Army of the Roman Dominate (284 - 480 CE)

During the Principate, the primary distinction between different units of the Roman army was whether they were designated as legionaries (legiones) or auxiliaries (auxilia). The legions of the Principate were based on the levies of the Republic, built around a semi-flexible phalanx of main battle infantrymen, with some ancillary units of infantry scouts, artillerists, and cavalrymen who were drawn from among the members of each legion and detailed to fulfill different tactical functions. Emphasis was laid on the self-sufficiency of each legion—although legions could be combined to create larger armies, each legion was considered an army unto itself, capable of acting alone, and as such the size of each legion was large (approximately 5,000 men). However, Rome’s first emperor—Octavius Augustus (27 BCE-14 CE)—recognized that the empire he inherited from the Republic had strategic needs that could not be met by stretching the legions too thinly, and the population of Roman citizens (cives Romani) was not large enough to be able to sustain a massive buildup in the number of legions fielded by the Empire (especially considering the heavy casualties among the citizen legionaries suffered during the civil wars that ended the Republic). Not to mention the fact that the Republic had significantly under-utilized its human resource base by not allowing provincials (peregrini) to serve in the regular military, and thereby be given a stake in sustaining the Roman socio-political order (i.e., by providing a viable path to Roman citizenship). Thus, from the time of Augustus, the Roman legions were supplemented with auxiliary units recruited from among the provincial populations of the empire. Although it is important to note that some auxiliary units were designed to expand the tactical repertoire of the Roman army with specialist units—particularly various types of cavalry, slingers, and bow-armed infantry—most of the auxiliary infantry units of the imperial Roman army seem to have been equipped and trained in a virtually identical manner to legionary infantrymen. The primary distinction between legionaries and auxiliaries was social and not tactical. Even those specialist units of auxiliaries that expanded the tactical repertoire were mostly of types that had existed in the Republican legions, but the delegation of these roles to auxiliaries both expanded the number of men that could fill these roles and allowed the legions to increasingly specialize in their primary tactical role as close combat infantrymen. I think it is instructive to note that the literal translation of auxilia is “helpers,” and this seems to encapsulate how the auxiliary units were conceived—as adjuncts to the legions that supplemented their numbers. However, in 212 CE the emperor Caracalla issued an edict—the Constitutio Antoniniana—that made all free-born men within the borders of the Roman Empire full Roman citizens, and this would have erased the main distinction between legionaries and auxiliaries in the military.

 

Soon after Caracalla’s edict, the empire fell into what historians call the Third Century Crisis (235-284 CE). Unfortunately, the many calamities suffered by the Roman population during this period, which led to huge demographic and institutional changes, are accompanied by a corresponding drop in the number and reliability of primary sources for the history of the period. The army of the Dominate that emerged at the end of this period seems to have been the result of the culmination of a large number of initially ad hoc adaptations that were institutionalized by the Illyrian emperors (ca. 268-305 CE). Infantry units still seem to have used the designations, legiones and auxilia, although these seem to have become entirely semantic distinctions (i.e., they perhaps reflected a given unit’s origins as a legionary or auxiliary unit during the Principate, but there was no discernible social, strategic, or tactical difference). The army of the Dominate still had a basic division between two classes of soldiers, but this division reflected the changing social and power dynamics of the new regime that was consolidated under Diocletian (284-305 CE). In the army of the Dominate, the primary distinction was between units that were part of the mobile armies (comitatenses, “companions”) or that were part of the frontier armies (limitanei, “border men”). As in the army of the Principate, however, we have no direct evidence that there were any consistent differences in the equipment and training of the constituent units that fell into each of these categories. As a general principal, it is believed that the units that were classed as comitatenses were likely better equipped and more professional than those that were classed as limitanei, simply by virtue of the fact that financial and military resources in the Dominate were concentrated around the figures that commanded the comitatenses, especially the emperors themselves, while limitanei were essentially provincial militiamen dependent on limited local resources (material and manpower). But some of the empire’s provinces were wealthier and more urban than others, and thus had greater manpower and material resources, and there were differences in the inclinations of various military commanders to finance and maintain their comitatenses or limitanei—some were military men that were scrupulous in the maintenance of the forces under their command, others were essentially social climbers for whom a military command was simply another rung on the ladder to socio-political elevation who neglected various aspects of their post. The imperial civil service was supposed to observe the performance of commanders and hold them accountable to the emperor(s), but corruption was widespread.

 

The late Roman army was essentially designed with an eye toward defense-in-depth and internal security, and thus its constituent elements were dispersed throughout the empire, and so each of the two broad groups of soldiers had a different strategic function—basically, limitanei were expected to suppress banditry, nip nascent rebellions in the bud, and repel small-scale raids by barbarians, but in the face of large-scale invasions they were mostly expected to garrison strong-points (mainly walled cities) and/or fight a guerrilla-style campaign against enemy forces until the comitatenses could arrive and draw the enemy forces into a decisive engagement (or to crush persistent rebellions). In addition to policing the provinces in which they were headquartered and assisting the limitanei, comitatenses were generally used in larger numbers in armies that campaigned beyond the borders of the empire. Basically, limitanei were viewed as being static garrisons, while comitatenses were conceived as being mobile armies. Often, when the imperial army was campaigning beyond the borders of the empire, only a few units of limitanei were brought along, and those that were seem to have been designated as pseudocomitatenses (“false companions,” i.e., in the sense of not being true comitatenses). Nevertheless, these pseudocomitatenses were often chosen from the best of the available limitanei units, or units that had a tactical stance that was viewed as being essential to the mission at hand (e.g., if the army was going to be campaigning on plains, more cavalry units would be included). However, as with the legionaries and auxiliaries of the Principate, we have no direct evidence that there were any broad tactical distinctions between similarly classified units of the army of the Dominate, regardless of whether they were comitatenses or limitanei (e.g., legions were close combat infantry formations regardless of whether they were comitatenses or limitanei, as were auxilia).

 

During the Principate, the standard size of a Roman legion (legio, plural, legiones) was approximately 5,000 men, the standard size of an auxiliary infantry cohort (cohors, plural, cohortes) was approximately 500 men, and the standard size of an auxiliary cavalry wing (ala, plural, alares) was approximately1,000 men. There did exist some auxiliary units that had a larger compliment of soldiers, but these were generally exceptional. However, the peripheral deployment of the bulk of the empire’s military during the Principate, coupled with the concentration of relatively large numbers of troops in each legion, made it difficult for the Roman army to quickly react to destabilization (e.g., rebellions) within the empire, or to react to incursions or invasions that entered the empire at remote locations (and most small-scale incursions tended to avoid the legionary and auxiliary fortresses). It also often placed a large amount of material and manpower resources in the hands of a handful of generals who could easily make a play for the throne. The private army of the emperor—the Praetorian Guard (cohortes praetoriae or Praetoriani)—only partly compensated for this, since most of the Praetorians were stationed in Italy and their primary task was to protect the emperor. Some emperors were reluctant (for various reasons) to dispatch Praetorians to the provinces, and in general the Praetorians usually only took the field when the emperor was in command, but Italy was also far away from many areas of the empire, and thus the Praetorians were also generally not particularly well-placed to quickly react to rebellions or invasions in distant provinces, even if the emperor was willing to use them in such a capacity. In order to avoid abandoning their post on the frontiers, many Roman generals chose to detach sub-groups of legionaries and auxiliaries to form semi-independent “flag units” (vexillationes, singular, vexillatio) that could travel to a hot-spot to suppress a small-scale rebellion or repel a small-scale incursion. These vexillations would be commanded by one of the general’s adjutants or sub-commanders, and the unit would be given its own battle standard (a flag called a vexillum, from which the type of unit took its name). The size of these vexillations seems often to have been around 1,000 men. Initially, these ad hoc commands were short-term—once a specific mission was accomplished, the surviving members of the unit returned to their unit of origin—but as the challenges faced by the empire gradually developed into the calamities of the Third Century Crisis, vexillations seem to have increasingly become more-or-less permanent. The net result was that the manpower resources of the legions and auxiliary cohorts were reduced and dispersed to different locales within the empire, while any distinction between legionary and auxiliary soldiers evaporated, both because vexillations often included both legionary and auxiliary soldiers, and because Caracalla’s edict (cited above) made all free-born men within the empire full Roman citizens. This shift seems to have become institutionalized in the army of the Dominate, in which the primary organizational formation became the vexillatio of approximately 1,000 men (although the Romans do not seem to have been consistent in this, and vexillation sizes actually ranged between about 800 and 1,200 men). The bulk of both cavalry and infantry units, comitatenses and limitanei, were organized into vexillations. Sub-groups of the vexillations were identified as ala (cavalry), cuneus (cavalry and infantry), or cohors (infantry). These smaller formations could be sub-units of a vexillation or detachments of a vexillation or fully independent units within the strategic command structure. Their size varied between about 80 and 500 men. Vexillations were generally commanded by a tribunus, cavalry alae and cunei were commanded by a decurion, and infantry cunei and cohortes by a centurio princeps. Internally, most of these units were divided into sub-groups of 100 men each commanded by a centurio.

 

Most of what we know about the strategic organization and structure of the high command of the late Roman army comes from a remarkable document called the Notitia Dignitatum (“List of Offices”). Compiled in the late fourth and early fifth centuries CE, the Notitia gives us a snapshot of the military organization of the late Roman army—it lists several thousand imperial court and provincial civil service offices and enumerates their resources and duties, and it lists all of the army commands, and has pictures of the emblems (indicia) that were used to identify different units (see below). Unfortunately, we are often left guessing about the tactical stance of each of these units, and their size, although some clues to the tactical stance of the unit are given (e.g., pictures of the weapons and armor utilized by the unit are sometimes provided), and corroborating evidence from other sources sometimes provides us with greater details. In the Western Empire (Patres Occidentalis), there were eight mobile armies (comitatenses) located at Ravenna (Italy), Milan (Italy), Paris (Gaul), Sirmium (Illyria), Carthage (Tunisia), Tingis (Mauritania), Merida (Spain), and London (Britain). All were commanded by a comes (plural, comites)(“companion” or “associate” of the emperor), except the mobile army at Paris, whose commander held the title of magister equituum (“master of horse”), and the senior mobile army commander at Milan, who held the title of magister utriusque militiae (“supreme commander of the army”). The emperor directly commanded the units at Ravenna, while the senior army commander at Milan served as a kind of military viceroy with authority over the other mobile army commanders, who in turn held authority over the frontier army commanders in their administrative region (diocese) of the empire. The Western Empire had fourteen frontier armies (limitanei) located in Gaul (x6), Illyria (x4), Tunisia (x2), and Britain (x2). The commanders of each of these frontier armies held the title of dux (plural, duces)(“commander”), except for the Comes Litoris Saxonici (“Count of the Saxon Shore”) who oversaw the littoral defenses of Britain, and the Comes Tractus Argentoratis (“Count of the Silver Fort”) who guarded the silver mines in the vicinity of the modern city of Strasbourg on the Rhine frontier. In the Eastern Empire (Patres Orientalis), there were seven mobile armies located at Constantinople (one in the capital and two in nearby Thrace), Sirmium (Illyria), Marcianopolis (modern Bulgaria), Antioch (Syria), and Alexandria (Egypt). Only the commander of the mobile army in Egypt held the title of comes (Anglicized as "count"), while the remainder held the title of magister militum (“master of the army”). The emperor directly commanded the mobile army at Constantinople. The Eastern Empire had 14 frontier armies located in Illyria (x2), Thrace (x2), Anatolia (x1), Syria (x7), and Egypt (x2). All the commanders of these frontier armies held the title of dux (Anglicized as "duke"), except the one in Anatolia, who held the title of Comes Isauriae (“Count of Isauria,” a border province along the Taurus Mountains in southeastern Anatolia). The Comes Isauriae also answered directly to the emperor in the eastern command hierarchy, while the duces were junior to the mobile army commanders in their administrative region (diocese) of the empire. Each of these commands had a different mix of regular army units and other units (see Irregulars below), and there were was no standard size for mobile or frontier armies (the smallest mobile army had 2,450 men, and the largest had 32,000 men; while the smallest frontier army had 250 men, and the largest had 12,500 men). The overall size of the military during this period of Roman history fluctuated wildly, from around 260,000 to 581,000 men, depending on whether you include irregulars, whether the empire was experiencing a temporary population contraction or expansion, and of course taking into account the gradual collapse of the Western Empire and its consequent gradual reduction in manpower resources. The snapshot provided by the Notitia gives us an overall army size of approximately 366,550 men (late fourth-early fifth centuries CE), with 212,700 men in the eastern army and 153,850 men in the western army.

 

Keep in mind that during most of the period of the Dominate (ca. 284-480 CE), the Roman Empire was divided into western and eastern halves, each with its own—although mostly parallel—civil and military hierarchies. Only during the Constantinian Dynasty (ca. 307-364 CE) did the empire revert to a single emperor of both east and west, and although there were four emperors during the Tetrarchy (ca. 284-306 CE), the east-west division was maintained with a senior emperor (augustus, plural, augusti) ruling each half of the empire, each with the assistance of a junior imperial colleague (caesar, plural, caesares). This does not seem to have affected the command hierarchy or division of army forces much. Each emperor, whether there was one, two, or four, had his own units of bodyguards (see below), while the size and composition of the mobile armies (comitatenses) remained fairly consistent, as did the frontier armies (limitanei).

 

Imperial Guard Units and Mobile Armies

The Praetorian Guards (Praetoriani) were the primary bodyguards and personal army of the emperors during the Principate. However, the Praetorians frequently played a political role—making and breaking emperors—so that by the third century CE, the Illyrian emperors (ca. 268-305 CE) largely sidelined the Praetorians, basically treating them as the garrison of Rome (most of the Illyrian emperors only traveled to Rome for ceremonial occasions, and neglected to bring the Praetorians with them on campaign). Under the Tetrarchy (ca. 286-337 CE), Rome remained the symbolic capital of the empire, but the Tetrarchs (i.e., the four colleague emperors) established their administrative capitals at Trier (in Gaul), Milan (in northern Italy), Thessalonica (in Thessalia), and Byzantium (in Thrace). Nevertheless, Maxentius (306-312 CE) came to power as a usurper with the support of the Praetorians in Rome, and when Constantine I (307-337 CE) defeated and killed Maxentius at the Battle of the Milvian Bridge (28 October 312 CE), Constantine thereafter disbanded the Praetorians. By that time, Diocletian had already set a precedent by elevating two of the senior legions of his army—the Herculiani and the Joviani—to the status of palatini (“palace-men”), and the other Tetrarchs followed suit with their own legionary and/or auxiliary palatini units. Constantine took this a step further, promoting several cavalry units to the status of scholae palatinae (“Palace Schools,” i.e., as in a school or squadron of fish, not a place where you learn your letters). These palatinate units formed the core around which the comitatenses (singular, comitatus) of the Dominate were built. Eventually, there were legionary infantry, auxiliary infantry, and cavalry units that were accorded either the honorific of palatini (infantry) or scholae (cavalry), which marked these units out as members of a comitatus. By the end of the Constantinian Dynasty (ca., 307-364 CE), the comitatenses had been divided into two groups—the mobile armies that were headquartered at strategic locations within the empire, and the units that became the personal bodyguards of the emperors. The mobile armies continued to be referred to as comitatenses (and some continued to be known as palatini or scholae), while the bodyguard units were referred to as candidati (singular, candidatus), which literally means “clothed in white.” The designation, candidati, may refer to the white uniforms of the soldiers in the candidati units, although this would be a little puzzling—the entire Roman army of the Dominate used a new set of livery for identification of units that was based on those of the Sassanian Persians (see below), and white was the predominant color of most of the uniforms worn by most of the soldiers of the empire at this time (both comitatenses and limitanei), so identifying the imperial bodyguards as candidati would not seem to have been based on any great visual distinction. It is possible that the candidati were required to rub their uniforms with white chalk to intensify the white color—a practice used by Roman politicians running for office—but no period source explicitly says this. However, from the time of Constantine I, the imperial court had become Christianized, the emperors were seen as the Hand of God on Earth, and the color white was associated with sanctity and purity, so it might be that the term candidati/candidatus was simply intended to emphasize that the bodyguards of the emperor were performing a sacred duty in service to God and empire.

 

Cavalry units in the comitatenses generally fell into four categories—armored close combat specialists that carried substantial shields and rode unarmored horses; heavily armored lancers that carried no shields and rode armored horses; unarmored skirmishers that wielded javelins as a primary weapon, carried small shields, and rode unarmored horses; and unarmored skirmishers that wielded bows as a primary weapon and rode unarmored horses. The majority of cavalry units in the comitatenses wore mail (hamata) or scale (squamata) short-sleeved hauberks (lorica) and ridge helmets (cranus), carried relatively large oval body shields (scuta), and used thrown javelins (lanceae or verruta) as a prelude to initiating close combat with spear (hasta) and long sword (spatha). They rode unarmored horses. There were several terms used in reference to these cavalrymen—scutarii (“shield-men”), armaturae (“armored”), or gentiles (“natives,” i.e., not barbarians). Armored cavalry lancers that imitated the equipment and tactics of Sarmatian cavalry first appeared in the Roman army during the Principate, although their armor gradually became heavier, including the introduction of horse armor, probably under the influence of the Parthians, Syrians, Armenians, Georgians, and Sassanian Persians. They generally wore a mail hauberk with a lamellar or scale vest, their helmets often included plate visors or a mail coif (i.e., a tight-fitting hood with a veil for the face), and they had segmented leg and arm defenses (manica, “sleeves”). Horses were armored in scale. The Romans used two different terms for these armored lancers—cataphractarii (“covered-men,” i.e., covered in armor) or clibanarii (“oven-men,” i.e., wearing heavy armor that would cause them to bake in the hot sun). There were no stirrups at this stage in history, so Roman lancers eschewed use of a shield in order to wield a long two-handed lance called a contus (“barge-pole”). A special four-horned saddle helped them keep their seat atop their mount. Unlike the other close-combat cavalrymen, Roman lancers generally charged straight into close combat (i.e., no prelude of thrown javelins), and used the spatha as a sidearm. During the Principate, mercenaries from North Africa (Numidia and Mauretania) and the western Balkans (Illyria and Dalmatia) gained a solid reputation as cavalry skirmishers, serving as scouts, to screen the deployment of the main battle contingents, and as flankers. These units made such a positive impression on the Romans that several formed the core around which the emperors of the Late Principate built some auxiliary cavalry squadrons. By the dawn of the Dominate, units modeled on these mounted skirmisher squadrons could be found in both the comitatenses and the limitanei. Although these units had originally been ethnically distinct from the other units of the Roman army (they were barbarian mercenaries), by the time of the Dominate recruits could be of any ethnicity. However, the Romans continued to refer to many of these units as equites Illyricani (“Illyrian cavalry”), equites Dalmatae (“Dalmatian cavalry”), or equites Mauri (“Moorish cavalry”). A more general term was equites promoti (“running-ahead cavalry”), sometimes rendered as equites promoti indigenae ("indigenous running-ahead cavalry," i.e., not barbarian cavalry of this classification). The troopers of these units generally wore no body armor (although they sometimes had helmets) and rode unarmored horses, although they usually carried a small round buckler (clipeus). Their primary weapons were a collection of javelins, although they also usually carried a spatha as a sidearm. They usually used hit-and-fade tactics to harass opponents, but would normally try to avoid close combat. Filling a similar tactical role were units of unarmored horse archers—called equites sagittarii (“cavalry archers”) or equites sagittarii indigenae (“native cavalry archers,” i.e., not barbarian cavalry). The equites sagittarii units had initially been developed from Syrian, Parthian, and Armenian mercenary units in Roman service during the Principate, but as with the equites promoti, by the time of the Dominate these horse archer units took recruits of any ethnicity and trained them to fight as part of a disciplined Roman unit using Roman uniforms and equipment. They generally did not carry shields, although most would have carried a spatha as a sidearm.

 

By the time of the Dominate, the general term for close combat infantrymen was milites (“soldiers”), although sometimes the older Latin term, pedes (“foot-men”), was trotted out. Milites could be legionary (legionarii) or auxiliary (auxilia), although as I stated previously, there were no social or tactical differences between the units that bore these descriptors, and they utilized the same array of equipment. The sole exception were units of auxilia palatina sagittarii (“auxiliary palace archers”)—as the name implies, they fought primarily in ranged rather than close combat, and used equipment appropriate to that role. As with most of the cavalry of the comitatenses, most of the milites of the comitatenses were organized into vexillations, with only a few independent smaller units (cohortes, alae, or cunei). The legiones and auxilia of the comitatenses wore mail or scale hauberks and ridge helmets, carried the large oval body shield called a scutum, and fought primarily with spear (hasta), long sword (spatha), and/or short sword (semispatha). Legiones and auxilia usually preceded close combat with a prolonged period of throwing javelins (lanceae or verruta) and lead-weighted war darts (plumbatae). Auxilia palatina sagittarii were similarly equipped, although they did not carry shields, spears, javelins, or darts (and sometimes didn’t wear helmets), avoided close combat (although they still carried a sword as a sidearm), and fought with a bow (the Romans used a composite recurve bow called a flecto). Such units were mostly utilized as massed archery battalions, delivering withering volleys of arrows. The late Roman army also had distinct units of infantry skirmishers, called ventores (“swift ones”), exculcatores (“trackers”), exploratores (“scouts”), or praeventores (“interceptors”). There does not seem to have been any consistent distinction between the units described by these terms—they were generally unarmored (some may have worn helmets), javelinmen carried the small round clipeus shield but bowmen and slingers did not, and they fought primarily with ranged weapons (thrown javelins and darts, bows and arrows, crossbows and bolts, slings and stones)—and in fact we do not know whether they were organized into vexillations like most of the other infantry, although the Notitia Dignitatem clearly lists them as distinct units. As with cavalry skirmishers, they served as scouts, to screen the deployment of the main battle contingents, and as flankers (cavalry skirmishers were generally used in these roles on relatively level ground, while infantry skirmishers were used in more difficult terrain). Also as with cavalry skirmishers, their main role was to harass enemy troops, and they generally avoided close combat.

 

During the Principate, each century of a legion or auxiliary cohort was provided with a ballista (a large bolt- or stone-throwing crossbow), and each cohort was provided with a catapulta (torsion stone-throwing engine). During the Dominate, only some legionary and auxiliary units utilized a handful of hand-held crossbows (manuballistae)—usually in the hands of sharpshooters that fought as snipers—while there were independent brigades that specialized in tension and torsion artillery. These units were called ballistarii, even though they generally included a mix of ballistae and catapultae. As in the Principate, some of the smaller ballistae were sometimes mounted on mule-drawn carts (carroballistae) for greater mobility and to allow the crew to position themselves behind friendly infantry and provide direct-fire support over their heads. Unfortunately, we do not know the size of ballistarii units or the complement of engines, although they seem to have been used in a kind of brigade system—when on campaign, they traveled as a distinct unit, but when the army deployed for battle (and when deemed tactically appropriate), sub-units of the ballistarii were generally attached to other units (comitatenses or limitanei) to provide artillery support.

 

Frontier Armies

As previously stated, most modern historians assume that, on average, the units that composed the late Roman frontier armies (limitanei) were probably less well-equipped than those of the mobile armies (comitatenses), although the evidence indicates that units with similar designations in the two different classes of troops nevertheless filled similar tactical roles, at least during major battles. However, there are some easily identifiable differences. In the comitatenses, most of the units seem to have been vexillations. In the limitanei, there were generally only a few larger units classed as vexillations, but there were many more independently deployed smaller units—cohortes (infantry), alae (cavalry), or cunei (both infantry and cavalry). There were no units of cataphractarii/clibanarii (i.e., lancers) or ballistarii (i.e., artillery) in the limitanei. Oddly, since you would expect such units to have been particularly useful to frontier troops, the Notitia does not list any distinct units of infantry skirmishers as being part of the limitanei (although there were units of the two types of cavalry skirmishers). This might be the strongest evidence for a potential difference in the tactical stance of infantry units in the mobile and frontier armies—it is possible that, unlike the comitatenses, where there was a clear distinction between close combat infantry and infantry skirmishers in separate units, limitanei infantry units may have occupied a tactical stance somewhere in-between, with lighter equipment, a heavier emphasis on ranged weapons, and a quasi-skirmishing style of combat. Unfortunately, we just don’t know. In keeping with the new strategic doctrine of defense-in-depth, legionary and auxiliary infantry, as well as auxiliary infantry bowmen, were mostly strung along the border (limes) in small units, manning small forts, and patrolling the frontier. At strategic locations within each of the border provinces could be found the larger forts of the larger infantry and cavalry vexillations. These units generally policed the province and backed up the border garrisons as necessary.

 

Irregular Units

Due to the exigencies of the Third Century Crisis, many of the great nobles (Latin, patricia, Greek, patrikioi, "patrician"), military and civil office holders, and other notables of the Roman Empire had been granted the right to recruit and maintain units of household troops that were known as bucellarii (“biscuit-eaters”). The origins of the term, bucellarii (singular, bucellarius), are obscure. Biscuits were a form of hard tack eaten by both regular army soldiers and irregular household troops (not to mention some federates and mercenaries), and so it would not seem to be a term that would have particularly distinguished any single group of soldiers. The first use of the term is traced to the reign of the western emperor Honorius (395-423 CE). Prior to that date, they seem to have been considered mercenaries (numeri). However, beginning with the reign of Honorius, such troops were required to swear allegiance to the reigning emperor (i.e., not just their employer), and they were increasingly recognized as an official adjunct to the regular military (considering the many financial economies of the period, this was likely due to a desire to shift some of the expense of the military onto the shoulders of the nobles, many of whom did not pay a fair share of the taxes that paid for the military). This practice was a double-edged sword—by most accounts, the household troops of most notables were better paid and more lavishly equipped than their counterparts in the regular army (especially the limitanei), but despite the imperial oath their primary loyalty usually lay with their employer. Initially, most of these units were small (less than 100 men), but by the fourth century, some of the great nobles and high military commanders developed substantial troops that amounted to private armies (several thousand strong). Cavalrymen were favored, usually equipped as scutarii/armaturae/gentiles and having a similar tactical stance, although sometimes barbarian mercenaries were recruited that fought in their native fashion (e.g., Alanic lancers or Hunnic horse archers). Some main battle infantrymen were equipped in the manner of milites, or bowmen in the manner of auxilia sagittarii, and formed units with a corresponding tactical stance.

 

The emperors of the Dominate continued the policy of settling barbarian peoples within the empire as “federates” (foederati). Several terms were used in reference to these peoples—laetus/laeti (“lucky one(s)”), dediticius/dediticii (“surrendered one(s)”), or tributarius/tributarii (“tribute-payer(s)”)—although the relationship of all these peoples to the empire was based on a treaty (foedus) that laid out their rights and responsibilities, essentially as a people collectively working toward gaining full Roman citizenship (a process that normally took about one or two generations). Most of the federate treaties stipulated that the signatories would provide a set number of young men annually for recruitment into the regular units of the Roman army, but the leaders of federate peoples were also expected to maintain sufficient household troops and/or militiamen to essentially act as limitanei in the region where they were settled. Federate colonies were generally located in land that had become depopulated by native Roman peoples (gentiles) for various reasons (e.g., war, disease, famine), or that had formerly been wilderness (mostly lands considered marginal by the native Roman population), and they were often located along the frontier, on both sides of the official border. There was a difference between the ethnic groups that were settled in the Western Empire and the Eastern Empire, based on what groups of barbarians could be found along the frontier of that half of the empire—in the west, West, East, and North Germanic peoples were predominant, as well as Alanic peoples, although in Britain there were some Pict (Picti) and Irish (Scoti) federates, as well as Berbers (Gauramantes and Mauri) in North Africa, while in the east, East Germanic, Sarmatian, Syrian, Armenian, Georgian, and Arab (Saraceni) federates predominated, with some Cushitic groups in Egypt (Blemmyes and Nubians). The irregular troops fielded by these federate peoples, sometimes in conjunction with Roman limitanei or comitatenses, sometimes independently, generally fell into two types—household troops (bucellarii or comitatus) and militiamen (foederati or laeti). Although federate colonies were financially subsidized by the imperial government, these funds often fell victim to the grasping hands of the Roman officials that were supposed to distribute them, and although federate troops were supposed to have access to local state arms manufactories (fabricae), the officials that ran the manufactories also often overcharged federate troops and gave them inferior equipment, so many federate peoples continued to produce some of their own equipment according to their native traditions. Thus, federate troops probably had a very motley appearance that had elements drawn from both sides of the border. Similarly, most federate troops fought under their own leaders, and units usually had a tactical stance that was closer to their native traditions than to regular Roman army practice. Germanic household troops would mostly have conformed to a tactical stance similar to Roman scutarii/armaturae, although they were often viewed as being more impetuous, and they also tended to prefer to dismount to fight as infantry when and where it was practical. They wore short-sleeved mail hauberks, Germanic spangenhelm or Roman ridge helmets (the two styles were very similar), carried large round Germanic or oval Roman body shields, and wielded a combination of sword, spear, and axe. Germanic federate militiamen were generally equivalent to the common warbands of unassimilated Germanic groups—organized into shock infantry units, usually with less armor than the household troops, although carrying the same types of shields and weapons as the household troops. A minority probably fought as skirmishing infantry bowmen. Although the Alanic federate peoples were ethnic Aryans, and came from a culture with much stronger cavalry traditions than those of the Germanic peoples, most had begun to become heavily Germanicized by the time they settled in the Roman Empire. We sometimes hear of comites Alani lancers, but these mostly seem to have been recruited from Alanic groups originating outside the empire. Thus, federate Alanic cavalry generally consisted of two types—those that were similar in equipment and tactical stance to Germanic household troops and Roman scutarii/armaturae, and those similar to Roman javelin-armed equites promoti. However, unlike Germanic cavalrymen, Alanic cavalrymen clearly preferred to fight mounted (i.e., they generally never dismounted to fight on foot, except perhaps in sieges). The Alans tended to field large, but poorly equipped, groups of infantry militiamen in support of their cavalry (mainly spearmen). In the eastern empire, Syrian, Georgian and Armenian federates had long been part of the empire’s eastern frontier defense system, although Arab groups (known collectively as Saraceni, “tent-people”) began to supplant many Syrian groups by the end of the third century CE. The Syrian, Georgian, and Armenian household troops were generally divided between cataphracts (equipped like Roman cataphractarii/clibanarii and with an identical tactical stance) and horse archers (once again, identical to those used by the Romans), while the militiamen were mostly close combat infantry armed with spear and sword (wearing mail and carrying wicker shields bound in leather), backed up by unarmored bowmen and slingers that usually performed as skirmishers. There were some infantry archer regiments that used mass archery and even became part of the household troops of various federate leaders. Although Roman sources often refer to Arab household cavalrymen as clibanarii, images from the period never show them riding armored horses and no archeological find has revealed horse armor to be a part of the panoply of Arab cavalrymen in Roman service. Arab household cavalrymen were often armored in mail, sometimes including a mail coif, but they carried both a small round shield and a long spear/lance that was used for thrusting. Despite the lack of stirrups, the evidence seems to point to the fact that these Arab cavalrymen had a tactical stance that lay somewhere in-between Syrian and Persian cataphracts on the one hand, and Roman scutarii/armaturae on the other hand—one-handed lancers that did not use the two-handed contus so that they could still utilize a shield (albeit a small one). Swords and axes were carried as sidearms.

 

Mercenaries and Allies

The precise dividing line between federates (foederati or laeti), allies (amici, singular, amicus), and mercenaries (numeri, singular, numerus) in service to the Roman Empire was never clear. All three groups were subject to the terms of a document called a treaty (foedus) that stipulated the rights and responsibilities of both signatories (i.e., the empire and its federates, allies, or mercenaries). However, federates generally were groups that were considered to have become subjects of the empire—as the terms, dediticius and tributarius (see above) indicate, the Romans preferred to enter into federate contracts with groups that were clearly subordinated to the empire—while allies were considered to be independent sovereign states that had interests that were aligned with those of the Roman Empire. Mercenaries were viewed as akin to prostitutes, actors, and musicians—individuals with a particular set of skills (in this case, martial skills), whose services were purchased for a finite period of time, as defined by the foedus. Generally speaking, most federates lived within the borders of the empire and were in the process of being assimilated into Roman society, but allies and mercenaries were mostly unassimilated groups that lived outside the empire. Pretty much the same groups served as allies and mercenaries in the Roman army as served as federates, so I won’t repeat the descriptions of the troop types here, although there was one important addition—the Huns. The Huns entered Europe in about 370 CE and had reached the Danube frontier of Rome by about 376-378 CE. They reached the Rhine by about 430 CE and went on to consolidate an empire that stretched from the Rhine to the Don, north of the dominions of Rome. Although the Huns sometimes allied themselves with the Roman Empire against various Germanic opponents—notably, the alliance of Attila and Aetius against the Burgundians in 437 CE—and Hunnic mercenaries became highly prized in the Roman Empire, no significant settlement of Hunnic federates within the empire was made until the sixth century CE, beyond the scope of this gallery (but see my history of the Black Huns in my Hunnic Peoples gallery). Hunnic mercenaries were invariably horse archers. Many were unarmored, but under the influence of the Germanic peoples and the Romans, and due to the amassing of tribute and plunder, armored Hunnic horse archers became increasingly common (usually mail or scale hauberk and spangenhelm or ridge helmet and small round clipeus shields). The Huns had an improved composite recurve bow that quickly became renowned in Europe and fostered the development of imitations, particularly in the eastern Mediterranean and Black Sea regions.

 

Uniforms

During the Principate, the typical clothing of the Roman soldier was a single-piece, short-sleeved tunic (tunica) that reached the knees, coupled with hobnailed sandals (caligae). Common soldiers generally wore un-dyed tunics, although for ceremonial occasions they would bleach their tunics (thus, most were probably a kind of off-white in color). Centurions, decurions, and optios (i.e., lower-ranking officers) wore red tunics, sometimes with red crests or dyed feathers on their helmets. Higher-ranking officers were usually identified by the wearing of a white tunic with a purple band, the width of the band indicating rank. When the empire expanded into northern Europe (Gaul and Britain in particular), long-sleeved tunics, trousers (bracae), socks (worn inside the caligae), and laced boots replaced the Mediterranean clothing in winter. For unknown reasons, this winterwear style of clothing proliferated in the third century and had become the standard clothing worn by soldiers throughout the empire by the time of the Dominate. Europe did experience a broad climatic cooling trend at that time—a shorter growing season is believed to have been one of the reasons many Germanic peoples migrated south toward the Roman Empire, in search of better weather for agriculture—and this may also explain the change in Roman fashions, although the introduction of large numbers of Germanic and Celtic recruits into the regular army might also have played a role (probably both things). Late Roman uniforms were often highly decorated, with embroidered strips (clavi) along the hems, circular applique roundels (orbiculi) (usually over the shoulders), or square applique panels (tabulae) (usually on the front of the tunic where it hung down over the thighs) added to tunics and cloaks, and these uniforms were generally worn by both civil service and military personnel. A brimless hat known as the Pannonian cap (pileus pannonicus) was often worn by civilian officials, off-duty soldiers, or soldiers that did not have a helmet. Some effort does seem to have been made to coordinate the colors and designs of these uniform decorative elements with the colors and symbols (indicia) of a given soldier’s unit, although as in most things the Romans tended to lack the modern sense of uniformity in kit, and there do not seem to have been strict guidelines or any effort to visually mark out different ranks. Informally, higher-ranking officers usually had higher pay and thus could afford more colorful clothing, while rankers generally seem to have preferred more utilitarian styles with a minimum of ostentation.

 

While these new uniforms make a striking contrast to the relatively plain garb of soldiers in the Principate, they have also sparked controversy among modern historians. Much Roman art from the period depicts Roman soldiers wearing these uniforms, and only rarely depicts soldiers wearing armor. When armor is shown, it is usually on cataphractarii/clibanarii or scutarii/armaturae, and only rarely on close combat infantrymen (usually comitatenses/palatini/candidati). Coupled with a relative dearth in archeological finds from the period, this has led many modern historians to posit that many late Roman soldiers—especially infantrymen—did not wear armor, perhaps as an economy or to increase their mobility and agility. However, the Notitia Dignitatum clearly shows pictures of armor among the equipment of many units and in the equipment provided by many state manufactories, and battle accounts from the period often mention armor (e.g., the men were tired and overheated from marching in full armor in the hot sun, the rebels stripped the dead legionaries of their armor, etc.). Further, we know that Roman uniform usage was inspired by—and continued to be influenced by—the development and use of uniforms in the Sassanian Persian Empire (see my Sassanid Persians gallery), and we know that the Persians often wore their uniforms over their armor to deflect much of the heat of the sun, to better display the heraldic devices on the uniforms, and to help dull the impact of arrows on armor (especially mail). It seems perfectly reasonable to assume that the Romans may have done the same thing. We know that Roman artists often used tropes in portraiture—particularly of soldiers—in order to make their subjects easily identifiable to viewers (essentially, visual shorthand). For example, during the Principate, legionaries were often portrayed wearing segmented armor (lorica segmentata), auxiliary infantry wearing mail (lorica hamata), and auxiliary cavalry wearing scale (lorica squamata), even though we know from archeological finds that all three groups usually wore mail or scale, and that auxiliary infantry and cavalry sometimes seem to have worn segmented armor as well. Thus, it is likely that some of what we are seeing in late Roman art is also due to tropes that were used for purposes of visual shorthand—cataphractarri/clibanarii and scutarii/armaturae were probably shown with armor to differentiate them from other types of cavalry, while uniforms were often shown because they themselves had insignia that could be used to identify different groups. I should also mention that there are some modern analysts of the art of this period that believe they can discern clues that the uniforms portrayed in late Roman art appear to be worn over armor—the Romans usually sought after a realism in their art, and uniforms often appear to be too bulky for the body of the wearer, possibly indicating another layer beneath.