By the last quarter of the seventh century the Byzantine areas of Italy had experienced over a century of upheaval. Within decades of their first invasion of Italy in 568 the Lombards had established a powerful kingdom consisting of the territories north of the river Po, Tuscany and the two outlying duchies of Spoleto and Benevento. The empire was confined to the areas of Rome and its duchy, Ravenna, and the neighbouring areas of the exarchate and the Pentapolis, approximating to the present-day Romagna and Marche, and a few coastal areas elsewhere. The Byzantines had only been able to hold on to their possessions by initiating a thoroughgoing militarisation of society, which involved the concentration of land in military hands and the concentration of authority in the hands of the commander-in-chief in Ravenna (the exarch) and his subordinates (duces and magistri militum at a provincial level and tribuni in the localities). In many areas, such as the Roman Campania, this process was accompanied by a steady shift of population, as settlement became concentrated on military strongholds and refuges, usually located on promontories. Although the pressure eased somewhat in the seventh century, Liguria and most of the remaining settlements on the Venetian mainland were lost to the Lombards in the reign of King Rothari (636–52), and the duchy of Benevento made continual encroachments in the south, accelerating after the unsuccessful expedition of Emperor Constans II (641–68) to southern Italy in 663–8. Internal tensions were reflected in a series of revolts, the determined opposition led by the papacy to Constans II’s monothelite doctrines and a bitter conflict between the sees of Rome and Ravenna over the same emperor’s grant of ecclesiastical autonomy (autokephalia) to the latter in 666. In two letters addressed to his successor, Pope Agatho (678–81) bemoaned the dislocation caused by the ‘gentiles’ and complained that lack of food forced the clergy to work the land.

In the present state of our knowledge it is not difficult to describe the physical setting for pre-Islamic Arabian history, and new archaeological discoveries in Saudi Arabia, Yemen, Jordan and the Gulf are producing much valuable evidence. Over the past century a vast body of epigraphical material – some 50,000 north and south Arabian inscriptions and the inscribed sticks now emerging by the hundreds in northern Yemen – has provided a wealth of information on the societies of the peninsula, especially the bedouins. But all this seldom provides a coherent picture of the course of events, as opposed to vignettes and bare details, and thus does not replace a literary historical tradition. There are external epigraphic records of the Arabs and Arabia, and historical sources – especially in Greek and Syriac – are often helpful. But this information too is profoundly discontinuous, and in any case represents the perspective of outsiders who regarded the Arabs as barbarian marauders and most of Arabia as a menacing wasteland.

The early medieval societies of Byzantium and western Europe that emerged from the late Roman world shared more than a few institutions, traditions and religious experiences. They sometimes rubbed shoulders in ways we overlook. Rome’s clerical elite was so hellenised that the pope who reigned at Charlemagne’s birth spoke Greek as his mother tongue. Under Charlemagne’s grandsons, members of the Byzantine missionary Methodios’ entourage wrote Greek majuscules in the memorial book of a German monastery to record their stay; Methodios was himself a native of Thessaloniki, formerly a Byzantine imperial official in Macedonia and a monk in Bithynia (see above, p. 300). Conversely, Franks served in the Byzantine emperor’s military household and figured at palace banquets.

Anyone wishing to unravel the history of the relationship between Byzantium and Armenia from late antiquity into the eleventh century has to confront a series of historical and historiographical challenges. The most immediate, and intractable, of these is one of definition: what does ‘Armenia’ mean? Although Armenia is used to express a territorial entity in contemporary texts, both Armenian and non-Armenian in origin, its precise meaning varies according to the date and the context in which it is used. Far from finding a single, stable definition of Armenia, one discovers multiple ‘Armenias’. Thus a seventh-century Armenian geographical compilation depicts ‘Great Armenia’ as comprising not only regions currently recognised as Armenian but also those with historic associations. Successive provinces of Armenia were imposed and superimposed by external powers, each with a particular scope. The kingdom of Armenia, re-established in 884, bore little relation to its Arsacid precursor and increasingly represented only the Bagratuni kingdom centred on Ani, excluding rival kingdoms in Vaspurakan, Siwnik‘ and elsewhere.

Byzantines were perhaps more concerned than most medieval people with the insecure business of measuring time and defining authority. There was not much they could do about either, but naming is a taming of the forces of nature and anarchy, and placed the humblest in relation to the stability of God. Byzantines called this order taxis. They craved taxis all the more in the fifteenth-century anno domini (ad), because for orthodox Christians, who counted by the anno mundi (am), it was, quite simply, the end of the secular world. For subjects of either, or both, emperor and patriarch in Constantinople, the world was created on 1 September 5508 bc. Gennadios II Scholarios (1454–6, 1463, 1464–5), Sultan Mehmed II’s (1451–81) first patriarch after the fall of Constantinople to the Ottoman Turks on 29 May 1453, put matters in cosmic proportion by foretelling doomsday on 1 September 1492, the end of the seventh millennium am. In 1393, the first year of the last century of the world, Patriarch Antony IV (1389–90, 1391–7) put matters in taxis. Grand Prince Vasilii I of Moscow (1389–1425) had remarked that although there was a church, there did not seem to be a credible emperor in Constantinople. The patriarch replied: ‘it is not possible to have a church without an emperor. Yea, even if, by the permission of God, the nations [i.e. the Turks] now encircle the government and residence of the emperor … he is still emperor and autocrat of the Romans – that is to say of all Christians.’

The last seventy years of the ninth century were an era of disorder and continued crisis in southern Italy. The government of the principality of Benevento, which ruled over most of the south of the peninsula, was riven by succession disputes which led to the formal partition of the principality in 849. But far from ending the contention, this division gave only a brief pause in the internecine strife. Muslim attacks from Sicily and North Africa threatened to swamp a feeble and divided Christian defence, and the local rulers were far more intent on their internal power struggles than on making any coherent and effective stand against the invader. However, the years round about 900 marked a very significant change, with regard both to the internal stability of southern Italy and also to its relative freedom from external threat – or at least from the threat of conquest rather than sporadic raiding. For much of the tenth century the land was not exactly peaceful, but freed at least from the dreary litany of civil war and the establishment of territorial footholds for further Muslim advance that had made the previous period a troubled one, the impact of which had been reflected in the pessimism of contemporary chroniclers such as Erchempert, and in the number of charters mentioning relatives or fellow monks captured by the Saracens.

Byzantine emperors desired stability and security in the peripheral regions of the empire so as to continue controlling and exploiting the productive lands which provisioned the principal cities, most importantly Constantinople; these also yielded tax revenues to support the apparatus of government. In the Balkans the vital regions were the rich lands of Thrace and the hinterland of Constantinople in the east, and Thessaly and the lands around Thessaloniki in the west. Security required direct supervision of major communication routes, by land and water, and of strategic cities across the peninsula, but only a stabilising influence in the mountainous interior, the north-eastern plains and the north-western littoral. Control of the Black Sea ports between Constantinople and the lower Danube, notably Anchialos, Mesembria and Varna, was considered essential, as was command of the major mountain passes through the Haemus mountains. Minor paths remained in the hands of locals, largely Vlachs, whose allegiance was assiduously cultivated.

As earlier chapters have shown, the empire’s military situation was alleviated by political upheavals in the Muslim world and the abatement of hammer blows directed by the Abbasid leadership. The caliphate itself had more recourse to diplomacy, recognising Ashot I Bagratuni (‘the Great’) (884–90) as paramount prince among the Armenians and bestowing a crown on him. Soon afterwards, Basil I (867–86) responded with démarches of his own towards Ashot. The later ninth century probably saw the elaboration of the basileus’ diplomatic web eastwards, drawing in political elites in central and eastern Caucasia such as ‘the chiefs of Azia’, lords of the Caspian Gates. By the reign of Leo VI (886–912) the court was maintaining well-to-do Turks from the Fergana valley as well as Khazars, and these young men were making substantial down payments of gold in order to receive annual rogai as members of a unit of the imperial bodyguard. The chinks in Muslim power were shown up in other forms, such as the prisoners-of-war kept at court. The more prominent among them were enrobed in the white garments of catechumens at the emperor’s Christmas and Easter banquets, as if to affirm willingness to adopt the religion of the Christians. Triumphal parades of Basil I, as of Theophilos (829–42), celebrated with spectacular props the emperors’ occasional forays into Muslim-held regions, and a poet could write of Basil as a new David, who with God’s help will vanquish the enemy hosts.

Throughout the political history of western Europe, there have been few periods of such dramatic change as the fifth century. In 400 the borders of the Roman empire in the west, by then distinct from the eastern empire which was governed from Constantinople, stood reasonably firm. They encompassed all of Europe south of the Antonine wall in Britain and the Rhine and Danube rivers on the continent, extending eastwards of the Danube’s confluence with the Drava; they also included a band of territory along the African coast, stretching two-thirds of the way from the Straits of Gibraltar to the Nile. But within a hundred years this mighty entity had ceased to exist. North Africa had come under the power of groups known as Vandals and Alans; Spain of Visigoths and Suevi; and Gaul of Visigoths, Franks and Burgundians. The Romans had withdrawn from Britain early in the century, leaving it exposed to attacks from the Irish, Picts and Anglo-Saxons, while in Italy the last emperor, Romulus Augustulus, was deposed in 476 by a general, Odovacer. The supplanter of Romulus was himself deposed and murdered in 493 by Theoderic the Ostrogoth (493–526), who established a powerful kingdom based on Italy. While the empire had weathered the storms of the fifth century largely unscathed in the east, in the west it had simply ceased to exist. Western Europe, one might be excused for thinking, had moved decisively into a post-Roman period, and the middle ages had begun.

Between the death of Alexios I Komnenos and the establishment of the Latin empire of Constantinople, eight emperors ruled in the eastern Roman capital. Their reigns were as successful as they were long: under John II Komnenos (1118–43) and Manuel I Komnenos (1143–80) Byzantium remained a wealthy and expansionist power, maintaining the internal structures and external initiatives which were necessary to sustain a traditional imperial identity in a changing Mediterranean world of crusaders, Turks and Italian merchants. But the minority of Manuel’s son Alexios II Komnenos (1180–83) exposed the fragility of the regime inaugurated by Alexios I. Lateral branches of the reigning dynasty seized power in a series of violent usurpations that progressively undermined the security of each usurper, inviting foreign intervention, provincial revolts and attempted coups d’état. Under Andronikos I Komnenos (1183–5), Isaac II Angelos (1185–95), Alexios III Angelos (1195–1203), Alexios IV Angelos (1203–4) and Alexios V Doukas (1204), the structural features which had been the strengths of the state in the previous hundred years became liabilities. The empire’s international web of clients and marriage alliances, its reputation for fabulous wealth, the overwhelming concentration of people and resources in Constantinople, the privileged status of the ‘blood-royal’, the cultural self-confidence of the administrative and religious elite: under strong leadership, these factors had come together to make the empire dynamic and great; out of control, they and the reactions they set up combined to make the Fourth Crusade a recipe for disaster.

A chapter dealing with Iranian feudalism in a distinguished series dedicated to The rise and fall of the Roman world bears the title ‘Iran, Rome’s greatest enemy’. This title is more than merely a justification for the inclusion of a chapter on Iran in a work devoted to the history of the East Roman empire. It also reflects a host of fears and prejudices fostered for long centuries in the Roman world, since the trauma of Crassus’ defeat by the Parthians at Carrhae. Not even extended periods of decline and internal disarray within the Parthian monarchy, during which it was repeatedly invaded by the Roman army, could dispel the myth of the uncompromising threat posed by Iran to the Roman order. The replacement of the Parthian Arsacid dynasty by a vigorous new one, based in Fars, namely the Sasanian dynasty, at a time when the Roman empire itself was facing one of its severest crises, only aggravated its inhabitants’ deeply rooted fear of Iran. Ancient writers in the Roman oikoumenē passed on this attitude to modern western scholars.

The so-called Byzantine iconoclast period is a ‘dark age’ whose obscurity is only randomly illuminated by the few remaining sources, and even these are difficult to interpret. Apart from in Italy, no archives have been preserved. The contemporary sources comprise two chronicles – that of Theophanes the Confessor, covering the period up to 813, and the Breviarium of Patriarch Nikephoros, which stops at 769 – and an account of Leo V’s reign whose author is known as the ‘Scriptor incertus’. The only near-contemporary chronicle for the reigns of Michael II and Theophilos is that of George the Monk, probably completed in 846 and reworked in 871–2, with Theophanes Continuatus being the most important of the later chroniclers to cover this period. Other sources include the Acts of the second council of Nicaea (787); these contain several extracts from the ruling of the iconoclast council of Hieria (754), which they set out to refute. Further sources include a legal code called the Ecloga (741); the Farmer’s law (or Nomos georgikos) – though this is not dated with precision; the Taktikon Uspensky (842–3); the correspondence of the monk Theodore the Stoudite, and of Bishop Ignatios of Nicaea (known as Ignatios the Deacon) from the first half of the ninth century; numerous saints’ Lives; and the polemical anti-iconoclast literature. We can add to these sources others of Arab, Syriac, Armenian and Greek origin from the caliphate, as well as several inscriptions and numerous seals.