persistence of public fiscality affected the statutory classification of the various categories of dependents and holdings and the very conditions governing the mobility of land and men. Compared to the Latin West, it is appropriate to note the relatively high rate of monetization of the rural economy, which is in part attributable to the paying of tax and its effect on the vigor of the local economy. This was so in a period when, by a sort of inverse symmetry, the erosion and fragility of the mechanisms of the so-called feudal coinage obliged the landlords, in the West, to favor rents in kind and sharecropping contracts.

Another element that reveals the monetary vigor peculiar to the Byzantine agrarian economy was the importance of the agricultural workforce, far greater than in the West at any rate until the thirteenth century, and of salaried labor in the rural economy generally. The existence of this salaried labor is attested by a specific terminology that has no counterpart in the West, and that designated all the kinds of paid rural labor, whether that of peasants whose holding was insufficient to provide them with a livelihood and who thus belonged to the lowest level of poor and ill-paid laborers, or, on the contrary, that of workers assigned to specialized tasks (millers, shepherds, muledrivers, etc.) as part of the direct exploitation of the domanial reserve. Their appearance in the documentation points to a real absorption of this workforce into the good management of the estates, in the context of a rural economy that, in the case of the eleventh century, Paul Lemerle did not hesitate to qualify as “essentially monetary.”23

22For an excellent critical analysis of the aberrations of the “fiscalist” theses, see Ch. Wickham, “La chute de Rome n’aura pas lieu,” Le Moyen Age 99 (1993): 107–26.

23P. Lemerle, Cinq ´etudes sur le XIe sie`cle byzantin (Paris, 1977). The reference to an economy that was “essentiellement mone´taire” may be found on p. 189 (commentary on the typikon of Gregory

388 PIERRE TOUBERT

Here again, this is in clear contrast with the ministeriales of the western estates who belonged to, or came from, the servile class, even if this latter category is not entirely absent from contemporary Byzantine documentation.24

The enduring capacity of the Byzantine state to intervene, in a way that was both regulatory and practical, in the agrarian economy is all the more striking in that several of the fields of intervention were precisely those that, in the West, were most hostage to the arbitrary behavior of the great private landowners. As an example one may mention the Byzantine state’s power to administer or reassign untenanted properties and the role of clasmatic lands, different from that of the absae or vacantes lands of the great western estates of the ninth and tenth centuries.25

More generally, the tributary nature of the Byzantine state meant that its officials measured the surface of fields and orchards, indeed of space; described holdings or listed their tenants; and defined and periodically adjusted the basis of the land tax. In short, the employee of the fisc became a familiar figure in rural society. The use of computers to process the quantitative data supplied by the fiscal registers or by the accompanying praktika has enabled us to grasp the principles and formal rules followed by the public officials. For the student of the medieval West, there is a twofold lesson to be drawn from the range of research that, in this field, has enriched Byzantine studies. First, the absence of this type of document, or even of any mention of it, in the West, from the eighth to ninth century on, and the private nature of those western documents whose content is most comparable—polyptychs and rent-rolls—is in itself indicative of the disappearance of the fiscal yoke from the Carolingian period on.26 From then, and until the thirteenth century, the demarcation of cultivated land and the measurement of parcels of land became, in the West, the business of notaries or scribes acting exclusively on behalf of private individuals. But a second conclusion is of greater interest: the rebirth of the state and, thereby, of its tributary domination, observable in Europe since the thirteenth century, had the effect of bringing about the establishment of detailed fiscal registers, first of all in the large Italian communes. The oldest surviving fiscal cadasters, such as the great cadaster of Orvieto of 1292, have been analyzed with the help of the computer. The analysis shows clearly that they, like the equivalent Byzantine documents, followed principles of bureaucratic formalization that, in themselves, were responses to the needs of the state and to social conditions.27

Pakourianos). The author also notes that the typikon presents us with “une image valable pour un tre`s grand nombre de cas semblables” (p. 181).

24As, for example, the will of Eustathios Boilas, commented on in Lemerle, Cinq ´etudes, 15–63. For the western equivalents of these “poches d’esclavage rural” in the world of estate ministeriales, one can find all useful references in A. Verhulst, “The Decline of Slavery and the Economic Expansion of the Early Middle Ages,” Past and Present 133 (1991): 195–203.

25See J.-P. Devroey, “Mansi absi: Indices de crise ou de croissance de l’e´conomie rurale du haut Moyen Age?” Le Moyen Age 82 (1976): 421–51.

26In addition to the study by Wickham, quoted above, note 22, see also J.-P. Devroey, “Polyptyques et fiscalite´ `a l’e´poque carolingienne: Une nouvelle approche,” RBPH (1985): 783–94.

27On the subject of the cadaster of Orvieto, which names more than 6,300 taxpayers and, it should be borne in mind, lists more than 18,500 parcels, see E. Carpentier, Orvieto et son contado `a la fin du

A more thorough comparison of the measuring and registering practices of the Byzantine fisc with the registration practices of the extimatores of thirteenthto fifteenthcentury Italian communes would be most enlightening. This is a subject for the future.

It is also well known that the forms of intervention of the tributary state are not limited to the levying of taxes. It is apparent that the Byzantine state played a motive role in numerous areas of the life of the countryside. Without being exhaustive, one could mention the weighty role of the state in the definition of the juridical status of the various categories of tenants, in the mobility of rural populations, and in important transfers of property. The state also established population policies to meet local or regional military needs and influenced the organization of regional or interregional exchange structures through the preferential treatment of urban demand. In all these fields, recent historiography shows clearly how the particularities of the tributary state, as defined by historical anthropologists, make a comparison with Islam more pertinent than a reference to the feudal West.28

Finally, I would like to raise a question, central to Byzantine rural history, that is very revealing, from the western point of view, of its particularities: the variable or permanent coexistence of two apparently distinct structures, the village and the estate. As a form of settlement and of the organization of cultivated space, and also as a social unit, the village is certainly the primary agrarian structure. However unequal and socially differentiated it may be internally, the village is a community. It distills the forms peculiar to peasant society, rich perhaps in internal conflict but also in its ability to soothe these antagonisms and to create local custom. Until the eleventh century the state, in contrast with the West, was committed to the protection of the village. From the eleventh century on, however, this structure, always vulnerable, became more clearly prey to the estates and to the greed of the landed aristocracy. The latter, however, is seen to have been quite different from the contemporaneous aristocracies of the West, clearly because of its links with the state. But this was also because of the particular nature of its family networks, its mobility and firm establishment within the empire, and its ability, at least until the period of the Palaiologoi, to absorb new men. The great estates paralleled these developments and were directly affected by them. Until the eleventh century, the structure of the estate is in fact quite reminiscent of that of the western curtis or villa; it engendered similar forms of settlement: isolated farms and scattered hamlets with dependent settlers. It was run along similar lines of economic management and social structure: the division into lots of the areas reclaimed by agriculture, the strong presence of the estates’ steward whose image, as

XIIIe sie`cle: Ville et campagne dans le cadastre de 1292 (Paris, 1986). Also some very useful contributions in the work of J.-L. Biget, J.-C. Herve´, and Y. The´bert, eds., Les cadastres anciens des villes et leur traitement informatique (Rome, 1989).

28 Some of the most stimulating works on this subject are S. Amin, Sobre el desarollo desigual de las formaciones sociales (Barcelona, 1974); P. Chalmeta, “Al-Andalus: Socie´te´ ‘fe´odale’”? in Le cuisinier et le philosophe: Hommage `a Maxime Rodinson (Paris, 1982), 179–90; and above all the exceptional text by P. Guichard, Les Musulmans de Valence et la reconqueˆte, XIe–XIIIe sie`cles Damascus, 1990–1991, esp. 2:247ff.

390 PIERRE TOUBERT

portrayed by the epitropos of the Geoponika, mirrors on all points that of the villicus of the great Frankish estates, described in Charlemagne’s Capitulare de villis. The occasional presence of elements of fortification designed to protect—really or symbolically—the estate and the preferential investment in certain activities (e.g., mills, vineyards, and olive orchards) are all elements of similarity.

Thus there is no fundamental difference between the structure of the great estate of Byzantium and that of a type of large estate that is easy to find in the ninth to tenth centuries in the Mediterranean Latin West,29 even if this statement cannot be extended to include the great Frankish estates in general. But from the eleventh century on, in conditions that there is no reason to repeat here, the system underwent a series of profound transformations. As a result, what had been a vigorous and quite clearly defined economic unit became a more complex management structure by virtue of the absorption, into the sphere of economic control of the estates, of rural communities henceforth submitted to the hybrid demands of subordination to the state and the domanial context. This development reveals new distinguishing features. The establishment, from the twelfth century on, of a real culture of estate management and accounting afforded the Byzantine landlord a clearer picture of his interests. The documentation clearly reveals a concern for reinvestment of part of the land revenues in the agricultural sector, a form of economic behavior unknown at that time in the feudal West. There can be no doubt that the smallholding allocated mainly (but not exclusively) to a nuclear peasant family continued to be the rule. However, the forms of integration of the rural communities into estate control, known from case studies permitted by the documentation, for example, the studies on the villages of Mamitzon and Radolibos, point to situations entirely irreconcilable with the patterns of western agrarian history of the same period. For example, the well-known case study of the village of Mamitzon in the 1320s presents the picture of a structure that combines a considerable domanial reserve given over to grain cultivated through the tenants’ corve´e labor, together with revenues of fiscal origin levied on the peasants and their plots. It would be fruitless to search for a system of economic organization so complex, and based on such a level of institutional synthesis, in the western seigneurial structure.

These final observations point to a conclusion. For the student of western medieval Europe who is fond of comparative history, the possibilities raised by Byzantine agrarian history are rich in content and full of potential, but on one condition. We must always be careful, if we are to make valid comments on the similarities and differences, to examine the rural realities of the East from two different perspectives. The first allows one to describe the conditions (be they environmental, technical, demographic, and also social, in part) that created a sophisticated agrosystem and one that, in many respects, was developed at the same time and at the same pace as in the western Mediterranean. The second should, on the other hand, point up the factors proper to By-

29 See P. Toubert, “L’Italie rurale aux VIIIe–IXe sie`cles: Essai de typologie domaniale,” in I problemi comuni dell’Europa nel secolo VIII (Spoleto, 1973), 95–132 and 187–206. Here I refer to the great estates belonging to Type I of the typology mentioned.

zantium: political (in the broad sense), institutional, social, and cultural. These two perspectives, as I have attempted to demonstrate briefly, should be complementary. Only a proper awareness of this complementarity allows a healthy comparative approach that would follow the principles and wishes once expressed by Marc Bloch in a fine programmatic essay that has lost nothing of its relevance.30

30 M. Bloch, “Pour une histoire compare´ee des socie´te´s europe´ennes,” Revue de synthe`se historique 46 (1928): 15–50, repr. in idem, Histoire et historiens, 94–123.

The Urban Economy, Seventh–Twelfth Centuries

Gilbert Dagron

Introduction: Cities and Their Economy

Toward a Definition of the City and of the Urban Economy

To counter the excessive confidence of historians who tend to ascribe universal applicability to the urban phenomenon, a number of sociologists have offered a radical critique not only of any uniform definition of “city,” but also of any typology of the city. By these lights, the notion of the “city” is a pure abstraction, and models of cities constitute little more than window dressing.1 It is true that the characteristics ascribed to cities by historians are often more evocative than they are coherent: for example, definitions of cities as fortified sites and market settlements (the Byzantine kastron and/or emporion); centers of production versus consumer or “parasitic” cities; “patrician” and “plebeian” cities; “bourgeois” towns and “rural towns” (the Akerbu¨rgersta¨dte put forth by Max Weber, whose cautions on the limits of the contrast between urban and rural are apposite here). Without venturing into excessively theoretical terrain, it is useful to recognize that the classical origins of our culture have encouraged us to view the city as a microcosm in which social institutions ( juridical, economic, political, and cultural) are concentrated, and to identify it with the “polis”—a juridical entity that represents the antithesis of the countryside (over which it holds sway) and that functions as both intermediary for and counterbalance to the state.2 To the schematics of antiquity, the Middle Ages added its own, envisioning the town as a pocket of resistance against a land-based feudal system, an environment in which ties of dependence dissolve (Stadtluft macht frei; the city is a “savonnette `a vilains”; the city, in short, gives freedom) and in which a specific kind of political relations can develop. Max Weber put particular emphasis on how medieval towns spawned new ties of solidarity, treated as suspect by

This chapter was translated by Charles Dibble.

1See in particular P. Abrams, “Towns and Economic Growth: Some Theories and Problems,” in

Towns and Society: Essays in Economic History and Historical Sociology, ed. P. Abrams and E. A. Wrigley (Cambridge, 1978), 9–33.

2Such are the issues discussed in La Ville, Recueils de la Socie´te´ Jean Bodin 6–8 (Brussels, 1954–57).

394 GILBERT DAGRON

virtue of their being outside the norm and based on the conjuratio. These solidarities could be confraternal and corporative (representing a mechanism of social leveling), or aristocratic (representing the de facto power that leading citizens could gain). Finally, and above all, the medieval and the modern city were considered the locus in which an economy could ultimately cast aside religious taboos and state control to establish its own rationality, deriving from itself a regulatory apparatus over which the state need no longer intervene.3

As doubt proverbially follows certainty, so, with respect to the medieval East, can we formulate a modest set of facts. With the exception of a few large centers, there was little pronounced distinction between the “urban” domain and a rural world that was doubtless home to between 90 and 95% of the population and that dominated the region’s economy and fiscal system. Byzantine towns are perhaps to lesser extent the successors of ancient cities than they are of fifthand sixth-century large rural agglomerations, which had been fortified relatively early in their history in order to resist invaders, and in which basic cottage industries developed. Their population consisted mostly of peasants, who farmed the adjoining land. Thessalonike, the empire’s secondlargest city, lost most of its population at harvest time. The phenomenon was even more pronounced in small towns, in which 1,000–2,000 inhabitants lived essentially off the land, or in mid-sized communities, which mark Benjamin of Tudela’s itinerary and which al-Idrisi mentions in his Geography, with populations of no more than 5,000– 15,000.4 A twelfth/thirteenth-century document, recently studied, reveals Lampsakos as a cluster of approximately 1,000 inhabitants, lacking both specific economic institutions and specific economic organization; its status as something more than a village was a function solely of its substantial trade with Constantinople.5 Although interpretable data are regrettably scarce for the Byzantine period, Ottoman records are revealing. In 1464–65, Serres had a population of approximately 6,000; Drama had fewer than 1,500.6 In the prosperous Asia Minor of the sixteenth century, a mid-sized town comprised 3,000–4,000 inhabitants, a large town between 10,000 and 15,000, a large “city” three or four times that.7 This shows the difference between Constantinople, which would certainly have had 250,000 inhabitants in the wake of its demographic

3See M. Weber, Die Stadt, excerpted from Wirtschaft und Gesellschaft, reworked by Weber, and published separately in 1921, several months after his death, in the journal Archiv fu¨r Sozialwissenschaft und Sozialpolitik, and since published as a separate work in an English translation by D. Martindale and G. Neuwirth, The City (Glencoe, Ill., 1958), and in a French translation by P. Fritsch, La Ville (Paris, 1982). An analysis of the text appears in Figures de la ville: Autour de Max Weber, ed. A. Bourdin and M. Hirschborn (Paris, 1985).

4The Itinerary of Benjamin of Tudela, ed. M. N. Adler (London, 1907); La ge´ographie d’Edrisi, trans. P.-A. Jaubert, 2 vols. (Paris, 1836–40; repr. Amsterdam, 1975).

5G. G. Litavrin, Vizantiiskoe obshchestvo i gosudarstvo v X–XI vv. (Moscow, 1977), 110–27; idem, “Provintsial’nyi vizantiiskii gorod na rubezhe XII–XIII vv.,” VizVrem 37 (1976): 17–29.

6P. S¸. Na˘sturel and N. Beldiceanu, “Les ´eglises byzantines et la situation ´economique de Drama,

7 L. T. Erder and S. Faroqhi, “The Development of the Anatolian Urban Network during the Sixteenth Century,” Journal of the Economic and Social History of the Orient 23 (1980): 265–303, cited by A. Harvey, Economic Expansion in the Byzantine Empire, 900–1200 (Cambridge, 1989), 198–99.

recovery in the second half of the ninth century and nearly 400,000 under the Komnenian emperors,8 and the rest of the empire’s urban centers.

This relativization of the urban phenomenon during the Middle Ages and the unequivocal contrast between mid-sized towns and the empire’s singular megalopolis have evident consequences for how we define the urban economy. Max Weber established a sensible distinction among three levels of artisanal activity.9 (1) At a basic level, there is “demiurgical” activity, concentrated in the village and corresponding to household economy—only somewhat, if at all, specialized and more or less self-sufficient.

(2) At the intermediate level—undoubtedly the most significant with respect to medieval towns—is the production of items for sale by the artisan himself or someone close to him: a relative, a friend, a member of the household, an employee; this presupposes a relatively higher degree of technical specialization, but a local market, or, at the limit, a narrowly circumscribed regional distribution. (3) Finally, there is a level of production that, even if not conducted on an industrial scale, exceeds local or regional demand and is put into the stream of commerce by a merchant rather than by the producer himself; thus ends the straddling of production and sale.

This cautious approach serves as a reminder at the outset that artisanal production is not synonymous with industry. In rural areas—whatever the location, whatever the era—peasants have, with the assistance of neighbors more skilled than they, built or repaired their houses; their wives have crafted homespun clothing, made pottery, tanned skins to make leather; any tools needed could be forged by a blacksmith. Even when these specific skills acquired the status of a specialization, it was not with the intent of creating markets, but in the interest of a complementarity that sought to minimize recourse to monetary exchange. Such was Kekaumenos’ view in the eleventh century when he counseled a large landowner living on his own estate to have mills and workshops so that he would be dependent on no one;10 such was also the intent of the anonymous author of De obsidione toleranda, in listing the artisans that a town needed to resist a siege and that should be brought in from a neighboring region in the event of external threat.11 The objects and rudimentary installations found in the fortified villages of Dinogetia on the lower Danube, or Rentina at the mouth of the Strymon, are associated with a system of social complementarity and not with an urban economy in the strict sense.12 It is at Max Weber’s second level that the basis of activity in Byzan-

8See P. Magdalino, “Medieval Constantinople: Built Environment and Urban Development,” EHB; idem, Constantinople me´die´vale: Etudes sur l’e´volution des structures urbaines (Paris, 1996), 55–57.

9M. Weber, Wirtschaftsgeschichte: Abriss der universalen Sozialund Wirtschaftsgeschichte, 5th ed. (Berlin, 1991), 110ff; French edition and translation by C. Bouchindhomme, Histoire ´economique: Esquisse d’une histoire universelle de l’e´conomie et de la socie´te´ (Paris, 1991), 140ff; English edition and translation by F. Knight, General Economic History (Glencoe, Ill., 1927; repr. New Brunswick, N.J., 1981), 115ff.

10Cecaumeni Strategicon, ed. B. Wassiliewsky et V. Jernstedt (St. Petersburg, 1896), 36–38 (§§ 88–91); Sovety i rasskazy Kekavmena, ed. G. Litavrin (Moscow, 1972), 188–92 (chap. 35).

11Anonymus, De obsidione toleranda, ed. H. van den Berg (Leiden, 1947), 47–48.

12I. Barnea, “Dinogetia, ville byzantine du Bas-Danube,” Byzantina 10 (1980): 239–86; N. Moutsopoulos, “ Anaskafhj` sto`n ojcuro` buzantino` oijkismo` th'" RentiJ´nh",” Prakt. Arcj. EtJ. (1986) [1990]: 154–55.

396 GILBERT DAGRON

tine towns is to be situated, given the importance of itinerant sale and the system of the ergasterion, which most often functioned as both workshop and retail store. The main challenge to the Byzantine economy was advancing to the third level, that of the rationality of exchanges; such is one reading of the Book of the Eparch, whose regulations, discussed below, seek to impose a more pronounced differentiation among products and economic activities.

More recently, Karl Polanyi and his school have stressed the danger of applying the type of analysis that is appropriate to the contemporary world to ancient civilizations, or to civilizations that find themselves outside the modern mainstream; in these cultures, the economy is closely embedded in social relations and has not yet acquired its proper rationality or autonomy.13 Under this theory, one should distinguish the exceedingly rare cases during the Middle Ages in which trade was guided by a sort of self-regulation and influenced production and currency, from the much more frequent situations in which the market responded to basic demands without ever attaining a “national,” much less “international,” level, and in which money represented a means of exchange rather than a true standard of valuation. Particularly in highly centralized states like Byzantium, these two forms of exchange—the one closed and local, the other open, long-distance, and involving professional, often foreign, merchants— could coexist. It is essential, then, to examine the role of the state in regulating the system as a whole, and its policy of providing controlled access to privileged places of exchange (the port of trade) and sustaining a monetary system whose purpose was international. Such questions suggest a multilevel model of the Byzantine economy and link Constantinople, the quintessential port of trade, to a highly specialized role in relation to its unique demography, its urban structures, and its status as capital of the empire.

These theoretical approaches find immediate applicability to the analysis of the two great breaks in Byzantine history: the crisis in the seventh century—which, in the wake of the Slavic invasions and the Arab conquest, provoked a decided retrenchment of urban civilization and mapped out a new urban geography—and the turning point in the eleventh century, which manifested itself initially as an economic development in a climate of peace and territorial expansion, and subsequently as a recession. One might well ask whether this recession was economic or purely military and political at a time when the Turkish advance once again lopped off large portions of the empire and the emperor granted privileges to Venetian, Pisan, and Genoese merchants. Everything indicates that these events, separated by several centuries, modified a number of fundamental equilibria, and it is to be expected that they might serve to mark the broader history of Byzantium. One should nonetheless ask whether this periodization, so useful in tracing out the thread of events, remains applicable to the study of economic and social mechanisms over time.

13 K. Polanyi, C. M. Arensberg, and H. W. Pearson, Trade and Market in the Early Empires: Economies in History and Theory (Glencoe, Ill., 1957); useful studies of the work of Polanyi by L. Valensi et al. appear in AnnalesESC 29 (1974): 1309–80. See the analysis of A. E. Laiou, “Economic and Noneconomic Exchange,” EHB 681–89.

The “Break” of the Seventh Century

The facts are known. In the Balkans, the frontier began to give way when the khagan of the Avars took the city of Sirmium in Pannonia in 582 and furthered the advance of the Slavs into Thessaly, Epiros, Achaia, and Hellas. The Danubian limes, reestablished around the year 600 by Maurice, gave way for good around 613–615, as the last fortified points succumbed: Naissus, Sardis, Justiniana Prima. Thessalonike alone resisted the numerous Avaro-Slavian sieges (in 586, 615, 618). Byzantium retook control of the territories it had lost, but only slowly and partially: Justinian II reached Thessalonike with difficulty in 688/89; Constantine V launched a decisive campaign against the Sklaviniai in 758–759, but the surrender of the Slavs of Thessaly, Hellas, and the Peloponnese was secured only by the patrician Staurakios heading a large army in 782–783; a few pockets of resistance were defeated by the strategos Skleros in 805.14

In Asia Minor and in the East, the crisis began with the assassination of Maurice in 602. A little later, the Persian armies brought about the fall of the eastern provinces and opened access to Asia Minor. The counteroffensive launched by Herakleios beginning in 624 resulted in the capture of Dastagerd, the collapse of the Persian Empire, and the recovery of the purported relics of the True Cross, restored to Jerusalem on 21 March 630.15 The Arab conquest began almost immediately thereafter, however, and met with little resistance from an empire that was by now exhausted. The cities, which had acquired a de facto autonomy, most often preferred to bargain and to open up their gates. In 636 the battle of Yarmuk took place, and within four years Syria, Palestine, and subsequently Egypt were lost for good.16 The period that followed was but a slow consolidation, lasting more than a century, of a new frontier that consisted of the Taurus Mountains and Mesopotamia. The towns, reduced in both number and size, began to fortify themselves, and their social structures assumed a military character. Both camps, envisioning a state of permanent war, organized defensive networks: thu¯gurs on the Arab side, small border themes on the Byzantine side. With occasional advances and occasional retreats, this equilibrium lasted until the frontier was again breached, starting in 962, with the great campaigns of reconquest launched by Nikephoros II and Leo Phokas, John I Tzimiskes, and Basil II.

The turmoil did not spare Constantinople, which was severely affected by plague in 542 and subsequently forced to endure food shortages that began in 618, with the cessation of wheat imports from Egypt. The city resisted a united assault by the Avars

14For the chronology, see P. Lemerle, Les plus anciens recueils des Miracles de Saint De´me´trius et la pe´ne´tration slave dans les Balkans, vol. 2 (Paris, 1981). Cf. A. E. Laiou, “Exchange and Trade, Seventh– Twelfth Centuries,” EHB 700–702.

15Shahrbaraz crossed the Euphrates in 610; Antioch was taken in 611; Cilicia fell in 613. Shahin captured Caesarea in Cappadocia in 611 and reached Chalcedon. Shahrbaraz subsequently took Palestine (Jerusalem was captured in 614 in a bloody battle) and Egypt (Alexandria was taken in 619). A chronological analysis of the sources appears in B. Flusin, Saint Anastase le Perse et l’histoire de la Palestine au VIIe sie`cle (Paris, 1992), 2:67–181.

16On the Arab conquest, see F. M. Donner, The Early Islamic Conquests (Princeton, N.J., 1981).