An Alternative to the Thucydides Trap: The Buffer Zone of Byzantium and Sasanian Persia

Abstract The return of contemporary great power rivalry has renewed interest in the idea of a ‘Thucydides trap’ within IR and security studies. Missing, however, are discussions of great power conflict between Thucydides and 1648. The following paper seeks to open new ground by examining the bipolar rivalry between Byzantium and the Sasanid empire. This four century rivalry rested upon a fragile buffer zone along the eastern Roman frontier. Its longevity was based on a stable offense-defense balance that eventually collapsed in the seventh century. The Thucydides trap narrative presents a bipolar clash with an air of inevitability, this interdisciplinary case study however, develops a more nuanced narrative, which shows how the Byzantium and the Sasanid rivalry existed across a buffer zone that only collapsed once imperial ideologies led to competing irreconcilable visions of international order.


Introduction
The return of great power conflict and major inter-state war has become a key concern within contemporary security debates. The prospect of a new Cold War, proxy wars or outright conflict between the United States and China, or even Russia, raises the question of whether the world is entering a 'Thucydides trap' of a war between a rising and a declining power. 1 Our understanding of the Thucydides trap and great power conflict however, is often taken from just a few lines of the Melian dialogue and a series of case studies limited to either early modern or 19th and twentieth century European history. 2 This creates a limited interpretation of great power confrontation based solely on a utilitarian and hyper-rationalist view of modern Western experiences. 3 There is a need therefore to widen the engagement of historical examples within international relations (IR) and security studies, to include a broader range of case studies which can highlight the fraught difficulties of great power tension. The following article aims to break new ground by examining a relatively underexplored area of great power conflict within IR, the four centuries bipolar relationship (A.D. 224-651) between Eastern Rome, or Byzantium, with the Sasanid empire of Persia.
The Sasanid empire founded in A.D. 224 competed incessantly with the eastern portion of the Roman empire for four hundred years. This period of competition included phases of war, detente, co-operation and drives for conquest. The two empires shared many similarities with an unclear frontier between the two great powers, but who each relied upon numerous proxies, client states and subordinate powers to compete against one another. The competition between these two great powers rested upon a buffer zone across Mesopotamia stretching from the mountains of the Caucasus region in the north to the deserts of Arabia in the south. This buffer zone remained stable for large periods of time but was also fraught to sudden wars of manoeuvre and raiding missions, sieges and diplomatic bargaining, and collapsed entirely in the seventh century with the ensuing power vacuum leading to the rise of Islam, one of the most significant events of Middle Eastern and global history. 4 Pre-modern great power war is often neglected within IR and an interdisciplinary examination of this kind must of necessity focus on broad brush strokes of history. The aim of this case study however, is not to provide a detailed historiography of Persian-Byzantine relations, but to examine the clash of empires along the eastern Roman frontier, which although fluctuated between the Tigris and Euphrates rivers, nevertheless remained relatively stable until the seventh century.
The relationship between these two empires warrants attention due to its longevity and the strategic competition over fortified cities along a porous border. The eastern Roman frontier can be understood as a buffer system, 5 or as Nicholas Spykman argued as a 'demarcation between territorial power structures'. 6 Security studies and IR scholars usually only focus on buffer states since the nineteenth century, particularly, with examples such as Afghanistan and Belgium. 7 The buffer system of late antiquity however, provides a case study of imperial rivalry which was intertwined with the ontological security of the Byzantine and Sasanid empires. Each power sought to use captured cities and fortifications as linkage to stabilise relations and define a hierarchy of international order, with those outside the frontier as 'barbarians'. Conflict between Persia and Byzantium was often experienced across this buffer zone in frequent low intensity warfare in which the siege constituted over half the military engagements. 8 The construction of this buffer zone between the two empires developed over time and reflected the offense-defense balance of the two powers whereby the advantages for the offensive were minimal due to technological limitations and a terrain with few natural barriers. 9 Warfare in late antiquity also favoured defensive strategies for these two sedentary empires and both sides developed to resemble one another, with the Byzantines adopting more cavalry, while the Sasanians adopted Roman siege warfare techniques. 10 The significance of this rivalry is that it shows how imperial geopolitical competition can develop through periods of co-operation, detente and great power war. Looking to late antiquity also broadens our strategic debates to include a wider range of cases from world history outside of the traditional focus on modern European nation states. The following paper therefore develops by making a case for why IR and security studies should consider pre-modern conflicts outside of the whiggish teleological narrative of 1648 that negates examples before modernity. The second section to this paper, then outlines a broad picture of the Byzantine and Persian frontier, before in the third section, examining how nomadic tribes provided a common foe for the two sedentary empires. Section four outlines how the eastern frontier system broke down as ideology shifted the offense-defense balance. Overall, this paper argues that the Thucydides trap narrative presents a concept of a clash between two superpowers as a question of 'destined for conflict' with an air of inevitability. But the four-century rivalry between Byzantium and the Sasanids shows that other narratives are possible. The example of Byzantium and the Sasanid empire therefore offers a more nuanced approach to great power conflict beyond the Thucydides trap analogy.
The 'missing years' and great power rivalry The Byzantine empire was the Greek speaking and eastern half of the Roman empire which existed as the continuum of Rome for a thousand years. At the height of its power, Byzantium stretched from Gibraltar in the west to Syria in the east. The western Roman empire collapsed between the 3rd and 5th centuries and traditional Eurocentric historiography sees this collapse as the oncoming of the 'dark ages'. The existence of the Byzantine empire has therefore always posed a puzzle for Western historians, as Byzantium does not fit easily into a 'Western' narrative and is seen as 'outside' the development of the Western state and international state system. 11 Within IR, Byzantium is often used only as an analogy for critiquing a bureaucratic Soviet Union or deceptive methods of diplomacy. 12 But as the following explains, looking to pre-modern history, and polities which do not easily fit the mould of a modern neo-Weberian nation state, can illustrate the complexities of gray zones and proxies in warfare, the blurring of war and peace, and imperial hierarchical relations.
Ancient history and pre-modern discussions of great power rivalry within IR, often 'begin and end with Thucydides'. 13 Conventional realist accounts describe how Thucydides uncovered a general law in the dynamics of hegemonic conflict, in which the distribution of power and shifts in power can change the structure of the international system. 14 A 'cottage industry' has therefore grown within IR to examine Thucydides and his lessons for modern great power rivalry. 15 Much of the discussion of Thucydides however, focuses on just a few lines from the Melian dialogue and often overlooks the imperial context. 16 As Barkawi and Laffey argue, the Athenians were trying to convince the Melians to surrender without fighting. It is portrayed by modern realists, as a clear example of a great power enforcing its will on the weak. However, Thucydides' account can also be read as showing that it led to an expensive siege, and that the destruction of Melos, was the first step towards the eventual fall of Athens as an imperial power. 17 Less attention is paid within IR to other classical authors of antiquity, such as Tacitus, Polybius, or Herodotus. Nearly all other wars of antiquity and pre-modern history therefore remain completely outside the bounds of IR theorizing and are left to specialist historians. 18 As one commentator has noted, while we 'all know the name of Thucydides. The same cannot be said of Procopius, the great historian of the wars of Justinian', the eastern Roman emperor from the early sixth century. 19 This problem arises because IR tends to only pay lip service to Thucydides and the Athenian polity before drastically jumping in chronology to the Italian Renaissance and finally, 1648. This leap in chronology creates a vast period of 'missing years' within the discipline of IR because pre-1648 polities rarely resemble the Eurocentric and ahistorical conception of states as units within an international system. 20 The exclusion of historical case studies between Thucydides and 1648 results in the neglect of Byzantium and Persia's rivalry. Instead, strategic debates have become overly focused on understanding all rising and declining powers through the prism of a Thucydides trap. Graham Allison portrays the Thucydides trap as an oncoming 'collision' which is analogous to the Cuban missile crisis or the First World War. The US and China are therefore at risk of sleepwalking into an unwanted war by a fatal mistake. 21 Jonathan Kirshner however, argues that while an unwanted war between the US and China is a risk, a deeper reading of Thucydides shows that neither Athens nor Sparta made any serious attempts to even avoid war. 22 Further commentators have engaged with the Thucydides trap by warning of a 'tug of war' between the US and China, or the possibility of a 'Churchill trap' of long term rivalry and the need for a 'new-type' of great power relations. 23 This would suggest that there is no clear consensus on how a bipolar rivalry is likely to evolve and that there is a need to widen our use of historical case studies. Looking at the rivalry between the Byzantines and Sasanians presents an under-examined case study of hegemonic competition that goes beyond the simplistic 'Thucydides trap' assertions within IR, that a rise of material power for state A leads to war with state B. 24 While this hypothesis may at times be valid, there are also many other characteristics of pre-modern history which are particularly relevant for widening our understanding of contemporary security issues.
Few scholars have sought to examine great power war prior to 1648 through an IR lens. Two of the most prominent exceptions however, are Edward Luttwak's work on The Grand Strategy of the Roman Empire (1976), and The Grand Strategy of the Byzantine Empire (2009), along with Arthur Eckstein's Mediterranean Anarchy, Interstate War, and the Rise of Rome (2006). What these accounts show is that history prior to 1648 can provide a rich array of case studies that illustrate a range of security challenges. As Jakub Grygiel has argued, pre-modern history shows characteristics of limited deterrence abilities, blurred lines between war and peace, the proliferation of strategic actors, and ungoverned spaces that are prone to violence in which a state monopoly of violence is hard to ensure. 25 The premodern international system was a harsh environment in which violence was pervasive, the technology of war was relatively accessible, and there were numerous strategic actors which made deterrence difficult to achieve and created a 'diffused and unclear frontline'. 26 For Grygiel, this suggests premodern history can be used to understand modern strategic concerns for ungoverned spaces, porous borders and the multiplication of strategic actors. He argues that the ancient world offers lessons for today in approaching challenges of unstable frontiers, prolonged rivalries and the decentralized nature of violence. 27 The four centuries of great power conflict in late antiquity can therefore be used to examine warfare and imperial relations beyond the tired analogies of the rise of Athenian power and the clash with Sparta. The following sections therefore seek to broaden our discussions within IR and security studies to examine great power conflict in late antiquity.

Examining the Roman-Persian wars
The rivalry between Byzantium and the Sasanid Persians is significant because it was the first time that the Romans came up against an equal sovereign power that they could not overcome. The Roman empire had first come across a rival power in the east with numerous conflicts against the Parthian empire since the first century A.D. Roman expeditionary forces had been able to frequently penetrate deep into the heartlands of eastern Mesopotamia under the Parthians, but had not been able to fully subdue the Persian's territory. The rise of the Sasanid dynasty however, in overthrowing the Parthians in the mid-220s, created a new Persian rival which implemented restructuring policies and centralized power that for the first time created a competitor to match the power of the Romans. Whereas the Roman empire had previously considered the international only in terms of supplicant kings who bowed down to the glory of Rome, the Romans of late antiquity had for the first time met a rival in the Sasanids in which they had to recognise another state as an equal power with a right to exist. 28 Conflict began between the Romans and Sasanids in the 230 s with the Persian king Shapur I's invasion of northern Mesopotamia and it took a century and a half of minor conflicts and raids before each power recognised the other as essentially an equal in a treaty in 387. 29 Relations between the two sides showed recognition of meeting an equivalent power with Constantius II and Shapur II referring to each other as 'brother' in diplomatic exchanges. 30 The eastern Romans, or Byzantines and the Sasinids existed in a bipolar relationship which recognised one another as above lesser barbarian rulers and nomadic peoples. 31 Across a four-century rivalry this bipolar system developed a frontier buffer zone that stayed remarkably stable and in many ways resembled the trench warfare of the First World War with little to gain and few decisive breakthroughs. 32 Relations between the two empires can be viewed as having developed in three main phases. The period of A.D. 230-363 is characterised by a series of aggressions by the Sasanian empire towards Rome; a second phase between A.D. 363-500 saw a period of relative coexistence and cooperation as each side recognised the other and the buffer zone between them became more formalised. Finally, a third phase during A.D. 500-628 saw increasing levels of warfare, with major outbreaks of conflict occurring between 572-591 and 602-628. The final conflicts in the seventh century led to the exhaustion of the Sasanian empire and are viewed by historians as the last great war of antiquity. 33 The rivalry between these two powers is significant because the frontier or buffer zone which evolved between these two blocs was a contest over territory and spheres of influence. Contemporary international thought has centred on territory and views sovereign power as synonymous with territorial control. 34 The distinction of sovereign control and territory is however, a modern idea. While territory had significant importance for security and material resources, the Roman's sought to subject and control peoples not lands. The Roman empire developed by conquering 'client kings not client kingdoms'. 35 The eastern frontier therefore was an exception in which control over territory and corridors of access was a strategic priority as each side sought to develop fortified settlements and walls to project power and establish authority over trade routes and cities.
The study of geopolitics within IR tends to focus on frontiers shaped by natural boundaries, such as the stopping power of water or mountain ranges, but the frontier between Byzantium and the Sasanids was an exposed space of relatively homogenous terrain with few natural markers or barriers. The buffer system that developed between them was therefore a space which blurred the nature of war and peace, and of borders and exchange. Modern conceptions of territory privileges a view of sovereign power over a singular homogenous area, but it may be more conceptually useful to understand the premodern state as exercising control over a dynamic network, which consists of resource nodes, each joined by formal or informal corridors of power projection. 36 Byzantium was therefore 'an empire of cities' in which siege warfare was the main offensive tool of expansion and fortified cities along the frontier acted to create a network of 'hard points' that could defend the interior. 37 The nature of the siege during this period and the blurred boundaries of the frontier meant that cities which were besieged, could either wait in hope of a relief army, or seek to compromise. It seems likely that the desire to keep provinces functioning means that captured cities could often keep existing local forms of governance. 38 The Sasanids may have also had limited objectives of extracting ransom and plunder from cities as a way to boost the wealth of the Sasanian king and to claim tributary status from external populations. 39 Capturing a city along the eastern frontier therefore became a tool for strategic gain as each captured city was used as 'bargaining counters' in negotiations. 40 The frontier was a network of fortified cities, but defences against the nomadic people of the steppe could also take on more traditional lines of defence. In the north and to the east, the Caucasus approaches, such as the Dariali Pass, were defended by a series of walls constructed by the Sasanians known as the Caspian Gates. These defences against the Hephthalite Huns included the Gorgan Wall in the north east which stretched across from the Caspian sea for 200 km and included a series of fortifications and irrigation systems that served a dual purpose as further defenses. 41 Other parts to the frontier included the territory of Lazica, now part of modern Georgia, which existed as a vassal state of the Romans, while Armenia was partitioned between the two imperial powers in A.D. 387. The division of Armenia was even facilitated by nobility within Armenia who each favoured different kings, one appointed by Rome, another by the Persians. 42 Across the frontier of northern Mesopotamia the Romans developed a series of watch posts and fortifications for frontier control. 43 It is also likely each side maintained a system of fire signals to alert the towns in the interior of the enemy's approach. 44 The frontier was also populated by a series of walled cities such as Nisibis, Singara, Maiozamalcha, Pirisabora and Amida which acted as frontier forts to control flows of trade and territory. 45 In the south, each imperial power supported client groups. The Sasanians supported a client dynasty of the Lakhmids based around the city of Hira to manage the tribes of north east Arabia. 46 While the Romans had a long history of tributary relations with the Christian Ghassanids. Although the southern desert region was of a lower priority in which Byzantium withdrew from the region in the sixth century to employ local Arab forces to provide security. 47 The city of Nisibis, in the Tigris basin, was perhaps one of the most significant parts of the frontier system. Once a Roman city, the Byzantine loss of Nisibis to the Persians in 363 created a relatively stable border for over two and a half centuries. 48 Despite the militarisation of the frontier zone, there could be significant trade and cultural exchange between the two sides. The buffer system was essentially a permeable boundary between the two sides and there was significant interest among the Romans in ensuring trade continued as they sought silk, spices, incense, jewels and ivory from the Far East. 49 The Byzantines began an increased building project in the sixth century, particularly, with the construction of the city of Dara which was just a mere 20 km from the city of Nisibis. The fortified city of Dara showed increased Byzantium determination to project power and served as a frontier stronghold, able to act as a staging post for campaigns, as a custom post or as an armoury and place of refuge. 50 The construction of Dara was a vital concern for the Sasanids who even requested a reduction of troops stationed at Dara and that the local commanding Dux (Duke) at Dara to be moved further back from the frontier during negotiations in 532. 51 The construction of fortifications and an organized strategy of defence in depth also suggests that, unlike the western Romans, the Byzantines no longer had 'any expectation of significant territorial gains', but rather than withdraw from the east they established an organized frontier to project power. 52 The advantages of the Byzantine defences meant that the Persians also had limited options for outright conquest. They could aim to take a single stronghold through siege warfare in a campaigning season. Or, they could aim to bypass a defended site and seek to plunge deep into the interior seeking plunder or ransom from frightened cities. 53 Essentially, this meant that without significant changes in the offense-defense balance the result was often a long war of attrition.
The fortifications between the two sides were of a dual nature. They acted to prevent invasions against nomad invasions, but also had value for the ontological security of each side. Permanent walled cities and defensive structures symbolised who was civilized and who existed outside as 'barbarians'. The fortifications along the frontier ensured some stability between the two empires, yet also served to increase conflict as they 'became a central issue in the friction between the two major empires'. 54 The following section therefore explores how fortifications and the presence of the steppe nomads shaped the bipolar rivalry between these two empires. The frontier of outposts and walled cities created frequent contestation across an east-west direction, but the threatening presence of nomadic migrations created some cooperation along a north-south axis, as each empire had an interest in preventing the overrunning of defences by nomadic groups.

The Caspian Gates and the steppe nomads
The Sasanid empire was based on the rich agricultural lands of Mesopotamia and covered the areas of modern-day Iran and Iraq. It existed as a hierarchically organized state, albeit dependent on noble support, in which the monarch, or Shahanshah, was seen as the 'king of kings'. The Sasanid empire also developed an elaborate system of defensive walls and extensive irrigation systems which suggests a high level of effective state control. 55 The geopolitical position of the Sasanid empire was relatively exposed as a central continental power which left the state vulnerable to numerous pressures particularly steppe nomads with increasing threats from the Huns around the fourth century onwards. 56 The Byzantine and Sasanids competed as the only two great imperial powers, but nomadic actors could still play a significant part within this rivalry, as the following explains.
The Romans and the Persians recognised each other as equal powers above the less civilized barbarian peoples, but waves of nomadic groups threatened the stability of the buffer zone developed along the imperial frontier. The major nomadic groups who shaped relations on the eastern frontier were the Huns, Hephthalites (also known as the White Huns) and the T€ urks. The first major Hunnic raids or invasion in 395 and those that followed in 397 and 398 shifted the priorities of the two empires to search for stability on the eastern frontier. Despite being rivals, a predictable buffer zone between them became a strategic interest when they were both attacked by the same foe of the Huns. 57 This shared strategic interest between the two empires can be seen with their cooperation over defences along the passes crossing the Caucasus. The Dariali Pass known as the Caspian Gates defended the main route across the upper Caucasus. The Sasanians constructed a series of linear defensive walls directed towards defending attacks from the north. The most significant of which was the Darband and the Gorgon walls, with the Gorgon wall as the largest single investment in military infrastructure by the Persians in either classical antiquity or the middle-ages. 58 These Caspian Gates contained a series of fortifications, ditches and walls to defend against steppe nomads. In the southern Caucasus the Darband wall is separated by 120 km to the next wall (known as the Ghilghilchay wall), and a defensive fort is positioned a further 30 km south. The Darband wall even included southern facing towers, which suggests that the defences were made with the deliberate intention of creating an arena for combat or killing ground to defend in depth against nomadic invaders. 59 Between 363 and the beginning of the sixth century, relations between the two empires were less volatile, at least partly due to the threat of nomadic groups. The defences of the Caucasus passes however, were a significant investment by the Sasanians in resources and manpower. The cost of these defences created cooperation and friction with the Romans. Several instances demonstrate this fraught relationship, such as the Persians sending a delegation to the Pope in A.D. 464/5 requesting financial support against the Huns. 60 The Byzantine's refusal to pay towards defending the Caucasus passes also led to conflict at times, such as when the Persian King Kavad (488-531) besieged and captured the city of Amida in 503 in retaliation for non-payment of subsidies. 61 Emperor Justinian's conflicts with Persia also resulted in treaties in 545, 551, 557 and 561/2, which all sought to regulate various aspects of frontier interaction, including Justinian agreeing to pay 30,000 solidi per annum for Caucasus defences. 62 The treaty of 561/2 also made a direct stipulation that the Persians would guard the pass and prevent nomad incursions into Byzantine territory in exchange for further subsidies paid by Byzantium. 63 The concern for frontier defences reflects a common fear of ontological and physical insecurity for empires in which there is an 'enduring sense of paranoia' when it comes to their peripheries. 64 The role of nomadic actors within the Byzantine-Persian rivalry is particularly apparent in the way each actor sought for strategic advantages at opportune moments. Domestic strife, nomadic pressure, or war on multiple fronts could all serve as catalysts for one of the imperial powers to seek an advantage along the eastern frontier. The Sasanid King Khosrow I (A.D. 531-579) for example, had sought to take advantage of Byzantine struggles against the Goths in Italy in order to attack the eastern Roman possessions in the Caucasus region. 65 A period of detente was then filled with increasing competition leading to the Byzantine-Sasanian War of A.D. 572-591. This conflict again proved largely inconclusive with little change along the frontier. The outbreak of war in 603 however, began the last great war of antiquity and saw the greatest shift in power along the eastern frontier in several centuries. The climax of this conflict was the Sasanid march to Constantinople with Avar support in 626. The Avars were a nomadic people centred on the Hungarian planes who joined the Sasanid campaign by attacking the Byzantines western front and laying siege to the Byzantine capital. The eastern Roman empire therefore faced complete collapse in 626 in struggling to fight both its Persian rival and the nomadic Avars; before a miraculous turnaround in fortunes with the counter-offensives of Emperor Heraclius (610-641).

The last great war of antiquity
The following section aims to provide a broad overview of the final conflict between the Byzantines and the Sasanids. The aim is not to list a detailed history, but to examine how the stable eastern frontier dramatically changed leading to a power vacuum and the collapse of the Sasanid empire.
The war began after domestic turmoil within the Byzantine empire as the Emperor Maurice was murdered in a coup in 602, with Heraclius eventually emerging as the unchallenged emperor in 610. This period of turmoil was enough for the Sasanids to take advantage in conquering frontier towns along the Euphrates river and the Caucasus region. The offensive of the Sasanids began a slow gradual process that eventually reshaped the entire eastern frontier. Sasanid expansion was aided by weakened Byzantine defenses due to turmoil over the imperial succession and seeing this opportunity, led to attacks in Syria with the capture of Jerusalem in 614, and the further expansion into Egypt with the capture of Alexandria in 619. The conquest of Alexandria and reaching the Mediterranean created a significant shift in the material resources available to the Persians. This changed the offense-defense balance of the imperial rivalry and for the first time gave the Sasanids real reason to believe they could dissolve the Roman empire. 66 The Sasanid expansion into Palestine and Egypt changed the conflict over the eastern frontier into a great power war for survival. The Emperor Heraclius had sought to negotiate with the Sasanids and seek to end the conflict soon after coming into power. But the Persian Shahanshah, Khosrow II (590-628), decided in 615 against any negotiated settlement to the conflict. 67 The Persian-Avar assault on Constantinople in 626 was however, not the end of the Byzantine empire but the beginning of a miraculous turning point in the war. The Avar siege of Constantinople failed and the two allies were never able to win decisively due to the strength of the city's walls and a lack of supporting naval power. 68 This led to Heraclius launching a decisive counter-offensive, first, defeating the Persian army at Nineveh in 627, before marching on the Persian capital of Ctesiphon and agreeing a peace treaty in 629, in which the Sasanids agreed to return peacefully all of the occupied territories and restore the original eastern frontier.
This remarkable shift of fortunes for the Byzantine empire was due to a series of factors. Firstly, the Byzantine's own nomadic allies the T€ urks intervened in 627 driving through modern Azerbaijan and into northern Iraq. The T€ urk forces with the Byzantines then broke the Persian hold in Transcaucasia leading to the incitement of revolts in the north western provinces. 69 The dynastic power of the Sasanid regime was based on a claim to uphold the status of 'king of kings' and it never seriously recovered from this grave damage to the prestige of Khosrow's right to rule. 70 The Sasanid drive for conquest from 603 was also led by Khosrow's 'blind pursuit of imperialistic aims' which drained the empire of resources and turned noble Persian families against his regime. 71 Khosrow's fanatical drive for imperial conquest developed into an increasingly tyrannical regime inciting revolt at home. 72 Finally, the Persian conquest of Jerusalem in 614 had turned the conflict into a religious war. Historians debate the extent to which Heraclius' counter-offensive can be seen as a 'proto-crusade', but the campaign was coupled with extensive propaganda appealing to Christians in Sasanid territory. The Emperor Heraclius also made appeals to martyrdom through promising heavenly rewards to those who died fighting the infidel. 73 The appeal of ideology and nationalism is recognised within security studies as increasing the will to fight, 74 and the appeal of Christianity in this period cannot be dismissed. The loss of Jerusalem to the Sasanids was viewed as divine judgement and the Persian removal of Holy Relics, such as the True Cross, became important for its symbolic significance. The motivation to avenge and retake the cross played a significant role in Heraclius' campaign. After the defeat of the Sasanids and treaty in 629, the Emperor restored the True Cross to the Church of the Holy Sepulchre and even changed his imperial title to reflect his position as defender of the Christian faith. 75 Edward Luttwak's analysis of Byzantine strategy takes religion merely as an instrumental tool, 76 but this downplays how religion also provides the meaning and ideals which define the goals that are sought. 77 The campaigns of Heraclius became enveloped with religious significance and the ontological security of the empire which saw itself as the defender of Christian civilization.
The counter-offensive and reconquests of Emperor Heraclius led to anarchy in the Sassanid state with 8 rulers in the years A.D. 628-632. 78 Despite the return of the occupied territories and the recognition of the eastern frontier the resulting power vacuum led to the rise of the Arab conquests and the expansion of Islam. 79 For four centuries, the eastern frontier had fluctuated but remained relatively stable. The ideological ambition of Khosrow and then the religiously motivated counterattack was perhaps one of the main reasons why this frontier stability broke down. The final section to this paper therefore seeks to examine how the stability of this buffer zone and its eventual collapse provides an alternative case study to the prevalence of using Thucydides as the only example of ancient warfare within IR and security studies.

The east roman frontier in late antiquity
The logic of the security dilemma within IR theory, posits that states will pursue expansion until they can secure a frontier or buffer zone from attack. 80 The frontier of Byzantium and Persia was not, however, an automatic outcome of a bipolar system, but a gradual process which developed through a series of practices that created a recognition of a frontier zone between the two empires. This relationship only collapsed after the ideological hardening of both sides created irreconcilable imperial visions of order. As the following shows, the frontier acted as both a source of contention and co-operation.
The fifth century saw a period of detente between Byzantium and Persia which was sustained through a gradual process of reciprocal recognition. Peace negotiations in A.D. 387 froze the defensive systems of the two sides and prevented the construction of any new frontier fortifications. This mutual agreement of a ban on new defensive works lasted for over a century. 81 The ban on further militarisation of the frontier was also coupled with a growing body of diplomatic practices including formalised rituals and gift giving. 82 The formal development of diplomacy between the two sides can be seen with the practice of sending an embassy on the ascension of each new monarch or emperor and the exemption of ambassadors from customs duties. Such details suggest that there was an agreed upon protocol between the two sides. Attempts to restrict trade along the frontier to Nisibis, also demonstrates a reciprocal relationship developed to formalise the frontier which helped to strengthen deterrence between the two sides. 83 Interempire competition in this period also saw minor flashpoints, but these were often limited incursions or raids reflecting a desire to maintain a balance of power along the frontier rather than wars of conquest designed to overturn the international order. 84 The rivalry along the frontier shifted significantly however, during the sixth century. The increased militarisation of the frontier was intertwined with tensions on the central issue of Roman payments and subsidies to the Persian monarchy. The Sasanid King Kavad captured the city of Amida in 503 in response to the stopping of Roman subsidies for Caucasus defences, eventually selling the city back to the Byzantines for 10,000 lbs of gold. 85 This spurred on the Byzantines to revisit frontier defences, leading to the construction of Dara in 505 in conjunction with a wider 20 year programme of progressively extending and reinforcing fortifications along the northern half of the mutual frontier. 86 Dara's construction was a violation of previous treaties and was attacked repeatedly by the Persians who finally dismantled the walls when it was taken in 606. 87 The question of payments and subsidies for Caucasus defences exposed greater divisions between the two sides in their conceptions of international order. For the Byzantines, payments were used to buy peace, to encourage other actors to refrain from attacking Byzantine cities, or to secure additional nomadic allies. For the Sasanids, the payment of subsidies was intertwined with an understanding of political order and the role of the Persian Shahanshah as king of kings. 88 The payments towards Persia were of symbolic value showing that the Sasanid King had suzerainty over other powers. The Emperor Justinian had sought to finally settle the issue of the frontier with several treaties including the 'Eternal Peace' in 532. Previously, the Romans had accepted the right of Persia to ask for assistance in defending the Caucasus passes against nomads. Justinian's treaties however, were the first which accepted a regular payment of annual subsidies to Persia. Later Byzantine emperors sought to stop annual payments and the use of annual subsidies to Persia caused resentment among Romans who wanted to negotiate on 'equal terms' without any payments. 89 The T€ urk Khagan even interpreted Roman subsidies to mean that they were subjects of the Persians. 90 Occasional payments for Caucasus defences had been seen by the Romans as pragmatism, but the increasing demands for annual payments showed their conflicting worldviews. The Roman refusal to make regular payments after Justinian had already done so may have been a reason for the increased offensive ambitions of the Sasanids. Khusrow II's campaigns of total conquest were therefore a 'fundamental departure' for a regime that had previously been willing to accept payments as tributary status. 91 Coupled with the Persian's increased demand for recognition of a universal monarchy as the king of kings was the steady growth of Christianity and the Byzantine belief of Christianity as intimately connected to their Roman identity. The question of Christianity was still unsettled and evolving during the fourth century, but by the sixth century there was a prominent shift in the importance of Christianity. The Byzantine scholar Procopius even used Biblical allegories to detail the wars of Justinian 92 and a common belief of the time was the Christian legend that the city of Edessa could not fall to barbarians, which spurred on the attempts by Khusrow to take the city. 93 The growth of Christianity, coupled with the question of subsidies to a king of kings, increased tensions between these two incompatible worldviews which slowly grew further apart as conflict increased in the sixth and seventh century finally leading to the last great war of antiquity.
The example of an imperial buffer zone in late antiquity shows that a buffer system can act both as a passive tool of conflict management as well as an active area of intervention. The buffer system of late antiquity acted as a 'low-cost alternative' to direct military intervention and occupation. 94 Once the Byzantines made peace under Emperor Jovian in 363 there was never a serious attempt to reconquer lands lost to Persia. 95 As this paper has shown, trade, diplomacy and a network of bases along the frontier helped to stabilize the inter-imperial rivalry. The collapse of this buffer system occurred only after each power increasingly became convinced of success in war as linked to their own sense of imperial order and international hierarchy.

Conclusion
The Byzantine and Sasanid rivalry is notable for its longevity and relative bipolar stability, which still saw frequent clashes as each side sought to use the capture of fortified posts and cities as part of a wider bargaining strategy to test resolve and the extent of power projection. The eastern Roman frontier also highlights the porousness of borders, in which fortified cities were used both as a defensive position and as a means to project imperial power. It further shows that spheres of influence, which have become a 'great taboo' in modern foreign policy debates, reflect the behaviour of great powers to exert hierarchical ordering of weaker actors, 96 as each empire sought to use nomadic allies and client states to shore up their power projection and control along the frontier. The Thucydides trap presents an impression that the bipolar rivalry 'locks in' two great powers to the 'exclusion' of other strategic relationships, 97 but as this paper has shown, while nomadic tribes were unlikely to rival the material capabilities of Byzantium and Persia, they were significant strategic actors who impacted this rivalry. The growth of an incompatible world view between the two empires also parallels modern concerns for an ideological conflict between civilizational states. 98 Primarily, the Byzantine and Sasanid frontier offers a case study for security studies that goes beyond simplistic interpretations of just a few lines of text from Thucydides. The teleological jump and 'missing years' in IR, between Thucydides and 1648, suggests that there is much wider scope for interdisciplinary examinations of great power conflict before the benchmark date of the Peace of Westphalia.