CHAPTER 32
A. D. LEE
DURING late antiquity the Roman Empire faced many new military challenges. Of those challenges, one of the most serious, and certainly the most consistent, was that presented by Sasanian Persia to the east. In the mid-220s the Sasanians—an aristocratic family from the southern region of Persis—overthrew the Parthian Arsacid regime and quickly established themselves as a force to be reckoned with (Frye 1984: 291–6). They were to remain so for the next four centuries, until they in turn succumbed to the Arab invasions of the 630s which also had such grave repercussions for the Roman Empire. Warfare between the Empire and Sasanian Persia warrants special attention both because of the longevity of their relations and because this was the only context in which the post-Augustan Empire had to confront another state with a reasonable claim to be its equal in terms of military capabilities and mobilization of economic resources.1
Figure 32.1 Breastplate of the Prima Porta Augustus, showing detail of the Parthian return of lost Roman standards. Vatican Museum, Rome. Photo Credit: Alinari/Art Resource, New York.
The significance of the advent of the Sasanian dynasty for the Roman Empire can be exemplified by a comparison of two images. The first (figure. 32.1) is the cuirass worn by the emperor Augustus on the statue from Prima Porta, whose central focus is a trousered and bearded Parthian surrendering a legionary standard to a Roman in military garb (Tiberius or Mars?: see further Galinsky 1996: 155–64, 396 n. 54, 410 n. 35). In his Res Gestae (29.2) Augustus claimed:
I compelled the Parthians to restore to me the spoils and standards of three Roman armies and to ask as suppliants for the friendship of the Roman people. (Brunt and Moore 1967: 33)
Although the legionary standard was a potentially risky reminder of past Roman humiliations at Parthian hands, most famously the defeat of Crassus in 53 B.C., Augustus was able to present the return of the standards in 20 B.C. as a major foreign policy success and the overriding message of the scene is of course one of Roman dominance, which was indeed to typify the balance of power between the two states over the next two centuries or so (i.e., the three Roman invasions in the second century which resulted in the capture of the Parthian capital at Ctesiphon: cf. Isaac 1992: 28–31, 50–3; see further Culham, 250–1). The second, contrasting image (figure.32.2) derives from beyond the borders of the Empire, from the rock cliffs at Naqsh-i Rustam near Persepolis in southern Iran, on which a relief carved in the mid-third century depicts the Persian king Shapur I on horseback grasping the wrists of one Roman emperor standing before him, while a second emperor kneels in obeisance—a powerful commemoration of his victories over Roman forces in the 240s and 250s, culminating in his capture of the emperor Valerian in 260 (debate about the identity of the Roman figures: MacDermot 1954; Ghirshman 1962: 154–61; Göbl 1974: 7–31; Shepherd 1983: 1080–4; Herrmann and Mackenzie 1989). Although this image of Persian ascendancy and Roman subservience is not a fair guide to the balance of power between the two states across the whole of late antiquity, or even across the third century, it does vividly encapsulate how events outside the Empire in the 220s dramatically redressed the balance of power to the Empire’s disadvantage, and set the stage for a much more volatile military situation along the eastern frontier during late antiquity.2
Figure 32.2 Victory monument of Shapur, over Roman emperors—Naqsh-e Rostam, Iran. Photo Credit: Ali Ganjei.
This image of the triumphant Shapur provides a rare insight into the Persian perspective on the Empire’s eastern frontier, for the sources and evidence on which knowledge of the military history of the region during late antiquity relies are heavily skewed toward the Roman viewpoint.3 From the Persian side there are a number of important inscriptions, most famously the great trilingual record of the achievements of Shapur I at Naqsh-i Rustam, but otherwise contemporary Persian accounts have only survived embedded in later Arabic sources, and disentangling genuine Sasanian data from the latter is notoriously problematic.4 On the Roman side the most important narrative sources are the histories of Ammianus Marcellinus and Procopius and the chronicle attributed to Joshua the Stylite, supplemented by a variety of other less detailed writers. The authors of these three main sources all had direct experience of Roman warfare with Persia. Ammianus served as a junior officer in the army in the 350s and early 360s, with active involvement on the eastern frontier during the tumultuous events of Shapur II’s invasion of 359 and Julian’s Persian expedition in 363, and the surviving books of his history include detailed narrative of these events (in Latin); the unknown author of the chronicle attributed to Joshua the Stylite (conventionally referred to as “Pseudo-Joshua”) lived in northern Mesopotamia during the major Roman-Persian war in the early years of the sixth century and recorded these events (in the Aramaic dialect of Syriac); and Procopius, in his capacity as secretary to the general Belisarius, observed at close hand a number of major battles between Roman and Persian forces in the early 530s, and included an account of them in his history of the emperor Justinian’s wars (in Greek). The reliability of the histories of Ammianus and Procopius is compromised to some degree by their literary aspirations (specifically their concern to emulate the great classical historians of earlier centuries) and by their personal loyalties (to Julian and Belisarius, respectively), with the result that their descriptions cannot automatically be taken at face value. Such agenda are much less evident in the case of Pseudo-Joshua the Stylite, which enhances the value of his account accordingly.5
Other textual sources with smaller contributions to make are too great in number to be easily summarized, beyond noting their diversity in terms of genre (orations, church histories, saints’ lives, military treatises, administrative registers, inscriptions) and language (Greek, Latin, Syriac, Armenian; cf. Dodgeon and Lieu 1991 and Greatrex and Lieu 2002). To these can be added visual and material evidence, particularly the archaeological remains of forts, of which those at Dura-Europus are especially significant for this subject.
The subject of material evidence and archaeological sites raises the question of geographical context. Conflict between the two powers was played out across an extensive and varied frontier region, stretching from the mountainous terrain at the eastern end of the Black Sea in the north, to the sands of the Syrian desert in the south (in the sixth century, the Arabian peninsula and Red Sea region also became contested areas in Roman-Persian relations). However, both powers relied to a considerable extent on allies to fight on their behalf in the less hospitable environments at the northern and southern extremities of the frontier, with direct confrontation largely reserved for the central corridor of northern Mesopotamia, with its urban communities and arable farmland sustained by the waters of the Euphrates and Tigris and their tributaries, and it is on events in this region that the following discussion will concentrate. First, however, the question of why the advent of the Sasanian dynasty had such a significant impact on the Empire’s eastern frontier warrants discussion.
PERSIAN MILITARY POWER AND THE ROMAN RESPONSE
The Sasanian regime was able to pose a more serious military threat to the Roman Empire compared with its predecessor partly because it proved able to harness the economic resources at its disposal more effectively. This feature of Sasanian rule is principally evident in the results of archaeological survey work in lower Mesopotamia, which show that the irrigation network was developed significantly in scale and complexity during late antiquity (including a contribution from the labor of Roman prisoners of war in the mid-third century). This enhancement of agricultural infrastructure in turn facilitated increased food production and demographic growth (see especially Adams 1965, 1981; also the overview and commentary by Howard-Johnston 1995: 198–203). Sasanian economic resources were further augmented by more active promotion of Persian trade in the Indian Ocean (Whitehouse and Williamson 1973; Howard-Johnston 1995: 203–5). Although explicit data on the size of Persian armies in late antiquity are lacking, circumstantial evidence implies forces broadly able to match the Romans numerically (Howard-Johnston 1995: 165–9), which in turn is consistent with this general picture of economic developments under the Sasanians.
As for the composition of Persian forces, the following “snapshot” from Ammianus, embellished though it may be for literary effect, provides some sense of what Roman forces confronted:
About dawn an immense host of Persians appeared with Merena, the master of the horse, two of the king’s sons, and many other grandees. All their troops were clad in mail; their bodies were covered with plates so closely fitting that the stiff joints of the armour conformed to the articulation of the limbs beneath, and representations of human faces were so skilfully fitted to their heads that the whole man was clad in metal scale. The only spots where a weapon could lodge were the tiny holes left for the eyes and nostrils, which allowed some degree of vision and a scanty supply of air. Those of them who were to fight with pikes stood so still that they might have been fixed to the spot by metal ties. Close by, the archers, practiced from the very cradle in a skill in which that people especially excels, were bending their flexible bows. Their arms stretched so wide that while the point of the arrow touched their left hand the string brushed their right breast. By highly skilful finger-work the shafts flew with a loud hiss, dealing deadly wounds. Behind them were the gleaming elephants, whose awful aspect and gaping jaws inspired almost unbearable fear, and whose noise, smell, and strange appearance terrified our horses even more than ourselves. (Amm. Marc. 25.1.11–14 [tr. W. Hamilton])
Important features of this battle array entailed significant elements of continuity from the Parthian period, notably the use of heavy armored cavalry, referred to by the Romans as cataphract(ari)i and clibanarii,6 and significant reliance on the projectile power of archers. The archers, who were usually mounted, allowed the “softening up” of the enemy at a distance prior to direct engagement, which the heavy cavalry could then potentially exploit through the impact of a massed charge. Persian cavalry strength reflected the aristocratic ethos of Persian society, as did too the tradition of expertise in archery, both epitomized in the elite’s passion for hunting which can be observed in so much surviving Sasanian art (see above figure 26c.1).7 It may be that the Sasanians exploited these traditional strengths with greater effectiveness through more rigorous training of their forces, since Ammianus comments that “their military training and discipline, and their constant practice of maneuvers and arms drill, which I have often described, make them formidable even to large armies (Amm. Marc. 23.6.83 [tr. W. Hamilton]; these comments may find indirect corroboration in the evidence for Sasanian military treatises, on which see Inostracev 1926).
Notable by its absence from the above description is any reference to infantry, an area of military endeavor in which neither the Parthian Arsacids nor the Sasanian Persians ever developed any prowess, in contrast to the Roman tradition of the citizen soldier. The lowly status of Persian infantry is reflected in Ammianus’s observation that “their infantry are armed like gladiators, and obey orders like soldiers’ servants. These people follow behind their masters in a mass, condemned as it were to perpetual servitude and never remunerated either by pay or presents” (Amm. Marc. 23.6.83 [tr. W. Hamilton]).
As for the elephants, these do appear to be a novel feature of Sasanian military organization, compared with the Parthians, but their highlighting in the passage above in the context of battle probably has more to do with a desire to entertain audiences with the exotic than with their actual influence on the outcomes of engagements. It is likely that elephants made a more substantial contribution to the Persian war effort in the context of siege warfare, which was as important a feature of Roman-Persian conflict in late antiquity as pitched battles (see further Rance 2003 for discussion of the role of elephants in warfare on the Empire’s eastern frontier in late antiquity).
The prominence of siege warfare on the Empire’s eastern frontier in late antiquity is the most significant difference from the prevailing pattern during earlier centuries vis-à-vis the Parthians. As one might have expected of a people from a nomadic background, there is little evidence of the Parthians having the capacity to undertake effective sieges. There are references to their deploying siege machines of some sort, apparently drawing on the knowledge of subject peoples, but these cases are rare (e.g., Adiabeni, with ladders and siege machines [machinamenta] used unsuccessfully during the Parthian attempt on Tigranocerta in 61 [Tac. Ann. 15.4]). The Sasanians, on the other hand, undertook numerous sieges of Roman forts during late antiquity, developing a strong reputation in this field: “They are formidable when laying siege, but even more formidable when besieged” (Ps.-Maurice Strategikon 11.1 [lines 9–10] [tr. G. Dennis]. This assessment appears in a Roman military treatise of the late sixth century, but it is clear that they were proficient from a relatively early date in the mid-third century.8 This raises the intriguing question of why they achieved greater proficiency in sieges than the Parthians so quickly. There are no straightforward solutions to this conundrum, but one likely element in the equation is borrowing from the Romans. They are known to have done so in other areas of military organization and to have made use of captured Roman siege artillery in the mid-fourth century, and the many Roman prisoners captured in the early 250s must have included individuals with the requisite knowledge.9 Another likely element is superior Sasanian logistical organization which will have facilitated their sustaining the investment of cities for longer periods of time (cf. Lee 1993: 19–20). It is worth adding that the Sasanians themselves proved adept at the construction of fortifications, creating “a defensive zone of a depth unmatched in any Roman frontier region (at least before the sixth century) to provide effective protection for the Mesopotamian alluvium, the [Sasanian] empire’s political and economic heartland, which had suffered repeatedly at Roman hands in the second and third centuries” (Howard-Johnston 1995: 196).
The Roman response to the increased threat on the eastern frontier can be seen in a number of areas. Sasanian siege capability resulted in increased Roman investment in the fortification of cities and towns on and near the frontier. The emperor Diocletian (284–305) in particular strengthened the Empire’s frontier defenses in this region. According to one chronicler, “Diocletian built fortresses on the frontier from Egypt to the Persian borders and stationed troops in them, and appointed commanders for each province to be stationed further back from the fortresses with a large force to ensure their security” (John Malalas, Chronicle: 308). This statement has prompted controversy in relation to larger arguments about the relevance of the concepts “grand strategy” and “defense in depth” to the Roman world (Isaac 1992: 162–71), but its general import is clear and has found corroboration through inscriptions and archaeological remains, particularly relating to the so-called strata Diocletiana, the military road which linked forts between Sura on the Euphrates and Damascus (Kennedy and Riley 1990: 77–8, 181–4; Isaac 1992: 163–6; Millar 1993: 180–6). Persian attacks were obviously not the sole concern here—control of nomadic movement and raids from the Syrian desert was another consideration—but combating the Persian threat was clearly an important factor in Diocletian’s thinking from the Euphrates northwards, as is apparent in Ammianus’s observations about the site of Circesium, which occupied an important point in the communications network, at the confluence of the Euphrates and the Khabour. “In earlier times it was a small and insecure place, but Diocletian, when he was organizing the inner frontiers on the actual borders of the barbarian territory, surrounded it with walls and high towers to prevent the Persians making an inroad into Syria of the kind which had occurred some years before and inflicted great damage on our provinces.”10 Nisibis was central to communications across the northern Mesopotamian plain and was appreciated as such by Diocletian (Diocletian specified it as the point for Roman-Persian trade in the peace treaty imposed on the Persians in 299: Peter the Patrician fr.14). It is likely that its fortifications were strengthened at this time, even if explicit evidence for this has not been identified (cf. Millar 1993: 181, with Lightfoot 1988: 109–10 for a summary of what is known from written and archaeological evidence), while the remains of the Roman fortifications at Singara have been assessed as early fourth century (Oates 1968: 106; cf. Kennedy and Riley 1990: 125–31). Constantius II (337–361) supplemented Diocletian’s program by building a fort at Constantia (Pollard 2000: 291) and undertaking work at Amida on the Tigris: he “enclosed it with strong towers and walls to provide a safe place of refuge for the neighborhood, and established it as an arsenal of engines to repel an assault” (Amm. Marc. 18.9.1; evidence indicates that after the Persian sack of the city in 359, the walls were restored during the reign of Valens [364–378]: Matthews 1989: 66, 489 n. 32).
The Roman response can also be seen in the types of troops deployed on the eastern frontier in late antiquity. Although Sasanian use of archers and heavy cavalry was not a novelty, their more aggressive and effective deployment of these troops must account for the presence of significant numbers of Roman units of mounted archers and heavy cavalry. The administrative register known as the Notitia Dignitatum, which records troop dispositions for the eastern half of the Empire toward the end of the fourth century, includes details of six units of equites catafractarii and seven units of equites clibanarii in the eastern half of the Empire (conveniently tabulated in Eadie 1967: 169, 171), and arsenals manufacturing cavalry armor were located in Antioch and Caesarea in Cappadocia.11Although the history of Roman heavy cavalry can be traced back to the early second century in the context of Roman encounters with Sarmatian cavalry on the lower Danube, “the concentration of these units in the east [in the Notitia listings] differs markedly from the earlier pattern of distribution in the western provinces and probably reflects a policy revision in the fourth century—perhaps in response to the challenge of Persian cavalry” (Eadie 1967: 169). The Notitia also records substantial numbers of units of mounted archers in the provinces near the frontier with Persia.12
ROMAN-PERSIAN WARFARE: A HISTORICAL OUTLINE
The parameters of this chapter do not permit a detailed chronological narrative of Roman warfare with Sasanian Persia during late antiquity, but a brief overview can be presented before turning to some case studies of specific encounters which will provide greater specificity for some of the points noted above, and illustrate them in a dynamic context.
In the early decades of Sasanian rule, attempted Roman invasions by Severus Alexander in 231/232 and by Gordian III in 243/244 were turned back, and then followed by two major Persian invasions of the Roman east led by Shapur I (241–272?), the first, in 252, culminating in the sack of Antioch, and the second, in 260, in the capture of the emperor Valerian and the penetration of Persian forces into Anatolia. These were the victories celebrated in the reliefs of Shapur at Naqsh-i Rustam and elsewhere.
In the turmoil which followed the unprecedented events of 260, the initiative was wrested from Shapur, not by a new Roman emperor, but by Odenathus, ruler of the city of Palmyra, an important and wealthy trading entrepôt on the edge of the Syrian desert. Odenathus acknowledged Roman suzerainty, but after his death in 267, his widow Zenobia pursued a more independent policy which resulted in the brief establishment of a “Palmyere empire” in the east. The restoration of central control by the emperor Aurelian in the early 270s, and the death of Shapur around the same time, permitted a gradual reassertion of Roman power in the region, culminating in the emperor Carus invading Persia and sacking Ctesiphon in 283. After a brief setback in the mid-290s, Roman forces under the emperor Galerius inflicted a major defeat on the Persian king Narses which permitted Galerius’s senior colleague Diocletian to impose a humiliating peace settlement on Persia in 299 and extend Roman territory to the Tigris. This was to be the high point of Roman power vis-à-vis Sasanian Persia until the early seventh century.
The death of the emperor Constantine in 337 signaled a new phase in Roman-Persian warfare during which another energetic Sasanian ruler, Shapur II (309–379), endeavored to reverse the settlement of 299. Much of the fighting during these decades of the mid-fourth century involved Persian attempts to capture Roman strongpoints in northern Mesopotamia, above all Nisibis, which was besieged (unsuccessfully) three times between 337 and 350. Although his careful but unexciting defensive strategy was much criticized by contemporaries, Constantine’s son, Constantius II, proved adept at preventing Shapur II from repeating the exploits of Shapur I (cf. Blockley 1989), and the wisdom of his approach was confirmed when his brash successor Julian embarked on an ill-advised invasion of Persia in 363 which failed to capture Ctesiphon. During the retreat, Julian himself was killed in skirmishing and his hastily appointed successor, Jovian, was only able to guarantee the safe passage of the surviving Roman forces out of Persia by agreeing to relinquish the territory gained in 299 and to surrender important Roman strongpoints including, notoriously, Nisibis.
Although there was further jockeying between the two powers for preeminence in Armenia between the 360s and 380s, and two brief and inconsequential wars in the fifth century, the eastern frontier was markedly less volatile between 363 and the beginning of the sixth century, at least partly because both the Roman Empire and Sasanian Persia were preoccupied with problems on their northern frontiers arising from the activities of Asiatic nomads. The Roman refusal to provide financial help to Persia toward defending the Caucasus passes against this threat became the excuse for a renewal of war on a significant scale at the start of the sixth century when the Persian king Kavad (488–531) besieged and captured Amida in 502/503. Significant Roman forces were deployed in northern Mesopotamia to recapture Amida and force the withdrawal of the Persian army, which was eventually achieved in 505. When Roman commanders complained to the emperor Anastasius that their efforts had been hampered by the lack of a fortified frontier base to compensate for the loss of Nisibis, a new fortress was constructed at Dara, although it in turn became a bone of contention between the two powers, with the Persians interpreting it as a forward base for Roman offensive operations into Persian territory. Further efforts to strengthen Roman frontier defenses in northern Mesopotamia by the emperor Justinian in the late 520s resulted in renewed Persian aggression, culminating in a famous Roman victory outside Dara in 530 and a dismal Roman defeat the following year at Callinicum on the Euphrates.
The accession of a new, young Persian king in 531, Khusro I (531–579), facilitated a peace settlement in 532 which freed Justinian to pursue his ambitions in the western Mediterranean. However, increasing Roman preoccupation in the west proved too tempting an opportunity for Khusro and in 540 he invaded Syria, sacking many cities and extorting money from others. In the following years Persian forces invaded Lazica in the western Caucasus, prompting Roman fears that the Persians might try to establish a naval presence on the Black Sea. Intermittent conflict continued until a peace settlement was finally agreed in 561/2, although only at the price of substantial annual Roman payments to the Persians. Justinian’s successor, Justin II, soon decided that these payments were an unnecessary expense, and renewed war with Persia in 572 with the immediate strategic aim of regaining Nisibis. Justin’s hopes, however, were rapidly dashed as Persian forces relieved the siege of Nisibis and then proceeded to capture Dara. Another two decades of intermittent and inconclusive warfare ensued until events within Persia offered a resolution. Another new, young Persian king, Khusro II (590–628), was driven from the throne by one of his generals and fled for sanctuary and help to the Roman Empire, where the emperor Maurice, adopting a far-sighted perspective, agreed to assist in his restoration (successfully achieved in 591, with appropriate territorial compensation for Roman efforts, including the restoration of Dara).
This should have set the stage for a lengthy period of peaceful relations between the two powers. However, in 602 Maurice himself was overthrown and killed in a military mutiny on the lower Danube led by an army officer, Phocas, and Khusro seized the opportunity to invade Roman territory. His ostensible justification was avenging the death of his mentor and friend Maurice, but it is difficult to avoid concluding that Khusro was as much concerned to demonstrate to the Persian aristocracy that he was not a Roman puppet. At any rate, because of the chaotic conditions within the Roman Empire engendered by Phocas’s usurpation, Persian forces made dramatic advances into the Empire’s eastern provinces, gradually gaining control of Syria, Palestine, Egypt, and all of Anatolia, and eventually besieging the capital at Constantinople itself in 626. In the meantime, Phocas had been overthrown in turn by the governor of North Africa, Heraclius, who proceeded to conduct a daring and risky campaign behind Persian lines, eventually descending from the Caucasus on Ctesiphon in 628 and transforming potential disaster into emphatic victory (a poorly documented phase of Roman-Persian warfare, but analyzed in Howard-Johnston 1999, Kaegi 2003: 100–91). Once again, this should have set the stage for a lengthy period of peaceful relations between the two powers, but within less than a decade, Arab armies energized by Islam inflicted fatal defeats on the Persians and a crippling defeat on Heraclius in Syria which resulted in the loss of the Empire’s wealthiest province, Egypt, and the need for a radical reordering of the Empire’s priorities and organization.
FACE TO FACE WITH THE ENEMY: BATTLES AND SIEGES IN ROMAN-PERSIAN WARFARE
It will have become apparent by now that the two most significant and frequent contexts in which Roman and Persian forces confronted one another were pitched battles and sieges. While there is a wealth of material available for detailed study of the latter, pitched battles pose more of a problem. As the preceding survey has indicated, it was the third and seventh centuries when warfare between the Romans and the Persians was at its most dynamic, but these are also among the least well-documented phases. It is possible to reconstruct the broad outline of events, but detailed accounts of battles from these crucial periods are lacking. As already indicated, the most detailed surviving sources for Roman-Persian warfare are the narrative histories of Ammianus and Procopius, and the chronicle of Pseudo-Joshua. Ammianus’s narrative contains much detail about Roman-Persian relations between 354 and 363, but the most detailed episode he recounts from Constantius II’s reign is the siege of Amida in 359, while Julian’s invasion of Persia in 363 has more about sieges than pitched battles since Shapur II seems deliberately to have avoided risking a major pitched battle and to have relied instead on a strategy of wearing the enemy down through a combination of burning crops, flooding land (by destroying irrigation dykes), and hit-and-run attacks. Since Pseudo-Joshua’s account also focuses on sieges, it is Procopius who provides the most detailed accounts of major pitched battles between Roman and Persian forces.
The first of these took place before the walls of Dara in June 530, with Belisarius in command of about twenty-five thousand troops, facing Persian forces of about twice that number under the command of a Persian noble, Peroz (Procop. Pers. 1.13–14, with discussion in Greatrex 1998: 169–85, Haldon 2000: 28–35). The deployment of Roman forces in front of the walls of the fortress meant that the Persians could not easily outflank the Romans and attack them in the rear, and Belisarius aimed to offset his numerical disadvantage further by digging trenches behind which he stationed his infantry. As a result the outcome of the engagement was largely determined by cavalry. Interestingly, in the preliminaries to the battle there were two Persian challenges to single combat, both won by the Roman respondent, with positive consequences for Roman morale (for details see Greatrex 1998: 177). The battle proper commenced with exchanges of archery fire before Persian cavalry engaged the Roman cavalry stationed on the wings of the infantry (Procopius’s reference to the arrows as “a vast cloud” [Pers. 1.14.35] is symptomatic of the influence of classicizing stylistic influences on his narrative). The favorable outcome for the Romans seems largely to have hinged on the timely intervention of units of Hunnic cavalry stationed at the pivotal points between the infantry and the other cavalry units on the wings. When the Roman left wing began to fall back under the pressure of the Persian advance, these Hunnic units were able to redress the balance by attacking the flanks of the advancing Persians, and then at a later point they moved across the front of the trench to reinforce the Roman right wing at a critical juncture.
The battle at Callinicum in April 531 involved a significantly smaller Persian force of about twenty thousand, confronting a comparable number of Roman troops, once again under Belisarius’s command (Procop. Pers 1.18; John Malalas, Chronicle: 462–5, with discussion in Greatrex 1998: 200–7). Procopius portrays Belisarius as reluctant to engage the Persians who were by this stage in the process of withdrawing from Roman territory down the course of the Euphrates, but there remains a suspicion that this may be part of Procopius’s strategy for exculpating Belisarius from the subsequent defeat. Just as the walls of Dara had created an immovable object along one side of the battlefield the previous year, so this time the Euphrates lay along the left flank of the Roman forces. In contrast to Dara, however, Belisarius seems this time to have placed his cavalry in the center with infantry on the wings. This deployment did not work, since, after an initial, protracted exchange of arrow fire, the Persians strengthened their left wing and were able to overpower the infantry forces on the Roman right. The Hunnic cavalry which had played such a crucial role at Dara tried to prevent this Persian advance but when their commander Ascan was slain in the fighting, they began to give ground, and the remaining Roman forces found themselves turned and pinned against the Euphrates. The surviving cavalry dismounted and continued to offer stout resistance alongside the infantry, until nightfall offered the opportunity for withdrawal upstream. Procopius’s account places the blame for the defeat on the desertion of allied Arab forces on the right wing and the poor quality of the infantry units stationed next to them. Intriguingly and unusually, another detailed account of the battle appears in the Chronicle of John Malalas, also a contemporary, and an inhabitant of nearby Antioch. His version places the blame for the defeat on Belisarius who is presented as running for safety as soon as Persian pressure began to tell, and since Malalas appears to have based his account on the official report into the defeat it has been thought by some to be more credible. Certainly Belisarius was dismissed from his post in the aftermath—even Procopius reports this, albeit at later points in his narrative where its significance was less apparent (Procop. Pers. 1.21.2, 3.9.25)—but it has also been pointed out that its basis in an official inquiry does not necessarily guarantee the reliability of Malalas’s version, which could reflect the need to identify a scapegoat, as well as resentment against Belisarius on the part of some of his officers (cf. Greatrex 1998: 194–5). From a tactical point of view, however, although archery and cavalry once again played important roles, the ability of the remaining Roman infantry and dismounted cavalry to stand firm against Persian attacks in the latter stages of the battle shows that infantry still had an important role to play: “cavalry have seldom defeated steady infantry by frontal attack.”13
The other major set-piece occasions were sieges, which may well have been a more frequent occurrence in Roman-Persian warfare than pitched battles. Nisibis stands out as an important Roman stronghold which successfully resisted Persian sieges on three occasions in relatively close succession in the mid-fourth century, but little detail survives concerning the first two, while the sources for the third are problematic in a variety of ways, not least inconsistencies between them (Lightfoot 1988). The best documented sieges are those of another important Roman fortress, Amida on the upper Tigris, in 359 and in 502/503, as recounted by Ammianus and Pseudo-Joshua respectively—both contemporaries, with the former a direct participant (see Amm. Marc. 19.1–8, with Matthews 1989: 57–66; Pseudo-Joshua Chronicle 50–3, with Greatrex 1998: 83–94; Trombley and Watt 2000: 53–63). The sieges lasted seventy-three days and about three months respectively—testimony to Sasanian ability to sustain lengthy sieges—and ended with the taking and sack of the city by the Persians. The Persians used various methods in their efforts to effect the city’s capture. In 359 Shapur’s first tactic was apparently to try to overawe the inhabitants into surrender by displaying the size of his forces and having them surround the city walls, motionless and silent. He then tried frontal assaults, before finally resorting to the slower but more methodical approach of deploying siege towers with catapults and the construction of a siege ramp. In 502 Kavad had a siege ramp built at an early stage, up which a battering ram was brought to weaken the walls. It was also in the context of sieges that the Persians found the most useful application of elephant power, as mobile siege towers (Rance 2003: 368–9; they feature in this capacity at Amida in 359: Amm. Marc. 19.2.3). In addition to the obvious use of catapults against the besiegers, the responses of Roman forces within the city included (in 359) the construction of a counter ramp and the making of sallies to try to disrupt the besiegers’ progress, and (in 502) counter-tunneling (with some success) to undermine the enemy ramp.
Amida was eventually captured in 359 because Roman attempts to counter the Persian siege ramp eventually failed. By contrast, the city’s capture in January 503 was more anticlimactic: after three months of siege and with the onset of winter, the Amidenes apparently grew lax in their monitoring of enemy activities and failed to anticipate a simple Persian nighttime foray with ladders which allowed them to seize control of a section of the city walls. There were suspicions that treachery had played a part in 503, and a deserter in 359 who helped Persian archers gain temporary control of a tower one night almost succeeded in bringing that siege to an earlier end. Of course as the length of time the city was invested dragged on, so growing food shortages took an increasing toll on the physical stamina of soldiers and civilians, with disease also breaking out in 359—a reminder that sieges were the occasions in which warfare impinged most directly on the lives of noncombatants (Lee 2007: 133–8). That impact was of course felt even more keenly when the city was finally captured: the length of both sieges meant that on neither occasion was the Persian king inclined to restrain his soldiers from random slaughter. Civilians who survived were carried off for resettlement in Persia (cf. Lieu 1986, Kettenhofen 1994).
In addition to the literary accounts of Ammianus and Pseudo-Joshua, archaeological evidence also has an invaluable contribution to make to an understanding of siege warfare on the eastern frontier in late antiquity. This evidence derives from the fort of Dura-Europos, originally a Parthian outpost on the Euphrates which was then taken over by the Romans during the second century (Millar 1993: 467–71 provides a brief overview of the site’s history). The fort was eventually captured by Persian forces in 256/7 and promptly abandoned, so that when French and American archaeologists undertook excavations in the 1920s, they unearthed a wealth of material illuminating the history of the site in Roman times—not just the structures of buildings, but also wall paintings, inscriptions, and even papyrus documents (see Hopkins 1979 for a history of the excavations). Excavations also revealed the remains of Persian tunneling activities during their siege of the fort in the mid-250s, and the attempts of the defenders to counter those efforts with their own tunnels. What has given these tunnels particular interest is that they were found to contain human skeletons wearing armor and bearing weapons.
In the part of the gallery extending in front of the mud brick embankment to the east the bodies of sixteen or eighteen soldiers were found with the remains of their armour and clothing…. In the easternmost part … the skeletons lay in contracted positions as if the men had tried to save themselves from a cave-in or had been crushed in positions of defense. One man appears to have been seated, his spinal column being markedly curved. Another lay, thrown backward, with his legs spread wide apart and folded under him as if he had made an attempt to rise. (du Buisson 1936: 194–5)
The original excavator suggested a macabre scenario in which Persian soldiers had broken into the Roman counter tunnel and in their haste to prevent Persian entry to the fortress, defenders had sealed up the counter tunnel while some of their own men were still inside, the tunnel subsequently collapsing on top of them. A recent reexamination of the evidence has prompted a revised interpretation involving Persian tunnelers deliberately collapsing their tunnel on top of Roman pursuers after a grisly underground fight in the dark. Whichever version is correct, however, there can be no doubting the overall conclusion: “These gruesome deposits bring us as close as archaeology ever has to the immediacy, and the real horror, of ancient combat” (James 2005: 204).
The eastern frontier of the empire during late antiquity offers some of the most intriguing insights into warfare in the Roman world, not least because of the diverse and detailed character of the surviving evidence and because it presents a rare opportunity to observe Roman military power confronting an opponent with a comparable level of capabilities and resources. Since the particular episodes highlighted above have (not by deliberate design) mostly shown Roman forces on the back foot, it is worth concluding by emphasizing that both sides experienced their fair share of successes and setbacks and that, apart from the significant but short-lived Persian inroads in the mid-third and early seventh centuries, the frontier between them was characterized by a remarkable degree of stability across such a substantial period of time. That stability can be attributed to a range of factors, including developments in military technology and tactics on both sides which cancelled out the other’s advantages, but also to the steady evolution of diplomatic mechanisms for the defusing and resolution of disagreements—a valuable reminder of the wider context within which warfare needs to be considered.14
BIBLIOGRAPHY
Adams, R. M. 1965. Land behind Baghdad: A history of settlement in the Diyala plains. Chicago.
———. 1981. Heartland of cities: Surveys of ancient settlement and land use on the central floodplain of the Euphrates. Chicago.
Bivar, A. D. H. 1972. “Cavalry tactics and equipment on the Euphrates frontier.” DOP 26: 273–91.
Blockley, R. C. 1989. “Constantius II and Persia,” in C. Deroux (ed.), Studies in Latin literature and Roman history. Vol. 5. Brussels, 465–90.
———. 1992. East Roman foreign policy: Formation and conduct from Diocletian to Anastasius. Leeds.
Börm, H. 2007. Prokop und die Perser: Untersuchungen den römisch-sasanidischen Kontakten in der ausgehenden Spätantike. Stuttgart.
Brunt, P. A., and J. M. Moore (eds.). 1967. Res gestae divi Augusti. The achievements of the divine Augustus. Oxford.
Cameron, A. 1985. Procopius and the sixth century. Berkeley.
Cameron. (ed.). 1995. The Byzantine and early Islamic near east. Vol. 3: States, resources and armies. Princeton.
Canepa, M. P. 2009. The two-eyes of the earth: Art and ritual of kingship between Rome and Sasanian Iran. Berkeley.
Coulston, J. N. C. 1986. “Roman, Parthian and Sassanid tactical developments,” in Freeman and Kennedy 1986: 59–75.
Dennis, G. T. (trans.). 1984. Maurice’s Strategikon: Handbook of Byzantine military strategy. Philadelphia.
de Souza, P., and J. France (eds.). 2008. War and peace in ancient and medieval history. Cambridge.
Dignas, B. and E. Winter. 2007. Rome and Persia in late antiquity: Neighbours and Rivals. Cambridge.
Dodgeon, M. H., and S. N. C. Lieu. 1991. The Roman eastern frontier and the Persian Wars A.D. 226–363: A documentary history. London.
du Mesnil du Buisson, R. 1936. “The Persian mines,” in M. Rostovtzeff, A. R. Bellinger, C. Hopkins, and C. B. Welles (eds.), The Excavations at Dura-Europos: Preliminary Report of Sixth Season of Work, October 1932–March 1933. New Haven, 188–205.
Eadie, J. W. 1967. “The development of Roman mailed cavalry.” JRS 57: 161–73.
Freeman, P. M., and D. L. Kennedy (eds.). 1986. The defence of the Roman and Byzantine east. Oxford.
Frye, R. N. 1984. The history of ancient Iran. Munich.
Galinsky, K. 1996. Augustan culture: An interpretive introduction. Princeton.
Ghirshman, R. 1962. Iran: Parthians and Sassanians. Trans. S. Gilbert and J. Emmons. London.
Göbl, R. 1974. Der Triumph des Sasaniden Sahpuhr über die Kaiser Gordianus, Philippus und Valerianus: die ikonographische Interpretation der Felsrelief. Vienna.
Greatrex, G. 1998. Rome and Persia at war 502–532. Leeds.
———, and S. N. C. Lieu. 2002. The Roman eastern frontier and the Persian wars. Part II. A.D. 363–628: A narrative sourcebook. London.
Haldon, J. F. 2000. The Byzantine wars: Battles and campaigns of the Byzantine era. Stroud.
Hamilton, W. (ed. and trans.). 1986. Ammianus Marcellinus. The later Roman empire (A.D. 354–378). With Introduction and Notes by A. Wallace-Hadrill. Harmondsworth.
Herrmann, G., and D. N. Mackenzie. 1989. The Sasanian rock-reliefs at Naqsh-i Rustam: Naqsh-i Rustam 6, the triumph of Shapur I. Berlin.
Hopkins, C. 1979. The discovery of Dura Europos. New Haven.
Howard-Johnston, J. 1995. “The two great powers of late antiquity: A comparison,” in Cameron 1995: 157–226.
———. 1999. “Heraclius’ Persian campaigns and the revival of the east Roman empire, 622–30.” War in History 6: 1–44.
Inostracev, C. A. 1926. “The Sasanian military theory.” Journal of the K. R. Cama Oriental Institute 7: 7–52.
Isaac, B. 1992. The limits of empire: The Roman army in the east. Rev. ed. Oxford.
James, S. 2005. “The deposition of military equipment during the final siege at Dura-Europos, with particular regard to Tower 19 countermine.” Carnuntum Jahrbuch 2005: 189–206.
Kaegi, W. E. 2003. Heraclius, emperor of Byzantium. Cambridge.
Kennedy, D., and D. Riley. 1990. Rome’s desert frontier from the air. London.
Kettenhofen, E. 1994. “Deportations in the Parthian and Sasanian periods.” Encyclopaedia Iranica, VII. 3, 297–308.
Lee, A. D. 1991. “The role of hostages in Roman diplomacy with Sasanian Persia.” Historia 40: 366–74.
———. 1993. Information and frontiers: Roman foreign relations in late antiquity. Cambridge.
———. 2007. War in late antiquity: A social history. Oxford.
———. 2008. “Treaty-making in late antiquity,” in de Souza and France 2008: chap. 6.
Lieu, S. N. C. 1986. “Captives, refugees and exiles: A study of cross-frontier civilian movements and contacts between Rome and Persia from Valerian to Jovian,” in Freeman and Kennedy 1986: 475–505.
Lightfoot, C. S. 1988. “Facts and fiction—The third siege of Nisibis (A.D. 350).” Historia 37: 105–25.
MacDermot, B. C. 1954. “Roman emperors in the Sasanian reliefs.” JRS 44: 76–80.
Matthews, J. F. 1989. The Roman empire of Ammianus. London.
Mielczarek, M. 1993. Cataphracti and Clibanarii: Studies on the Heavy Armoured Cavalry of the Ancient World. Trans. M. Abramowicz. Łódź.
Millar, F. 1993. The Roman near east, 31B.C.–A.D. 337. Cambridge.
Oates, D. 1968. Studies in the ancient history of northern Iraq. London.
Pollard, N. 2000. Soldiers, cities and civilians in Roman Syria. Ann Arbor.
Potter, D. S. 1990. Prophecy and history in the crisis of the Roman empire: A historical commentary on the thirteenth Sibylline oracle. Oxford.
Rance, P. 2003. “Elephants in warfare in late antiquity.” Acta Antiqua Hungarica 43: 355–84.
———. 2005. “Narses and the Battle of Taginae (Busta Gallorum) 552: Procopius and sixth century warfare.” Historia 54: 424–72.
Shepherd, D. 1983. “Sasanian art,” in E. Yarshater (ed.), The Cambridge history of Iran. 3 vols. Cambridge, 3, 2: 1055–1112.
Smith, R. 1999. “Telling tales: Ammianus’ narrative of the Persian expedition of Julian,” in J. W. Drijvers and D. Hunt (eds.), The late Roman world and its historian: Interpreting Ammianus Marcellinus. London, 89–104.
Trombley, F. R., and J. W. Watt, 2000. The chronicle of pseudo-Joshua the stylite. Liverpool.
Whitby, M. 1988. The emperor Maurice and his historian: Theophylact Simocatta on Persian and Balkan warfare. Oxford.
———. 1994. “The Persian king at war,” in E. Dąbrowa (ed.), The Roman and Byzantine army in the east. Cracow, 227–63.
Whitehouse, D., and A. Williamson. 1973. “Sasanian maritime trade.” Iran 11: 29–49.
Widengren, G. 1976. “Iran, der große Gegner Roms: Königsgewalt, Feudalismus, Militärwesen.” ANRW II.9.1, 219–306.