Sicily between Byzantium and the Islamic World

The Hariri Ship, the first known picture of an Arab sailing vessel.

Map of the Arab–Byzantine naval conflict in the Mediterranean, 7th–11th centuries.

Map of southern Italy in the 10th century. Byzantine provinces (themes) in yellow, Lombard principalities in other colours.

Afterwards, the Saracens who had sailed from Rome came to Sicily, where they occupied the aforementioned city and slaughtered many of the population who had taken refuge in fortifications or in the mountains and, taking with them lots of booty or bronze, they returned to Alexandria.

Vita of Pope Adeodatus II

During the Byzantine centuries, Greek and Latin travelers to and from Sicily were examples and, indeed, agents of the complex web of connections between the Latin and Greek Christian worlds as they overlapped on Sicily. From the seventh century onward, Sicily also began to be drawn into the Islamicate world, as represented primarily by the political center of Qayrawān and the many seaports of Aghlabid Ifrīqiya (modern Tunisia) and Egypt. Long before Sicily became a Muslim province in the ninth century, in fact, considerable travel and communication were conducted between the island and the dār al-Islām, making the island increasingly important as a zone of interaction between Muslims and Christians, both Greek and Latin. Although, as in the sixth and seventh centuries, economic movements cannot be quantitatively reconstructed from the remaining data, by the eighth century, there is clear evidence of semiregular ship travel between the shores of Sicily and Aghlabid Ifrīqiya. While most of this traffic was of a military nature—with regular raids on Sicily’s southern shores starting in the seventh century of the common era—evidence also points to both diplomatic and, perhaps, even commercial transactions occurring between Sicily and Muslim North Africa while the island remained under the administrative control of Constantinople.

The introduction of Muslim powers into the western Mediterranean thus expanded the communication networks in which Sicily participated, in effect broadening the island’s place in the region rather than constricting or isolating it. New networks were opened while preexisting ones were maintained, even if altered. It is true that the relative amount of travel along each of the routes shifted and rebalanced over time, as Sicily conceptually drew closer to Muslim Africa and drifted farther from the Greek eastern Mediterranean. As the central Mediterranean Sea became populated with more and more Muslim-sailed ships, the waters around Sicily came to be linked more closely with northern Africa. At times we see ships from the Christian world encountering difficulties when sailing into hostile waters, but these voyages did not cease. The island, at the nexus of these three worlds, continued for some time to be a place of interaction and connection between Muslims and Christians, even if a preponderance of these interactions, as they appear in the sources, were hostile. Even violent interaction—and especially regularly recurring violent contact, such as that which took place during the nearly annual Muslim raids against Sicily—is a type of exchange that requires travel and the infrastructure of travel, and that connects peoples and spaces, drawing them closer together in terms of communications.

Even while communications with Muslim North Africa were increasing, Sicily remained in contact with the Greek East and with the Latin West. That is, the entry of Muslim polities into Sicilian affairs caused a relatively slow shift southward—rather than a break—of the communication networks of the island, concurrent with the persistence of many of the connections between Constantinople, Sicily, and Rome. The traditional periodization of Sicily’s history draws a firm line between the Greek Byzantine era and the Muslim period, with historians of Byzantium and the Middle East divvying up their examinations of the island. If, instead, we look across these centuries, at the transition period itself, our view of Sicily’s history and role within Mediterranean systems is very different. By placing the conquest of Sicily by Muslim forces in the middle of our examination rather than at the beginning or the end, we see that Muslim North Africa’s involvement with Sicily transformed the island’s communication networks rather than simply replacing one set of networks with another. Viewed across the period of the conquest, from the start of Muslim involvement in Sicily in the seventh century through the ninth–century conquest and into the tenth century, as Byzantine forces continued to try to retake Muslim Sicily—and by examining a variety of sources in Greek, Arabic, and Latin—political control did not necessarily determine the extent and range of the communications that defined Sicily’s regional affinities and its place within those local systems. Sicily was and remained broadly interconnected within the Mediterranean system, with Muslims and Latins as well as Greek Christians, even as the shapes and meanings of these connections shifted.

At the same time that military engagement was the most often recorded type of interaction between Sicily and Africa, the sources also allow glimpses of less martial communications between Greek Christians and Muslims. At times, those interactions took place because of or in the midst of battle, and at other times they could arise from diplomatic exchanges aimed at the stabilization of political and military tensions. Just as Byzantine Sicily was the site of diplomatic negotiations and the transfer of information between Greek and Latin Christian officials, so too did diplomats and envoys travel between Greek Sicily and Islamic North Africa, carrying both news and negotiations for peace. For example, the semiannual military incursions from Ifrīqiya were several times halted by truces that were officially concluded between embassies traveling between Syracuse and Qayrawān. Likewise, economic connections between the two may also have developed at this time. Because direct evidence for trade between Sicily and Ifrīqiya at this time is scarce, we can only assume the existence of economic connections that might be implied in the source record. Ships sailing back and forth within the Sicilian Strait between Ifrīqiyan ports and those of Sicily could have easily made the trip without meriting record in textual sources, and there are some suggestions that Sicily’s economic conditions were attracting the attention of Qayrawān. The Arabic chronicles, although written much later than the events they describe, detail the raids on Sicily carried out from Ifrīqiya and list all of the items gathered from the island, which suggests that the Aghlabid emīrate was taking an increasingly economic interest in the island of Sicily. Even if these lists of valuable items reflect a nostalgic image of a lost island of wealth, they demonstrate that the memory of Sicily’s conquest was tied closely to the perceived value of the products to be gained there. While the collection of war spoils was a regular part of this type of military strike, and a common way to reward soldiers for their service, it is the prolonged interest paid to the details of this booty by the later chroniclers that merits our attention. On the other hand, the products mentioned were exclusively high-value items—bejeweled icons and human slaves, for example—rather than more mundane trade items such as grain or textiles, which may also have proved attractive. At any rate, the Arabic chroniclers’ focus on these spoils indicates that they associated the conquest (and, therefore, also the loss) of Sicily with the annexation of an opulent and prosperous society.

The precise reasons that in the ninth century these regular raids for the collection of booty turned into an outright conquest of Sicily are not perfectly clear. The sustained interest that North African Muslims had taken in Sicily for many years suggests that the conquest was not simply the result of a sudden revival of jihād ideology or a desire to expand Islamic rule into Italy. Likewise, the conquest of Sicily should not be understood as part of the same process that brought North Africa and Iberia into the Islamic world, although those conquests do provide a background for this one. The conquest of Sicily was a major undertaking that occurred more than a century after the conclusion of the initial period of Muslim expansion into the Mediterranean region, and it arose from unique impulses relating to the nature of the Byzantine-Muslim frontier in the central Mediterranean. As with the later Norman Latin takeover of Sicily, outright military conquest followed many years of involvement in the island’s affairs. Sicily had been slowly entering the orbit of North Africa for several centuries prior to the ninth-century takeover. Then, as the boundary line between Byzantine and Muslim territory in the Mediterranean became more porous, the balance of power tipped far enough in Muslim favor that the outright military conquest of Sicily appeared to be advantageous for the Aghlabid administration of Ifrīqiya.

Indeed, it is the permeability of the Sicilian borderland itself that created the shift in relative power between Muslim and Christian authorities in the region. Much work has been done on the relationship between the Byzantines and the Muslims along the eastern frontier between Anatolia and Syria, and on the importance of that border zone for the health and wholeness of the Byzantine Empire. Far less has been written about the western frontier, partly because the Muslim-Greek battles that took place in Syria and the eastern Mediterranean were closer to the heartlands of both civilizations, and partly because that region produced the preponderance of sources about the Muslim-Christian conflict. But Sicily operated within the Byzantine Empire of the sixth through tenth centuries as an equally important frontier for Constantinople: one that both connected and separated the Greek world from the Latin Christian world and, as we will see here, one that did likewise with the Muslim world. Sicily was not simply a point on the dividing line between polities or religiopolitical civilizations; it also connected cultures in a zone of contact and conflict. The paradigm for discussing the relationships between Byzantines and Muslims has also tended to be that of conflict—both rhetorical and militarized. But some more recent work has also located shared traditions and a high degree of continuity between the Roman past and both the Christian and Islamic Middle Ages. Likewise, the three cultures that overlapped in the border region of Sicily and southern Italy indeed did so with violence and war, but also with shared reliance on the Roman tradition and through diplomacy, trade, and interpersonal interactions in the midst of warfare.

During the centuries of Byzantine control, Sicily was a region where fluidity of communications made it possible for Greeks, Muslims, and Latins to contest their control over a coveted locale while also maintaining the diplomatic and economic ties that were important to all of the parties involved. That is, this boundary zone between the Latin, Greek, and Muslim worlds was a disputed area, but one where various parties could meet, rather than a solid line of demarcation between Christians and Muslims. Sicily was often considered—by both Constantinople and local powers in Italy—an extension of Constantinople’s authority and, at the same time, was an important venue for managing relationships between local Muslim powers and the Greek Byzantine world. These multifaceted relationships along the Sicilian borderland will here be viewed by means of military, political, diplomatic, and economic communications between Byzantine Sicily and Muslim North Africa, along with the consequent population transfers that wrought demographic changes in the region, which would themselves also help shape future communication networks on and around the island.

‘Better the Turkish Turban than the Papal Tiara’

proverbial saying attributed to Loukas Notaras, grand admiral in the years 1444–53

Between the recovery of Constantinople from the Latins in 1261 and its fall to the Ottomans in 1453, Byzantine foreign policy was dominated by the question of the union of the churches. Political considerations required emperors to pursue this policy because they desperately needed military help from the West to combat the Turks, and the spiritual leaders of the West had made the reunion of the churches, with Constantinople subordinated to Rome, a precondition of any assistance. After the crusaders’ actions in 1204, many in Byzantium considered this abhorrent, if not heretical, and consistently refused to support it. The Palaiologan emperors therefore found themselves in a cleft stick: if the price of an alliance with effective western military forces was reunion, then they had to find an ecclesiastical policy of compromise and agreement. But any such policy would be condemned by those concerned with correct theology, and by the great majority in Byzantium who remained devoted to their own church, icons and ideas of orthodoxy. Most Byzantines wanted support not subordination.

As the Christian oikoumene had expanded in the early medieval West, contracted in the East under the pressure of Islam, and then reunited during the crusading period, specific features of liturgical practice emerged as major differences. For the Byzantines, any change in the wording of the creed was always considered incomprehensible and unacceptable. The primacy of St Peter, as interpreted by his successors – the bishops of Rome – jarred with the eastern concept of the pentarchy, the rule of the five patriarchs. And in the different forms of the Eucharistic bread (leavened or unleavened), all Christians could appreciate a very obvious visual divergence. Whether all clerics were obliged to maintain celibacy, and whether all Christians fasted on the same days, was perhaps less of a problem. Similarly, geography accounted for the use of Greek or Latin in the liturgy and certain unfamiliar habits, which had given the churches distinct histories within the world of Christendom.

Nonetheless, there was a fundamental desire to sustain Christian unity, especially in the face of Muslim beliefs. Bishops of Old and New Rome traditionally accorded each other great respect and ensured that prayers for the other were included in their services. Despite a breakdown in these relations in the ninth century under Patriarch Photios and Pope Nicholas, and again in 1054, mutual excommunication did not last beyond the lifetimes of the individuals involved. When Alexios I Komnenos appealed to Pope Urban II to support the Christians of the East against the infidel Turks, he did so precisely because they shared a common faith. Whatever the divergences in their practices, the First Crusade was duly preached on this basis and Christian control over Jerusalem was restored.

The events of 1202–4, however, deepened the sense of profound difference and left both parties hostile and wary. The orthodox were particularly outraged by the crusaders’ occupation of their churches and monasteries, not to mention the desecration of Hagia Sophia. From the new centres established after the Fourth Crusade, Greek prelates denounced the Latin bishops and friars appointed to ‘their’ sees and monasteries in the occupied capital and conquered territory. Yet in the empire of Nicaea, John III Vatatzes and Theodore II Laskaris had supported contacts between Latin and Greek representatives, finding the western friars less dogmatic than Cardinal Humbert. Serious discussions took place about rebuilding unity among the Christians. After 1261, Michael VIII Palaiologos determined to intensify these contacts.

Political developments, however, continued to impede the process. When the last Latin emperor Baldwin II fled from Constantinople, Pope Urban IV received him at Rome and promised to restore him to his throne, a policy actively supported by his successor, Clement IV (1265–8). At Viterbo in 1267, the pope gave his blessing to a formidable anti-Byzantine alliance led by Charles of Anjou and sealed by political marriages: Charles married his daughter to Baldwin’s son, and his son to the daughter of William Villehardouin, prince of Achaia. Fortunately for Byzantium, Clement IV died, and after a papal interregnum of three years, Gregory X was elected in 1272. The new pope’s overriding concern was to plan a new crusade against the Muslims, and to this end he announced a general council of the Church which would impose ecclesiastical reforms and reunite the western and eastern churches.

This promising declaration encouraged Michael VIII to try to win over the clerics in Byzantium who had expressed doubts and even denounced the idea of reunion: Patriarch Joseph, numerous bishops and monks who were opposed to ‘submission’ to Rome. Against the emperor’s wish ‘to spare the Greeks the terrible wars and effusion of blood threatening the empire’, they considered his proposal for rebuilding Christian unity unacceptable, because it conceded the primacy of St Peter over all churches and the Latin wording of the creed. They had additional concerns, but because the declaration of faith was held to be a critical method of teaching and preaching Christianity, any disagreement over the text was bound to cause splits. During twelfth-century debates between western and eastern theologians, the filioque regularly formed a stumbling block: both Peter Grossolano and Anselm of Havelberg wrote about this after their visits to Constantinople and Thessalonike. In response, Niketas ‘of Maroneia’, later Archbishop of Thessalonike, wrote six dialogues which defended the western interpretation, although perversely he refused to add the clause to the creed.

With full knowledge of this background of disagreement, Michael VIII began a campaign to win over the Byzantine opponents of union. In 1273, he imprisoned Patriarch Joseph and obliged John Bekkos, archivist of Hagia Sophia and later patriarch, to spearhead the campaign. But very few clerics were won over to the Latin position by John’s treatise on the subject, even though the emperor and his son and heir declared their personal adherence to the Roman definition of the faith. It even became difficult to find high-ranking clerics to represent the Church of Constantinople at the General Council which Pope Gregory X had summoned to meet in Lyons in 1274. The Byzantine delegation was led by George Akropolites, the head of the government, the former Patriarch Germanos III (who held the authority very briefly in 1266) and Archbishop Theophanes of Nicaea. It was much stronger on the civilian than the ecclesiastical side.

After a difficult journey, in which all their gifts of icons and church treasure intended for the pope were lost at sea, they arrived at Lyons, where the Council had been opened with tremendous fanfare and ceremonial on 7 May 1274. In their two weeks at the council, the filioque, papal primacy and a relatively new aspect of western theology – the existence of Purgatory – were debated. Since the 1230s, theologians on both sides had been discussing what could happen to sinners who did not have time to repent before death. Pope Innocent IV (died 1254) and Thomas Aquinas in 1263 had elaborated on the purging of minor sins in the fire mentioned in the Gospels. But the Orthodox Church had no notion of an alternative post-mortem existence, as the soul would ultimately be judged and sent either to heaven or to hell, so it was unwilling to accept the new western definition. As a result, the compromise wording adopted in 1274 did not refer to Purgatory, though it stressed the power of masses, prayers and pious almsgiving to assist the souls of the departed, which both sides accepted.

At Lyons, the three Byzantine delegates signed the profession of faith previously agreed with the emperor, George Akropolites swore an oath of loyalty to the pope and the Roman version of the creed, and the Council duly accepted the ‘submission’ of the Emperors Michael VIII and Andronikos. The reunion of the churches was celebrated on 6 July 1274 in the cathedral of Saint Jean, and Pope Gregory welcomed the Greeks back into the fold. The Council was interpreted by Rome as the submission of the entire Orthodox Church, rather than of its rulers; in the East, Michael VIII was legitimized and could demand Christian support against the infidel, but he could not persuade the orthodox to accept the terms of union. After 1274, the emperor begged the pope that their church

be permitted to recite the sacred creed as it had been before the schism and up to our time, and that we may remain in observance of the rites we had before the schism – these rites not being contrary to the faith declared above.

In their later professions of faith sent to Rome, however, both Michael VIII and his successor Andronikos II accepted the existence of Purgatory, citing ‘penalties of purgatory or purification’.

The union was duly celebrated in Constantinpole by John Bekkos, who became patriarch in place of Joseph I, but George Metochites, one of the Byzantine delegation, recorded serious opposition:

Instead of a conflict of words, instead of refutative proof, instead of arguments drawn from the Scriptures, what we envoys constantly hear is, ‘You have become a Frank’. Should we who are pro-unionists… be called supporters of a foreign nation and not Byzantine patriots?

From Constantinople, where memories of the sack of 1204 were still vivid, to Epiros, where the despot presented himself as a true representative of orthodox tradition, an anti-unionist party was created. Serbia and Bulgaria also supported this view, which conveniently combined their political antagonism to Byzantium with correct theology. Nor did the union produce the promised military results, partly because Gregory X died in 1276 and Charles of Anjou continued to campaign for the restoration of the Latin empire. Eight years after the Council, when Michael VIII died, his unpopular policy was immediately abandoned. Andronikos II (1282–1328) took vengeance on John Bekkos, who was deposed, brought to trial and imprisoned; three years later the new patriarch, Gregory II, repudiated the union.

From texts that circulated in the East, it is clear that opposition to the union was based on numerous differences between Latin and Greek Church practice. On the issue of what bread should be used in the Eucharist, raised bread or the western wafer called azymes, because it lacked yeast, zymos, the Byzantines believed:

Those who still partake of the azymes are under the shadow of the Law and eat of the table of the Jews, not of the reasonable and living table of God nor of the bread which is both supersubstantial and consubstantial to us men… For indeed the azymes plainly are lifeless, as the very nature of things even more plainly teaches.

Later on this anonymous tract asks:

Why do you priests not marry?… The Church does not forbid the priest to take a wife, but you do not marry. Instead you have concubines and your priest sends his servant to bring him his concubine and puts out the candle and keeps her for the whole night.

The same text criticized the Latins for not venerating icons, calling the Theotokos ‘Santa Maria’, i.e. simply a saint, using two fingers to cross themselves from the other side, eating ‘strangled meat’ and numerous other habits which seemed strange and wrong to the Greeks. These differences would all re-emerge in the 1440s as the population of Constantinople disavowed the Union of churches negotiated by John VIII.

Although the attempt to achieve the union of churches had failed in 1274, the hope that western Christian forces with papal blessing would eventually come to the aid of the Byzantines was kept alive by a growing interest in Latin theology and the first translations of Latin Fathers by Greek scholars. Knowledge of medieval Latin in Byzantium, as well as the vernacular tongues spoken by merchants, crusaders, diplomats and pilgrims, had expanded from the eleventh century onwards. When the scholar and monk Maximos Planoudes (c. 1255– c. 1305) began to translate classical Latin authors and St Augustine, he revolutionized Byzantine understanding of the West. His complete prose version of Ovid’s Metamorphoses, Heroides and some amatory verses; Boethius’ Consolation of Philosophy; Cicero’s Rhetoric; Macrobius and sections of Augustine’s City of God: all of these made some fundamental Latin texts available to a Byzantine audience for the first time. The brothers Demetrios and Prochoros Kydones and Manuel Kalekas extended this work, while Gregory Chioniades demonstrated the importance of Islamic astronomy through his translations from Persian into Greek.

This was a new development in Byzantine culture which reflects an awareness of the value of foreign, non-Greek learning. It marked a departure from assumptions of intellectual superiority in all fields and shows that Byzantium could adapt and learn from both sides in the arguments over church union. Most of those who translated from Latin into Greek had learnt the new language from friars in Byzantium. Like the ‘Apostles to the Slavs’, they used their linguistic skills to enhance Byzantine culture. Planoudes also served as ambassador on a diplomatic mission to Venice in 1296. He drew upon a very broad interest in ancient Greek culture. He made an edition of Diophantos’ theorems and other mathematical works, as well as copying and adding to the Anthologia Graeca, the late antique collection of epigrams. In contrast, two generations later, Demetrios and Prochoros Kydones were primarily concerned with theology and were directly involved in fourteenth-century church politics. The brothers were responsible for translating St Thomas Aquinas’ Summa contra gentiles and Summa theologiae into Greek, works which infused new vigour into the unionist camp. They also devoted attention to the works of Augustine, Boethius and St Anselm of Canterbury, and translated a Refutation of the Qur’an by Ricoldo de Monte Croce.

Despite growing interest in western philosophy, the style of Aristotelian logic adopted in the nascent medieval universities of Europe did not make a great impact in Byzantium. The educational system had its own traditions, based on the original texts of Aristotle and enriched by many later commentaries devoted to metaphysics, cosmology, ethics and logic, which had always been taught in the East. Another reason may lie in the growth of hesychasm and the teaching of enlightenment through spiritual contemplation, which owed more to Plato than Aristotle. The hesychast monks of Mount Athos proved to be implacable opponents of church union on the terms negotiated at Lyons. On the other side, those Byzantine intellectuals who favoured union were more impressed by western argumentation based on Latin translations of Aristotle – a tradition of logic that ignored the eastern commentaries.

John V Palaiologos, however, sought to realize plans for western military cooperation against the Turks by making a personal conversion to Catholicism. His travels to Hungary and Italy in 1366–9 culminated in his submission to Roman authority. While this remained his own decision and did not involve the Byzantine Church, he hoped it would secure military assistance. But the fact that on his way home the Venetians arrested him for debts revealed the precarious situation in Byzantium. His son Manuel was forced to ransom him, and before John V could return the island of Tenedos had to be handed over to Venice in lieu of money owed. Although the promised military intervention took shape under Serbian leadership, the Turks defeated this Christian force at the Marica in 1371 and the emperor abandoned his pro-western policy. Several leading Byzantine intellectuals nonetheless converted to Roman Catholicism and continued to urge the reunion of the churches as the only way of defeating the ever-tightening Ottoman encirclement of Constantinople. One of them, Demetrios Kydones, wrote a treatise proposing terms for winning Latin help in 1389, but it was ignored. Divisions within the elite thus contributed to weakening Byzantium while the Turks concentrated on expanding into Europe.

In 1422, the capital survived a major siege, but eight years later Thessalonike was captured, enabling the Turks to surround Constantinople from the West as well as the East. In these parlous circumstances, John VIII Palaiologos began another attempt to reunite the churches and thus win a serious commitment to western military aid supported by the papacy. In 1438, he led a high-level delegation – including Patriarch Joseph II, the two chief spokesmen Mark Eugenikos of Ephesos and Bessarion of Nicaea, sixteen metropolitans, officials and monks, making up a party of over seven hundred – to Ferrara to meet the papal party. Patriarch Joseph was so incensed at the demand that he, like all officials, kiss the pope’s foot that he refused to leave the ship until the issue was resolved. As a result, Pope Eugenius IV accorded him only a private reception rather than a grand public ceremony. The Council opened officially on 9 April after twenty days of debate over where the thrones for the leading figures should be placed. After many delays and inconclusive meetings, an outbreak of plague and shortage of money forced the parties in January 1439 to move to Florence, where the Medici family supported the Council.

While detailed records of the long debates that preceded the agreement were kept, the most interesting account of the Council was written by Sylvester Syropoulos, a patriarchal official. His memoirs record impressions of the unofficial discussions which accompanied the negotiations: how the Byzantine participants argued among themselves (for there were major disagreements between John VIII and Mark Eugenikos) and picked topics for discussion which would not reveal these rifts (such as the existence of Purgatory); how it became ever clearer that if the Greeks knew no Latin, they could not debate with the western theologians, who countered every eastern text with an argument of their own, often drawn from unfamiliar writings.

The filioque addition to the creed remained a major barrier, both as an extra clause in the wording of the creed as agreed at the First and Fourth Oecumenical Councils, and as a statement of orthodox theology. After many months of Latin pressure, agreement was reached on the grounds that all saints are inspired by the same Holy Spirit, whether they are western or eastern, and their faith must therefore be the same in substance even if it is expressed differently in Latin and Greek. Disagreement over papal primacy proved more fundamental, however. While the words of the creed might be accepted, the power claimed by Rome meant subjection, which the Church of Constantinople found much harder to bear. After centuries of elaboration and reinforcement through Rome’s judicial position in the West, popes had asserted superior authority over all churches based on their founder St Peter. They considered that patriarchs in Byzantium should submit to Rome before the union of churches could be celebrated. This not only implied inferiority, it also denied the tradition of the five leading sees meeting in Council as the highest authority in Christendom. While New Rome/Constantinople recognized Old Rome’s higher place of honour, the eastern theory of the pentarchy was hard to reconcile with Rome’s claim to overall primacy.

Under pressure from John VIII, the eastern clerics were persuaded nonetheless to agree a form of words which permitted the Union to be drafted. Remaining issues, like the use of leavened or unleavened bread, the marriage of lower ranks of orthodox clergy, and fasting and genuflecting habits were identified as local customs, which could be accepted. When the Act of Union was finally read in Latin and Greek in Florence on 6 July 1439, and acclaimed by all present, the churches were formally united in one faith. John VIII was commemorated in miniatures, bronzes and a medallion by Pisanello, which show him wearing the large peaked hat then fashionable. The process of negotiating the Union took nearly three years; the imperial party only returned to Constantinople in February 1440.

As a consequence, the princes of central Europe – Hunyadi of Transylvania, Vladislav I of Hungary and George Branković of Serbia – led a crusade into the Balkans which defeated the Turks in 1443/4. Murad II agreed to a ten-year truce, which might have been effective had not some of the western crusaders broken the terms at Varna. In November 1444, they attacked the city and were defeated. Constantinople was now abandoned to its fate; the ‘crusade of Varna’ was to prove the last. Although Hunyadi remained committed to the policy of assisting Byzantium, and Branković, who had not participated in the attack, remained a Christian ally, Constantinople’s essential weakness was symbolized when John VIII Palaiologos was forced to congratulate the sultan on his victory.

Only Mark Eugenikos of Ephesos and one other metropolitan had refused to sign the Union, and Eugenikos became the spokesman of resistance to it. Claiming that he had signed under duress, Syropoulos later joined the majority of Greeks who felt that both their beliefs and their traditions had been abandoned. In 1452, Pope Nicholas V sent Isidore of Kiev, who had converted and become a cardinal of the Catholic Church, to preach the Union in the beleaguered Byzantine capital. He arrived with a body of two hundred archers recruited at his own expense, which initially cheered the inhabitants. The Greek historian Doukas, reported, ‘Of the greater portion of the sacerdotal and monastic orders, abbots, archimandrites, nuns… not one among them assented to the Union. Even the emperor only pretended to do so.’ Nuns, he said, were particularly hostile and they implored Gennadios Scholarios of the Pantokrator monastery in Constantinople to support them. He finally wrote his tract in opposition to the Union and nailed it to his door: ‘Wretched Romans, how you have been deceived… Together with the city which will soon be destroyed, you have lost your piety.’ As these monks and nuns spread news of the resistance, the people called on the Mother of God to protect them against the Turks as she had done in the past against Chosroes and the Avars and the Arabs. They also implored her to ‘Keep far away from us the worship of the Azymites.’

On 12 December 1452, the Union was celebrated in desperation in Hagia Sophia, with the Turks encamped outside the walls of Constantinople. Although Isidore of Kiev reported to the pope that the liturgy was a triumph, Gennadios and other monks failed to participate, and the Union was not widely accepted in Byzantium. Nonetheless, Isidore himself fought on the walls in 1453, was wounded and taken prisoner. By disguising himself, he managed to escape to Crete and constantly mourned the loss of the city. Bessarion, the other major proponent of the Union, also continued to support efforts to regain Constantinople after the fall. As cardinals who served as papal legates, they were considered traitors by the orthodox. Both, however, encouraged humanist scholarship, wrote numerous works of theology and contributed to the growth of Greek libraries in the West. Bessarion’s legacy to Venice in 1468 ensured that his collection would remain intact as the core of the Marciana Library, while Isidore enriched the Vatican library with writings of his own and scholia in numerous manuscripts.

Among those opposed to the Union, Gennadios was also taken prisoner in 1453 but was discovered in the slave market, ransomed and installed as patriarch by Mehmed the Conqueror. His fierce loyalty to what was the original and true Christian theology reflects contemporary opinion voiced by Loukas Notaras, an adviser to the last three emperors: ‘Better the Turkish turban than the papal tiara.’ Byzantium could not accept the theory of papal primacy and the subordination of Constantinople to Rome. The Byzantines, however few, remained faithfully committed to what they understood to be orthodox. They preferred to maintain their own theology under Ottoman rule than to suffer union with the Church of Rome and western rule. This was surely an echo of the sacrilege of 1204.

Roman to Byzantine Army Transition Part I

In 330 ce, Constantine I, Emperor of the Romans, founded a new capital for his empire on the triangular peninsula of land that divided the Bosphorus from the Sea of Marmara, commanding the narrow water passage from the Black Sea to the Mediterranean. He named it Constantinople, and in time it grew to be not only one of the greatest cities of antiquity, but the center of one of the most impressive civilizations the world has ever seen: the Byzantine Empire.

Within 200 years, the Byzantines (or Eastern Roman Empire, as they styled themselves) had grown to massive proportions, controlling all of Italy, the Balkans, Greece, Asia Minor, Syria, Egypt, North Africa, and Southern Spain. Such an empire could be held together only by a strong and efficient military, and for several centuries the Byzantine army had no equal anywhere in the world.

Although the empire had expanded enormously through conquest, the basic role of the Byzantine army was defensive. Fortifying the long borders was out of the question, and since raiders and invaders could strike anywhere along the empire’s frontier, the army needed to be able to move quickly to meet these threats. Like their predecessors, the Roman legions, the Byzantine units formed a professional standing army which was trained to near-perfection as a fighting machine. Unlike the legions, however, the core of the army was cavalry and fast-moving foot archers. Speed and firepower had become the trademarks of the “new Romans.”

The stirrup reached the empire from China early in the fifth century, and increased the effectiveness of the cavalry enormously. Therefore, the core of the Byzantine army became the heavy cavalry. A typical heavy cavalryman was armed with a long lance, a short bow, a small axe, broadsword, a dagger, and a small shield. He wore a steel helmet, a plate mail corselet that reached from neck to thigh, leather gauntlets, and high boots. His horse’s head and breast might be protected with light armor as well. By the later empire, armor for both rider and horse became almost complete, especially in the frontline units. In a secondary role, unarmored light cavalry horse archers on smaller mounts supported the heavy units with missile fire, while other light cavalry armed with a long lance and large shield protected their flanks.

The infantryman who usually accompanied the cavalry in the field was either a lightly armored archer who used a powerful long bow, a small shield, and a light axe, or an unarmored skirmisher armed with javelins and shield. Because most Byzantine operations depended on speed, tactically as well as strategically, heavy infantry seldom ventured beyond the camps or fortifications. The heavy infantryman wore a long mail coat and steel helmet and carried a large, round shield. He used a long spear and a short sword. The Varangian Guard, the emperor’s personal bodyguards, were famous for their great two handed axes which they wielded with great effect. Their armor was almost complete plate and mail from head to foot.

To the Byzantines, war was a science, and brains were prized over daring or strength. Military manuals such as the Strategikon (ca. 580) and the Tactica (ca. 900) laid down the basics of military strategy that really did not vary for almost a thousand years. The army was always small in number (field armies almost never exceeded 20,000 men, and the total force of the empire probably was never greater than 100,000) and, because of its training and equipment, very expensive to maintain. Huge losses in combat could be catastrophic, and seldom were great winner take-all battles fought. The goal of any Byzantine general was to win with the least cost. If by delay, skirmishing, or with drawing the local population and their goods into forts he could wear out an invading force and cause it to withdraw without a costly pitched battle, so much the better. Bribing an enemy to go away was also quite acceptable.



Like all armies, ancient and modern, the Byzantines arranged their military apparatus hierarchically. The handbooks portray deep organizational structures, inherited from the Romans and persisting until the fall of the empire, with clearly delineated ranks to the level of five or four soldiers. The overall commander of the army was, of course, the emperor. In all cases emperors were expected to uphold the façade of military competence—even the most pacific possessed a smattering of training, could ride, wield weapons, and were literate in strategy and the structure of their forces. In many instances, the emperors were military men and possessed firsthand experience in the affairs of war. Since no head of state could manage security alone, even when he took the field himself, all relied heavily on practiced commanders.

Constantine appears to have made radical structural changes in military organization; he removed the prefects from command and made theirs an administrative post. He further removed the troops stationed in garrison, the frontier guards (limitanei or ripenses), from the emperor’s guard units (protectores) and the field army (comitatenses), which he expanded in size. Units were uprooted and pulled from their old third-century bases. The Master of Infantry (magister peditum) and Master of Cavalry (magister equitum) commanded those branches of individual field armies. We would equate the various magistri with marshals in more modern military parlance, with control over armies in a given theater. After Constantine the empire was once more divided between emperors in east and west and some mobile units transferred to the frontiers where they formed the core of campaign armies and an effective active defense supplemented by the limitanei. Such mobile regional field forces were under the command of a Master of Cavalry who commanded both the infantry and horse.

Prior to Constantine, Diocletian replaced the old Praetorian Guard—which had become infamous through fractiousness, rank insubordination, and regicide—with a new imperial bodyguard. Constantine further increased the new regiments, the Scholae (Latin: schools, group), which totaled twelve units, each with 500 men divided evenly between eastern and western halves of the empire. The magister officiorum (Master of Offices) led them. These units formed an elite guard for the emperor on campaign through the time of Theodosius I (379–95), but most units gradually declined to a civilian honor guard by the later fifth century. By the sixth century a count (comes domesticorum) commanded units of the scholae.

In the fifth century, the strategic disposal of forces and consequently the high command settled into the form it would resemble through the reign of Justinian. There were two imperial armies attached to the emperor’s person led by the magister militum praesentalis (Master of Soldiers of the Emperor’s Presence). These praesental armies comprised elite troops and mobile field forces that would form the core of any imperial expeditionary force. Five regional field armies (two praesental armies, Illyricum, Thrace, and the East) and their supporting frontier forces were under the command of the magister utriusque militiae (Master of Combined Forces [meaning of horse and foot]). His lieutenant, the vicarius, is known from the fifth century onward. There were frontier commands directed from the office of comes rei militaris (military counts) in Egypt and Isauria in mountainous and restive southern Asia Minor and thirteen dukes along the Danube, eastern frontier, and Libya. The magister commanded his field forces and also held authority over the armies under control of the comites and duces. The legatus (legate) or prefects held the reins of individual infantry legions. Infantry cohorts (regiments) of 500–600 still existed in the fourth century and their cavalry equivalent was formed of vexillations (vexillatio) or alae of up to 500 troopers. Tribune was the most common title for officers handling regiment-sized units, whether cavalry or infantry, but we also find the prefect in command of the cavalry vexilliations, alae, and among the limitanei. Another vicarius (hence our word “vicar”) was the lieutenant commander of the regiment whose duties and authority increased throughout this period. While much of the army underwent serious changes in organization and deployment, certain areas, such as Egypt, retained older structures and ranks.

Promotion within the ranks was a matter of service time or, not uncommonly, graft. St. Jerome (d. 420) provides a clear hierarchy of grades for enlisted men and noncommissioned officers in the early Byzantine period. He lists from lowest to highest grade: tiro; eques/pedes; circitor; biarchus; centenarius; ducenarius; senator; and primicerius.

A recruit was a tiro (pl. tirones) until he was trained, and such men did not draw full pay or rations. The anonymous author of a late fourth-century document, the De Rebus Bellicis (Military Affairs), recommended that cohorts maintain fifty or a hundred tirones so that losses could be quickly and cheaply replaced. Soldiers of the line were pedes (infantry) or eques (cavalryman). The semissalis seems to have been a senior ranker but below what we would consider noncommissioned officer status. At the base of the noncommissioned officer ladder of that time, the circitor at one time inspected sentries but little else is known of his authority or responsibilities. By the fourth century he may have been a junior biarchus, (mess-leader; sometimes called decanus or dekarch, “leader of ten,” even though he led eight soldiers, including himself) who commanded the contubernium, the squad or mess-group, which comprised eight to ten men who shared a tent and, as the name suggests, took meals together. By the fourth and fifth centuries the century numbered around eighty men, ten contubernia, commanded by the centurion with the rank of centenarius. The ducenarius, rather than commanding two centuries, was probably a higher-ranking centurion, since Vegetius stated that these men formerly commanded two hundred, an indicator that the title no longer reflected its old order. As historian Warren Treadgold argues, the senator likewise was probably a senior kind of noncommissioned officer with specialist duties, such as adjutor (clerk or scribal assistant), campidoctor (a centurion who drilled rankers and recruits), or actuarius (regimental quartermaster). Each regiment also had an optio (quartermaster), a surgeon, two heralds, two standard bearers, draconarii—named for the dragon-headed pennons known in the fourth century, a cape bearer, a trumpeter, and a drummer.

The five regional field armies possessed an extensive administration that handled correspondence, pay, logistics, and judicial matters. These large staffs, numbering up to three hundred, mirrored their civilian counterparts in the provinces. Military tribunals were more or less the same throughout the staffs of the magister militum, the dux, or the comes. The army judiciary was staffed by a princeps assisted by a commentariensis and an adiutor and a libellis; the latter dealt with judicial petitions. Deputy assistants (subadiuva) and shorthand writers (exceptores) handled the judicial clerking. Another bureau headed by a princeps with his assistant, the primiscrinius, two numerarii (principal accountants), and their support staff of scriniarii (clerks) dealt with financial and supply matters.

Scholars debate the tenor and role of the frontier forces (limitanei) who are sometimes characterized as “static” forces or even as “soldier-farmers” whose quality deteriorated in the fifth and sixth centuries. In a much-cited passage written no later than the year 550, Prokopios criticized Justinian for his elimination of their pay. While the loss of payment in coin may be true, frontier garrisons staffed by local troops continued to exist in some areas of the empire. An Egyptian known as Flavius Patermuthis (the name “Flavius” was taken upon entry into imperial service from the reign of Constantine to show one’s joining the imperial “family”) served as a soldier in Elephantine (modern Aswan, Egypt) from at least 585–613. Patermuthis and his comrades were prominent locals, indicating that in some places the limitanei had come to resemble local self-help forces rather than disciplined professionals. Elsewhere, the picture is somewhat different. Isaac argues that the limitanei were not soldier-farmers but simply the troops under the command of the duces of the provinces and as such they were mobilized for police duties and patrols, manned the frontier posts, and joined the field army on campaign. From papyri recovered in Nessana (modern Nitzana in southern Israel) we know of a numerus of dromedarii (camel riders) who patroled the desert routes around Gaza; these men appear as landowners and prominent members of the community until around 590, when the unit was either disbanded or transferred. Their duties were then probably assumed by allied Arab forces of the great confederation of Ghassan.

Federate soldiers (foederati) remained prominent in the Roman military structures of the fourth to seventh centuries. These troops served under a treaty (foedus) between the empire and tribes on the frontier. During the time of Diocletian and Constantine, federate troops served under their own commanders and were paid lump sums with which to provide for their soldiers’ needs. They also received annona: payment in kind of foodstuffs and fodder. By the sixth century, some tribal groups served under their own leaders in this fashion, such as the Ghassanid Arabs who guarded the eastern frontier from the Euphrates to the Red Sea. Others federates were enrolled in regular military units that appear to have been mixed Roman-barbarian contingents under the command of Roman officers. When not in the field these units were under the authority of the comes foederatorum, but for tactical purposes while on campaign they served under the magistri.

In 528, in light of new strategic realities in which the contest with Persia increasingly centered on Armenia and the Caucasus, Justinian divided the eastern command formerly under the magister militum per Orientem. He created a new command, the magister militum per Armeniam, headquartered at Theodosiopolis (modern Erzurum) whose army was drawn from both praesental units and the mobile forces of the old duces and comites of the frontier districts. Following their successful conquests, Africa, Italy, and Spain gained their own regional commands as well, which raised the number of army corps to nine, though there does not seem to have been a commensurate increase in troop numbers.

By the time of Maurice’s Strategikon in the late sixth or early seventh century, the army had changed considerably. The old guard units, the Scholae, Domestici, Protectores, and Candidati (originally a picked unit of the Scholae) became mostly civilianized but remained intact. The limitanei degraded and Justinian seems to have drawn down some of these frontier forces. The military returned to a purely decimal system of organization, with the main building blocks being the commands of ten and one hundred. A change in terminology reflects the decline of Latin in favor of Greek within the military, which was natural since the latter was the language spoken by most people in the eastern Mediterranean.

Book 1 of the Strategikon lays out the ideal officer structure of the Maurician army at the end of the sixth century. The general, now called by the Greek title strategos, held overall command of a given field army. A hypostrategos (lieutenant general) served as his second in command and led the meros (division) in the center of the battle line; this indicates that tactically the hypostrategos was important, since his forces anchored the army. The handbook also says that armies of medium strength were 5,000–12,000, thus representing groups of one to three meroi. A meros (Greek “part,”“portion”) was a division comprised of around 5,000 men, officered by a merarch. The division meros was built from multiple units called moira. The moira numbered 2,000–3,000 under the command of a duke, moirarch, or chiliarch. The units that replaced the cohorts of the older army were variously called tagma (not to be confused with the imperial mobile army which had taken on the name tagma or tagmata after the Greek for “order” or “ranks”), arithmos, or bandon. The tagma and its equivalents numbered 200–400 led by a count or tribune, with his second in command, the ilarch, a higher grade hekacontarch who commanded a hundred men. The hekacontarch then was the successor to the old legionary centurion. The lowest levels of command were the dekarch, pentarch, and tetrarch who commanded ten, five, and four men, respectively (including themselves).

The Strategikon provides the order of march for a 310–man cavalry tagma, probably a common strength (for a number of reasons, unit sizes were not uniform). The commanding officer (tribune or count) held under his command two hekacontarchs (or ilarchs), 27 dekarchs, 29 pentarchs, 31 tetrarchs, a standard bearer, a cape bearer, and a trumpeter, with 217 troopers. Treadgold hypothesizes that the tactical units mentioned in the text, ranging from 200–400, represent deployments from standard, 500 men regiments (tagma or bandon) whose remaining 100–300 men remained in quarters. This is a reasonable interpretation, given that unit sizes seem to have been based on decimal units grouped into thousand-man paper legions whose disposition varied according to the tactical situation.

The Strategikon names among mobile field meroi, the Optimates (“best men”), an elite cavalry regiment (bandon) unit of perhaps 1,000 men. In addition, elite cavalry units clearly owed their names to older Roman forces: the Vexillations, Illyriciani, and Federates, all mobile cavalry divisions that Treadgold estimates numbered around 5,000 each. Haldon sees there being only three elite cavalry units: the Optimates, Boukellarii, and Federates, all formed sometime after 575. These cavalry armies probably replaced the old praesental armies as the core of imperial campaign forces, since the author of the Strategikon envisions deployment of the three in the vanguard of an imperial campaign army.

The Persian War of Heraclius occupied more than a decade and drained the empire of men and resources. By the mid-620s the Romans had rebuilt their forces and attained victory, only to see them swept away by the armies of Islam. The Byzantines adapted to these exigencies by reconstituting their battered forces as best they could and billeting troops throughout the countryside of Asia Minor, the last large territory left in imperial possession. From the settlement of the military corps on the land evolved a new military and administrative apparatus called the theme system. Thema (theme) is a word of unknown origin, but may be derived from the army muster rolls or the tax rolls needed to support them. During the Persian campaigns of Heraclius the term simply meant headquarters of an army command. The earliest attested themes seem to date to the mid-or late seventh century. “Theme” as a territorial and army designator probably derived from the association in the minds of administrators with the cataloging of military men and corresponding territory and material needed to sustain them.

Roman to Byzantine Army Transition Part II


From the fourth through seventh centuries the Roman state ingested soldiers primarily in four ways: through native volunteers, through enforced hereditary service, by conscription, or by hire of foreign mercenaries. Native volunteers were the mainstay of the army and were generally sufficient to fill the requirements of the state. The hereditary obligation for sons to succeed their father in military service, introduced by Diocletian, was soon after abandoned for recruits to the comitatenses, but maintained among the limitanei. The era of Diocletian and Constantine witnessed annual conscription in the provinces in which state agents levied recruits based on regional resources as assessed in the minute reckoning imperial officials had made; villages and estates had either to furnish a set number of men based on their population and expected agricultural surplus or to buy out of their obligation. Slaves were not accepted. In the troubled years of the fifth century when the eastern army suffered from the aftermath of Adrianople and civil war, supplemental conscriptions fell upon elites who had to provide able-bodied men to serve or a cash payment of 30 solidi (the gold coin struck from 309 on at 72 to the pound)—a steep price, since a worker would have received around 12 solidi maximum annually. Unsurprisingly the draft was unpopular and seems to have been employed only in times of significant stress.

With the exception of the ranks of the limitanei, in which service was hereditary, the practice of conscription was generally abandoned. Justinian allowed slaves to join the army rather than resort to general forced levies, which were unpopular among elites and rustics alike.44 Limitanei did enroll in the regiments in which their fathers served until the end of their existence; there were incentives on both sides for the frontier guard to be maintained. For the state the provincial soldiery still served a useful role as garrisons and as logistics and police forces, even if those outside Syria and Mesopotamia rarely took part in campaigns. Soldiers still received payment, supplies, and certain tax and status privileges that somewhat offset the risks posed by service, which in places like Egypt was infrequent.

Although Justinian did eventually allow for slaves to be enrolled in the army (and these must have been provided as substitutions during episodic ad hoc conscriptions) volunteers usually staffed the mobile armies and imperial guards units. Justinian and his general Belisarios are good examples of this—both sought service as an escape from provincial obscurity. Volunteers continued to provide the manpower for the army through the reign of Phokas, though Maurice provided that sons of fallen soldiers would succeed their fathers in the comitatenses. This was a privilege rather than a burden that the soldiers welcomed—it assured their families salaries and support. When Heraclius found himself chronically short of manpower in the midst of the Persian War, he restored the old hereditary recruitment of all soldiers, something he managed to accomplish in a time of crisis.

Native recruits generally came from the rural, rough-and-ready regions of the empire. Illyricum (the modern eastern Adriatic coasts and mountains) provided an ample pool of military manpower. Countless troops and officers came from this and other regions south of the Danube from Diocletian’s time through the sixth century. Isauria, in the mountain lands of southeastern Anatolia, furnished large numbers of military men from the fifth century onwards, when the emperors were especially active in recruiting them to offset Germanic influence in the army. The rugged upland areas of Paphlagonia, Cappadocia, and Pontos also produced surplus men with martial prowess who helped to fill the legions.

Foreign recruits formed a major component of the army. Armenians provided excellent quality cavalrymen and infantry to both Rome and Sasanian Persia. Armenians dominated the imperial scholae after the fifth century. Hunnic horse archers provided a major tactical advantage for Byzantine armies of the sixth century—they were recruited in groups following a native leader and placed under Roman command. Iranian nomadic elements, such as Massagetae, also called “Huns,” and Alans in the sources formed another source of mercenary manpower. They fought as both cavalry and infantry. Three hundred “Hun” or Massagetae horse from Belisarios’s boukellarios proved decisive in the opening engagements of the battle of Ad Decimum (September 15, 533) when under the command of the Armenian adjutant John, they slaughtered the 2,000-man Vandal lancer vanguard and killed the king’s brother, Ammatas.46 Captured Sasanian Persian soldiers were brigaded into units that served among Byzantine forces, and some Persians or Armenian-Persians rose to high positions in the military command.

Germanic-speaking peoples also provided excellent warriors for the Roman army up through the sixth century. Among these groups we find the east-Germanic Goths, who dominated the ranks of the eastern field army after Adrianople and were still found in Roman service in the sixth century. The east-Germanic Heruls feature prominently in Prokopios’s description of Belisarios’s campaigns; they are often seen undertaking special missions and were brave to the point of reckless. Their east-Germanic neighbors, the Gepids, formed another tribal confederation that emerged from the shadow of Attila’s Hunnic Empire in the fifth century and also provided troops until their defeat and destruction by the Lombards. The west-Germanic Lombards provided significant manpower in Italy—5,500 of them served the Romans during the 551–54 campaigns of Narses.

The loss of most of the Balkans in the seventh century to Slavs and Avars deprived the Romans of some of their finest soldiery. This recruiting ground was replaced mainly with Anatolian Greek-speakers from the rugged interior. Armenians became especially important; at the beginning of the seventh century, the emperor attempted to transfer 30,000 Armenian troops with their families to Thrace. The army that Heraclius reformed in 621–22 was largely from native Roman troops—since the emperor was in the midst of an empire-wide collection of loaned church plate to melt down to coin money, there was little cash to pay foreigners. It was at this moment when Haldon proposes that the emperor made military service once more hereditary, as it certainly was by the end of the century.


During the era of the Tetrarchy, soldiers’ pay was rendered largely in-kind. This was a result of the rampant inflation that plagued the empire in the third century. Since the time of Septimius Severus (193–211) the empire had levied a tax in-kind to support the troops, the anonna militaris and accompanying capitus to supply animal fodder. The state issued clothing, arms, and horses to soldiers. Pay was measured in annona, rations paid annually to rankers. Prior to Anastasios (491–518) each annona was reckoned at 4 solidi. Officers received multiple annona; the primicerius of the fourth-fifth century legions typically earned five annonae. During the reign of Diocletian annual pay in coin continued but was modest to say the least—perhaps 7,500 denarii a year plus donatives and special payments made on accession dates of the emperor and other imperial holidays. Fourth-century pay has been calculated as equivalent to about 12 solidi plus arms and equipment, but by the mid-fifth century had fallen to the equivalent of 9 solidi. To provide some frame of reference, a stone cutter in contemporary Egypt might earn something less than 12 solidi per year. Upon their accession and in anniversaries of their reign, emperors paid substantial bonuses called donatives; Julian paid 5 solidi and a pound of silver, a standard sum offered through the sixth century. Donatives paid every five years from the emperor’s accession were about five solidi for soldiers of the line. But over time, by reckoning arms issuances and equipment in annona, the state deeply cut soldiers’ pay while theoretically maintaining their ability to fight. One wonders how such issues worked, since a soldier could have hardly worn out a spear or sword in a normal year; possibly these allowances were convertible to food or fodder.

In the fifth century the cumbersome and easily abused in-kind system was replaced by payments in coin; the stability brought by the fourth-century creation of the gold solidus and economic recovery of the empire permitted a remonetization of military pay. Anastasios seems to have spread the five-year donatives out as annual payments and offered cash instead of supplying arms and equipment; prior to his reign soldiers in the field army received something like 9 solidi plus equipment. Under Anastasios field troops earned 20 solidi annually, an increase of as much as two-thirds; the raise was probably a response to a lack of recruits and the general poor condition of the soldiery. By the beginning of the reign of Justinian, soldiers in the comitatenses were then well paid when compared with the average worker.

Limitanei received far less, perhaps 5 solidi and an equipment allowance. Justinian’s pay scale for the African limitanei survives. The dux earned 1,582 solidi, the cavalry primicerius 33 solidi, infantry centurions 20 solidi, and their cavalry counterparts 16.5 solidi while infantry rankers earned 5 solidi and cavalry 9. It is probable that even this modest wage was eventually cut by Justinian and that the state paid frontier troops only annona payments in-kind in equipment and capitus issuance for their mounts. Allied units on the frontier, like the Ghassanid confederacy, received annona in cash and kind. But like their comrades in the mobile armies, limitanei received tax exemptions for certain family members and were exempt from corvée labor, among other burdens.

In response to the fiscal and military crisis sparked by the Persian War, in 616 Heraclius seems to have ended the cash allowances for uniforms and equipment, which amounted to reducing pay by one-half. The state returned to issuing clothing and equipment to the soldiery. Constans II (641–68) apparently cut this salary in half again and probably replaced the lost salary with grants in land from which soldiers could support themselves. Annual base pay for the rank and file was thus around 5 solidi during the Dark Ages. To put into perspective this abysmal remuneration, we should note that a carpenter in eighth-century Egypt might earn 16 solidi per year.

By the tenth century, the situation had improved and cash payments in gold had expanded. Officers in the tagma were well paid by contemporary standards. In the mid-ninth century average pay had doubled to about 10 nomismata (singular nomisma, the Greek term for solidus). A tagmatic commander earned 144 nomismata, a topoteretes 72 nomismata, a pentekontarchos 24 nomismata, and a ranker in the tagma 9 nomismata.

The health of state finances and the fineness of the nomisma declined sharply in the middle of the tenth century. Alexios I replaced the nomisma with the hyperpon (pl. hyperpyra) a coin inferior in fineness to the solidus/nomisma of the past. As most soldiers were by this time native and foreign mercenary professionals, they earned cash payments and donatives. The limited data suggest that soldiers in service in the late Byzantine period were well paid. In 1272 a soldier in Asia Minor earned 24–36 hyperpyra, well above the salaries of common workers, such as cooks or domestic servants (10 hyperpyra each) or doctors (16 hyperpyra). Even though the currency was further inflated by the fourteenth century, the 288 hyperpyra paid to a Catalan mercenary cavalryman even though he had to equip himself, was exorbitant.

Many of the soldiers of the Palaiologan allagia served on the basis of pronoia grants. The origin of these grants is obscure but, like the settlement of troops in the themes centuries earlier, they served to shift the burden of maintaining troops from the central government to the provinces. Pronoia grants included tax revenues or rents from dependent peasants—a system often likened to the “feudal” customs that supported the landed aristocracy of the West. Unlike the medieval western arrangements, however, the pronoia were at first held for the lifetime of the grantee; they became hereditary under Michael VIII. In contrast to the medieval west, the state remained the owner of the land and in control of the fiscal mechanisms by which the pronoia were administered.

Over the centuries the Byzantines showed a continuous tradition of army organization that evolved from the Roman imperial system but was adapted to the strategic and tactical realities with which the empire was confronted. Until the twelfth century, the organizational structure of the army was relatively conservative—were he to view the army of the eleventh century, the emperor Maurice from some five centuries prior would have recognized many units and their officer structure. There was, however, adaptation and reorganization in response to the defeats at the hands of the Arabs, but the seventh-century wars did not expose the system as completely broken and thus most structures continued, albeit in modified form. There was a generally deep command structure present in the organization, with officers down to the level of four or five soldiers which undoubtedly preserved discipline and offered considerable tactical flexibility.

On the whole the state managed the well-being of the soldiers reasonably well—service was often dreary, unpleasant, and dangerous. Only during times of severe crisis, such as the inflationary era arrested by Diocletian and Constantine, and the seventh-century military collapse faced by Heraclius, did the empire economize at the expense of its troops. Even during the worst of the crisis, cash payments were never halted, though they were apparently sometimes paid in copper and in arrears. Since the military was by far the largest governmental expense, it was frequently the only place that such economies could be enacted. However, once the crisis of the Dark Ages ended, pay rates climbed to an average well above those of most workers.

The Recovery of Constantinople [1205–61]

In contrast to Doge Dandolo, who now proudly styled himself ‘Lord of a Quarter and Haifa Quarter of the Roman Empire’, the Emperor Baldwin cut a sorry figure. He was left with just a quarter of the territory that had been ruled by his immediate predecessors; and even this was contested. Boniface of Montferrat, furious at having been passed over, refused the Anatolian lands he was offered and seized Thessalonica, where he established a kingdom extending over a large part of Macedonia and Thessaly.

The new rulers were, not surprisingly, detested. The Franks, staunch upholders of the Church of Rome, unhesitatingly imposed the Latin rite wherever they could. Many Greeks left their ancestral lands and moved to the Byzantine successor states in which the national spirit and the Orthodox faith were still preserved. Of these states, the largest and by far the most important was the so-called Empire of Nicaea, where Alexius III’s son-in-law Theodore Lascaris was crowned in 1208. It occupied a broad strip of land in western Anatolia, extending from the Aegean to the Black Sea. Although the official capital remained Nicaea – the seat of the Patriarch, where the imperial coronations took place – Theodore’s successor John III was to establish his residence in the Lydian city of Nymphaeum; and for most of the fifty-seven-year period of exile from Constantinople this was the seat of government. The two other successor states, situated one on the Adriatic coast and the other at the south-eastern extremity of the Black Sea, were too remote to exert much influence. The Despotate of Epirus was founded soon after the capture of Constantinople by a certain Michael Comnenus Ducas, great-grandson of Alexius I Comnenus. From his capital at Arta he controlled the north-west coast of Greece and part of Thessaly – a domain substantially increased in 1224 by his half-brother Theodore who captured Thessalonica from the Latins and was crowned Emperor as a rival to John III in Nicaea. Unlike Nicaea and Epirus, the Empire of Trebizond was not the result of the fall of Constantinople. It had been founded in April 1204 by Alexius and David Comnenus, grandsons of the Emperor Andronicus through his son Manuel, who had married a Georgian princess. After the fall of Andronicus in 1185 the young brothers had been brought up at the Georgian royal court. Determined to continue the Comnenus dynasty, they had captured Trebizond in April 1204. For the greater part of its 257-year history the Trapezuntine Empire was confined to a coastal strip between the Pontic mountains and the sea.

As ruler of Byzantium in exile, Theodore I Lascaris of Nicaea faced a multitude of problems. Even within his own borders, petty Greek principalities were declaring themselves; then, in autumn 1204, a Frankish army led by Baldwin himself began to move against him. Theodore was still hopelessly unprepared; and on 6 December a calamitous defeat at Poimanenon gave the Franks control of the whole Bithynian coast as far as Brusa. But Baldwin’s arrogance soon caught up with him. The Greek landowners in Thrace offered the imperial crown to the Bulgarian Tsar Kalojan in return for driving the Latins from Constantinople. Earlier in 1204 Kalojan had been crowned King (though not Emperor) by Innocent III’s envoy and had accepted the jurisdiction of Rome; but he was as anxious to get rid of the Crusaders as were the Byzantines themselves. On 14 April 1205 he destroyed the Frankish army outside Adrianople, taking prisoner Baldwin himself who died soon afterwards. Just a year after the capture of Constantinople, the power of the Latins was broken. In all Asia Minor, only Pegae on the Marmara remained in Frankish hands.

Now at last Theodore could forge his new state, never for a moment doubting that his subjects would be back, sooner or later, in their rightful capital. He followed the old Byzantine pattern in every detail; thus, after his coronation in 1208, there were two Eastern Emperors and two Patriarchs, the Latin in Constantinople and the Greek in Nicaea, each initially determined to destroy the other. In the following year Baldwin’s brother and successor Henry of Hainault, swallowing his Crusader scruples, concluded a military alliance with the Seljuks, who also saw the new Greek state in Asia Minor as a threat; and in 1211 he inflicted a serious defeat on Theodore, pressing on to Pergamum and Nymphaeum; but he was too hard pressed by the Bulgars in his rear to be able to pursue his advantage. In late 1214 the two Emperors agreed to a treaty of peace: Henry would keep the north-west coast of Asia Minor as far south as Atramyttion; the remainder as far as the Seljuk frontier would go to Theodore.

The young Empire had finally obtained formal recognition by the Crusaders of its right to exist. Almost simultaneously, the Latin Empire began once again to decline; in June 1216 Henry died at Thessalonica. In barely a decade, by respecting the rights and religion of his Greek subjects and achieving a balance of power with Nicaea, he had saved an apparently lost cause. He died childless; and to succeed him the Frankish barons elected Peter of Courtenay, husband of his sister Yolanda. Peter, who was then in France, set out for the East in the first weeks of 1217. Unfortunately he stopped at Durazzo to recover the city from the Despot of Epirus, but his attempt ended in fiasco. He was captured, thrown into prison and never heard of again.

The Empress Yolanda, who had wisely decided to travel out with her children by sea, meanwhile arrived without mishap in Constantinople, where she gave birth to a son, Baldwin. She then governed as Regent until her death in 1219, confirming her brother’s conciliatory policy by giving her daughter Mary to Theodore Lascaris as his third wife. News of this step, however, was received with horror in Epirus, where the star of the Despot Theodore was rising fast. He had never accepted the treaty of 1214; here, he claimed, was a further betrayal. The truth was that Theodore could never be satisfied with Epirus. As the legitimate great-grandson of Alexius I, he could boast a far stronger claim to the imperial throne than Lascaris. His immediate ambitions were now focused on Thessalonica; but Thessalonica was, in the eyes of Theodore Angelus Ducas Comnenus, little more than a stepping-stone to Constantinople itself.

Since the death of Boniface of Montferrat in 1207, Thessalonica had been governed by his widow, acting as Regent for her son Demetrius; but after the arrival of the Empress Yolanda it could no longer rely on firm support from Constantinople. It was already plain that its days as an independent state were numbered; and in the autumn of 1224 it fell. Theodore of Epirus now ruled supreme from the Adriatic to the Aegean. Soon afterwards, in open defiance of Lascaris, he was crowned by the Bishop of Ochrid as Emperor of the Romans. Thus it was that, in place of the single Empire that had existed little more than a generation before, there were now four three Greek and one Latin. And not far away there loomed a fifth: for the Second Bulgarian Empire was growing rapidly. Tsar Kalojan had already extended his rule over much of Thrace and Macedonia; his second successor, John II Asen, also coveted Constantinople. By far the weakest of the powers was the Latin Empire itself, by 1225 reduced to the capital itself, the region immediately to the north and west, and a small area of Asia Minor south of the Marmara. Yolanda had died in 1219; her son Robert was a feckless youth, totally outclassed by Theodore, John Asen and John Vatatzes, who had inherited the Empire of Nicaea from his father-in-law Theodore Lascaris in 1222. After a punishing defeat by Vatatzes, the capture of Thessalonica was too much for him. From that moment on he gave himself up to a life of pleasure and dissipation, dying in January 1228.

Robert left no legitimate children; and since his brother and successor Baldwin II was still only eleven the barons therefore turned for a Regent to the most distinguished of living Crusaders: the former King of Jerusalem, John of Brienne. Though now nearly eighty years old, he was still remarkably spry – he had a daughter of four – and no one else could match his record. He made, however, a number of conditions. The young Emperor must immediately marry Maria, his own four-year-old daughter, who must receive a suitable territorial dowry; he himself must be recognized as basileus in his own right, with Baldwin succeeding him on his death; and at the age of twenty Baldwin, if not yet Emperor, should be invested with the Empire of Nicaea, together with all Frankish possessions in Asia Minor. He was still in no hurry: only in the autumn of 1231 did he finally appear off the Golden Horn. A few days later he was crowned Emperor in St Sophia.

During this three-year interregnum, the balance of power in the Balkans suffered a radical change. In April 1230 the Emperor Theodore Comnenus had been defeated and captured by John Asen. To be sure, his brother Manuel was allowed to stay on in Thessalonica with the title of Despot; but this was only because he was married to Asen’s daughter. He was a puppet of his father-in-law and made little pretence of being anything else. The Latins had been saved from almost certain destruction – and by a nation that they had previously spurned. But they now had to watch John Asen advance unopposed across the Balkans, from the Adriatic to the Black Sea.

The effective elimination of the fourth participant in the struggle for supremacy led inevitably to a radical realignment among the other three. To John Asen, Vatatzes now seemed a far more useful ally than the Latins, particularly since he was about to abandon the Church of Rome. Western Christianity had never really taken root among the Bulgars; besides, any future offensive against the Latin Empire would be a lot easier to justify if the Tsar were seen to be acting against heretics. In 1232 the break was made. A Bulgarian Orthodox Patriarchate was once again established; and three years later John Asen signed a treaty of alliance with Nicaea, which was subsequently sealed by the marriage of his daughter Helena to the son of John Vatatzes, Theodore II Lascaris. In the late summer of 1235 the combined forces of Orthodoxy were besieging Constantinople by land and sea.

Old John of Brienne fought like a tiger for the defence of his Empire, and Venetian ships provided invaluable support; but Constantinople was saved only by a change of heart on the part of John Asen, who suddenly realized that an energetic Greek Empire would constitute a far more serious threat to Bulgaria than an exhausted Latin one and called off the attack. Almost at once, however, disaster struck. His own capital fell victim to a furious epidemic, which carried off his wife, one of his sons and the recently-installed Patriarch. To John Asen, this was the judgement of heaven; immediately he made his peace with Vatatzes. Soon, however, he began to look for a new wife; and somehow his prisoner Theodore of Thessalonica managed to persuade him to marry his daughter Irene. As the Tsar’s father-in-law, Theodore was then released from his captivity and returned to Thessalonica, where he deposed his brother Manuel and enthroned instead his own son John, restoring to him the title of Emperor.

The year 1241 proved a watershed. Before it was over, three of the protagonists were in their graves: John Asen of Bulgaria, Manuel of Thessalonica and Pope Gregory IX, one of the most redoubtable champions of the Latin Empire. That same year also saw a Mongol horde sweep through Hungary into the Danube basin, leaving the Bulgars little opportunity to undertake further adventures to the East: another once formidable nation was thus effectively eliminated. The power of Thessalonica had already been broken. The Latin Empire, which now consisted of little more than the city of Constantinople itself had survived only thanks to dissension among its enemies. Of those enemies, there remained but one: the Empire of Nicaea, whose ruler John Vatatzes continued to prepare for its reconquest. He still had the problem of Thessalonica to settle. Its Emperor John was a weak and pious figurehead; the real power was back in the hands of Theodore, as ambitious as he had ever been. Thus it was Theodore whom in 1241 John Vatatzes invited to Nicaea as his guest. The old man accepted, and was received with every courtesy; only when he came to leave was it politely explained to him that his departure would unfortunately not be possible. He remained a prisoner until the following summer, when Vatatzes escorted him back to Thessalonica and then sent him as an envoy to his son to negotiate a treaty. The result was that John exchanged the title of Emperor for that of Despot, and acknowledged the supremacy of Nicaea.

While Vatatzes was still in Thessalonica, the Mongols invaded Asia Minor. In June 1243 they defeated the Sultan Kaikosru II at the battle of Kösedağ the Emperor of Trebizond, who had been a vassal of the Sultan, suffered much the same fate. Fortunately the Mongols moved away again, leaving a broken Sultanate behind them but the Nicaean lands untouched. The Bulgar Empire too had been crippled by this most recent of the barbarian invasions; while the death in 1246 of Coloman, John Asen’s twelve-year-old son, and the succession of his still younger half-brother Michael, further troubled the waters in which Vatatzes cheerfully intended to fish. By the autumn of that year he had occupied a good deal of western Macedonia. He was still encamped there when a group of Thessalonians arrived with a proposal. If he would guarantee to the city the continuation of its ancient rights and privileges, it would be surrendered without a struggle. Vatatzes agreed at once. In December he entered Thessalonica unopposed, exiled old Theodore and left as his Viceroy his distant kinsman Andronicus Palaeologus.

One more enemy was left for him to conquer before he could concentrate on Constantinople. Some nine years before, Epirus had separated from Thessalonica and set itself up once again under Michael II, an illegitimate son of its original founder Michael I. John Vatatzes did not attack it: instead, in 1249 he concluded a treaty of friendship with Michael, sealing it by betrothing his granddaughter Maria to Michael’s son Nicephorus. Theodore, still unreconciled, persuaded his nephew to take up arms once again against the Nicaean Empire; but John Vatatzes was taking no more chances. Early in 1253 he forced the Despot’s surrender. Michael ceded much of his territory; his son Nicephorus was carried off to Vatatzes’s court as a hostage for his future good behaviour. As for the old, blind, insufferable Theodore, he was shipped off to end his days in the prison he so richly deserved.

The Latin Empire was tottering. Already in 1236 the young Emperor Baldwin, now nineteen, had left for Italy in a desperate attempt to raise men and money; it was not until early 1240 that he returned to the Bosphorus. This chronic shortage was also responsible for another decision, deeply demoralizing to Greeks and Latins alike: the pawning to Venice of Constantinople’s most hallowed possession, the Crown of Thorns that Christ had worn on the Cross. The Emperor being unable to redeem it, the opportunity was seized by St Louis of France, who built the Sainte-Chapelle to receive it.

For Baldwin, even cap in hand, the courts of Europe must have been vastly preferable to life in gloomy, beleaguered Constantinople. In 1244 he was off again – to Frederick II; to Count Raymond in Toulouse; to Innocent IV in Lyon; to St Louis in Paris; and even to London, where King Henry III made a small and grudging contribution to his funds. But he returned in October 1248 to find himself in such straits that he was obliged even to sell off the lead from the roof of the imperial palace. He would never have reigned for another thirteen years if his enemy in Nicaea had survived; but on 3 November 1254 John Vatatzes died at Nymphaeum. During the last ten years of his life his worsening epilepsy had seriously unbalanced him: it was clear to everyone at court that he was rapidly losing his grip.

He had, nevertheless, been a great ruler. He had inherited from his predecessor a small but viable state; when, thirty-two years later, he left it to his son Theodore II, its dominions extended over most of the Balkan peninsula and much of the Aegean, its rivals were crippled or annihilated, and it stood poised to achieve the purpose for which it had been established. At home, John continually reminded his subjects that they lived in a state of emergency, and that sacrifices were required of them until Constantinople should be theirs. Foreign imports were forbidden; self-sufficiency was now the watchword, and he himself set an example by running a profitable farm, using the profits from his sales of eggs to buy his wife Irene her ‘egg crown’ – a jewelled coronet, which he publicly presented to her as proof of what could be achieved by efficient husbandry. The gift was well-deserved. Thanks to the two of them hospitals and orphanages were established, art and literature encouraged, and the foundations laid for the spectacular cultural revival which was to occur in the reign of their son Theodore, under whom Nicaea would become a dazzling centre for Byzantine culture. In consequence John and Irene were genuinely loved by their subjects.

John knew as he lay on his deathbed that the day towards which he had worked all his adult life could not be long delayed, despite some doubts that he may have entertained about his only son and successor. Not that the young Theodore II Lascaris was altogether unworthy of the throne. He was an intellectual who produced in the course of his short life a whole corpus of literary, theological and scientific works; and he never allowed these interests to deflect him from the business of government. Unfortunately he had inherited his father’s epilepsy in an even more serious form. This was dangerous enough in Constantinople; when he was with his army in the field it was potentially disastrous. He ruled, nevertheless, with a strong hand. Instinctively distrustful of the aristocracy, he relied on a small group of bureaucrats, chief among them being his protovestiarius George Muzalon and George’s two brothers, Theodore and Andronicus; and he enraged the clergy by appointing as Patriarch a bigoted ascetic named Arsenius, annihilating at a stroke his father’s old dream of union with Rome.

Theodore signed a peace treaty with Bulgaria in 1256, and relations were further improved when the Tsar Michael Asen was murdered shortly afterwards and succeeded by a boyar named Constantine Tich, who married Theodore’s daughter Irene. Another dynastic marriage was that of John’s daughter Maria to Nicephorus, son of the Despot Michael II of Epirus. This unfortunately proved counter-productive, Theodore having unwisely made a last-minute demand for Durazzo and the Macedonian city of Servia as a condition of the marriage. The bridegroom’s mother, who had accompanied her son to the imperial camp on the Maritsa, was intimidated into agreement; but when she returned to tell her husband that she had given away two of his most important cities, he immediately launched a furious campaign against Thessalonica, encouraging the Serbs and the Albanians to support him. Within days, Macedonia was up in arms.

The man best qualified to handle the situation was a young general named Michael Palaeologus; the Emperor, however, had always been jealous of this handsome young aristocrat, who seemed to possess all the gifts he himself lacked. He also mistrusted him. Earlier that year he had accused him – quite unjustifiably – of high treason, threatening him to the point where the young general had been obliged to take refuge with the Seljuks. Michael had since sworn fidelity to the Emperor; nevertheless Theodore had decided only hesitantly to entrust him with the new command. Fearing, presumably, that his general might turn against him, he also gave him too small an army to be of any real use. Michael and his men fought bravely, penetrating as far as Durazzo; but they were unable to stem the tide. By summer the Despot was at the gates of Thessalonica, and Michael Palaeologus, disgraced and shortly afterwards excommunicated, was languishing in a Nicaean prison. This shameful treatment of the Empire’s outstanding general confirmed the people of Nicaea in their conviction that their basileus was no longer capable of responsible government; and there would surely have been a military revolt had not Theodore suddenly and conveniently succumbed to his disease in August 1258, aged thirty-six. His eldest son John being a child, he had appointed the hated George Muzalon as Regent. On his deathbed he had forced the leading members of the aristocracy to swear allegiance to John and George together, but in the course of a memorial service held nine days later they murdered Muzalon at the high altar and hacked the body to pieces. A palace revolution ensued, the result of which was to nominate the hastily-liberated Michael Palaeologus – who had probably instigated the plot – in his stead.

Michael, now thirty-four, was in many respects the obvious choice. He could claim kinship with the houses of Ducas, Angelus and Comnenus, while his wife Theodora was a great-niece of John Vatatzes. His complicity in Muzalon’s murder should have been seen as a stain on his character; but the protovestiarius had been so hated that it was overlooked. He remained immensely popular with the army and was well thought of by the clergy. He was awarded the title of Grand Duke (megas dux) and soon afterwards that of Despot. Finally in November 1258 he was raised on a shield and proclaimed co-Emperor, his coronation taking place at Nicaea on Christmas Day. He and Theodora were crowned first, with imperial diadems heavy with jewels; only afterwards was a narrow string of pearls laid upon the head of his young colleague, John IV.

Few of the congregation doubted that it was Michael VIII Palaeologus who would lead his subjects back into their capital. First, however, there was one more enemy to be faced. Early in 1258 Manfred of Sicily, bastard son of Frederick II, had invaded Epirus and occupied Corfu. The Despot Michael had joined with him against Nicaea, offering him the hand of his eldest daughter Helena. Manfred had accepted and had sent his new father-in-law four hundred mounted knights from Germany. Soon afterwards the new alliance was joined by William of Villehardouin, the Latin Prince of Achaia, who married Michael’s second daughter Anna. The ultimate object of the expedition was Constantinople, but this would clearly involve the capture of Thessalonica on the way.

Thus, at the time of the accession of Michael Palaeologus, virtually the whole of the Greek mainland was ranged against him. Fortunately he had dispatched a large expeditionary force to the Balkans, commanded by his brother, the sebastocrator John Palaeologus, and the Grand Domestic Alexius Strategopulus; and early in 1259 he ordered them to advance against the enemy. The two armies met at Pelagonia; and almost immediately the coalition fell apart. The Despot Michael and his son Nicephorus, wrongly suspecting that their allies were planning to betray them to the enemy, deserted the camp and fled. Another son, John the Bastard, taunted by Villehardouin over his illegitimacy, joined the Nicene forces out of pique. By the time the battle began John Palaeologus found only the cavalry of Villehardouin and Manfred ranged against him; and they proved defenceless in the face of his Cuman archers. Manfred’s knights surrendered and were taken prisoner, as – subsequently – was Villehardouin himself, who was found hiding in a haystack near Castoria and was recognized only by his protruding teeth. John then advanced through Thessaly, while Alexius marched straight to Epirus and captured its capital, Arta. The victory was complete.

It was by now plain that the recapture of the city could only be a question of time, and a short time at that. Of all Baldwin’s allies, there remained only the Papacy and Venice. Pope Alexander IV was uninterested; that left the Venetians, who had been largely responsible for the Latin Empire, and whose fleet of thirty ships still patrolled the Bosphorus. But soon the value of even Venetian support began to appear problematical, for on 13 March 1261 Michael Palaeologus, desperate for a navy, signed a treaty with Genoa whereby, in return for their help, the Genoese were promised all the concessions hitherto enjoyed by Venice, with their own quarter in Constantinople and the other principal ports of the Empire and free access to those of the Black Sea. For Genoa it was a historic agreement, laying as it did the foundations for her commercial empire in the East.

The recovery of Constantinople eventually came about almost by accident. In the summer of 1261, Michael VIII had sent Alexius Strategopulus to Thrace with a small army to indulge in a little mild sabre-rattling, sounding out the city’s defences at the same time. At Selymbria, Alexius learned that the Latin garrison was absent, having been carried off by the Venetians to attack the Nicaean island of Daphnusia, a harbour controlling the entrance to the Bosphorus from the Black Sea. They also told him of a postern gate in the walls, through which a handful of men could easily pass into the city. The opportunity seemed too good to miss. That night a detachment slipped into the city, surprised the guards and threw them from the ramparts. They then quietly opened one of the gates. At dawn on 25 July 1261 the army poured in.

Baldwin, awakened by the tumult, fled for his life. Making his way on foot to the little harbour of the Bucoleon, he escaped on a Venetian merchantman to the Latin-held island of Euboea. Meanwhile Alexius and his men set fire to the entire Venetian quarter so that the sailors returning from Daphnusia, finding their houses destroyed and their terrified families huddled on the quayside, would have no real choice but to sail back to their lagoon. Among the remaining Franks – perhaps a thousand all told – there was widespread panic. Some hid; some fled to monasteries; a few even resorted to the sewers; but there was no massacre. Gradually they emerged from their various refuges and made their way down to the harbour where the thirty Venetian ships were waiting These too sailed for Euboea – not, apparently, even pausing to take on provisions, since it is recorded that many of the refugees died of hunger before reaching their destination.

The Emperor Michael was two hundred miles away, asleep in his camp at Meteorium in Anatolia, when the messengers arrived. His sister Eulogia woke him and told him the news; but only when he was handed Baldwin’s abandoned regalia did he believe her. Immediately he began his preparations; and on 15 August 1261 he made his entry into the capital. Entering by the Golden Gate and preceded by the great icon of the Virgin Hodegetria – She who points the way’ painted, as everyone knew, by St Luke himself he proceeded on foot along the traditional route through the city as far as St Sophia, where a second coronation ceremony was performed by Patriarch Arsenius. This time, however, he and his wife were crowned alone, their baby son Andronicus being proclaimed as heir presumptive. As for John Lascaris, Michael’s ten-year-old co-Emperor, he had been left behind in Nicaea, ignored and forgotten. A little over four months later, on Christmas Day, his eyes were put out. It was, as it happened, his eleventh birthday.

From the start, the Latin Empire of Constantinople had been a monstrosity. In the fifty-seven years of its existence it had achieved nothing, contributed nothing, enjoyed not a moment of distinction or glory. After 1204 it had made no territorial conquests, and before long it had shrunk to the immediate surroundings of the ruined and ravaged city. The only wonder is that it lasted as long as it did. Of its seven rulers, not one made the slightest attempt to understand his Greek subjects, let alone to learn their language. Meanwhile its knights trickled back to the West, its allies turned away, its treasury lay empty. And its fall was, if anything, even more ignominious than its beginning – overpowered by a handful of soldiers in a single night.

But the dark legacy that it left behind affected all Christendom – perhaps all the world. For the Greek Empire never recovered from the damage, spiritual as well as material, of those fateful years. Nor, with its loveliest buildings reduced to rubble and its finest works of art looted or destroyed, did it ever succeed in recovering its morale. Before the Latin conquest the Empire had been one and indivisible, under a single basileus, Equal of the Apostles. Now that unity was gone. There were the Emperors of Trebizond, still stubbornly independent on the Black Sea shore. There were the Despots of Epirus, always ready to welcome the enemies of Constantinople. How, fragmented as it was, could the Greek Empire continue as the last great eastern bulwark of Christendom against the Islamic tide?

But Christendom too was changed. Long divided, it was now polarized. For centuries before and after the Great Schism, the differences between the Churches had been essentially theological. After the sack of Constantinople this was no longer true. To the Byzantines the barbarians who had desecrated their altars, plundered their homes and violated their women could not be considered, in any real sense, Christians at all. Future attempts to force them into union could never succeed for long, simply because anything appeared to them preferable to the idea of submission to Rome. ‘Better the Sultan’s turban than the cardinal’s hat,’ they used to say; and they meant it.

Warfare in the Age of Byzantine Reconquest I

Raids and Razzias

As well as being characterized by the sort of larger-scale offensive and defensive strategy exemplified in the campaigns of, for example, 838 or 863 described already, the period up to the middle of the tenth century saw a style of frontier fighting and skirmishing, of guerrilla tactics and raiding, that had developed over the centuries from the period when the frontiers became more-or-less stable in the first part of the eighth century. Quite a lot is known about this style of fighting both from historians’ accounts of campaigns and battles, as well as from a number of military handbooks, some of them written by serving soldiers. The late ninth- or early tenth-century Tactica of the emperor Leo VI, for example, shows that warfare along the eastern front followed a well-established pattern. By the 950s and 960s, this was changing, as imperial successes in pushing forward the frontier rendered the traditional system of defensive warfare more or less redundant.

The eastern frontier was guarded by a chain of lookout posts, with small units of irregulars acting as scouts and informants along the frontier, particularly covering the various points of ingress into imperial territory. The frontier was a broad band of territory, and the location of such lookout posts seems to have changed according to the situation, while raids and counterraids intended to destroy enemy outposts or more important local fortresses and bases frequently altered the pattern of local strategy.

An anonymous treatise written in the 960s sets out the key aspects of this type of warfare. First, the local commanders should make sure that the networks of watch-posts and lookouts are in order. Scouts should be recruited from among the local population, men with experience, a good knowledge of local routes and the different qualities they possess. They should work on a fifteen-day rotation, and be dispatched in small groups to watch the roads and routes that might be used by the enemy. Local commanders should make extensive use of spies, including merchants and others on genuine business in the enemy’s land – a long tradition in Byzantine strategic thinking. The call-up of registered soldiers should be strictly observed, and the scouting parties should be checked by an officer from time to time. They should also change their location in order to avoid capture. There were pre-planned schemes for evacuating the non-military population of the regions through which an enemy raiding party would pass, once its route had been ascertained, in order to preserve the local population and at the same time to deprive the enemy of the chance to collect provisions and easy booty.

The most important aspect of this frontier defensive strategy was ‘shadowing’. Following and harassing the enemy by exploiting one’s own knowledge of the local terrain was one aspect; keeping a close watch on his column and especially his encampment, in order to attempt ambushes on forage parties and other isolated groups, was another. Crucial to all operations was the idea of bringing together several smaller forces, leading eventually either to a full-scale confrontation, but with the imperial forces at a numerical advantage, or to a pincer movement, designed to encourage the enemy force to give up and return home. In this case, it was usually planned for imperial troops to have occupied the passes or exit routes which the enemy commander would follow. The subsequent surprise attack or ambush, which could result in the recovery of all or most of the booty, and certainly with the destruction and rout of the enemy army, was the ultimate aim. The possibility that his own forces might themselves become the victims of shadowing and ambush was ever-present, however, and the Byzantine commander was urged to use scouts and outriders in order to prevent this from happening.

One of the distinguishing features of this treatise is the focus on the judgement and independence of the local commanders. Not only should they themselves organize regular, small-scale raids over the border (unless the empire had made a formal truce with the Arab emirs or the Caliphate itself); they should be prepared to attack an invading force whenever an appropriate opportunity arose, and not necessarily wait for the arrival of reinforcements or the local senior commander.

The author of the treatise distinguishes three types of enemy raid, differentiated by size or by timing. Small, rapid raiding parties of cavalry, which might invade Roman territory at any time, and whose entry should be communicated to the local commanders as quickly as possible by the border scouts and watch-posts, should be shadowed, met, ambushed or hemmed in, and turned back, and if possible without any substantial gains in booty. Secondly, there were major raids, generally in August and September, consisting of substantial forces made up of volunteers for the jihâd as well as regular troops from the Arab borderlands – Malatya, Aleppo, Tarsos and Antioch. Such raids had both an economic and an ideological function, first in terms of the desire for booty, and to damage the Roman economy, and second in respect of the desire of many Muslims to participate in the jihâd. The local commander was enjoined to use every means at his disposal to find out when such raids would begin, by which route, and how numerous the enemy host would be. The invading force should be shadowed, along with any accompanying raiding parties which were sent out once the main force had reached Roman territory. The invaders’ logistical difficulties should be maximized by the removal of livestock and crops, or even their destruction in extreme cases. The enemy force should be subject to constant harassment as it moved, foraged for supplies, set up camp, or attempted to collect booty. The passes through which it would return should be occupied and ambushes laid; the water-supplies should be held by Byzantine forces. The enemy should be attacked as they returned, laden with booty. Naturally, the Romans were not always able to respond successfully to such attacks, and there are many examples where Roman preparations failed to produce the desired results, or where the Roman commanders were unable to outwit and out-general their adversary.

The local commander also had to be on his guard against surprise raids, launched before the local population had been evacuated or any sort of ambush or shadowing-party sent out. In an effort to delay the enemy, various measures could be applied, such as a feint attack to distract the enemy from pillaging the villages while they were being hastily evacuated. Once the local troops were in the field, the strategy of harassment and ambush, by day and by night, came into play. While he was one of the empire’s most successful antagonists, the emir of Aleppo was ambushed on at least three occasions using this strategy, barely escaping with his life on one occasion.

This sort of warfare could also be offensive. Local commanders were advised to maintain bands of raiders, whose task it was to raid deep into enemy territory in order to foment insecurity and uncertainty. One of their most important tasks was to take prisoners, so that Byzantine commanders might learn of enemy troop movements and intentions. Similar arrangements seem to have operated in the Balkans at times, although these were not always regular soldiers, but drawn from semi-independent peoples whose marginal situation between the two cultures suited them ideally for this task. Such sources also provided some of the regular light cavalry, in view of their detailed knowledge both of regular routes as well as side-paths and concealed tracks, watering- and camping-places in the mountains.

In the following, a typical penetrative raid into eastern Asia Minor by a medium-sized Arab army is described. The account comes from an eyewitness and contemporary of the leader of the expedition, the famous warrior emir of Aleppo, Sa’if ad-Daulah, and provides a wealth of topographical detail which neatly complements the information and the description of the Byzantine strategic response to such raids found in the tenth-century military handbooks.

Sa’if ad-Daulah’s Raid of 956

The raid in question was launched in early spring of that year, intended as a booty-collecting expedition, as an attack on the Roman frontier provinces, and as a distraction intended to draw off the Roman forces that had been sent to raid the lands to the east of Sa’if’s own base at Aleppo. Its object was the district of Anzitene, recently conquered by the eastern Roman armies and incorporated into the thema of Mesopotamia. This was a rich district and had always been of strategic importance, both in the wars between east Rome and the Persians and later in those between Romans and Arabs. Its strategic location attracted attention, for it commanded access to Armenia from the south and south-west and across the Euphrates. Whoever held Anzitene had a springboard for attacks in either direction, and the Romans were now using it for precisely that purpose, as their campaigns to roll back the Islamic emirates along their south-eastern flank progressed. Sa’if had received information that the military governor of Anzitene, based at the fortress town of Harput (mod. Elaziz), had set out to raid the upper Tigris region, part of Sa’if ’s domain, leaving his home territory more-or-less undefended, presumably because Sa’if was himself still far to the south-west in his capital at Aleppo.

Gathering his forces of both cavalry and infantry (who nonetheless, as was customary with Muslim raiders throughout this period, were mounted so that they could keep up with the fast pace set by the cavalry units), Sa’if set out on Monday 28 April first for the town of Harran, where he negotiated the support of the local Beduin, the Banu Numair. Rather than marching east to deal with the Byzantine raiders in and around Amida, however, he then turned north and marching past several of his own fortresses, entered enemy territory north of the fortress of Hisn Arqanin (mod. Ergani) – the marches of Anzitene – some twelve days after leaving Aleppo, on Saturday 10 May. The castle controls the southern end of the Ergani pass, and it is likely that Sa’if ’s forces controlled the rest of the pass, which cuts through the mountains to enter the plain of Anzitene near the lake now known as Hazar Gölü (anc. Lake Thospitis). Upon receiving news of the Muslim raid, the eastern Roman commander and his forces withdrew from Amida, and began the march back to their own territory. Sa’if ’s raid had clearly taken them by surprise.

Sa’if set up camp on the shores of the lake. The nearest major Byzantine base was at Arsamosata (mod. Haraba), some distance to the north-east at the end of the valley of the Arsanias river, which flows westwards down to the Euphrates. Sa’if ’s force was, therefore, safe from attack for a while. Here the Arab cavalry ravaged the surrounding countryside, carrying off much booty and many captives. The next day – Sunday 11 May – Sa’if himself sent a small raiding force to scout ahead as far as the Arsanias, following himself once it had been deemed safe, and encamped in a small village at the base of the hill on which the provincial capital of Harput was situated – the absence of most of the local forces gave him essentially a free hand in the region. The area was thoroughly ravaged, before he set forth again marching this time to the north-west, where he bridged the river with materials he had carried with him (dismantled rafts and boats) and, having sent across a small cavalry vanguard in advance to secure the bridgehead, crossed with his main army three days later (on Thursday 15 May), destroying the governor’s residence. This was an entirely unexpected action on Sa’if ’s part, and as well as burning the residence of the governor other undefended settlements were also destroyed. The whole area was thoroughly ravaged and an enormous booty in people, livestock and materials was collected.

Having completed his action on the northern bank of the river, Sa’if withdrew to his base and then marched south, pillaging as he went, until he reached the important fortress of Dadima, to which he lay siege. The garrison was quite unprepared for an attack, and would have surrendered had Sa’if not received news at this point of the occupation of the passes through which he was expected to retire by the returning Roman forces. Rather than head for the obvious exits, however, on Friday 23 May he marched south east and encamped not far from Arsamosata, whence he made his way not to the pass of Ergani, by which he had entered Anzitene, but to the pass of Baq’saya, to the south of Arsamosata and east of the Ergani pass. Here he found a small eastern Roman force blocking his way and, in a fierce battle late on Saturday 24 May, he managed to drive it off with considerable loss, seizing its baggage train and killing several prominent leaders. By the evening of Sunday (25 May) he was back in Amida, where he received, of course, a hero’s welcome. In three weeks of campaigning he had penetrated deeply into the Byzantine frontier region, caused a great deal of damage and dislocation to the local population and the military command, totally outmanoeuvred his enemy, outwitted them in a short, sharp field action, and returned safely laden with booty. The raid is a classic of its kind, and also illustrates the problems faced by eastern Roman commanders when they failed to follow the strategy of shadowing warfare enunciated in the treatise we have discussed above and left their own territories inadequately protected. And although there can be little doubt that Sa’if was by far one of the most outstanding of the Muslim emirs with whom the Byzantines had to contend in the region, the account of this raid provides an excellent insight into the character of the warfare along the eastern frontier of the Roman world from the eighth until the tenth century.

Sa’if was not always so successful, however, and indeed there is enough evidence to suggest that the Byzantine tactics – permitting the enemy force to enter Roman territory but blocking their exits – proved sufficiently successful to deter all but the bravest raiders. In 950, for example, he had mounted a similar large-scale raid with a substantial force – supposedly some 30,000, although this may be exaggerated. Leaving Aleppo in the spring he had pushed up through the Taurus, marched north to cross the Halys, and ravaged the area before meeting and defeating a smaller Byzantine force. On his return, however, the Roman forces had correctly assessed which defiles he would use and, permitting the vanguard to pass their pickets unopposed, had then blocked the pass and fallen on the main body and the rearguard. The raiders panicked and fled, suffering substantial casualties, the booty was recovered more-or-less in its entirety and Sa’if, having been abandoned by all but a few of his men, barely escaped with his life. Similar tactics were employed again in both 958 and 960, when Sa’if found his way blocked and suffered a substantial and, on the last occasion, nearly fatal defeat, with the loss of all the booty and his own baggage train.

Warfare in the Age of Byzantine Reconquest II

Offensive Strategy and Tactical Change

From the seventh to the 12th centuries, the Byzantine army was among the most powerful and effective military forces – neither Middle Ages Europe nor (following its early successes) the fracturing Caliphate could match the strategies and the efficiency of the Byzantine army. Restricted to a largely defensive role in the 7th to mid-9th centuries, the Byzantines developed the theme-system to counter the more powerful Caliphate. From the mid-9th century, however, they gradually went on the offensive, culminating in the great conquests of the 10th century under a series of soldier-emperors such as Nikephoros II Phokas, John Tzimiskes and Basil II. The army they led was less reliant on the militia of the themes; it was by now a largely professional force, with a strong and well-drilled infantry at its core and augmented by a revived heavy cavalry arm. With one of the most powerful economies in the world at the time, the Empire had the resources to put to the field a powerful host when needed, in order to reclaim its long-lost territories.

As a result of the increasingly aggressive warfare carried on by the empire from the second quarter of the tenth century, particularly on the eastern front, the need to recruit more professional soldiers, and the need to operate effectively on campaigns which demanded more than the seasonally available forces provided by the traditional thematic armies, a number of important changes appeared in the tactical structure and in the arms and armour of Byzantine troops. A number of important technical treatises on strategy and tactics were written in the middle and later tenth century, and the narrative accounts of contemporaries, both Byzantines and Arabs, corroborate much of what they say. The changes can be enumerated briefly as follows: (1) the revival of a corps of disciplined, effective heavy infantry, able to stand firm in the line of battle, confront enemy infantry and cavalry, support their own cavalry, march long distances and function as garrison troops away from their home territory on a permanent basis; (2) the introduction of a corps of heavy cavalry armed with lances and maces, which could operate effectively alongside infantry, adding weight to the Byzantine attack and thus substantially enhancing the aggressive power of the Byzantine cavalry; (3) the development of field tactics in which these arms operate in a complementary way, offering the commanding officer a flexible yet hard-hitting force which could respond appropriately to a range of different situations.

Evidence for these changes comes partly from the contemporary sources, especially the military handbooks already referred to, but also from the startling successes marked up by Byzantine armies in the process of reconquest and expansion from the 950s onwards. In a tract known as the ‘Recapitulation of Tactics’, a new formation of infantry soldiers is described, consisting of troops wielding thick-stocked, long-necked javelins or pikes, probably similar in form to the Roman legionary pilum. Their task was to confront and beat back enemy heavy cavalry attacks. According to the ‘Recapitulation’ there should have been about 300 soldiers equipped in this manner, arrayed in the intervals between the infantry units making up the main battle line. They were deployed in either line or wedge formation to break up an enemy attack. In a treatise known as the ‘Military Precepts’ compiled some twenty years later, the tactic had evolved further, so that there were in each major infantry unit of 1,000 men 100 soldiers so equipped, integrated with 400 ordinary spearmen, 300 archers and 200 light infantry (with slings and javelins). Their task remained unchanged.

This important change in the role of infantry was reflected in the changed political and military situation of the tenth century. Whereas the sixth-century Strategikon presents its sections on infantry drill and formations after those (more detailed) dealing with cavalry, the tenth-century texts give infantry formations equal or even preferential treatment. Infantry had now become a key element of the army both numerically and tactically, outnumbering cavalry by 2:1 or more, in contrast to the normal situation in the preceding centuries. Contemporaries note the greatly improved discipline and training which such troops displayed. The importance of infantry is demonstrated in the fact that a special commander for the infantry division in each army was appointed, the hoplitarch (hoplitarches), in charge of training, discipline and fighting skill. The new tactics were embodied in a new formation, in which infantry and cavalry worked together, essentially a hollow square or rectangle, depending on the terrain, designed to cope with encircling movements from hostile cavalry, as a refuge for Byzantine mounted units when forced to retreat, and as a means of strengthening infantry cohesiveness and morale.

These new formations mark a real change in the role of infantry, no longer drawn up in a deep line with only a limited offensive role, but actively integrated into the offensive heavy cavalry tactics of the period. Infantry units now represented a sort of mobile marching-camp, with a traditionally rather unreliable force given new strength as a defensive field formation on the one hand, to provide security in defence and on the march, a mobile base and refuge for lighter troops and cavalry, and on the other as a formation which could be transformed into a solid attacking formation at a few simple commands. One important aspect of this change was a focus on the recruitment of good infantry from warlike peoples within the empire, especially Armenians. The demand for uniformity in tactical function and therefore equipment and weaponry meant that the Byzantine infantry of this period were more like their classical Roman predecessors than anything in the intervening period.

The cavalry also evolved at this time. New formations of ‘super heavy’ cavalry appear, called klibanophoroi, heavy cavalry troopers armed from head to foot in lamellar, mail and quilting, whose horse was likewise protected – face, neck, flanks and forequarters were all to be covered with armour to prevent enemy missiles and blows from injuring the cavalryman’s mount. Very few in number (because they were so expensive to maintain) these became the elite strike force within each field army. Drawn up in a broad-nosed wedge, their primary function was to smash through the enemy heavy cavalry or infantry line, disrupt his formation, and open up the enemy battle order to allow the supporting horse to turn the enemy’s flanks. The ‘Military Precepts’ mentions a formation of just over 500 such troops for a large wedge, two-thirds of whom would be real klibanarioi/kataphraktoi, the rest equipped as lightly armed mounted archers.

Both Byzantine and Arab writers of the time comment on the impressive effects of this formation on their foes. One Arab writer notes that the horse-armour of the heavy cavalry mounts made them appear to be advancing without legs. Some of these changes were probably due to the general, later emperor, Nikephoros Phokas, who became commander-in-chief in the East in the 950s, and immediately embarked upon a programme of training and drilling the soldiers in an attempt to re-establish good discipline, fighting spirit and good battlefield skills. His success is evident in the effective warfare waged by Byzantine forces over the following fifty years or more.

A Feigned Retreat in 970

In the autumn of the year 965, shortly after the conquest by Byzantine armies of the islands of Crete and Cyprus, as well as the destruction of the Islamic power in Cilicia and its incorporation into the empire, Bulgarian envoys arrived at the court of the emperor Nikephoros II Phokas. Their purpose was to request the payment of the ‘tribute’ (or ‘subsidy’ from the imperial perspective) paid by Constantinople to the Bulgar Tsar as part of the guarantee for the long-lasting peace which had been established after the death of Tsar Symeon in 927. But the situation of the empire had changed radically in the course of the preceding half century, and rather than pay, the emperor Nikephoros, outraged by the presumptive demand of the Bulgarian ruler, had the envoys beaten and sent home in disgrace. He despatched a small force to demolish a number of Bulgarian frontier posts, and then called in his allies to the north, the Kievan Rus’, to attack the Bulgars in the rear.

The steppe region stretching from the plain of Hungary eastwards through south Russia and north of the Caspian was the home of many nomadic peoples, mostly of Turkic stock. It was always a key principle of Byzantine diplomacy to keep these peoples well disposed towards the empire. Following the collapse of the Avar empire in the 630s, Constantinople had been able to establish good relations with the Chazars whose Khans, although converting to Judaism, remained a faithful ally of most Byzantine emperors. Chazar Khans often took up Byzantine invitations to attack the Bulgars from the North, for example, when war broke out in this region. And they served also to keep the imperial court informed of developments further east. The Chazar empire contracted during the later ninth century, as various peoples to the east were set in motion by the expansion of the Turkic Pechenegs. These newcomers clashed with both the Chazars and the Magyars, establishing themselves in the steppe region between the Danube and Don. Their value to the empire as a check on both the Rus’ and the Magyars was obvious, particularly in the wars of the later tenth century, but they were a dangerous and frequently unreliable ally. The Magyars (Hungarians) had been established to the north-west and west of the Chazars since the middle of the ninth century, settling in what is now Hungary by about 900 AD. Both Chazars and Magyars served as mercenaries in Byzantine armies, particularly against the Bulgars, although the establishment by the later tenth century of a christianized Hungarian kingdom on the central Danube posed a potential challenge to Byzantine power in the region, which became especially acute during later centuries. But the growing power of the Kiev Rus’ during the later ninth and tenth centuries introduced important changes to this situation and to Byzantine diplomacy.

The Rus’ were the product of an amalgamation of Scandinavian settlers and indigenous, largely Slavic peoples, based along the rivers of central and western Russia. Their dominance over the neighbouring steppe and forest peoples had made them an important political power. By the middle of the ninth century their longships were entering the Black Sea, and by the early tenth century they had established trading agreements – not without some serious fighting between the two parties – with the empire. An alliance had been established from the middle of the tenth century; and when Nikephoros II asked for their help in 966, the warlike and ambitious Prince Svyatoslav, who had already established a considerable reputation through his successful warfare with and defeat of the Chazars, was only too willing to agree. In 968 he arrived on the Danube and easily defeated the Bulgarian forces sent against him. In 969 he had to return to Kiev to repulse an attack from the Pechenegs, but he returned later in the year and quickly occupied most of northern and eastern Bulgaria, deposing the Tsar, Boris II, and incorporating Bulgaria into his own domain. (See Map 9)

This was not part of the emperor’s original plan at all. Nikephoros tried in vain to establish an alliance with the Bulgars, but late in 969 he was assassinated, leaving his successor, John I Tzimiskes, with the difficult task of removing this potentially far more dangerous foe. To add to his problems, some of the Bulgar nobility now saw a chance to recover their independence of the Byzantine state and its culture by working with the Rus’. Svyatoslav sent the new emperor an ultimatum to evacuate all the European provinces and confine the empire to Asia alone. Then, in the spring of 970 a large Rus’ force invaded Thrace, sacking the fortress of Philippoupolis (mod. Plovdiv) and moving on down the road to Constantinople. John was forced to take action.

The emperor was not able to march immediately against the Rus’, for the majority of the effective field units were still in the east, where they had been campaigning in the region of Antioch and beyond, consolidating Roman gains after the recent fall of that city to the forces of Nikephoros II. In response, however, John appointed Bardas Skleros, together with the patrikios Peter, both experienced commanders, to take command of a small force and reconnoitre the enemy dispositions in the occupied territories. Their mission was, in addition, to exercise the troops and to prevent, as far as they were able, enemy raiders committing widespread damage on imperial territory. They were also to send spies, disguised in Bulgarian and Rus’ costume, deep into enemy-held territory to learn as much as possible about Svyatoslav’s intentions and movements.

It was not long before the Rus’ leader was informed of the imperial army’s presence, and he despatched a considerable force, consisting of both Rus’ and Bulgar troops as well as a powerful detachment of Pechenegs, whom he had temporarily managed to bring onto his side with the promise of booty and pillage, to drive the Romans off. Bardas immediately collected an elite force of some 10-12,000, sending one of his officers ahead to keep an eye on the enemy army and find out how many they numbered and where they had encamped. On receipt of the information that the enemy force was quite close, in the region of Arkadioupolis (mod. Luleburgaz), Bardas divided his force into three: two divisions were concealed in the rough scrub and woodlands on either side of the track leading towards the enemy position; the remaining division he led himself, launching a furious surprise attack against the Pecheneg contingent in the enemy force. Although heavily outnumbered – Bardas can have had only 2,000-3,000 men with him – he was able to draw the enemy out of their encampment and feign a gradual withdrawal. The fortunes of the battle swung back and forth, and it seemed at times that the small Byzantine force must be overwhelmed. Yet their discipline and training told, and Skleros finally ordered the pre-arranged signal to be given for the whole force to fall back. At the same time, however, the two divisions which lay in ambush prepared themselves, and as their comrades drew level with them and then past them, they too launched themselves upon the unsuspecting enemy from both flanks and the rear. Within a few minutes the Pechenegs had received such a savage mauling that they turned and fled, while their allies, the Rus’ and Bulgars, who had been hastening to catch them up in their pursuit of the supposedly defeated Romans, were caught in the panic and suffered similarly heavy casualties as the rout became general. According to a contemporary, the Romans lost some 550 men and many wounded, as well as a large number of horses, which fell to the archery of the Pechenegs. The combined enemy force, however, lost very many thousands. The action won the emperor John valuable time and also provided him with essential information about the make-up, morale and fighting prowess of his enemy, information which he put to excellent use in his campaign the following year.

This brief encounter, although the sources offer few details of the order of battle of either side, gives some idea of the possibilities for a properly trained eastern Roman force, when well-led and prepared, to defeat an enemy vastly superior in numbers. There is no doubt that the troops Bardas Skleros had under his command were well-trained and disciplined, as their careful withdrawal and feigned retreat under very difficult conditions, heavily outnumbered and under constant fire from accurate enemy archery, demonstrates. The short battle admirably reflects the esprit de corps, training and morale of the armies of this period.

The Battles of Dorostolon 971

Varangian Guard 10th Century

Plan of the battle of Dorostolon.

Immediately following this victory the emperor ordered more units from the Asia Minor forces to cross into Thrace and prepare for the coming campaign. He also began to build up a strong siege train and amass supplies adequate for the powerful force he intended to assemble in order to drive the Rus’ forces from Bulgaria entirely. The expedition was to be accompanied by a powerful naval force which was to carry troops and supplies and be ready to harry the invader from the coast or support the imperial land forces once they had reached the Danube. For this was John’s intention, and his careful preparations make it clear that he was hoping to drive the Rus’ forces back across the mighty river. Diplomacy also played a key role, with the defeat near Arkadioupolis being used to persuade the Pechenegs to withdraw their support, and John claiming that his intention was to replace the deposed Boris on his throne.

In April 971 the emperor’s army – a reasonable estimate based on contemporary sources suggests it may have numbered as many as 30,000 – set off and passed through the Bulgarian frontier regions. The passes which had so often proved the site of Byzantine reverses were undefended – possibly because the Rus’ were engaged in suppressing Bulgarian rebellion to the north – and the army passed safely through, to issue onto the plain in which the Bulgarian capital, Preslav, was located. After a series of short but vicious encounters the Russian forces were defeated and driven off, and the city was recovered, along with the important political prize of the deposed Tsar himself. The Rus’ forces withdrew to the north and joined the remaining Rus’ troops under Svyatoslav himself, who had established his base in the fortress of Dorostolon (Dristra, modern Silistra) on the south bank of the Danube.

The campaign proceeded with the emperor’s march northward. As he went the Bulgarians began to slip away from Svyatoslav, which prompted the latter to execute a number of high-ranking Bulgars he had with him at Dorostolon and imprison many others. The emperor’s advance was virtually unopposed, and the army quickly took a series of smaller fortresses and strong-points defended nominally by Bulgarian troops who, however, offered no resistance when the emperor offered them terms. Only in the approach to Dorostolon itself was there some hostile action: a small Rus’ force had set an ambush in the dense woodland on the approaches to the fortress, and were able to surprise an advance party of imperial cavalry, sent ahead to scout the terrain and, possibly, locate a suitable site for the various divisions of the imperial army to set up camp. As the main body of the vanguard, together with the emperor, arrived to find the bodies of those who had been killed, troops from the imperial bodyguard were despatched to comb through the woods and eradicate the threat. A number of prisoners were taken and as a token of his policy to the enemy the emperor ordered them immediately put to the sword.

Once clear of the woodland, the imperial forces established a base camp where the baggage and siege trains were drawn up in a defensive position with a small detachment to guard them. Shortly afterwards the scouts returned to inform the emperor that the Russians were drawn up in battle order on open terrain before the fortress. Their formation was reported as a long, dense line bristling with weapons, awaiting the imperial assault and confident in their numbers. Although the figure of 60,000 given by one of the sources is an exaggeration (the area of the medieval fortifications within which the Rus’ forces were quartered shows that this figure is far too large), it is clear that the Russian army was considerable, and had been built up over the three years that Svyatoslav had been active in Bulgaria. The incremental addition of more soldiers – possibly including some from beyond the Rus’ lands themselves – had no doubt increased very considerably the original force that had accompanied Svyatoslav’s first campaign in 968.

To oppose the long, dense line of the Rus’, John drew up his force in three divisions over the same front. Both wings were reinforced by a reserve of heavy cataphract cavalry; and in a second line behind the first he placed infantry archers and slingers, who were instructed to maintain a constant hail of missiles on the enemy forces as long as they were in range. By shortly after midday the two forces were drawn up opposite one another. The Russians did not await the Roman attack, but with a long drawn-out battle roar advanced to meet the imperial forces. At the initial clash the Romans were able to halt the Rus’ advance and at one or two points break through the densely packed mass of warriors; but the latter were quickly able to regroup and re-establish the shield-wall with which they now opposed the Roman thrusts. The battle lines moved back and forth for more than an hour until both sides fell back and regrouped preparatory to a fresh assault. Again the Byzantine forces halted the Russian charge, and were even able to push their line back some distance, but it could not be broken. A contemporary eye-witness, Leo the Deacon, who accompanied the imperial high command on the expedition, remarks on the Russian refusal to concede defeat when they had a reputation for invincibility to maintain, as well as their personal honour as warriors; the Roman forces, unwilling to accept defeat at the hands of a barbarian nation who, to paraphrase Leo, could not even ride, were equally unwilling to give up. By late afternoon the battle appeared to have reached a stalemate, and it was at this point that the emperor gave the command to commit the heavy cavalry on both wings. Supported by a renewed push from the heavy infantry in the centre, the wedges of heavily armoured horsemen now advanced against the enemy wings, and with a concerted war-cry from the whole Roman front the imperial forces charged into the Russian lines, the cavalry crashing through the shield-wall with irresistible force, driving the Russian wings back towards their centre. Within a few minutes the Rus’ front collapsed, imperial units had penetrated the enemy line, and the Russians began to break and stream back towards the fortress for refuge. The rout was complete, and many were killed as they tried desperately to enter the fortress before the Romans caught up or the gates were shut on them.

The emperor recalled his troops, and as they began to return to their base camp, preparatory to establishing a siege of Dorostolon, Leo records that the soldiers sang a victory song. The emperor had selected for the imperial encampment a low eminence at some distance from the fortress. This was fortified by a ditch with the earth piled inside, upon which the troops were ordered to set their spears and lances and, propped against them in an unbroken wall, their shields. Leo the Deacon notes that this was the usual arrangement for a Roman encampment in hostile country.

The following day the imperial troops approached the fortress and launched attacks against various points around the defences, but were met with a hail of arrows and stones. Although they replied in kind, neither side seems to have suffered particularly badly from this exchange and, after being unable to make any headway against the defenders, the army was withdrawn to its camp, although the fortress was kept under constant surveillance. Towards evening the Rus’ made a sortie from the fortress and sent a mounted detachment to harry the Roman pickets. Leo notes that this was the first time they had seen the enemy on horseback. It soon became obvious that their lack of experience in mounted fighting would tell against them, and a detachment of Roman cavalry rapidly broke their formation and sent them helter-skelter back to Dorostolon.

The emperor seems at this stage to have decided not to press the fortress too closely but rather to try to tempt the Russian forces out to meet his own army in open battle, where victory might be had much more readily and quickly than through either a direct attack against the well-defended walls and towers of Dorostolon or a prolonged siege. But he had prepared for all eventualities, and it was at this point, about the third day after the Roman forces reached the fortress, that the fleet arrived, a fleet which included both supplies and reinforcements, as well as a number of warships equipped with liquid fire projectors. This weapon – a type of medieval napalm projected from tubes mounted on the bows of the vessels in question – was already known to the Rus’, for the warships of Svyatoslav’s father, Igor, had been destroyed by the very same weapon some thirty years earlier. The arrival of the fleet, which now sealed the Russians on the southern bank of the Danube and removed any chance they might have had to escape, was a great boost to Roman morale but must have seriously eroded the confidence of the Rus’ forces bottled up in Dorostolon.

The day after the arrival of the fleet, Svyatoslav led his forces out once again in an effort to draw the Roman forces into battle, hoping to defeating them and relieve the siege. As before, however, the two armies were evenly matched until the Roman heavy cavalry with their fearful iron maces once more drove the Russian line in on itself. Again the enemy fell back, at first in some order, then dissolving into rout and fleeing back inside the defences.

The emperor now set up his siege weapons around the fortress and began to attack both the walls and the troops within by a constant shower of missiles. In an effort to rid themselves of this annoyance, which was causing casualties and affecting morale, the Rus’ mounted a series of sorties with the aim of burning the Roman siege engines and other equipment. In one such incident the detachment guarding one of the emplacements was taken almost by surprise and its leader, a certain Kourkouas, who is reported to have been slightly drunk following his midday meal, was killed. Kourkouas was renowned for his luxurious and conspicuous outfit, and the Rus’ who killed him, thinking he was the emperor himself, hefted his head on a spear and mocked the Romans, though without the effect for which they had hoped. The siege machinery that they attacked remained undamaged, and the siege continued.

Pleased with this apparent success, however, the Russian army issued forth again the next day, and the emperor permitted them to draw up their line of battle. Once again the battle was joined, with the Roman forces drawn up in a single deep phalanx, the cavalry partly concealed behind the extremities on each wing and the missile troops behind the heavy infantry in the centre. On this occasion, a sharp charge by a Rus’ contingent succeeded in pushing deep into the Roman line, inflicting heavy casualties. But a counter-charge by the supporting cavalry, including units of the imperial guards, thrust it back, with the loss of its leader, Svyatoslav’s second-in-command, at the hand of Anemas, a member of the emperor’s own bodyguard. The Romans were immediately ordered to advance and push back the demoralized Rus’ line, which broke and ran, retreating once more in disorder to the safety of the fortress.

That night the Rus’ soldiers and their families opened the gates and came out onto the field of battle to search for their dead, for whom they then built funeral pyres and, having carried out the appropriate rites, committed their bodies to the flames. Whether the Roman forces watching the fortress were ordered not to interfere, or whether they allowed this activity to take place without consultation, is not stated. But Svyatoslav now summoned a meeting of the leading warriors to seek counsel. Some advised an attempt to break out by boat, across the Danube, braving the Roman warships which were on constant patrol; others advised opening negotiations with the emperor in an effort to save as many lives as possible – for, as one of the Rus’ leaders put it, ‘we are not accustomed to facing heavily armoured cavalry, especially after the loss of so many of our foremost warriors, upon whom the men depend for their courage and leadership’.

Svyatoslav was unwilling to follow either of these courses of action, however, and with the support, apparently, of the majority, voted to fight it out to the bitter end. The Rus’, he said, are not accustomed to give in, but would rather die in battle and go to Valhalla. And, as Leo the Deacon adds in his account, a Rus’ soldier has never been known to fall living into the hands of their enemies.

The same night, therefore, in a desperate attempt to obtain supplies, a Rus’ detachment of some 2,000 men left the fortress in small boats, and at some distance along the southern bank of the river crept ashore to begin their search for provisions. The naval detachments guarding the river should have detected this move, for the emperor had ordered that no Russian should be allowed to get out of the fortress. Instead, having succeeded in collecting some food for their beleaguered comrades, the Russian force made its way back by a different route from that which it had followed on the way from its boats, and stumbled by accident upon a small group of Byzantine cavalry soldiers sent out to water their horses and collect wood. Heavily outnumbered, the unsuspecting Roman troops were quickly put to the sword, and although the alarm was quickly raised, the Russians managed to get back with their supplies before they could be intercepted. The emperor was furious at this lapse in naval discipline, which had caused an entirely unnecessary loss of life and allowed the enemy, who were clearly in increasingly short supply, to replenish their stores. The naval commanders were threatened with execution if such a lapse recurred.

The next day was Friday 24 July, and having spent the day preparing for the battle, the whole Rus’ army sallied out towards late afternoon before the fortress. They formed up in a deep, solid phalanx protected by a solid shield-wall bristling with spears. The Roman forces were drawn up in the usual three divisions with the heavy cavalry behind the wings and with the emperor himself in the centre leading his own cavalry as a reserve. The location of the conflict was slightly nearer the fortress than hitherto, however, with the result that both forces were crowded together on a narrower front, with woodland on one flank and the marshy regions stretching back from the river on the other. The Russians, in order to counter the effect of the heavy cavalry on the Roman side, placed their archers on the wings where they could do most damage should the Roman cavalry attack them there. Led now by Svyatoslav himself, the desperate Rus’ warriors hurled themselves with fury against the Roman line, and soon began to push the centre back. Many cavalry mounts were injured or killed by the unusually heavy Rus’ archery. The emperor, observing that while his men were holding the line they were also rapidly tiring in the heat of the day, ordered the issue of water mixed with wine to the ranks, presumably on a rotational basis so as to leave the battle-lines undisturbed, a move which refreshed the imperial troops and enabled them to withstand the constant Rus’ charges.

Realizing that the overall tactical situation did not favour the Roman dispositions, however, the emperor then ordered a general withdrawal, and with admirable discipline and order the whole Roman line was able to pull back some distance from the fortress onto a broader plain which offered much greater room for manoeuvre and for more effective use of the Roman heavy cavalry. During the withdrawal one Roman officer, a certain Theodore of Misthia, was cut off and isolated from his unit – suggestive of the closeness of the Russian pursuit – and a fierce mêlée developed around him. At one point he is reported to have used a Rus’ corpse, which he lifted before him by the belt, as a shield. He was eventually helped by his comrades who rushed his attackers and got him safely back to their own lines. Some time after this, Anemas, who had already distinguished himself, moved with his unit of guards (probably the Athanatoi, the Immortals, a unit created by the emperor John shortly after his accession in 969) to reinforce the line. Anemas himself came close enough to the Rus’ leader to attack him and knock him down, but Svyatoslav’s armour prevented serious injury and within seconds Anemas was himself cut down by the spears and axes of Svyatoslav’s men.

Heartened by this the Rus’ in the centre attacked with renewed force and the Roman line began to give way. Some of the cavalry in the rear of the line began to waver and turn, and at this crucial moment the emperor decided that he must commit his reserve, his own bodyguard, to the fray. The emperor’s units, with the emperor himself in the forefront, moved to join the fray, and this sight steadied the Roman centre and allowed them to stabilize their position once more. At this critical point a strong wind blew up, accompanied by a fierce thunderstorm – not untypical of midsummer weather conditions in the region – and the Rus’ had the misfortune to find the wind blowing directly in their faces. Eye witnesses reported seeing a rider on a white horse, who appeared in the centre of the Roman lines and led a charge which smashed through the Russian shield-wall. Later stories had it that this was in fact St Theodore, the emperor’s own patron and a leading soldier saint in the Byzantine Church. In fact, it was almost certainly the emperor himself and his elite guards who, in a final move to open up the Russian shield-wall, led a charge of the heavy cataphract cavalry. Either way, the move was successful. The Roman centre now began to push forward once more, and the Rus’ warriors themselves, partially blinded by the rain and dust, were unable to organize themselves against this new onslaught. At the same moment the Roman heavy cavalry on one wing (the reports do not specify which) under the general Bardas Skleros completed an encircling movement begun moments before, crashing into the flank of the Russian wing and driving it back onto the troops in the centre. With this sudden turn of events the Russian lines dissolved and began to flee once more towards the fortress. The slaughter was enormous as the victorious Roman cavalry fell on the routing troops. The Romans lost some 350 against over 15,000 Russian casualties, an improbable figure but indicative nevertheless of very heavy losses. Some 20,000 shields and innumerable swords were taken from the field by the Romans.

This battle marked the end of the Russian attempts to break the siege and drive the Romans off. They had lost too many men and were rapidly running out of provisions to continue the struggle. Despite his earlier refusal to contemplate asking for terms, Svyatoslav now so no other option before him. The emperor accepted the Rus’ proposals for a peaceful withdrawal and the handing over of Dorostolon undamaged, with all the prisoners and booty the Rus’ had taken, and agreed to permit the remnants of the Rus’ army to cross the Danube without being attacked by the warships with the liquid fire projectors, which the Russians greatly feared. Indeed, John even agreed to resume friendly commercial contacts with the Rus’, once Svyatoslav had reached home. In the event, the Russian prince was never to do so, dying in an ambush set at the mouth of the Dnieper by his erstwhile Pecheneg allies.

In a brilliant four-month campaign, carefully prepared and supported by a well-oiled logistical organization – one of the greatest strengths of eastern Roman military administration – the emperor John I had crushed and driven off one of the fiercest enemies the Romans had yet faced. Leaving a strong garrison in Dorostolon, he returned to Constantinople, where the great victory was celebrated with an imperial triumphal procession. The deposed Tsar Boris was asked to surrender his crown, and the eastern parts of Bulgaria were absorbed into the empire as a province. The Bulgarian campaign of 971 highlighted again the discipline, order and effectiveness of the imperial armies of the second half of the tenth century.