Byzantine Cities, Villages and Fortifications

The fortress town of Theodoro-Mangup in the 15th century, the last remnant of the Byzantine Empire to resist against the Ottomans until being conquered in 1475.

Gevele Castle is a ruined castle located on the summit of Mount Takkeli in Konya Province, Turkey. The site was used as a fortified site during the Hittites, Hellenistic, Roman, Byzantine, Seljuk, Karamanid and Ottoman eras.

One of the most obvious effects of warfare is to be seen in the architectural heritage of a society, primarily in respect of fortifications and in shifts in settlement patterns and relationships between centres of consumption and areas of production. In the East Roman world such shifts are especially apparent during the seventh century and in the aftermath of the Persian and more particularly the Arab invasions. While these wars were in themselves neither the original stimulus for the transformation of urban life in the late Roman and early Byzantine period, nor the only factor affecting the evolution of fortified inhabited sites during the period from the seventh to the twelfth centuries, they were nevertheless a crucial factor in the form towns and fortresses took and in the pace of their evolution.

In fact, there had been a slow process of transformation in the pattern of late Roman urban society over the centuries preceding both the Persian wars and the Arab conquests which it will be worth very briefly summarizing here. During the Roman period cities—poleis or civitates—had held a key role in both social and economic relations, as well as in the imperial fiscal administration. They could function as market centres for their district or region or, where ports were concerned, as major foci of long-distance commerce. Some fulfilled all these roles, others remained merely administrative centres created by the state for its own fiscal administrative purposes. All cities were also self-governing districts with, originally, their own lands, and were made responsible by the Roman state for the return of taxes—indeed, where cities in their Mediterranean form did not exist, the Roman state created them, either establishing new foundations or amalgamating or changing the form of preexisting settlements, providing them with the corporate identity, institutional structure and legal personality of a civitas. All cities, with a few exceptions such as Rome and Constantinople, were dependent on their immediate hinterlands for their (usually highly localized) market and industrial functions, where these existed at all, as well as for the foodstuffs on which the urban populace lived. As the society of the empire evolved away from the relationships and conditions which gave rise to and maintained these urban structures, so the cities became the first key institution of the classical world to feel the effects of these changes.

The form which these changes took are complex, but mirror the effects of a growing tension between state, cities and private landowners to extract surpluses from the producers, and the failure of the cities to weather the contradictions between their municipal independence on the one hand, and on the other the demands of the state and the vested interests of the wealthier civic landowners. While many cities were able to maintain themselves and their fiscal role well into the first half of the seventh century in the east, it is clear already by the later fourth century that many did or could not. There were regional variations, but as a result, and over the period from the later fourth to the later fifth century (in the west until the empire disappears as well as in the east), the state had to intervene increasingly to ensure the extraction of revenues, so that the burden of fiscal accountability had been considerably reduced, if not removed entirely, during the reign of Anastasius (491–518). This may even have promoted the brief renaissance in urban fortunes which took place in some eastern cities in the sixth century, but it did not re-establish their traditional independence and fiscal responsibilities.

The physical structure of cities was transformed over the course of the later fifth and sixth centuries, and archaeological evidence has revealed an almost universal tendency for cities to lose by neglect many of the features familiar from their classical structure. Major public buildings fall into disrepair, systems of water supply are often abandoned (suggesting a drop-in population), rubbish is dumped in abandoned buildings, major thoroughfares and public spaces are built on, and so on. These changes may not necessarily have involved any substantial reduction in economic or exchange activity in cities, of course. On the other hand, the undoubted decline in the maintenance of public structures or amenities—baths, aqueducts, drains, street surfaces, walls—is suggestive of a major shift in the modes of urban living: of both the object of the investment of wealth, and of finance and administration in particular. And from the middle of the seventh well into the ninth century the only evidence for building activity associated with provincial urban contexts concerns fortification work and the construction or repair of churches or buildings associated with monastic centres.

By the early years of the seventh century all the evidence suggests that cities as corporate bodies were simply less well-off than they had been before about the middle of the sixth century. There may have been as much wealth circulating in urban environments as before, with the difference that the city as an institution had only very limited access to it, having lost their lands and the income from those lands. During the later sixth century in particular the local wealthy tended to invest their wealth in religious buildings or related objects (so that there was an evolving pattern of investment as much as there was a decline). In addition, the church was from the fourth century a competitor with the city for the consumption of resources. And however much their citizens might donate, individually or collectively, this can hardly have compensated for this loss. Indeed, such contributions became the main source of independent income for many cities. The archaeological data suggests a shrinkage of the occupied area of many cities during the sixth century, and even an increasing localization of exchange activity; but again, this does not have to mean a change in their role as local centres of such exchange.

The survival of urban settlements during and after the Arab invasions—thus from the 640s until the 750s—owed much to the fact that they might occupy defensible sites, as well as be centres of military or ecclesiastical administration. But endemic warfare and insecurity, economic dislocation and social change meant that the great majority played a role peripheral to, and derived from, the economic and social life of the countryside, and reflected if anything the needs of state and church. The invasions of the seventh century dealt what was simply the final blow to an institution that was already in the process of long-term transformation.

Fortifications serve several purposes: to protect populations and/or soldiers and their supplies, equipment and armaments, to act as refuges for civilian populations in times of need, and to provide safe bases for soldiers from which to protect the surrounding countryside or a particular route or crossroads of strategic value, as well as to serve as a deterrent to hostile attack and as defended watch-posts to warn of invasion and perhaps to delay the enemy advance, or to function as bases from which raids or attacks against enemy installations might also be mounted. Each of these functions demands different sorts of defensive works, of course, depending upon size, location, availability of supplies of food and water, proximity to similar defensive structures, the possibilities of relief when attacked, and so forth. The Roman state had a long and sophisticated tradition of fortification, and this was inherited without a break by its medieval East Roman successor.

During the period from the third to the sixth century the Roman world saw a generalized tendency to provide settlements of all sizes with walls and some form of defensive perimeter where there had hitherto been no such defences, a reflection both of a real threat in those areas most affected by external attack, and a changing set of assumptions about what a “city” should look like. In many exposed areas a move from a lowland site to a more defensible situation nearby, or the re-use of older pre-Roman hilltop fortified sites takes place, and although there are a number of reasons for this gradual process in the late Roman period, it increases very dramatically during the later fourth and fifth centuries in the Balkans as a result of the constant threat from Germanic and steppe nomadic barbarians, and again during the seventh century in Anatolia in response to the effects of the Persian and then particularly the Arab invasions and raids. But the contrast between the late ancient polis and the middle Byzantine kastron should not be exaggerated: of the large number of settled sites which can clearly be differentiated from undefended rural settlements, only a small proportion bore the official or unofficial characteristics of a polis in the classical sense. A far larger number were characterized already in the fourth and fifth centuries, and especially in the sixth century, by features normally identified archaeologically and topographically as characteristic of defended centres of population with administrative and military functions, exactly the same, in fact, as the later Byzantine kastron. The transformations which occurred did not, except in a relatively small number of cases, involve a universal abandonment of formerly urban sites (poleis) in favour of hilltop fortified sites (kastra). Rather, it involved a change in the way populations were distributed between such sites, their extent and how they were occupied.

With a handful of exceptions, such as Nicaea, Constantinople and Thessaloniki, most of the major classical cities shrank during the seventh century to the size of their defended citadels, even though the “lower city” of such towns—the main late Roman inhabited area—may have been in many cases still the site of smaller communities. Archaeological surveys suggest that Ancyra shrank to a small citadel during the 650s and 660s, the fortress occupying an area of 350×150 metres, the occupied upper town in which it was situated occupying an area not much larger; Amorion, which supposedly had a vast perimeter wall, was defended successfully in 716 by 800 men against an attacking army more than ten times larger, the area of the kastron occupying some 450×300 metres. The latter survey has also shown that, while the classical/late Roman site was indeed very extensive, with an impressive wall and towers, the occupied medieval areas were thus similar to those of Ancyra. Amastris, mod. Amasra, offers similar evidence, as does Kotyaion, mod. Kütahya, and there are many more formerly major centres which underwent a similar transformation. In some Byzantine texts, mostly hagiographical, there occur descriptions of “cities” with populations inhabiting the lower town. Excavations at Amorion and several other sites show that while the very small fortress-citadel continued to be defended and occupied, discrete areas within the late Roman walls also continued to be inhabited, often centred around a church. In Amorion there were at least two and probably three such areas. Small but distinct communities thus continued to exist within the city walls, while the citadel or kastron—which kept the name of the ancient polis— provided a refuge in case of attack. Many cities of the seventh to ninth centuries survived because their inhabitants, living effectively in separate communities or villages within the walls, saw themselves as belonging to the polis itself. In some cases, the walls of the lower town area were maintained—irregularly, for the most part—in order to provide shelter for larger than usual concentrations of troops. This may have been the case at Amorion, for example. Together with the large number of much smaller garrison forts and outposts of a purely military nature (although sometimes associated with village settlements nearby or below them), such provincial kastra (which were also called, confusingly, poleis by their inhabitants and by many writers who mention them) and frontier fortresses, generally sited on rocky outcrops and prominences, often also the sites of pre-Roman fortresses, typified the East Roman provincial countryside well into the Seljuk period and beyond, and determined the pattern of development of urban centres when they were able to expand once more during the tenth and eleventh centuries.

There is in the development of late Roman fortification a move from passive, linear defences sufficient to repel relatively primitive, barbarian attackers, to more complex, active defensive arrangements, with large numbers of towers providing intersecting fields of fire and complex gate arrangements. Byzantine fortresses after the seventh century generally involved combinations of protruding towers, angled gates, sometimes including a tower-fortress integrated into an inner curtain wall. The notion of a central stronghold that could continue to resist the enemy after the curtain had fallen and the “lower” defences were taken can be traced back to the Hellenistic period at least in some Anatolian fortresses, and was reflected both in the reoccupation and refortification of many ancient citadels and acropoleis within, or attached to, cities of the Roman period as well as in the construction of tower-fortresses where a natural defensive height was not available (as at Nicaea, for example). The Norman and western keep represents the same idea, given added stimulus in respect of technique and materials, especially in the use of lime mortar, by the Crusaders’ experiences in the Balkans, Asia Minor and Syria-Palestine. With the recovery of the empire’s economic stability from the ninth century on, many urban centres recovered their fortunes, although their physical appearance was very different from that of their late antique predecessors. On the eastern frontier especially the empire constructed a number of major fortified centres serving chiefly as strategic centres and military bases, rather than centres of local population, fortresses which have only recently attracted the attention of archaeologists and architectural historians and which clearly had a major role in both frontier defence and internal security. Such fortifications closely reflected the strategic networks of the regions in which they were established, both in respect of communications and routes of ingress and egress, as well as—depending upon the region—of economic activity and the movement of resources. Fortifications were an integral element of every town and, the recovery of substantial areas in western Asia Minor during the first half of the twelfth century owes much to the policies of Alexios I, John II and Manuel I in utilizing fortress towns as solid bases which, regardless of the frequency or damage caused by the raids of the Turk nomads from the plateau to the east, could control the countryside and maintain imperial political and fiscal authority. Warfare—and the events of the seventh century in particular —had a lasting effect on the pattern and form of concentrated settlement in both the Balkans and Asia Minor, a pattern that was further inflected in Asia Minor especially by the Seljuk invasions and the warfare of the twelfth century and after.



Over the millennium of its existence, the Byzantines faced a vast array of peoples who threatened its territory and people. Several of these proved militarily superior and dealt heavy defeats on the empire. In the end, however, the Byzantines generally gained the upper hand, often through decades or even centuries of defense, stabilization, assimilation, and counterattack. The Byzantines learned a great deal from their enemies; indeed the ability to adapt to the challenges posed by opponents was one of the great pillars of Byzantine military success.


The Gothic tribal confederacies posed the most serious challenge to the late antique Roman state of any Germanic group. The Goths comprised coalitions of tribal groups, mostly from the east Germanic peoples who by the third century A.D. inhabited a vast swathe of territory from the Oder and Vistula rivers to the southern steppes of Russia, the Crimea, and the Carpathian basin. East Germanic peoples had posed a significant threat to the eastern provinces of the Roman state from the third century. In 267, Goths and Heruls burst through the Danubian defenses and ravaged Thrace and much of the Balkans, sacking Athens before the emperor Claudius Gothicus dealt them a stinging defeat in 269. Following their defeat by the Huns, large groups of Goths migrated south to the Danube where they were admitted as suppliants to Roman territory. Their provisioning was bungled due to corruption, and an underdeveloped transportation response led to starvation among the Goths and rebellion that culminated in the armed confrontation at Adrianople. At the end of the sixth century, after its recovery from the Goths, the empire had to concede the loss of most of Italy to the newly arrived Lombard confederation, whose grip on the peninsula spread throughout the seventh century. The Byzantines also entered sporadic conflicts with the Franks from the sixth century and even fought against Charlemagne (801–10) for control of the Istrian and Dalmatian coasts.


The Goths were organized in decimal units with major groupings of “hundreds” (hundafaþs) as were their Germanic relatives, the Anglo-Saxons and their Roman neighbors, whose centurions were well known to the eastern Goths. Gothic mercenaries served in the Roman army throughout the late third century, and by the time of the emperor Constantine, Gothic elements were settled in Transdanubia. By the fourth century Gothic military organization had evolved at least in part under the influence of Roman practice. Gothic tribal raiders crossed into Roman territory and proved a sufficient nuisance to attract the interest of Constantine, who waged multiple campaigns against them. By now the Goths probably included and coexisted alongside elements of several ethnic Iranians (Sarmatians), Slavs, Romano-Dacians, and Getae. The remains of a commonly articulated material culture from the second through fifth centuries (the Chernyakhov culture) indicate broad contact and exchanges; such adaptations were not always peaceful and the transferal of knowledge from one people to another certainly included warfare. According to Maurice, the “fair-haired races,” especially the Lombards, grouped themselves not into numerically ordered units but according to kin group.

Methods of Warfare

The Goths fought as both cavalry and infantry. Until the last few decades, historians have viewed the Goths as primarily a cavalry army and attributed to this their shattering victory over the infantry legions in 378 at Adrianople. Their numbers were probably never as numerous as some Roman authors would have us suppose—Heather estimates that in sixth century Italy and Gaul there were about 15,000 Gothic elite males.1 When the Gothic king Theodoric reigned over the united Gothic territories in Spain, Gaul, and Italy, his Gothic subjects numbered about 200,000 people. However, although we have few contemporary sources, the majority of Goths seem to have often fought as infantry spearmen and swordsmen. Certainly the Goths served in large numbers in the legions as infantry. At Adrianople the Goths had perhaps 5,000 cavalry and probably twice as many infantry. According to Vegetius, the Goths possessed numerous archers, who fought on foot. In the sixth century, Prokopios provided a clearer picture of the Gothic army, which fielded a large cavalry component who fought in massed formations as lancers, while the infantry seem to have been mainly skirmishers armed with javelins and archers. Other infantry fought as spearmen and swordsmen equipped with a spatha and carrying shields. Given the high casualty rate caused by Roman archery among the Goths, it is doubtful that they were more heavily armored than their Roman foes. In fact, the Goths closely resembled their late Roman counterparts.

Byzantine Adaptation

Since after Adrianople the empire was too weak to destroy the Gothic confederacies, the Romans sought to neutralize them by treaty. The emperor Theodosius recruited numerous Goths into the Roman army, as an expedient means to replenish the devastated ranks of the eastern field forces, but also as a way to weaken the Goths, whose presence in the Balkans created a state of emergency. Theodosius recruited numerous Gothic federates who fought loyally for him and of whose lives the Roman high command was apparently none too careful—a contemporary panegyrist acclaims the emperor for using barbarian to fight barbarian, thus bleeding both of them. Nonetheless the Goths formed a sizable but not dominant portion of the eastern field army. By 400 A.D. the Gothic warlord Gainas dominated imperial politics in the capital of Constantinople, but his unpopular policies led to his downfall and a riot of citizens who trapped and massacred 7,000 of his Gothic troops. The Gainas affair marked the apogee of Gothic influence in the imperial center; the Romans countered Germanic elements in the army by recruiting Isaurian highlanders from Asia Minor. Finally, the last major elements not assimilated or settled within the Roman Balkans or Asia Minor were sent to Italy under Theodoric the Amal. Justinian renewed the Gothic conflict, invading Italy and conquering it. In 554 the Roman general defeated a Frankish-Alemanni force at Volturnus through his combined arms approach—horse archery again proved a major Roman tactical advantage over the Frankish infantry force. Though the Byzantines lost most of Italy to the Lombards in the later sixth and early seventh centuries, they created the exarchate of Ravenna with several dukes under its control to check the Lombard advance. The exarch held joint civil and military power and, as viceroy of the emperor, was free to respond to crisis without direct orders from Constantinople. These reorganizations helped the Byzantines maintain territory in portions of Italy until 1071.


The most sophisticated, rich, and militarily threatening power that the Romans faced in the early part of their existence was the empire of Sasanian Persia. Founded after victory in a civil war in 226 A.D., the Sasanian dynasty ruled territory stretching from Central Asia to the Persian Gulf and Mesopotamia. Their propaganda declared dynastic ties to the Achaemenid Persian Empire destroyed by Alexander the Great and consequently the rights to the former Persian territories of Asia Minor, Egypt, and the Mediterranean coast. While the Sasanians acted on these grand claims on only one occasion, during the mighty conflict that raged with Rome in 603–28, clashes over strategic borderlands and satellite peoples were frequent. The frequency and intensity of these conflicts rose from a simmer to a steady boil by the sixth century, culminating in the Persian conquest of most of the Roman east in the following century.


The Sasanian shah Kosrow I (531–79) reformed the Persian military and in doing so created several Roman-style structures. Kosrow divided the empire into four army districts in which he stationed army corps under the command of four spahbeds (field marshals). Along the border, the king established margraves, marzbans, who administered sensitive border districts and commanded the frontier forces stationed there. The Eran-ambaragbed, “minister of the magazines of empire,” was, like his Roman counterpart, the praetorian prefect, in charge of arming and equipping the troops. The general (gund-salar) led individual field armies on campaign; sometimes under the authority of the spahbed. By the sixth century the army was largely professionally recruited and paid; there was a professional infantry commander in charge of standing guard units, but in the sixth century the Persians apparently continued to rely on conscripts for a large portion of their rank-and-file infantry. Mailed cavalry units and the royal guard formed the crack troops of the empire; these were generally drawn from the Persian nobility or from aristocratic allied families, such as the Hephthalites and Armenians, with whom the Persians had close contacts.

Methods of Warfare

The proportion of infantry to cavalry in the Sasanian army is unknown, but the Persians relied to a large degree on heavy horsemen, who could both shoot the bow and strike with heavy lances. The Persians favored direct massed cavalry assaults to break up enemy formations; the shock of their horsemen proved decisive against the Romans on several occasions. Normally the Sasanians drew up their forces in three cavalry lines. The Sasanians occasionally employed elephants in combat, but though they made a great psychological impression, they were not an important part of their military. The left of the Persian line was traditionally manned by left-handed archers and lancers who could thus strike effectively across the face of the enemy formation (right-handed mounted archers especially had difficulty shooting to their right). The left of the host formed the defensive anchor, whose role was to avoid enemy flanking maneuvers and to support the offensive right of the formation, where were stationed the best noble cavalry. The Sasanian right typically tried to outflank the enemy left, though the heavily armed kataphracts, covered from head to toe in mail and bearing lances, could be used in frontal assaults on infantry and cavalry groups. Behind the center line of regular cavalry were stationed the infantry formation, which supported the cavalry and sheltered retreating horsemen in case their attacks failed. In addition to their archery and horsemanship, the Sasanians were outstanding siege engineers. From the fourth through seventh centuries they seized some of the best defended and most powerfully built Roman fortress cities.

Byzantine Adaptation

The Sasanians and Byzantines knew one another well and there was considerable exchange of military knowledge and practice across the frontiers. Militarily, each side came to resemble the other. In early twentieth-century excavations at Dura Europus, a Roman frontier city on the middle Euphrates taken by Sasanian assault in the year 256 archaeologists discovered the remains of at least nineteen Romans and one Sasanian attacker. The Sasanian wore chain mail, carried a jade-hilted sword, and wore a pointed ridge-type helmet with a prominent center piece whose rivets joined the two lobes of the helmet together. Such gear was typical in Roman armies by the third century. In 533 at the battle of Dara, Belisarios countered Sasanian superiority by limiting their cavalry and playing to their psychological sense of superiority. In subsequent battles he used the Sasanians’ wariness of his stratagems to force their withdrawal by aggressive posturing. The Persians, used to the traps and feigned retreats of their nomad enemies, could be made too cautious by aggressive maneuvers. They could be thwarted by the commander’s well-chosen battlefield that cut off the Persians’ ability to place their weaker elements on protective rough ground. The poor soldiers among the Sasanians did not fight with spear and shield, but seem to have been mainly skirmishers and archers. They were therefore susceptible to Roman cavalry charges delivered over level ground. The Romans thus relied on strategems, strategic maneuver, tactical coordination, and discipline to defeat the Persians. When Roman commanders selected the battlefield, they were able to neutralize or defeat these stubborn eastern opponents.


Throughout its existence, the empire confronted a vast array of steppe nomad military powers. The Byzantines fought major wars against the Huns, Bulgars, Avars, Khazars, Hungarians, Pechenegs, and Cumans and numerous minor conflicts with a host of other groups. Nomads were generally bent on plunder of imperial territory and rarely sought to settle on lands south of the Danube, only a small portion of which were suitable for the transient, cattle-herding life of pastoralists. However, both the Huns and Avars posed existential threats to the empire, as they sought to dominate the lands south of the Danube and to destroy the Roman power that contained them north of the river. Nomadic confederations formed under charismatic leadership or during periods of environmental or physical stress. Maurice stressed in the Strategikon that nomads typically fought in kin-based tribal or extended family groupings, and this contributed to the nature of their tactics.


Nomadic society was based on nuclear families and wider, extended kinship ties. Like other tribal societies, blood relation or imagined genealogical connections helped to smooth political dealings and allow for larger groupings or “super tribes” that made massive nomadic military enterprise possible. The Huns under Attila formed an effective monarchy and Maurice stressed that the Avars, unlike many nomads, possessed a kingship. Undoubtedly the power of the central figures within a hierarchy during the Hunnic and Avar episodes of Byzantine history bolstered the barbarians’ military effectiveness. After they settled north of the Danube in the late sixth century, the Avars conquered and coopted elements among the Bulgars, Slavs, and Hunnic and Germanic peoples in Transdanubia. The Byzantines portray a grim fate for those whom the Avars conquered, especially the Slavs who served as hard laborers and pressed soldiers during the siege of Constantinople in 626. According to Maurice, the Avars arranged themselves by tribe or kin group while on the march. Their social structure made them vulnerable to desertions and divisions within the ranks, which the Byzantines sought to exploit.

Methods of Warfare

Steppe nomads fought primarily as lightly armed horse archers. Speed and surprise were cornerstones of their strategic and tactical success. Their ability to swarm and the firepower they brought to bear could break up enemy formations and drive the enemy from the field. In the fourth century, when the Romans had little experience dealing with the tactical swarming attacks, war cries, strange appearance, and mobile horse archery of the Huns, these nomads struck terror into the hearts of many soldiers and won numerous victories across the length of the empire. In addition to horse archers, the Huns and Avars deployed heavier lancers who bore a resemblance to the Sasanian hybrid cavalry, armed with bow, sword, and lance. The Strategikon notes that the Avars carried a lance strapped on their back which freed them to operate their bows. In addition to the lance and bow, Avar warriors carried swords; they seem to have been more heavily armed than their Hun predecessors, as Maurice noted that they wore chain mail coats. The Avars wore long coats of mail or lamellar split at the crotch, with panels on each side to protect the leg. The famous Nagyszentmiklós Treasure includes a gold plate depicting what is probably an Avar or Bulgar warrior wearing such a coiffed mail coat, splinted greaves, helmet, and carrying a pennoned lance.

Byzantine Adaptation

The Byzantines relied on diplomatic means to buy off and deflect Hun designs on imperial territory. The defensive posture of the empire throughout the fifth century precluded decisive confrontations against a superior enemy in the open field, and the massive defenses of Constantinople shielded the eastern territories from Hun penetration and conquest, though most of their European possessions were ravaged and slipped from Byzantine control. Although our sources provide no insight into the exact mechanisms of the adoption of steppe nomad tactics and equipment, the Byzantines recruited Hunnic horse archers into their armies and probably from these and deserters derived the knowledge of horseback archery. By the sixth century, the hybrid horse archer and lancer cavalry among the armies of Justinian were the most important tactical elements within the Roman army. The Byzantines adopted the stirrup from the Avars and this provided Roman cavalry with a more stable fighting platform. Maurice’s Strategikon notes that the thonged Avar lance and Avar-type tents and riding cloaks were also adopted directly from their steppe enemies. Lamellar cavalry armor also became more prominent in the panoply of Roman soldiers in the sixth and seventh centuries and this, too, indicates that the Byzantines borrowed extensively from nomads. The use of the feigned retreat, while known to classical armies, was a common steppe nomad tactic that the Byzantines perfected under steppe influence and employed throughout their history. The adoption of nomadic equipment, tactics, and strategy were among the most important adaptations of the Byzantine army and proved critical to the long-term survival of the empire.



By the time of the rise of Islam in the early seventh century, the Romans possessed extensive military experience with the Arabs. Arab scouts and light troops had served as guides and auxiliaries almost from the beginning of Roman rule in the Near East. By the sixth century, the Roman system of paying subsidies to allied tribal confederations to maintain law and order along the frontier from the Red Sea to the Euphrates was integral to the governance of the eastern provinces. The powerful Christian tribal confederation of Ghassan, which included both settled and tribal elements, largely managed the eastern periphery of the empire, and despite the general hostility of Greek-speaking elites to their Arab allies, these clients were both effective and reliable. Ghassanid auxiliaries defeated their Persian-sponsored counterparts and provided valuable light cavalry raiders and skirmishers to the eastern field armies on campaign in Syria and Mesopotamia. At the Battle of Yarmuk in 636 the Ghassanids fought alongside their Roman masters and though many subsequently converted to Islam and remained in Syria, a sizable group migrated to Roman territory. The Muslim Arab victors at Yarmuk overran the whole of Syria, Mesopotamia, and eventually wrested Egypt, Libya and North Africa from Roman control. Muslim Arab attempts to conquer Constantinople and thereby destroy the remnants of the Roman Empire unfolded in the epochal sieges of the seventh and early eighth centuries in which the empire emerged battered but intact. With the overthrow of the Umayyad dynasty and shift of the locus of Muslim government to Mesopotamia, the threat to the existence of the Roman state diminished, and as the Abbasid caliphate unraveled politically, the Byzantines mounted a sustained counterattack to recover lost territories in the east.


Arab armies of the conquest were organized along tribal lines, though it is uncertain if these were grouped into units of 10–15 soldiers called ‘arifs known from just after the conquests. Muslim Arab armies were recruited mainly from Arabic-speaking family and tribal groupings. But soldiers were also raised from among Byzantine and Sasanian deserters, as well as non-Arab clients (mawali) dependent on regional Arab lords. Larger tribal groups fought under the banners of their tribal sheikhs in army groups of varying strength, usually numbering 2,000–4,000 men. On rare occasions, as at Yarmuk, combined commands could field as many as 30,000 or 40,000 soldiers. In 661, the Battle of Siffin was fought between the Syrian forces under Mu ‘awiya and the Iraqi Arabs led by the Prophet’s cousin and son-in-law ‘Ali, said to have comprised 150,000 and 130,000 men, respectively; these numbers are inconceivable and could probably safely each be reduced by a factor of ten. During the Umayyad era, when the Syrian army provided the main prop to the caliph’s authority, armies of 6,000 Syrian troops are commonly mentioned and these may represent standard field force groupings, not dissimilar in size and equipment from their Byzantine neighbors. In 838, the caliph al-Mu ‘tasim (d. 842) led an army of up to 80,000 men against Amorium, a number that represented a large force and among the largest the Byzantines ever confronted.

Methods of Warfare

Although the commonly held perception of early Muslim armies today is of swift-moving horsemen mounted on Arabian chargers, the armies of the conquest era were mainly infantry forces fighting as spearmen and archers. Arab archery was particularly deadly to both the Byzantine and Persian forces encountered during the first campaigns of the conquest. Early Muslim armies generally lacked heavy cavalry, and they eagerly accepted the Sasanian horse who deserted to their ranks following the initial encounters in Mesopotamia. Infantry continued to form an important part of Arab armies up to the end of their military encounters with the Byzantines. Nikephoros Phokas noted that the Arab raiders who penetrated the Byzantine borderlands included a mix of cavalry and infantry; like their Roman counterparts, the infantry formed a foulkon, a dense mass of infantry spearmen, and supported the cavalry who formed the major offensive wing of Arab armies. Regular Arab cavalry fought primarily as lancers, while missile support was provided by foot archers. The Arabs never mastered horseback archery and instead relied on Turkic troops to provide mobile fire. The light cavalry encountered by the Byzantines in their reconquest of northern Syria and Mesopotamia were Bedouin light horse riding swift Arabian mounts. Nikephoros advised to keep them at bay with archery rather than chase them, since even the best Byzantine horses, encumbered as they were with heavily equipped fighting men, would not be able to catch them and the danger of being cut off and overwhelmed was a persistent peril of pursuit. Well led and generally possessing superior numbers, training, and equipment, the Arab armies of the early medieval period repeatedly exposed Byzantine weaknesses. Decisive engagements nearly always ended with Arab victories; only when the empire recovered somewhat economically and demographically while the caliphate began to fragment did the initiative return to the Romans.

Byzantine Adaptation

Given the asymetrical nature of the encounter between the Byzantines and Arabs after the initial clashes of the early and mid-seventh century, Byzantine commanders responded in the only way they could, via a strategy of defense coupled with limited, punitive raids to keep the enemy from settling in the strategic Anatolian highlands and to maintain the appearance of Byzantine power among the populations of the border lands. Imperial troops, seriously degraded through the loss of many men in the defeats in Syria and Egypt, underpaid, poorly equipped, and scattered throughout the provinces, were scarcely a match for caliphal field armies. The Byzantines often found themselves paying tribute to convince the Arabs not to attack them—a humiliating concession that drained both the fisc and morale. But the sieges of 674–78 and 717–18 revealed that without achieving naval dominance the Arabs had to conquer the Anatolian plateau if they were to achieve their objective of outright conquest of the Christian empire. Yet, due to their organization of the themes, whose armies could shadow and harass Muslim raiding columns and sometimes defeat them, the Romans made penetration of their territory hazardous. Stubborn Byzantine forces, although no match for grand caliphal campaign armies, often held their own against raiding columns and themselves raided exposed regions when Arab field forces were engaged elsewhere. By the tenth century, the centuries of incessant warfare had helped to create a warrior caste among the frontiersmen of the eastern marchlands who would remake the Byzantine army based on their experiences fighting the Arabs. Their combined arms approach and their use of psychological terror, scorched earth, and incremental advancement of imperial territory by sieges marked the apogee of the practice of Byzantine arms in the medieval east.


The Turkic Bulgars appeared in the sixth century, first as a rump of the so-called Old Bulgarian Empire, the Kutrigurs, defeated by Belisarios outside Constantinople in 559, settled north of the Danube and were absorbed by the Avars. Following the collapse of Avar power in the eighth century, new Bulgar arrivals and existing elites in Transdanubia gradually formed the Bulgar khanate, which adopted Slavic language and customs. Given their cultural origins in the Eurasian steppe, it is unsurprising that throughout the medieval period the Bulgarian social elite fought mostly as heavy armed cavalry lancers. Bulgaria formed the most important state to the north of the empire. Though there were long stretches of peace between the two peoples and even alliance, Byzantine-Bulgar relations were strained by their fundamental conflicting goals—both empires sought to dominate the Balkans and each considered the presence of the other unacceptable. Thus the Bulgars sought to capture Constantinople or subjugate the Byzantines militarily, while the latter sought to contain or even annex Bulgaria outright.


Initially the Bulgars organized themselves along the lines of most steppe empires, with “inner” and “outer” tribes whose power relationships were articulated through marriage alliances, genealogies, and material exchange. Beneath the outer tribes in the pecking order were subject groups like Slavs, Greeks, and the mélange of Avar, Hunnic, and Germanic remnants that rendered the rich cultural matrix of the Danube basin. The khan stood at the pinnacle of an increasingly sophisticated hierarchy that developed under steppe and Byzantine influence. Senior “inner” nobles, called boilas (often Anglicized as “boyar”), and junior “outer” nobles, bagains, formed the elite of the Bulgar state and provided both the military leadership and elite troops of the khanate. The Bulgars matched their Byzantine foe with a strong hierarchical military organization with the khan in overall command while his leading generals, the tarqan, commanded his administrative regional center and presumably took the center of the battle line as well. The targan’s subordinates included komites (sing. komes), after Byzantine usage, who commanded the wings of the army. The highest-ranking Bulgar nobles were heavily equipped cavalry with barded mounts and relied on heavy household cavalry and lighter armed horse archers as did their steppe nomad ancestors.

Methods of Warfare

The Bulgars employed mass conscription to fill out the ranks for their armies. Fear was the main tool used to compel men to enlist and show up equipped for the occasion. Khan Boris Michael (d. 907) ordered that men who arrived for muster without proper equipment or unprepared for campaign were to be executed, as were those who deserted before or during battle. The rank and file included many Slavs who fought as light infantry, carrying shields and javelins. Bulgar cavalry resembled both their Byzantine enemy and other steppe nomads. The Bulgars were expert in their use of terrain, relying on ambush and surprise in their confrontation with the enemy. They demonstrated a high level of strategic planning, strong discipline, and military cohesion, and on numerous occasions were able to confront and defeat imperial field armies, as they did at Varbica in 811 when they trapped a large force led by the emperor Nikephoros I and destroyed it by hemming the Byzantines against a wooden palisade and surrounding it. The emperor himself was killed and his heir mortally wounded. The Bulgars were intimately acquainted with Byzantine military strategy and tactics and, unlike the fragmented Arab emirates to the east, formed a more unified foe unbowed by the shock of repeated defeats.

Byzantine Adaptation

The Byzantines dealt with the Bulgars via a full range of economic, diplomatic, and military strategies. Trade was limited by treaty to designated zones and monitored by imperial officials. Spies were maintained at the Bulgar court at Pliska; the Bulgar khan Telerig (768–77) tricked the emperor into revealing the identity of Byzantine agents among the Bulgars by the ruse of his promised defection, then slaughtered those in the pay of the empire. Byzantine failures against the Bulgars were often due to weakness in strategic and battlefield intelligence that resulted in the surprise of imperial field forces. Experienced and cautious commanders found warfare in Bulgaria perilous. Thus, in the ongoing dispute over control of lands in Thrace and Mesembria on the Black Sea coast, the emperor Nikephoros II Phokas mounted a brief campaign in which he found the the Bulgars’ skillful use of the mountainous terrain and difficulties of supply and communication hard to overcome. Nikephoros therefore induced Sviatoslav I of Kiev to invade Bulgaria; the Rus’ captured scores of Bulgarian towns and fortresses and overwhelmed Bulgar resistance, which led to a direct confrontation between the Rus’ and their new Bulgar subjects and Byzantium. John I Tzimiskes’s defeat of the Rus’ at Dorostolon in 971 opened the way for Byzantine annexation of Bulgaria. The subjugation of Bulgaria took decades, however, with persistent and arduous campaigning by the emperor Basil II, who reduced each quarter of the Bulgar state through sieges and attrition, finally grinding down Bulgar resistance. Bulgaria provided another test for Byzantine strategies of attritive warfare: imperial forces used sieges, scorched earth, and incremental capture-and-hold methods to gradually expand their bases of operations and finally wear out a formidable, skillful, and disciplined opponent. Although the empire possessed a dominant position in Bulgaria by the death of Basil II in 1025, serious resistance continued to the death of the Bulgarian tsar Peter II in 1041. Byzantine control of Bulgaria, won over decades of bitter warfare, lasted for nearly a century and a half.


The Normans arrived in the Byzantine world not as enemies, but as valued mercenaries esteemed for their martial prowess. The settlement of Scandinavian raiders created the duchy of Normandy, when the region was ceded to their war leader Rollo (d. ca. 931) by the Carolingian king Charles the Simple (898–922). Rollo’s descendants mingled with the local French population to create the Normans, a people thoroughly Christian, doggedly militaristic, and unfailingly expansionistic. Norman soldiers entered Italy around the start of the eleventh century where they served as mercenaries for various Lombard princes. By the 1050s large numbers of “Franks,” as the Byzantines called them, had served as mercenaries in Byzantine armies from Syria to Bulgaria, and Normans served as part of the standing garrison of Asia Minor. In the 1040s the Normans began the conquest of south Italy, establishing several counties in the south and finally invading and conquering Sicily from the petty Muslim dynasts there by 1091. Since the late 1050s the Normans had challenged Roman interests in Italy and Robert Guiscard led a Norman invasion of the Byzantine Balkans in 1081. In the ensuing conflict the Normans defeated Alexios I Komnenos, who expelled them only with great difficulty. Two more major Norman invasions followed over the next century, and the Norman kingdom of Sicily remained a threat to imperial ambitions in the west and to the imperial core until the Hauteville Norman dynasty failed in 1194. By this time all hope for the Byzantine recovery of south Italy and Sicily had vanished, thanks to Norman power.


The Normans served under captains who rose to prominence due to birth or their fortunes in war. Minor nobility like Tancred of Hauteville, who founded the dynasty that would conquer much of Italy and Sicily, was a minor baron in Normandy and probably the descendant of Scandinavian settlers. The warriors who carved out territory within Byzantine Anatolia seem to have been either petty aristocrats or simply successful soldiers. One such Norman was Hervé Frankopoulos, who in 1057 led 300 Franks east in search of plunder and territory. After initial successes around Lake Van, he was delivered to the emperor and eventually pardoned. Thus, Norman companies were of no fixed numbers, and it seems that each baron recruited men according to his wealth and status. Norman lords in Italy raised the core of their army from men to whom they distributed lands and wealth in exchange for permanent military service. Lords were required to provide fixed numbers of troops, either knights or infantry sergeants. Other Normans served for pay and plunder, including conquered lands to be distributed after successful occupation of enemy territory. The Normans that the Byzantines encountered were a fluid group—some fought for the empire and then against it; their interests were pay and personal advancement rather than any particular ethnic allegiance. In this the Normans who warred against the Byzantines resembled the later free companies of the late medieval period—variable in numbers, generally following a capable, experienced, and charismatic commander, and exceptionally opportunistic. As a warlord’s success grew, so did his resources. Thus Robert Guiscard rose from the leader of a band of Norman robbers to be Count and then Duke of Apulia and Calabria; in 1084, following his defeat of Alexios at Dyrrachium, Guiscard marched on Rome with thousands of infantry and more than 2,000 knights, a far cry from the scores or hundreds with which he began his career.

Methods of Warfare

The bulk of the Norman fighting forces were infantry, but they formed a largely defensive force that operated in support of the cavalry. Norman infantry fought generally as spearmen—the Bayeux Tapestry shows many Normans on foot wearing the nasal helm and mail hauberks, but it is unlikely that the majority were so armed. Most were probably unarmored and relied on shields for protection like most of their counterparts throughout Europe. Light infantry archers fought with little or no armor, and missile troops played a role in their Balkan campaigns as well—the Byzantine commander George Palaiologos suffered an arrow wound to his head in battle at Dyrrachium in 1082, but generally the Byzantines relied on superior Turkish archery in order to unhorse the Normans and immobilize the knights. Norman knights wore heavy mail hauberks and mail chausses with in-pointed mail foot guards, which Anna Komnene noted slowed the Norman cavalry down when they were unhorsed. These mounted men carried lances and swords. The weight of their mail made them relatively safe from the archery of the day. Norman knights usually decided the course of battle; it was the shock cavalry charge delivered by the Norman knight that delivered victory in battle after battle. Unlike the Turks and Pechenegs with whom the empire regularly contended and whose weaponry was lighter and who relied on mobility, hit-and-run tactics, and feigned retreat, the Normans preferred close combat. They fought in dense, well-ordered ranks and exhibited exemplary discipline. In an era when infantry were generally of questionable quality, most foot soldiers throughout Europe and the Middle East could not stare down a Norman frontal cavalry charge. Norman horsemen punched holes in opposing formations and spread panic and disorder that their supporting troops exploited. By the end of the eleventh century, Norman prowess on the battlefield yielded them possessions from Syria to Scotland.

Byzantine Adaptation

The Byzantines avidly recruited Normans into their armies. Though critics have unfairly blamed the medieval Romans for not adapting their warfare in light of the new western techniques and technologies to which they were exposed, fully equipped and well-trained kataphraktoi could match the skill and shock power of the Norman knight. What the Byzantines of the Komnenoi era lacked were the disciplined heavy infantry of the Macedonian period and combined arms approach of mounted and dismounted archery that could blunt enemy attack and cover infantry and cavalry tactical operations. Alexios I relied on Turkish and steppe nomad auxiliaries and patchwork field armies assembled from mercenaries drawn from the empire’s neighbors. As with other intractable foes, the Byzantines relied on a combination of defense and offense—the Normans were contained in the Balkans allowing space for an imperial recovery and the time to muster new forces following the heavy defeat late in 1081 of the Roman army at Dyrrachium on the Adriatic. Alexios allied with southern Italian nobles and the German emperor Henry IV (1084–1105) who menaced the Norman flanks. The death of Robert Guiscard in 1085 removed the most serious threat to Byzantine rule since the seventh century, but Guiscard’s son, the redoubtable Bohemund, renewed war against the empire in 1107–8. Alexios had learned from his twenty years of dealing with the Norman adversary and returned to the traditional Byzantine strategies of defense, containment, and attrition. The Byzantines relied on their Venetian allies to provide naval squadrons on the Adriatic that interfered with Norman shipping and resupply, and Alexios’s forces blocked the passes around Dyrrachium; the emperor forbade his commanders to engage in a large-scale confrontation with the Normans. In the skirmishes and running battles against Norman scouting and foraging parties Byzantine archers shot the enemy mounts from beneath their riders and then cut down the beleaguered knights. Hunger, disease, and lack of money undid Bohemund, who was forced to sign a humiliating treaty and return to Italy. Thus the ages-old Byzantine principles of indirect warfare proved triumphant against a stubborn and superior enemy.

Constantinople and Her Navy

Perhaps no defensive structure summarizes the truth of siege warfare in the ancient and medieval world as clearly as the walls of Constantinople. The city lived under siege for almost all its life; its defences reflected the deepest character and history of the place, its mixture of confidence and fatalism, divine inspiration and practical skill, longevity and conservatism. Like the city itself, the walls were always there, and for anyone in the eastern Mediterranean, it was assumed they always would be. The structure of the defences was mature in the fifth century and changed little thereafter; the building techniques were conservative, harking back to practices of the Greeks and Romans. They had no particular reason to evolve because siege warfare itself remained static. The basic techniques and equipment – blockade, mining and escalade, the use of battering rams, catapults, towers, tunnels and ladders – these were largely unchanging for longer than anyone could recall. The advantage always lay with the defender; in the case of Constantinople its coastal position increased that weighting. None of the armies camped before the land walls had ever succeeded in effecting an entry through the multiple defensive layers, while the city always took prudent measures as a matter of state policy to keep its cisterns brimming and its granaries full. The Avars came with an impressive array of stone-throwing machinery but their looping trajectory made them far too puny to breach the walls. The Arabs froze to death in the cold. The Bulgar Khan Krum tried magic – he performed human sacrifices and sprinkled his troops with seawater. Even its enemies came to believe that Constantinople was under divine protection. Only the Byzantines themselves were ever successful in taking their own city from the land, and always by treachery: the messy final centuries of civil war produced a handful of instances where gates were flung open at night, usually with inside help.

There were just two places where the land wall could be considered potentially weak. In the central section the ground sloped down a long valley to the Lycus River and then up the other side. As the wall followed the downward slope, its towers no longer commanded the high ground and were effectively below the level occupied by a besieging army on the hill beyond. Furthermore the river itself, which was ducted into the city through a culvert, made it impossible to dig a deep moat at this point. Nearly all besieging armies had identified this area as vulnerable, and though none had succeeded, it provided attackers with a vestige of hope. A second anomaly in the defences existed at the northern end. The regular procession of the triple wall was suddenly interrupted as it approached the Golden Horn. The line took an abrupt right-angle turn outwards to include an extra bulge of land; for 400 yards, until it reached the water, the wall became a patchwork structure of different-shaped bastions and sectors, which, though stoutly built on a rocky outcrop, was largely only one line deep and for much of its length unmoated. This was a later addition undertaken to include the sacred shrine of the Virgin at Blachernae. Originally the church had been outside the walls. With a typical Byzantine logic it had been held initially that the protection of the Virgin was sufficient to safeguard the church. After the Avars nearly burned it in 626 – the shrine was saved by the Virgin herself – the line of the wall was altered to include the church, and the palace of Blachernae was also built in this small bight of land. Both these perceived weak spots had been keenly appraised by Mehmet when he reconnoitred in the summer of 1452. The right-angle turn where the two walls joined was to receive particular attention.

As they patched up their walls under Giustiniani’s direction and paraded the sacred icons on the ramparts, the people of the city could be pardoned for expressing confidence in their protective powers. Immutable, forbidding and indestructible, they had proved time and again that a small force could keep a huge army at bay until its willpower collapsed under the logistical burden of siege, or dysentery or the disaffection of the men. If the walls were decayed in places, they were still basically sound. Brocquière found even the vulnerable right angle to be protected by ‘a good and high wall’ when he came in the 1430s. The defenders however were unaware that they were preparing for conflict on the cusp of a technological revolution that would profoundly change the rules of siege warfare.

‘Navigantium fortitudo mihi soli inest’, is a remark made by Nicephorus II (as reported by Liutprand). Nicephorus could make this boast in truth, since the emperors of High Byzantium succeeded gradually in building up a fleet of such power as to check the depredations of Arab pirates almost completely in the eastern Mediterranean. During the latter part of the eleventh century, however, the Venetians and the Genoans gradually caught up with the Byzantine marine power and, despite the strenuous efforts of the Comnenian emperors to increase their naval forces, these represented the stronger force by the death of Manuel I, while the Byzantine naval presence after the death of Michael VIII was derisory and in time it vanished altogether, so that John VIII had to make his way to the Council of Ferrara-Florence by hired craft. The fleet of the naval era was divided into two main sections: the Imperial fleet and the fleet of the themes. The former was organized into two divisions, one for the personal use of the Emperor and Empress, and for the defence of the capital, the other for use on regular military expeditions and for policing the seas against pirates.

The fleet of the themes was kept up at the charge of various maritime themes, particularly those of the Greek islands (Aegea, Samos, Cephalonia), Greece and the Cibirriote theme in Asia Minor. The regular servicemen from these themes were paid in feudal land, as were the land forces in the army of the themes. An alteration was however made in the reign of Manuel I, whereby the monies expended by the themes on the upkeep of the fleet were diverted straight into the Imperial treasury, and the Emperor assumed the direct responsibility for the maintenance of the whole naval service. This was probably intended as an assurance for the better order of the ships, but, as might have been foreseen, it proved a disaster, as the money was repeatedly spent on wasteful civil-service projects, while the navy was starved of even necessary funds.

The fleet often employed foreign mercenaries, and Russian or Varangians who entered the Imperial service often began their time in the navy, this being a form of service which would have suited the temper of the Norse seamen. From what is recorded of Haraldr Siguroarson we may deduce that his first period of Varangian service will have been spent thus. The strategos of each maritime theme commanded his section of the thematic fleet, while the supreme commander was the commander of the Imperial fleet, who was titled in the High Byzantine era the Droungarios, and was of the rank of patrician. This official appears to have been known in the reign of Alexius I as the Grand Duke (Megas Doux), and his deputy as the Thalassokrator, while later still Pseudo-Codinus refers to an Admiral of the (by then insignificant) Fleet. These supreme commanders had other officers under them, and officers of the Hetairia were set to command the foreign naval mercenaries. In the tenth century 77 ships constituted the thematic fleet against 100 in the Imperial fleet, while the force manning the latter was 23,000-24,000 strong, against 17,500 in the former.

The capital ships of the Byzantine fleet were the dromoi, which differed considerably in size, but were built on the same pattern, with a wooden castle (xylokastron) on the deck, and carrying various military engines. In the bows was a figurehead of gilt bronze, usually the shape of the head of a wild beast, the lion being a popular motif, in which were housed the siphon and pumping mechanism to spray out the Greek Fire, the terrible Byzantine secret weapon which burned alike on land and water. This substance was also carried in fragile bowls or spheres of glass, which could be hurled on to the enemy ships and which then set everything that the stuff touched ablaze. The rowers were arranged in two banks, with a normal complement of 25 to each row; there were also on average some 50 soldiers on each dromos. It is calculated that there will have been around 220 persons to the full complement on a capital ship, or even more, since the account of the Cretan expedition of 902 refers to 230 oarsmen and 70 others, or in all a crew of 300 on each dromos. The Chelandia were smaller vessels, one class of which were named Pamphyloi’, they were often manned by foreign mercenaries, and their complement would be 130-160 men. Finally there were the light supporting vessels, the so-called ousiai, on which Varangians were frequently employed; these were swift and easily manoeuvred ships, which were especially useful for coastguard duty or for chasing and overtaking pirate vessels. The Taktnkca speaks of 50-60 soldiers forming the complement of each of these ships, and their total crew will there- fore have been c. 110 strong. On formal expeditions two ousiai generally accompanied each major vessel.

The commander of each dromos bore the title of Kentarchos, while over each division of 3-5 capital ships there was placed a homes? though the titles komes and droungarios are later used without discrimination of the captains of single ships. The fleet had its banner, the sign being a cross surrounded by four fire-siphons.

It appears that admission to the Imperial fleet, and especially appointment to one of the ships based on the capital, or in the personal service of the Emperor and his court, was very much sought after by personnel in the other divisions of the Byzantine navy. As the pay was higher, and the serving personnel could more easily obtain high court distinctions in these ships, this is understandable, and it is very likely that it was necessary to purchase such appointments in the same way as ones in the land Hetairia. It is, however, even more difficult to calculate the naval rates of pay than those of the land forces, though some inkling may be derived from the above mentioned narratives of the two naval expeditions to Crete. The pay of the Russians and Varangians in the sea-forces will certainly have been far smaller than that of those in the Hetairia. If it is true, however, that the commanders of the coastal protection vessels were entitled to keep a considerable proportion of the goods confiscated from pirate vessels, then this could obviously make a very sizeable difference to their emoluments. It is noted in Haraldar saga Siguroarsonar that he was to pay the Emperor 100 marks for every pirate vessel that he was able to capture, but could keep the rest for himself and his men. This could obviously be a very valuable source of income.

Warfare Frankish Greece 1204-1380 I

Frankish Greece, 1204–61

The Fourth Crusade (1202-04), consisting of a land army composed of French and Italian troops and a powerful Venetian naval fleet, had originally been planned as an offensive against Egypt. Through a combination of greed, political intrigue and mutual distrust, the expedition ended up attacking the Greeks instead. After the Crusade captured Constantinople in 1204, those members of the expedition who chose to stay in Greece undertook the conquest of what remained of the Byzantine empire. This resulted in the creation of several new Latin states around the Aegean. The Latin empire itself encompassed Thrace and the northern fringes of Asia Minor, while Macedonia and parts of Thessaly formed the kingdom of Thessaloniki. Further south, Byzantine Attica and Boeotia were turned into the Frankish duchy of Athens, and below it the principality of Achaia covered the Peloponnese. Meanwhile the Venetians established a number of important trading colonies around the Aegean, particularly at Modon, Coron, Constantinople and Crete. They also dominated numerous islands which lay either directly within the Venetian sphere of influence or were colonised by individual Venetian citizens. These included Corfu, Cephalonia and Euboea, as well as the Cyclades, which formed the duchy of the Archipelago under the Sañudo dukes of Naxos. Eventually other islands came to be controlled by rival powers, most notably the Genoese, who held Chios and several neighbouring trading posts between the fourteenth and sixteenth centuries. In addition, from 1306 onwards the Hospitallers subjugated Rhodes.

Leading Greeks who had fled from Constantinople during the Fourth Crusade also created three new states situated on the fringes of the Byzantine world. The capital cities of these states lay at Trebizond along the Black Sea, at Nicaea in Asia Minor and at Arta in Epiros. All three of these territories, and in particular the latter two, saw themselves as the natural heirs to the Byzantine empire and consequently often clashed both with each other and with their unwelcome new Latin neighbours. To the north of Constantinople, the Franks were also threatened by the Bulgars, who had been opponents of the Byzantine empire before 1204 and now continued to launch frequent attacks into northern Greece in conjunction with an aggressive tribe of horsemen known as the Cumans.

This highly splintered political situation inevitably led to almost constant fighting in the Aegean region and ultimately resulted in the destruction of virtually all the Latin states which had been set up after 1204. The first to disappear was the ephemeral kingdom of Thessaly, whose conquest by the despots of Epiros was effectively completed with the fall of Thessaloniki in 1224. At this time the Epirote Greeks also made some significant advances against the duchy of Athens to their south. Meanwhile, the Greeks of Nicaea made good progress in the east. By the end of the 1220s the Franks had lost their possessions in Asia Minor and the Latin empire itself had been reduced to the area immediately around Constantinople, leaving the rulers of Nicaea and Epiros to fight it out for control of northern Greece. Indeed, as the Nicaean Greeks crossed the Sea of Marmara and began to advance westwards, those of Epiros became so alarmed that they formed an alliance with their former enemies, the Franks of Achaia. In 1259 this alliance was ended by the Nicaean ruler Michael VIII Palaiologos (1259-82), who inflicted a crushing defeat on William II of Achaia (1246-78) at the battle of Pelagonia in southern Macedonia. As a result of this encounter Epirote power in northern Greece was weakened, and Prince William found himself in captivity until 1262, when Michael finally released him in exchange for the Achaian fortresses of Mistra, Old Mania and Monemvasia. These powerful castles were located in the south-eastern corner of the Peloponnese and provided Michael and his successors with a perfect bridgehead from which to reconquer Achaia from the Franks. This process continued sporadically until the 1420, when the last Latin remnants of the principality were swallowed up and incorporated into the Greek province of Mistra.

Two years after the battle of Pelagonia, the Nicaean Greeks also managed to recapture Constantinople, thereby wiping out the Latin empire. This made them masters of a newly recreated Byzantine empire stretching from western Asia Minor to Thessaly, and enabled Michael VIII and his son Andronikos II (1282-1328) to apply further pressure on the despotate of Epiros, the principality of Achaia and the duchy of Athens. While Franks in the Peloponnese struggled to contain attacks from Mistra, the 1270s witnessed the temporary conquest of Euboea by Licario of Karystos, a Veronese adventurer in Michael VII’s employ. Meanwhile, the duchy of Athens came under attack from the north, and in order to halt these advances and to strengthen his position in relation to other Greek lords in Epiros and Thessaly, Duke Gautier I of Athens (1308-11) eventually decided to seek external assistance. This led to the arrival of a ferocious band of Catalan mercenaries in central Greece. These soldiers of fortune had previously been used by Andronikos II to fight the Turks in Asia Minor, but after falling out with their former employer they had gradually moved west through the Byzantine empire, pillaging and looting as they went, before they eventually found themselves in the pay of Duke Gautier. However, after some initial successes an argument broke out between the duke and the volatile Catalans over pay, and as a result they took to the field against Gautier and his Achaian allies. This decisive encounter, traditionally known as the battle of Cephissus but actually fought at Halmyros (southern Thessaly), took place in March 1311 and led to the death of the unfortunate Gautier along with virtually all of his Athenian knights. In its wake Gautier’s lands were conquered by the Catalans, who then colonised and ruled the duchy of Athens until the late fourteenth century. During this period the Catalans acknowledged the overlordship of the kings of Aragon, who held the ducal title and ruled Athens through representatives whom they regularly sent out to Greece.

The dramatic advances made by the Greeks against the Latins had an important effect on the internal history of Achaia. Originally the Frankish rulers of this crusader state had recognised the Latin emperors as their overlords, but after Michael VIII recaptured Constantinople in 1261, and gained control over Mistra, Old Mania and Monemvasia the following year, it became essential for the Villehardouin princes of Achaia to find a powerful new suzerain who could aid them against the Greeks. During the 1260s, therefore, William II of Villehardouin (1246-78) allied himself with Charles of Anjou, brother of Louis IX of France and ruler of Naples (1266-85). As part of this alliance it was agreed that, if William did not produce a male heir, Achaia would eventually pass under Charles’s control, which was exactly what happened following William’s death in 1278. These events turned out to be a mixed blessing for the Peoloponnese, for on the one hand Charles and his Angevin descendants did provide sporadic military support for the Peloponnese (as William had hoped), but on the other relatively few of them ever visited the region, preferring instead to send royal representatives from Italy or to grant parts of Achaia to their own followers. Perhaps the most important such figure was the wealthy Florentine lord Niccolo Acciajuoli, who acquired large parts of northern Morea, including the castellany of Athens, between the 1330s and 1350s. Toward the end of the fourteenth century Niccolo’s descendants also made significant advances against the Catalans of Athens, who were detested in France and Italy because of their close links with the Aragonese, arch-enemies of the Angevins and the Avignon papacy.

The fact that the Angevin rulers of Achaia resided in Naples rather than Greece gradually had the effect of weakening central authority in the Peloponnese, so that the region’s late fourteenth- and early fifteenth-century history was reduced to a series of internal clashes between various local factions. In the north the Florentine lords of Corinth, allied with a new company of Navarrese mercenaries who had arrived in the 1380s, fought against the Catalans. In the south east the Greeks of Mistra continued to advance, whilst Achaia itself sporadically found itself disputed between rival Angevin claimants. As the Turkish threat grew, the Venetians and the Hospitallers also became more involved in the politics of Latin Greece, sometimes placing entire areas under their own protection. Indeed, from the 1390s onwards, local Christian squabbles became little more than an irrelevance as the Ottoman Turks began to overrun the entire region regardless of whether it was controlled by Greeks, Italians, Franks or Iberians. By 1460 the Byzantine empire had disappeared and most of the Greek mainland had been incorporated into the Ottoman empire. During the next two centuries all remaining Christian islands in the eastern Mediterranean also fell to the Turks.

During this period the military strength of the Latins and their opponents varied considerably. The relatively short life span of the kingdom of Thessaly and the Latin empire, for example, indicates that here western settlers found themselves under extreme pressure from the very beginning. It has been estimated that the total number of cavalry, including both knights and mounted sergeants, available to the first Latin emperor, Baldwin of Flanders (1204-5), only came to between 500 and 1000 horsemen. Contemporary accounts bear this out, for in late 1204 a mere 120 knights left Constantinople, crossed the Bosphorus and began to conquer as many lands as they could from the Greeks of Nicaea. Similarly, Baldwin’s successor Henry (1206-16) campaigned against the Bulgars with a total of 400 knights in 1206, and was accompanied by 260 knights when he advanced against the Nicaean Greeks in 1211.  Initially these figures do not seem small when compared with Greek forces, for the largest reliable total given for a contemporary Nicaean army is that of 2000 cavalry (including 800 Latin mercenaries) at the battle of Antioch on the Meander, an encounter with the Seljuk Turks which took place in 1211.  It should also be noted that the advances of 1204 and Henry’s campaign of 1211 both ended in Frankish success, suggesting that Latin and Nicaean forces were evenly balanced at this time, particularly when the Greeks were also struggling with their Seljuk neighbours to the east.

To some extent the same may have applied to the Cuman and Bulgar troops who threatened the northern border of the Latin empire and the kingdom of Thessaly. In 1208, for example, a mere 2000 Latins allegedly confronted 33,000 Bulgars in northern Greece, but the fact that this encounter ended in a Frankish victory certainly casts doubt on the latter figure and suggests that the Bulgars could not possibly have enjoyed such an overwhelming numerical superiority. Similarly, the combined armies of Nicaea and Bulgaria which reached the walls of Constantinople in 1235 must have been smaller than western sources imply, otherwise they would have annihilated the 140 Latin knights who emerged from the city and successfully drove them off under the leadership of John of Brienne. Here we are confronted with the usual problem of trying to establish accurate troop numbers from medieval chroniclers who tended to exaggerate (or simply invent) the size of hostile armies. Whilst the number of western knights mentioned by these sources appears to be realistic, they also fail to provide us with any detailed information about the quality and quantity of other troops such as sergeants, infantry and archers, even though these soldiers presumably made up the bulk of Latin armies. We can only assume that an apparently miraculous victory inflicted on tens of thousands of opponents by a few hundred western knights actually represented a more equally matched contest between a Frankish army which was really somewhat larger and a Greek, Bulgar or Cuman force which must have been considerably smaller.

Warfare Frankish Greece 1204-1380 II

Battle of Halmyros, a battle between the lightly armed but battle-hardened Almogavars of the infamous Catalan Company and the French Knights of the Duchy of Athens.  By Darren Tan

It still appears that Latin armies in northern Greece were generally outnumbered, and certainly too thinly spread out to defend both this region and the frontier with Nicaea at the same time. When in 1205 Kalojan of Bulgaria (1197-1207) invaded Frankish territories and encouraged the Greek population of Thrace to rebel, the Latin emperor Baldwin had to persuade his brother Henry to abandon ‘all that he had conquered’ in Asia Minor to cross the Bosphoros and help to defend the empire’s European borders. Even these drastic measures failed to prevent Kalojan and his Greek allies from occupying virtually all of Thrace. Consequently, even if we dismiss Villehardouin’s exaggerated claim that there were an additional 14,000 Cumans alongside all the other troops in Kalojan’s army, we must surely agree with his general conclusion that the Latin emperor ‘could not raise enough troops to defend his territories’. These events also confirm that, apart from having to cope with their external enemies, Latin settlers were heavily outnumbered by a potentially hostile native population. Indeed, during Kalojan’s invasion some Franks continued to fear for their lives even after they had found shelter inside Tchorlu, a fortified city to the west of Constantinople, ‘because they doubted the people of the town’.

It was not the overwhelming superiority of any one opponent but the number of different enemies, the unreliability of the local people and the sheer size of the territories which needed to be defended that contributed to the early demise of the Latin states in northern Greece. It is perhaps hardly surprising that the crusader states set up in south-western Greece and on the Aegean islands survived much longer. The duchy of Athens, the principality of Achaia and the Venetian colonies were all smaller in size and were protected by seas, mountains or other natural features such as the isthmus of Corinth. Moreover, the further Latin conquerors travelled from Constantinople, the more they benefited from the increasing regionalism which had been apparent within the Byzantine empire since before the Fourth Crusade. When Boniface of Montferrat, the man who established the kingdom of Thessaly, undertook his conquest of northern Greece in 1204, he received a warm welcome from some locals who regarded Constantinople as a distant and corrupt absorber of taxes, and consequently felt that it was best simply to come to terms with the Latins. Similarly, when the Franks reached the Peloponnese local Greek lords (‘archons’) were often allowed to keep their incomes provided that they recognised their new Frankish masters. By contrast, the volatile political situation in Constantinople itself before and during the Fourth Crusade, followed by the flight of many leading Greeks to Nicaea or elsewhere, left little room for a quiet transition of power in northern Greece.

The relative stability which existed in south-western Greece during the first half of the thirteenth century ensured that the rulers of Achaia were more prosperous and enjoyed greater military power than their Latin neighbours in Thessaloniki and Constantinople. The total number of knights settled in the region has been estimated at between five and six hundred, and according to one rather romanticised source Prince Geoffrey II of Villehardouin (1228-46) kept eighty knights just at his own court. These figures are confirmed by the Chronicle of Morea, if we can believe its claim that, at the battle of Tagliacozzo (1268), Geoffrey’s successor William II (1246-78) fought alongside his overlord Charles of Anjou with 400 cavalry brought from Achaia. This implies that, even though it was smaller, the military strength of Achaia was almost equal to that of the original Latin empire, and that its rulers had enough resources to intervene in conflicts beyond the Peloponnese. Indeed, in 1236 Geoffrey II of Villehardouin sent help to Constantinople whilst it was being besieged by the Bulgarians and the Greeks of Nicaea. In 1249 Geoffrey’s brother William II joined Louis IX’s crusade against Egypt with a fleet of twenty-four Achaian ships. This episode took place at roughly the same time as the Latin empress of Constantinople made a desperate plea to Louis for 300 knights to help her and her husband defend their capital, and therefore highlights the contrast in military strength between the Franks of Achaia and those of northern Greece.

The military strength of the principality of Achaia should not be overestimated, however, for it still took the Villehardouins many decades to subjugate the Peloponnese fully, and they did not capture the impregnable Greek outpost of Monemvasia, situated in the south-east corner of the principality, until 1249. Moreover, the days of sending troops to foreign conflicts came to an end after the 12605, when Monemvasia, Mistra and Old Mania were returned to the Greeks and subsequently used by them to reconquer the entire Frankish Morea. It is difficult to calculate the number of troops involved in this conflict, but the fact that it took the Greeks until the early fifteenth century to recover all of the Peloponnese suggests that local armies were fairly equally balanced. It has been estimated that even during the reign of Michael VIII (1259-82), which marked the zenith of late Byzantine power, major Greek campaign armies never contained more than 10,000 soldiers, and that most were much smaller. Bearing in mind that there never seem to have been more than five or six hundred western knights settled in Achaia, it seems unlikely that local clashes between Latin forces and the Greeks of Mistra ever involved more than a few hundred horsemen. When the Chronicle of Morea claimed that in 1262 300 Franks defeated a Byzantine force of 15,000 men, it may have given a reasonably accurate estimate for the Frankish army but must surely have exaggerated the scale of the Greek expedition. The Franks must have been less outnumbered than this, but the fact that the Greeks probably had a slight rather than an overwhelming advantage meant that the conflict dragged on for many decades. As the Franks gradually retreated toward the north and west of the principality, the land which they conceded was often devastated by regular yet indecisive fighting.

The inability of any one ruler to deal a swift death blow to his opponents also seems to have characterised the fighting further north. During the first half of the thirteenth century the Greeks of Epiros quickly recaptured Thessaloniki and wiped out many Latin conquests which had been made in Thessaly, but they were unable to overrun the duchy of Athens. Epiros eventually became a frequent ally of the Latins against the Greeks of Nicaea, who were themselves only successful in weakening rather than destroying Epiros and Athens. The exact details of this fighting are either poorly recorded or heavily exaggerated. The papal correspondence of Honorius III (1216-27) and Gregory IX (1227-41) indicates that during the 1220s the Greeks of Epiros attacked Boudonitza and Salona on the northern fringes of the duchy of Athens, and that by 1235 they had even reached Thebes. No details are given regarding the strength and nature of the opposing forces, although the concern caused by these events at the papal court is probably in itself indicative of Epirote superiority on the battlefield. By the second half of the thirteenth century, it is clear that this of the fighting are still impossible to come by. Thus in 1292 the Chronicle of Morea stated that the Epirote city of loannina was besieged by a Byzantine (formerly Nicaean) force of 30,000 infantry and 14,000 cavalry, but that this entire expedition retreated in panic when Florent of Hainault, prince of Achaia (1289-97) and ally of Epiros, approached with an army which probably only contained four or five hundred men! Once again, the Byzantine figures should clearly be disregarded, for the number of western knights living in this region suggests that most clashes between Greeks and Latins cannot have involved more than a few hundred horsemen, and perhaps only a couple of thousand men in total. Indeed, it has been suggested that during the entire late Byzantine period most Greek campaigns ‘probably involved hundreds rather than thousands of troops’.

Whilst the Nicaean Greeks eventually succeeded in conquering Achaia, we have seen that it was the Catalans who finished off the Frankish duchy of Athens. The ruthlessness with which they did so suggests that they were unusually well-trained and aggressive soldiers. According to the chronicler Muntaner the Catalan company consisted of 2500 cavalry plus 4000 Almogavers (light infantry) and 1000 other footsoldiers These figures seem high, although the total of around 2500 horsemen was also given by another source. On the other hand, Muntaner claimed that, when they defeated Duke Gautier I of Athens in March 1311, the Catalans overcame a Frankish army containing 30,000 infantry, plus 700 knights of whom all but two were killed. Even if all the lords of Athens and Achaia had turned up, these are impossibly high totals and unfortunately cast doubt on the other troop numbers given by this chronicler. The decisiveness of the Frankish defeat does at least suggest that Gautier had underestimated both the fighting skill and the size of the Catalan foe, even if the total number of combatants involved in this encounter was greatly exaggerated.

The Catalans also participated in another form of warfare which affected much of the Aegean area during the thirteenth and fourteenth centuries: piracy and raiding. Even before they arrived in the duchy of Athens the Catalans had a terrible reputation for such activities in Byzantine Greece, where ‘there was not a town or city that was not looted and burnt’ by them. From the mid fourteenth century onwards the Turks became an even greater threat, causing so much destruction that by the 1380s some Aegean islands had been completely abandoned by their original inhabitants. Even as early as 1358 a document recording the transfer of Corinth castle to the Florentine lord Niccolo Acciajuoli makes it clear that the Turks were having a devastating effect on the region: one of Niccolo’s primary duties was to persuade people to return to the area around Corinth, which had been overrun so many times by Greek, Catalan or Turkish raiders that superiority had largely disappeared in the face of Nicaean pressure, which forced Epiros to seek the friendship rather than the enmity of the Latins. Accurate details homelessness and starvation had forced entire communities to flee. In the thirteenth century there may not have been quite as much destruction, but piratical attacks, such as that carried out during the 1280s against Bartolomeo Ghisi, the Venetian lord of Tinos and Mikonos, nevertheless represented a common threat for the inhabitants of islands and coastal regions. There were also many clashes between Venetian and Genoese naval forces which had either been sent out by their respective cities or were operating independently in a semi-piratical manner. When the Venetians occupied Corfu and Crete in the years after the Fourth Crusade, they first had to get rid of Genoese privateers based on these islands.

The behaviour of the Catalans and the constant rivalry between Venice and Genoa also remind us that far from being united in their struggles against the Greeks or Turks, Latin settlers in the Aegean area were frequently embroiled in their own internal disputes. Warfare of this kind squandered precious resources and resulted in yet more suffering for those who already had to cope with the threat of piracy, looting and brigandage. Between 1255 and 1258, for example, Duke Guy de la Roche of Athens (1225-63), aided by his Venetian allies, fought a bitter civil war against William II of Villehardouin in a dispute over land on Euboea. William eventually won this conflict by systematically ravaging Guy’s lands until he surrendered; but, although this tactic worked, it must have caused untold suffering for the peasants and farmers who lived in the affected areas. Considering that most of these unfortunate rural inhabitants were Greek, it is hardly surprising to find that fighting also broke out occasionally between western newcomers and local natives, who sometimes had to put up with other forms of maltreatment such as unbearably high taxes. 36 Incidents of this kind must have been particularly unnerving as Latin settlers formed such a tiny section of the overall population. On Chios, for example, it has been estimated that there were 10,000 Greeks being ruled by a mere one to two thousand Genoese toward the end of the fourteenth century. Villehardouin claimed that, during the initial Frankish siege of Constantinople in 1203, the city’s inhabitants outnumbered their attackers by as much as two hundred to one, and that the crusaders were consequently reluctant to enter Constantinople even after they had seized the city walls. Such statistics help to explain why Latin conquerors in Achaia and elsewhere often tried to be as conciliatory as possible toward local Greek landholders.

One final form of combat which needs to be mentioned is that of naval warfare. Generally speaking, the Latins possessed far greater naval power than the Greeks; a factor which proved vital to the continued existence of some of the crusader states. In 1235 and 1236, when Constantinople was besieged by the combined forces of John Asen II of Bulgaria (1218-41) and John III Vatatzes of Nicaea (1222-54), the Venetians were able to break the naval blockades around the city and effectively save the Latin empire from destruction. Apart from the Italian city states, other Latin powers also had warships available to them. In 1236 many vessels from Achaia assisted in the relief of Constantinople and William II of Villehardouin later contributed a fleet of twenty-four ships to the crusade of St Louis. At the beginning of the fourteenth century the Catalans also brought their own fleet to the Aegean, using it in countless raids along the coasts of northern Greece. In 1305, for instance, after the Catalans had fallen out with their former employer Andronikos II, they sent five galleys out from their temporary headquarters at Gallipoli to attack neighbouring Byzantine targets.

The financial, administrative and political problems of the late Byzantine state often prevented the Greeks from raising adequate naval forces to deal with such aggressors. During the 1280s, for example, Andronikos II reduced the size of the Byzantine fleet in order to cut costs. However, it would be misleading to assume that the Greeks were entirely overawed at sea, for at times, and in particular during the reign of Michael VIII Palaiologos (1259-82), the Byzantine authorities were able to build up naval forces by relying on mercenaries. Thus in the 1270s Michael VIII employed the Latin pirate Giovanni de lo Cavo to act as his admiral in the Aegean. Latin seamen also did not have matters entirely their own way because of the growing threat of the Turks. Even the ruthless Catalans had to be aware of this danger, for during the winter of 1303-4 they sent their fleet to a secure winter anchorage at Chios ‘because the Turks, with barques, ravage all [the Aegean] islands’.

This summary of thirteenth- and fourteenth-century warfare around the Aegean makes it clear that there were many motives for the Latins who lived there to construct or occupy castles. Although exact troop numbers are almost impossible to establish, most of the Greek or Latin armies active in the region only contained between one and five hundred horsemen. Even when additional foot soldiers accompanied them, campaigns were rarely undertaken by more than one or two thousand men. Consequently the Latins who settled in Greece after the Fourth Crusade were not as heavily outnumbered by individual opponents as some of the contemporary chroniclers would lead us to believe. There were, however, ultimately too few Latins living in northern Greece to halt the combined attacks of many different Nicaean, Epirote, Cuman and Bulgarian armies. The fall of the duchy of Athens in 1311 suggests that Frankish settlement was so fragile that a single defeat in a pitched battle could seal the fate of an entire crusader state. Eventually the Franks proved equally incapable of halting the gradual loss of Achaia after the Greek acquisition of Mistra in 1262. In addition, the Aegean was almost constantly affected by some form of localised raiding, rebellion or piracy. While this did not necessarily threaten national security in the short term, it ultimately ground down the economic and military strength of states such as Achaia. In short, this was an extremely insecure world, and in order to protect themselves against it Latins needed fortifications to make up for their lack of troops and to defend their property against the constant threat of enemy attack





In 1162, the death of King Géza II (1141–62) presented the opportunity for Manuel I Komnenos (1143–80) to interfere in his neighbor’s realm. After a failed attempt to install an uncle of the reigning monarch, King Stephen III (1162–73), on the throne, the emperor reached a compromise whereby Géza’s youngest son Béla would live at the court in Constantinople and succeed Stephen as king. Béla married one of Manuel’s daughters, solidifying a Byzantine dynastic alliance. But Stephen continued to resist Byzantium in the Balkans, allying with the Holy Roman Empire under Frederick I Barbarossa (1155–90), Serbia, and the Russian principalities of Gallicia and Kiev. In violation of the treaty, Stephen designated his own son as his successor. In 1164, Stephen III and Duke Vladislav II of Bohemia marched to confront Manuel, who was stationed with his army on the Danube. Stephen agreed to cede to the empire the rich region of Syrmia, which was a family holding of Prince Béla, in exchange for the empire withdrawing its support for Stephen III’s uncle, also named Stephen, who had been fighting with Byzantine assistance to claim the throne. Later in the year, Stephen III seized Sirmium, a blatant act of war against the empire.

Manuel dislodged Frederick I Barbarossa from his Hungarian alliance, and pulled onto his side the Russian principality of Kiev, as well as Venice. Stephen’s forces busied themselves with the siege of Zeugminon (part of modern Belgrade, Serbia), which they seized by April 1165. Manuel led his forces northward in June 1165 and laid siege to Zeugminon. Manuel’s troops stormed the city on their third attempt and plundered the place mercilessly. In the meantime, Manuel’s general John Doukas had cut through Serbia and subdued the coastal cities and fortresses of Dalmatia, which Stephen III had also ceded as part of Béla’s holdings. In 1166 the Hungarians defeated Byzantine forces in Dalmatia and at Sirmium.

Manuel responded with the dispatch of his nephew, Andronikos Kontostephanos at the head of a strong Roman army, about one-third of which were mercenaries or allied foreigners. Roman scouts captured a Hungarian who revealed that the enemy force numbered 15,000 knights, bowmen, and light infantry. The Byzantine army was probably about equal in numerical strength. Kontostephanos drew up his marching order with Cuman and Turkish horse archers and a handful of western knights in the vanguard. Behind came three divisions of Byzantine regular cavalry and kataphraktoi, followed by units of allied Turkish and western mercenary cavalry. The last line comprised a mixed formation of Roman infantry and archers alongside a battalion of armored Turks, presumably also infantry.

Dénes, count of Bács, commanded the combined Hungarian-German force. Dénes drew up his mailed knights in the front, with infantry support to the rear. The historian Choniates noted that the Hungarian battle line was drawn up in a single, dense mass, in the shape of a tower; the cavalry fronted this deep formation. The Hungarian lancers presented an awesome sight—their horses wore frontlets and breastplates (these must have been padded or mail, since plate horse armor was uncommon in Europe prior to 1250) and carried riders mailed from head to foot. In short the Hungarian forces featured the best of modern western arms and equipment. They faced a lighter Byzantine force arrayed with the Turk and Cuman horse archers in the front of the formation. Behind, Andronikos divided his army into three divisions. On the left he stationed the regular Roman cavalry. In the center stood Andronikos, commanding elements of the Varangian Guard, Hetaireia imperial guard cavalry, Serbians, probably mailed cavalry, and Italian mercenary knights. The Roman right consisted of the third element of the line of march, with German mercenary knights and Turkish cavalry and Roman kataphraktoi cavalry. Behind the right and left wings of the army Andronikos stationed supporting troops, which presumably were mainly regular cavalry and infantry flank guards and outflankers who could also support the wings when pressured. That two of these supporting battalions were cavalry seems to be indicated by how the battle unfolded.

Andronikos opened the battle by sending ahead the Turk and Cuman horse archers and presumably the light infantry as well. They were instructed to send an arrow storm into the Hungarian cavalry and thus break up the formation. In the face of a Hungarian charge Andronikos instructed them to fan out to left and right and thus sweep to the side of the Byzantine force. The Byzantine left broke in the face of the Hungarian charge and fled toward the river Sava, but two battalions stood fast—these were likely the flank guards stationed behind the left wing. Dénes led a general charge into the Byzantine center, hoping to kill Andronikos; those in the center of the Roman formation sustained the heavy cavalry charge. The Byzantine right attacked the flank of the Hungarian cavalry formation, Andronikos’s men in the center of the line drew their iron maces and pressed forward for close combat, and the “routed” Byzantine left that had feigned flight returned to strike the Hungarian right flank. This envelopment broke the Hungarians, and thousands perished or were captured in the ensuing rout. Kinnamos reported that 2,000 cuirasses were taken from the dead, and countless shields, helmets, and swords came into Roman hands from the great number of fallen. The Battle of Sirmium was the greatest victory of Manuel’s reign; it demonstrated that tactical skill and great discipline were still to be found in the armies of the Komnenoi, as were commanders who were able to conceive and execute complicated battlefield maneuvers. As a result of Sirmium, Hungary became a client, and upon the death of Stephen III in 1172 Manuel easily installed his protégé Béla on the Hungarian throne, which remained at peace with the empire until 1180.

The campaigns of Manuel against Hungary that culminated in the Battle of Sirmium demonstrate that, when properly led, the Byzantine army remained the finest in eastern Europe, capable of defeating heavily armed and armored western knights. But these actions also show that the strategic situation of Byzantium had deteriorated significantly—with the coalescence of larger, more organized, and economically vibrant states on all sides, the empire faced extreme challenges to its territorial integrity. While Belisarios’s decisive victory over the Vandals a half millennium in the past had brought Africa under imperial control and established a peace that was largely maintained for a century, the “decisive” victory of Manuel at Sirmium delivered only twenty years of peace. In light of the capabilities of his enemies, it is small wonder that Manuel generally preferred attritive campaigns and small-war actions that wore down his foes and made enemy aggression too costly for them, rather than risking his limited forces in all-or-nothing engagements on the battlefield. In this sense, his failures are more telling than his numerous minor successes, since the emperor removed neither Sicily nor Hungary nor the Seljuks from their menacing positions along the frontiers. Instead, Manuel had to settle for a largely defensive posture in the territory he inherited from his father John.


The empire reached its largest medieval territorial extent under Basil II, who is considered by many to have been the greatest Byzantine emperor. While the view of Basil as a perfect sovereign who was wise in counsel and indomitable in war is largely a function of his effective propaganda, his campaigns against Bulgaria led to the annexation of vast territories in the Balkans and carried Byzantium to the apex of its medieval prestige and glory. He proved to be the bane of the Bulgars, in particular, and though the statement of the historian Skylitzes that he campaigned annually against them is exaggerated, Basil vigorously pursued their subjugation. Since the seventh century, when the Bulgars first settled between the Danube and the Balkan Mountains, the Byzantines and Bulgars had fought one another for control of the region. Severe clashes were interspersed with periods of simmering peace. In 708 Justinian II suffered defeat at Bulgar hands at the first Battle of Acheloos, but Bulgar allies played a critical role in staving off the Muslim attack on Constantinople in 717–18. Although imperial forces scored several important victories throughout the eighth century, the emperors could neither dislodge the Bulgars from their homeland, nor bring them under Byzantine political domination. In 811, the major expedition of the emperor Nikephoros I, the largest in centuries, met with disaster—the Bulgars destroyed the army, killed the emperor, and mortally wounded his heir. Though periodic conflicts followed, peaceful relations between the two powers dominated the ninth century, when the Byzantines were increasingly focused on the east and the Bulgars faced Frankish expansion and threats from the steppe.

Upon his ascent to the throne, the khan Simeon (893–927) pursued hostilities with Byzantium in the hopes of becoming emperor of a unified Byzantine-Bulgar realm. In 917, at the second Battle of Acheloos (Anchialos), Simeon’s forces ambushed and crushed the divided military command of Leo Phokas assisted by the fleet of Romanos Lekapenos. Simeon warred against the Romans for the rest of his reign and hostilities continued under his son and successor, Peter I (927–69), who suffered from the Byzantine-Kievan Rus’ alliance negotiated by Nikephoros Phokas. The invasion of Sviatoslav, prince of Kiev culminated in heavy Bulgar defeats in 968 and 969. Under John Tzimiskes, the Byzantines drove out their former Rus’ allies after their victory at the Battle of Dorostolon in the summer of 971. From this point on the Byzantines claimed rule over Bulgaria, but it would take decades of hard fighting for the empire to wear down their opponents and establish peace.

Following his suppression in 979 of the attempted usurpation of the Anatolian military magnate, Bardas Skleros, the young Basil II (he was just twenty-one at the time) sought to win his spurs against the Bulgars. Basil led a large imperial army northwest and struck Serdica (modern Sofia) and thus cut the Bulgar kingdom in half. The historian Leo the Deacon was present during the expedition in which Basil sieged Serdica for about three weeks but could accomplish nothing, allegedly due to the inexperience of his soldiers and the incompetence of the senior commanders. Clearly Basil was in large measure to blame—in all likelihood he excluded from the campaign seasoned veterans of the eastern wars who had fought for Tzimiskes a decade prior; perhaps these men had backed Bardas Skleros in his rebellion and consequently were stricken from the rolls. Whatever the case, as the army withdrew the Bulgars ambushed the Byzantines and routed them in a defile near present Ihtiman, in western Bulgaria. The imperial forces suffered heavy losses and withdrew. Little was accomplished in the war with the Bulgars since Basil, as a consequence of his internal military policies, faced renewed opposition from the Anatolian magnate families.

Only in 1001–5 could the emperor return to the theater. He made great gains, capturing Serdica in 1001 and besieging Vidin in the northwest of the kingdom at the confluence of the Sava and Danube rivers. In subsequent years Basil methodically campaigned, reorganized the political landscape by establishing Byzantine administrators, and undermined Tsar Samuel (997–1014) by dislodging his followers. In 1005 the Byzantine diplomatic offensive yielded the greatest of the low-hanging fruit of Bulgaria with the handover of Dyrrachium on the Adriatic by the influential Chryselios family who had previously acknowledged the overlordship of Samuel. Basil’s efforts in 1001–5 returned to imperial control the major trans-Balkan road, the ancient Via Egnatia, and provided the Byzantines a coherent strategic front on Bulgaria’s southern flank.

No sources detail action between 1005 and 1014, but when we next see the emperor in action, in 1014 at Kleidion, Basil faced a Bulgar army that blocked the passage of his army as it marched from the valley of the Strymon River in eastern Thrace to the valley of the Axios (Vardar). Samuel’s men had built a series of ramparts that blocked the trunk road between lofty mountains that led from Thessaloniki to Niš. Basil’s troops repeatedly assaulted the Bulgar earthworks, but the enemy repulsed these attacks and hurled missiles at the Byzantines from above. Basil was about to give up and depart for Roman territory when Nikephoros Xiphias, Basil’s senior commander and active campaigner with the emperor since 1001, hatched a plan: Basil’s forces would continue to attack the Bulgar wooden palisades while he picked infantry and led these troops to the south. Xiphias’s men pushed through the heavily wooded mountains and, via unknown trackways made their way to the Bulgar rear. On July 29, Xiphias fell upon the Bulgars from the heights behind them. Samuel’s men broke and fled as the Byzantines dismantled the makeshift fortifications. A vast number of Bulgars, said by contemporary sources to number as many as 15,000, were taken prisoner. The historian Skylitzes states that the emperor blinded these men and sent them back to Samuel with one-eyed leaders for each hundred men. Blinding was a treatment reserved for rebellious subjects, and this incident, apocryphal or not, shows Basil’s determination to bring to heel the Bulgar state and reflects the view of the emperor and those who later retold the story: the lands from Thrace to the Danube belonged to the empire. Although the final annexation of Bulgaria came in 1018 only after four years’ hard campaigning, the incorporation of the Bulgar realm within Byzantium was given its final impetus by the victory at Kleidion.