The sixth-century army of Justinian’s era, like its earlier counterparts, was an entirely professional force, but it no longer conformed to the patterns of the Roman army of Caesar or Augustus: overwhelmingly a force of heavy infantry, divided into legions composed of Roman citizens supported by non- Roman auxiliaries. The classic Roman legion of the early empire numbered about five thousand soldiers organized into ten cohorts, each commanded by a centurion, with more or less the same number of non- citizen auxiliary troopers organized in supporting infantry cohorts and cavalry alae (wings). The number of legions slowly increased from the time of Augustus until, in the Severan period in the early third century, it reached a grand total of thirty-three, implying a total paper strength, with an equal quantity of supporting auxiliaries, of around 350,000 men. More or less the entirety of this military establishment was distributed along the empire’s vast frontiers: in northern Britain, along broadly the rivers Rhine and Danube on the European continent, and in Mesopotamia and Armenia facing up to the Persians, while smaller forces patrolled the desert fringes of Egypt and the rest of North Africa as far west as modern Morocco. When larger forces were required for major campaigns, contingents were pulled together from all the legions within reach, but whole legions- each a small expeditionary force in its own right- were moved around the empire only occasionally. By the time of Justinian, the Roman army had changed out of all recognition under the pressure of two sequential periods of military crisis.
The nearest fully comprehensive listing of the Roman army’s order of battle to the time of Justinian is preserved in the eastern portions of the famous Notitia Dignitatum, dating to the 390s. Fifth-century legal materials dealing with military matters and some more episodic pictures of the East Roman army in action provided by fifth- and early sixth- century narrative sources make it clear, however, that the basic pattern of military organization did not alter in the intervening 130 years. Periods of heavy fighting could destroy individual units, and new threats demanded particular recruiting efforts. Sixteen regiments of heavy East Roman infantry were never reconstituted after their destruction at the battle of Hadrianople in August 378, and the Hunnic wars of the 440s both caused heavier casualties and occasioned major recruitment drives in Isauria (south- central Anatolia). But if individual units came and went, the over- all shape of East Roman military organization remained broadly stable. By the late fourth century and on into the mid- sixth, the old pattern of large legionary units stationed at intervals along the major frontier lines had given way to a more complex set of unit structures and dispositions. There were now three broad types of East Roman army grouping: in descending order by status, central (`praesental’) field armies, organized in two separate corps each with its own commanding general (Magister Militum Praesentalis); three regional field armies (one in Thrace, one in Illyricum, the third on the Persian front, each again with its own Magister Militum); and a whole series of frontier guard troops (limitanei) stationed in fortified posts on or close to the frontier line. The last were organized in more local, regional clusters each commanded by a dux (`duke’).
The number and type of military unit found within each grouping had also evolved. The word `legion’ survived in the title of many units, particularly of the limitanei, some of which were direct descendants of very old formations. Legio V Macedonica had originally been raised by Julius Caesar in 43 BC; it still existed in Egypt in the seventh century ad. Like all its late Roman peers, however, it had become a completely different type of unit, for which the standard term was now numerus in Latin, arithmos in Greek. No individual late Roman unit was anything like as large as the old legion of 5,000 men (about the size of a modern brigade). We don’t have exact information, but even the notional manpower of larger infantry formations was no more than 1,000 to 1,500 (much more like a regiment). There were also many more cavalry units in both the frontier limitanei and in the regional and praesental field armies than there used to be; these were smaller still, consisting of no more than 500 men.
The old binary divide between citizen legionaries and noncitizen auxiliaries, likewise, had been replaced by three main categories of soldiers, who received differing rates of pay and enjoyed varying grades of equipment. The highest- ranking palatini and second- ranked comitatenses were distributed across the central and regional field armies, while frontier forces were composed of limitanei and/ or ripenses. Differences in status materially affected military capacity. When a cavalry unit operating against desert raiders in Cyrenaica was downgraded from field army status (as comitatenses) to limitanei, it lost the right to the extra remounts and supplies, making it potentially much less effective against trouble- some desert raiders, much to the chagrin of Synesius of Cyrene. The men themselves presumably also didn’t much enjoy the resulting pay cut. But it is a mistake to write off the effectiveness of limitanei altogether. It used to be fashionable to envisage them as part- time soldier farmers who would have struggled to cope with anything more demanding than a little patrol- ling and the odd customs inspection. But while it is conceivable that their state of readiness and overall training may have varied substantially on different frontiers, the limitanei of the eastern and Danubean fronts were battle hardened. Warfare in the East largely took the form of extended sieges, and in this theatre the garrison forces of many of the major Roman fortresses were composed of limitanei. As such, they bore the brunt of much of the initial fighting in many campaigns. The same was also true of the Danubean front, where heavy fighting had been endemic throughout the fifth century. For really major campaigns, units of limitanei were sometimes also mobilized alongside designated field army formations.
Much of this reorganization can be traced back to the period of extended military and political instability known as the third-century crisis. The fundamental destabilizing factor here was the rise of Persia to superpower status under the Sassanian dynasty, which displaced its Arsacid rivals in the 220s and found new ways to unite the massive resources of what are now Iraq and Iran and turn them against Roman possessions in the East, with extremely negative effects upon the overall strategic position of the Roman Empire. The third- century Persian King of Kings Shapur I (ad 240/ 2- 270/ 2) set out the record of his achievements in a great rock inscription, the Res Gestae Divi Saporis.
I am the Mazda-worshipping divine Shapur, King of Kings . . . , of the race of the Gods, son of the Mazda- worshipping divine Ardashir, King of Kings. . . . When I was first established over the dominion of the nations, the Caesar Gordian from the whole of the Roman Empire . . . raised an army and marched . . . against us. A great battle took place between the two sides on the frontiers of Assyria at Meshike. Caesar Gordian was destroyed and the Roman army was annihilated. The Romans proclaimed Philip Caesar. And Caesar Philip came to sue for peace, and for their lives he paid a ransom of 500,000 denarii and became tributary to us. . . . And the Caesar lied again and did injustice to Armenia. We marched against the Roman Empire and annihilated a Roman army of 60,000 men at Barbalissos. The nation of Syria and whatever nations and plains that were above it, we set on first and devastated and laid waste. And in the campaign [we took] . . . thirty- seven cities with their surrounding territories. In the third contest . . . Caesar Valerian came upon us. There was with him a force of 70,000 men. . . . A great battle took place beyond Carrhae and Edessa between us and Caesar Valerian and we took him prisoner with our own hands, as well as all the other commanders of the army. . . . On this campaign, we also conquered . . . thirty- six cities with their surrounding territories.
It actually took the Roman Empire three political generations to recover from this cataclysm of humiliating defeats and restore balance to the eastern front and thereby to its own internal workings. The most immediate level of response, as you might expect, was a revolution in the empire’s overall military capacity. Some of this came in the form of new unit types. Persian elite forces of the third century AD characteristically took the form of heavily armed lancers-cataphracts-who were responsible for much of the carnage inflicted on the armies of Gordian, Philip, and Valerian. In response, Rome substantially increased the number of cavalry units at the disposal of its commanders and, in particular, created from scratch a number of heavily armoured cavalry units, the plate-mailed clibanarii. These units still formed part of the Eastern praesental field armies at the end of the fourth century.
For the most part, however, the response took the form of a huge expansion in the size of the traditional heavy infantry arm of the Roman military. Because the notional paper strength of the new unit types is far from certain, the exact scale of this expansion is impossible to calculate. But a whole range of evidence, from the size of extant barrack blocks to pieces of specific information, provides the basis for worthwhile calculation. From these materials, no serious student of the late Roman army thinks that its notional manpower strength increased by less than 50 per cent in the century after 230, and a pretty good argument can be made that it actually doubled in size. There could be no more eloquent testimony to the scale of the strategic problem posed by the emergence-better, perhaps, reemergence (Shapur’s great inscription was placed near the tombs of the great Achaemenid kings of antiquity, Darius and Xerxes)- of Persia as a rival superpower to Rome. As a result of this expansion, the Persian threat had been broadly contained by the turn of the fourth century. The first serious Roman victories came in the final decade of the third century, and while one side or the other often held a short-term advantage in subsequent years, the fourth century saw no repetition of the stunning victories recorded by Shapur I.
But the effects of increasing Persian power and of consequent Roman military expansion were felt not just on the battlefield. The rise of Persia to superpower status gave a new importance to the eastern front, which in the longer term destabilized existing political balances of command and control within the empire as a whole. Once Persian power became such a basic fact of life, it demanded imperial- level oversight be available more or less constantly for the eastern front, since only an emperor could safely command the kind of resources that war making in this theatre now required. In the Notitia Dignitatum, about 40 per cent of the entire Roman imperial army was positioned to deal with a potential Persian threat, and this was far too large a force to leave under the control of an unsupervised general, since few could resist the opportunity that such an army provided to bid for the imperial throne. Moreover, given the enormous size of the empire, stretching from Scotland to Iraq, and the catatonically slow speed of movement – Roman armies could move on average twenty kilometres a day for three to four days at a time before needing a rest day – this in turn meant, in practice, that an additional source of command and control had to be available for the empire’s other major European fronts, where a smaller but nonetheless significant increase in the level of threat posed by the new, largely Germanic- dominated confederations of the Rhine and Danube was another characteristic feature of the late imperial period.
After a long period of experiment in the third century, punctuated by repeated usurpations as under supervised generals made successive bids for the purple, the result was a general tendency in the late imperial period- for as long as the Western Empire remained in existence- for political power to be divided between two or more emperors. The knock- on political effects of military reorganization also help explain the relatively complex structure of central and regional field armies. Logistics meant that regional commanders required sufficient forces to respond to most `normal’ levels of threat. It generally took at least a year to concentrate the necessary food supplies and animal fodder and then move the actual troops required for major campaigns, and this was obviously far too long a delay for most frontier problems. But since army commanders also had a long track record of usurpation, emperors wanted to make sure that individual generals did not have so many troops at their disposal that they could easily make a bid for the throne. The field army organization of the fourth to the sixth centuries can be seen as compromise. It redistributed elite portions of the army to allow for quicker, more effective responses to the new strategic demands of the late Roman period but tempered the potential political consequences by carefully dividing units, even of the central, praesental field army, between two separate commanders whose political influence could be counted on more or less to cancel each other out.
The same kind of balance is also visible in another military innovation which had become a characteristic feature of East Roman armies by the time of Justinian. It is not clear when exactly it emerged, but by the sixth century field army generals, the Magistri Militum, all seem to have had substantial forces of officers and soldiers- `guardsmen and spearmen’, as Procopius calls them-who were recruited by them personally and tended to follow their generals on campaign even to far-flung corners of the Mediterranean. Belisarius’s guards served with him in the East, in Africa, and in Italy, and when the commanding general in Armenia was assigned to the Balkans in preparation for an Italian campaign, his guards came with him. The normal term for these soldiers is bucellarii, and the institution clearly grew out of the tendency of the great and good, military and civilian, in the late Roman world to maintain personal armed retinues. The bucellarii of the sixth- century Roman military, however, were different. They were supported at least in part out of state funds (although rich generals, such as Belisarius became, might also employ some of their own money in recruiting and equipping their men, just like richer ships’ captains in Nelson’s navy), and they swore an oath of loyalty to the emperor as well as to their own general. The state funding increased their numbers- at one point Belisarius’s guards amounted to 7,000 men, but 500 to 1,500 seems much the more usual range- and rather than think of them as an expanded personal retinue, they are better understood as elite striking formations whose permanent attachment to successful generals (successful at least in the sense of having been promoted to Magister Militum) meant that they enjoyed higher levels of training and equipment. It is also clear that by the sixth century, bucellarii were being recruited from both outside `barbarians’ and the empire’s own citizens. Here, too, we see the desirability of heightened military effectiveness being balanced by the necessity of preventing individual generals from becoming politically dangerous.
If the size, geographic distribution, and command structure of Justinian’s army can be traced back to the military convulsions of the third century, its unit forms and tactical doctrines had their origins in a quite separate crisis. From the late fourth century, the rise of Hunnic power in eastern and central Europe generated an unprecedented level of threat liability for service- but with the added proviso that the foederatii could preserve their own existing communal and political structures and would always serve under their own leaders. The use of mercenary contingents from beyond the imperial frontier, hired in for particular campaigns, also remained a regular feature of the sixth- century East Roman army. Procopius records a whole range of such contingents, from groups as diverse as the Germanic- speaking Lombards of the Middle Danube to the Turkic- speaking Bulgars (whom he calls Massagetae) from north of the Black Sea. But the empire continued to maintain largely autonomous groups of foederatii on Roman soil, too, even after the departure of the Thracian Goths for Italy in 488, with Heruli in particular playing an important role in Justinian’s campaigns.
In the long term, however, the most important military response to the era of Hunnic domination was tactical. The Romans first met the Huns as small- scale cavalry raiders equipped with a more powerful version of the reflex bow, which had long been a characteristic weapon of Eurasian steppe nomads. This gave different Hunnic groups sufficient military edge rapidly to establish hegemony over large numbers of the semi- subdued, largely Germanic- speaking clients of Rome- Goths and others- who controlled the territories beyond the defended imperial frontier. As a result, the military problem posed by the Huns in the era of Attila evolved into a much more complex one, since the great Hunnic warlord disposed of the combined forces of both the Hunnic core of his empire and of a host of conquered subject peoples: other steppe nomads, such as the Alans, and the largely infantry forces of Germanic Goths, Gepids, Suevi, Sciri, and others. The range of weaponry that Attila could deploy was accordingly varied; it encompassed mounted archers to heavy, mailed shock cavalry equipped with lances to dense groupings of infantry.
The full story of all the experimentation which underlay the Roman adaptation to new patterns of warfare in the Hunnic era cannot be recovered, but its overall effect upon the sixth-century army emerges clearly from the battle narratives of Procopius’s histories and contemporary military manuals, above all the Strategicon of Maurice. As seen in action in these texts, the East Roman army of the sixth century was characterized by a much greater reliance upon its cavalry arm. Now often deployed in the front of the battle line instead of just as flank protection (as had still generally been the case in the fourth century), it comprised two distinct elements. Occupying the van were the lighter cavalry (koursoures in the terminology of the Strategicon) characteristically armed with Hunnic- type reflex bows, whose archaeological remains, in the form of bone stiffeners, start to appear in Roman military contexts in the early fifth century. The koursoures were the first to engage an enemy, using their projectile weaponry at least to inflict some initial losses on an enemy or, at best, to spread disorder in his tactical formations. If this initial assault was successful, the heavier shock cavalry- defensores- could then be deployed almost literally to ram home the advantage. They were armed not only with bows but with cavalry lances to break up an opposition line. Alternatively, if the koursoures ran into trouble, the heavy cavalry would cover their retreat. Procopius’s battle narratives indicate that the new elite cavalrymen of the sixth- century army tended to be concentrated in the bucellarii of the Magistri Militum, but regular field army cavalry units, and some of the foederatii too, were intensively trained in the new battlefield practices.
The bucellarii of field army generals also provided the key military structure of institutional continuity which allowed new weaponry and the tactics to exploit them fully first to be developed and then passed on across the generations. This is partly an argument from silence. There were no officer training schools or military academies in the later Roman Empire where they might have been able to develop new doctrines by discussion in the classroom, which is how modern armies operate. But it is a bit more than that, too. The bucellarii, the new elite arm of the Roman army of the sixth century, enjoyed the highest rates of pay and best equipment on offer from the state factories (not to mention any extras that their often rich commanders chose to provide), so that they could generally attract the best recruits. The officer cadres of the bucellarii were also a source of new field army generals. At least two of Justinian’s initial tranche of his own appointees to the rank of Magister Militum in command of key field army formations in the late 520s- not only Belisarius who will play such an important role in this book but Sittas as well- had served in his bucellarii when the future emperor first held the rank of Magister Militum Praesentalis in the early 520s; several of Belisarius’s household and underofficers from the original African campaign would find promotion to the rank of magister in turn as the reign progressed. Not only were the bucellarii a key element in their own right of the new model East Roman army of the sixth century; they also trans- mitted military expertise across the generations.
Even if the most striking feature of this military revolution was its transformation of the role and equipment of Roman cavalry, it did also affect the battlefield operations of the infantry. Both the lighter and heavier cavalry units were trained to operate in integrated fashion with the infantry, which remained the largest element in every Roman field army and whose tactics and equipment had also been revamped accordingly. The latest interpretation suggests that defensive armour was indeed lightened- as the military commentator Vegetius complained in the later fourth century- but to increase the infantry’s battlefield mobility so that it could work in more integrated fashion with the rapidly developing cavalry arm. The range of infantry equipment was also increased to include more bows and other projectile weaponry so that foot regiments could perform a wider variety of roles: everything from reinforcing and driving home a tactical advantage created by successful cavalry assault to providing a strong covering force should the horsemen be forced to retreat. Experience of combat in the Hunnic era eventually taught Roman commanders that it was no use operating the infantry in dense, relatively static formations, since Hunnic- style horse archery was likely to cause mayhem within the massed ranks before the heavy infantry’s brute force could be brought tellingly to bear at close quarters. The infantry had to become more mobile and less vulnerable to sustained missile and cavalry attack and, by the time of Justinian, had been reordered accordingly. By this stage, it even operated with portable anti-cavalry barricades-munitiones, as an early sixth-century commentator labels them- to help protect it from the unwelcome attention of horse archers.
Two strategic crises, therefore, shaped the armed forces available to the Emperor Justinian on his accession to the throne in 527. The old heavy infantry legions which had conquered an empire had been forced to adapt: numerically, to the threat posed by a newly united Persian superpower in the third century, and tactically, to the intrusion of large numbers of steppe nomads into eastern and central Europe in the later fourth and fifth. Such was the importance of war making in both practical and ideological terms to the overall functioning of the empire that a military revolution on this scale was bound to have equally profound effects on the workings of its internal structures.