The upheavals of the seventh century had transformed Byzantium. The old urban fabric of the Roman empire largely gave way. Though the precise causes and chronology remain controversial, archaeological evidence shows that, in the long run, the cities of AsiaMinor and the Balkan peninsula fared little better than those ofwestern Europe.11 Despite streams of refugees, even the capital of Constantinople shrank dramatically in population.12 To the north, the old Danube frontier and much of the Balkans were overrun by Slavs, Avars and Bulgars, although the imperial government still clung to coastal strongholds like Thessaloniki or Monemvasia. The closing of the old military roads across the Balkans effectively sundered Byzantine Italy from the imperial centre in the winter months, when sailing was difficult.13 To the south, Rome and Ravenna hung by a thread as the Lombards expanded their power from the Po basin down Italy’s mountainous spine. The bold attempt of Emperor Constans II (641–68) to defend the empire’s southern flank by transferring his imperial headquarters back to Italy around 662 collapsed with his murder.14 At the same time, a cashstrapped government intensified the fiscal yield of its western provinces.15 That pressure may have reinforced tensions which had started over religious issues. For a government whose professional bureaucracy and military forces were sustained largely by a land tax levied on the provinces, the fiscal implications of such territorial losses were devastating, amounting to as much as three-quarters of revenues.16 Defeat and the fiscal crunch forced radical administrative and military reconfigurations in the empire’s besieged remnants. And conjugated disaster opened more than a political crisis in a society which lived and breathed its religious sentiment: the challenge of Islam was ideological no less than political and military. Was the sect in whose sign the Roman empire had conquered since Constantine’s conversion no longer stamped with God’s seal of success? Lifestyle and mental attitudes underwent a sea change as the amenities of late Roman daily life became a thing of the past outside the court’s island of archaism.17 By the seventh century, Greek had supplanted Latin as the characteristic language of the central administration. Outside the Latin-speaking outposts of Dalmatia and Italy, only the Latin lettering of coins and imperial documents, a fewfossilised acclamations and the massive presence of Latin loan words in the technical jargon of the state offered a faint linguistic echo of the old Roman past. Byzantine culture no longer coincided with the Byzantine polity. For a few generations, Constantinople ceded Hellenic cultural leadership to the empire’s geographic edges. John Damascene, the greatest Byzantine thinker of his time, wrote his Greek theological treatises under the Arab caliphs; the best Byzantine art adorned the shrines and pleasure palaces of the new Islamic empire, while remarkable Byzantine hagiography of the eighth century was produced in Italy or Palestine.18 Small wonder that one of the few pieces of contemporary Byzantine literature translated into Latin around 700 is an apocalyptic vision of the Arab conquest, the last Byzantine emperor’s return to Jerusalem and the impending end of the world!19 But events would follow an unforeseeable path. Transformed and reorganised, Byzantium was about to begin a remarkable resurgence. Bede, who had earlier succumbed to optimistic reports of the Roman reconquest of Africa, accurately reported the successful defence of Constantinople from the final Arab siege of 717–18.20 That victory inaugurated an era whose scarce sources cannot obscure the renewal of Byzantine civilisation which, by 900, stood on the threshold of its great medieval expansion. The changes that produced this revamped empire are much debated. The Byzantines themselves located the defining moments of their history in dynasties and doctrines, a vision which says as much about emperor and faith in Byzantine mentality as about historical trends. By these lights, confusion and usurpation followed the toppling and execution of Heraclius’ last descendant, Justinian II (685–95/705–11), until the usurper general Leo III (717–41) defended the capital from the Arabs and launched the ‘Isaurian’ or Syrian dynasty. The victorious Leo promoted a new cult practice whose affinity with Islam many observers feel is undeniable: he proscribed most religious images and their veneration as a form of idolatry. His dynasty championed iconoclasm almost to the end, uncovering powerful stresses within the Byzantine ruling class which succeeding generations memorialised as religious persecution (see above, pp. 279–84). Three generations later, the regent empress Irene recruited the support of Pope Hadrian I (772–95) to overturn the imperial doctrine at the second ecumenical council of Nicaea in 787 (see above, pp. 287–8). Charlemagne’s ambassadors witnessed the palace coup that ended Irene’s independent rule and the Isaurian dynasty in 802. This spell of short reigns, involving a toneddown reversion to iconoclasm, led to a coup by Michael II (820–9) who established the Amorian dynasty, named after his home town in AsiaMinor, where excavation has uncovered the material face of the age.21 Another regent, Empress Theodora (842–56), finally abolished iconoclasm in 843. Her son Michael III (842–67) and the Amorian house were overturned by a palace parvenu, Basil I (867–86). Down to Michael III’s time, soldier emperors predominated: Leo III and his son Constantine V (741–75) were particularly successful commanders. Reorganisation and re-establishment of control characterise this era. Survival required first and foremost the military stabilisation of the eastern front, where Arab incursions into the empire’s new agrarian heartland of western Asia Minor were increasingly checked thanks to new provincial defence systems, known as themes. These themata spread sporadically as events dictated. The word’s derivation is contested but it refers simultaneously to autonomous military units and to the large territorial districts in which they were permanently stationed and of which the empire was composed. They may have been inspired at least in part by the western exarchates, earlier administrative and military structures elaborated in reconquered Italy and Africa. By the time of Charlemagne and his son, themes and the generals or military governors (strat¯egoi) who headed them had everywhere ended the late Roman tradition that strictly separated civil and military administration, and government had shifted to a permanent war footing.22 The mighty themes of Asia Minor helped slow the Arab advance. The European themes straddled the capital’s western approaches and defended Constantinople from the rising power of the Bulgars. But the very concentration of power that facilitated the generals’ defensive tasks complicated the political structure of the empire, since strat¯egoi like the future Leo III often challenged the emperor resident in Constantinople (see above, p. 380). The last great revolt of the themes in particular had serious consequences. The civil war between Michael II in Constantinople and Thomas the Slav in 821–3 and the ensuing disarray contributed to the empire’s greatest territorial losses in our period: the Arab conquest of Crete (c. 824–8) and the beginning in 826 of the fall of Sicily, both of which had implications for imperial communications with western Europe. To counter their own provincial armies, the Isaurian emperors created a new, imperial army of cavalry and infantry, known simply as ‘the regiments’ (tagmata) and headquartered in the capital. The tagmata spearheaded offensive operations and played a key role in the Isaurians’ notable successes in the Balkans and AsiaMinor. At sea, the seventh-century Karabisianoi fleet, essentially conceived to defend the central coastal areas and sea approaches to Constantinople, was superseded by provincial fleets organised as maritime themes in the course of the eighth and ninth centuries. An imperial fleet equipped with Greek fire was stationed at Constantinople and chiefly destined for long-range intervention, flanked by elements of the thematic fleets.23 On the empire’s western flank, a naval squadron based in Sicily brought enhanced security to Italian waters in the 750s.24 Despite occasional setbacks, the newmilitary apparatus proved effective in preserving the empire. As surviving inscriptions attest, the emperors began refurbishing critical infrastructures across territories that had slipped out of their control in the seventh century. Whatever local discomfort the return of imperial tax-collectors may have brought to any provincial landowners who might have survived the century of storms, the centripetal dynamic was probably powerful: imperial armies brought coinage, administrators and bishops, who sent back to Constantinople the newly restored tax revenues; these, in turn, reinforced the imperial treasury which financed the bureaucracy and military apparatus and enabled the empire to extend its reach even further.25 In the capital, the few great late Roman ministers like the praetorian prefects or the masters of the offices, into whose offices various vertical chains of administrative institutions formerly converged, had disappeared. They were replaced by the omnipresent ‘accountants’ or logothetes ever vigilant for income and expenditure of a state straining against the abyss. These new sub-ministers reported directly to the emperor and so brought more direct lines of authority into his hands (see above, pp. 238–9, 273). Administrative structures were far more institutionalised than in the west, as professional bureaucracies looked after imperial finances and justice. Whatever survived or now emerged as a ruling class owed much to government service as the source and sign of its wealth and power. The stresses of a ruling class in the making mark the top echelon of society: frequent coups d’´etat, political shake-ups and church schisms start to stabilise only in the tenth century. From c. 800, Byzantine and Frankish sources yield the earliest glimpses of family names and clans like the Phokai or the Argyroi who would dominate the social scene at Byzantium’s apogee and who seem to ride a rising tide of economic and demographic recovery.26 A state hierarchy structured this emerging power elite, as imperial promotion granted life-long, nonhereditary state dignities like patrician or pr¯otospatharios and salaries to leading officials who trumpeted their titles on numerous surviving lead seals. Each official’s place in the hierarchy was communicated by his position in imperial ceremonies and delineated in official lists of precedence, the earliest surviving example of which (the Taktikon Uspensky) dates from 842–3.27 For all its factions, the power of this senatorial order was such that a prudent pope might demand that it confirm by oath guarantees issued to his legates by a shaky regency, and it is this social group that supplied most of the challenges to imperial authority, whether they came in the form of conspiracies, usurpations or doctrinal dissidence.28 Iconoclasm, the most lasting and disruptive doctrinal quarrel of the era, had many consequences. Resistance to the imperial heresy challenged the emperor’s power in matters of doctrine and, implicitly, in other matters as well. The considerable efforts subsequently devoted to restoring the emperor’s prestige and redefining relations with the ruling class are most visible in the refurbishing of imperial ceremonial. Iconoclasm affected the institutional history of the church even more deeply. The patriarchs resided only a stone’s throw from the imperial palace and were often under the emperor’s thumb. The secular church’s relative tractability with respect to imperial doctrinal shifts fostered internal conflict. Churchmen who sought to resolve conflict without throwing the ecclesiastical hierarchy into chaos clashed with zealots. A monastic party centering on the great cenobitic reformer Theodore the Stoudite (759–826) burned to root out any who had temporised with what had been the empire’s official doctrine over nine of the last twelve decades, factionalising the church in ways which paralleled and were perhaps connected with fissures in the lay aristocracy (see also above, pp. 288–9). In any case conflict spilled over into other issues and spawned a series of bitter schisms from the Moechian controversy – a dispute centering on Emperor Constantine VI’s decision to divorce and remarry in 795 – to the ‘Tetragamy’ in which the Italian-born patriarch and former imperial adviser, Nicholas I Mystikos (901–7, 912–25) bitterly opposed Emperor Leo VI’s (886–912) fourth marriage (see below, p. 503). Since partisans of each faction challenged their opponents’ ecclesiastical appointments, Byzantine bishops’ careers seemed noticeably unsettled in this era. Factionalism in the upper echelons of church and state provoked sudden political shifts which affected relations with the west. Since the days of Pope Leo I (440–61), whose memory the iconophile hero Theodore Graptos still celebrated, the Roman see and its doctrinal rectitude had enjoyed great prestige in the Constantinopolitan church. This prestige was only enhanced by Rome’s role in the earlier monothelite controversies and Pope Martin I’s (649–55) resistance, arrest and death in imperial custody which led the Byzantine church to venerate him as a martyr (see above, pp. 231–2). A Greek account of his suffering was composed in eighth-century Jerusalem or Rome, an alternative which is itself revealing. Rome had become the authority to which Byzantine religious thinkers under pressure appealed for support. That the duchy of Rome was slipping out from under the emperor’s effective administrative reach only increased its attractiveness, hence efforts to persuade western authorities to curtail the activities of eastern ´emigr´es at Rome.29 Culturally, four generations of theological debate for and against icons spurred renewed examination of the Hellenic theological and cultural heritage. The hunting out and recopying of old books – to uncover or rebut authorities on icons – marks the earliest stages in the birth of Byzantine humanism, the encyclopaedic movement. The political and economic recovery of a society based in large part on written administration equally invigorated literary culture. Imperial bureaucrats like the future patriarch Nikephoros I (806–15) figure prominently in the early phases of the revival.30 They were reinforced by intellectuals and others like Michael Synkellos or Patriarch Methodios (843–7) who migrated back to a recovering imperial centre from Arab-controlled Palestine or the Italian borderlands. As in the west so in Byzantium a new minuscule book script was the tool and hallmark of the new culture.31 So too the new Greek writing required and conditioned the phenomenon of transliteration. Ancient exemplars in the old script were sought out, compared and copied in the new script. That new writing has preserved in such Byzantine ‘editions’ most of what has survived of classical Greek and patristic literature. In the capital, the receding danger of Arab siege was replaced by the imminent menace of Bulgar attack. Repairs were made to the city walls.32 Behind their protective bulk, renewal stirred in a city where nature had reconquered much of the urban fabric. Though on a much smaller scale and with a more religious focus than the colossal monuments of the old Roman state, construction and redecorating were nonetheless significant by recent standards and invite comparison with contemporarywestern efforts. In 768, ConstantineVrestored the aqueduct ofValens, which had been interrupted since 626 and was essential to the water-starved site of Constantinople. Numerous churches were remodelled in the ninth century. Theophilos (829–42) built a new suburban palace, ‘Bryas’, modelled, significantly, on the Arab caliphs’, the new standard-bearers of luxury. Basil I constructed a splendid new chapel, the Nea, for the Great Palace, and the ebb and flow of icon veneration required redecoration of religious shrines according to the dictates of the moment.33 It was, then, a changing Byzantium which bordered onwestern Christendom. As the threat of political extinction receded, the reorganised empire reasserted control. The progress of imperial administration allied with an improving general situation and sporadic disarray amongst the empire’s most lethal enemies to allow renewed, if staccato, campaigns of intervention at the empire’s extremities, which despite all setbacks and reversals steadily extended outwards from Constantinople.