CHAPTER 4
SIMON JAMES
WARFARE comprises organized acts of physical violence, conducted against human bodies, victims’ possessions, settlements, and the landscapes in which they live. And these acts are perpetrated primarily through the dedicated material culture of lethality and destruction: arms, armor, and other “accoutrements of war.” Among the peoples of the classical world, and neighboring societies with whom they fought, further material preparations were widely undertaken both to facilitate and counter these actions.
State societies, especially, created infrastructure to support campaigning, facilities for assembling troops and storing supplies, plus the means to move them: harbors and shipping, roads and bridges, and also gathered the vehicles, draft and pack animals needed for armies to move. Means of defending territory, resources, and population also often involved construction of works, from small strongholds such as the cooling-tower-like drystone brochs of Scotland to the town walls of Numantia, Spain, the ingenious Hellenistic defenses of Syracuse, stupendous feats of engineering like the mighty double walls of Constantinople, or, largest of all, vast integrated frontier systems of forts and linear barriers, most famously Hadrian’s Wall, or the Gorgan Wall of Sasanian Iran.
The classical world—especially the sprawling Roman Empire—and adjacent regions have preserved many such remains of military infrastructure, of other martial material culture, and of warlike actions themselves, ranging from impact-bent javelins and sword-slashed skulls to battlefields and devastated cities. We also have the testimony of commemorative monuments, weapon-graves of warriors, and offerings of arms to the gods. This vast range of physical evidence is the domain of archaeology.
Especially associated with commemoration is another distinct category of ancient material evidence which makes its own unique contribution to our understanding of ancient warfare: contemporary visual representation. On victory monuments, figural tombstones, and in other contexts like figured pottery, some of the cultures concerned have left us depictions—sculptural, drawn, or painted—of their own and their foes’ arms and equipment, warriors, styles of combat, and particular wars. Like ancient texts, these visual images are constructed representations of the world, but are also material artifacts, often recovered by archaeological means.
This broad spectrum of material testimony directly surviving from antiquity reveals much about the beliefs, aspirations, and deeds of soldiers and generals, and how their societies valued them. It provides independent evidence for armies and warfare to stand alongside written records. Ancient images have their own rules of evidence and limitations, akin to those of texts in that they are confined to certain societies and, as another form of representation, they are subject to stylistic distortion and the filtering effects of ideology, rhetoric, and stereotyping. In these respects they are more the purview of art history than archaeology, but here I include their contribution to interpreting other material remains, and elucidating ancient practices.
My focus is on the contribution of archaeology to understanding aspects of ancient warfare, archaeological methodology, and its achievements and problems in the context of explaining how men fought and how armies were organized in the ancient world. A central aspect of archaeological evidence—arms and armor—is touched on, although covered in more detail elsewhere (see further Jarva and Campbell, pp. 395–437). I draw primarily on my own field, the Roman Imperial era, for much of which the evidence is exceptionally extensive, relatively rich, and well explored, both for the Roman military and for many of its enemies (for another view of the archaeology of Roman conflict, see Coulston 2001).1 Military remains of earlier periods, from archaic Greece up to and including the Roman Republic, are generally sparser, with far fewer dedicated military sites and, generally, less intensive deposition of martial equipment for archaeologists to find. Consequently its archaeological study is inevitably smaller scale and less developed.
STRENGTHS AND WEAKNESSES OF ARCHAEOLOGICAL EVIDENCE
Archaeology is no mere “handmaiden of history.” It can do much more than simply illustrate or refine what we already know from texts. It may indeed elucidate these, but also reveals entire domains far beyond the boundaries of surviving writings or invisible to their authors, highlighting their biases and misrepresentations. It is a central independent means of investigating past warfare even for historical contexts such as the classical world. For prehistoric eras it is the only means of direct study.
Figure 4.1 The tombstone of the Roman auxiliary cavalryman Insus, son of Vodullus, of the Gallic Treveri. Excavated in Lancaster, England, in 2005, a striking visual representation of a Roman soldier maintaining the tradition of his Gallic forbears: taking heads as trophies of victory (an act that the Romans themselves perhaps approved).
© Lancaster Museums.
As a major by-product of its activities, archaeology also makes direct contributions to text-based study, for fieldwork provides the main source of new texts for historians to work with. Many are military-related, especially from Roman times. They range from inscriptions on state monuments, milestones, and fort gates to soldiers’ epitaphs like that of the cavalryman Insus from Lancaster, United Kingdom (figure 4.1; Bull 2007). They also comprise graffiti and occasionally military “paperwork”: letters, dispatches, reports, lists of men and matériel on papyri, ostraca (potsherds), or wooden tablets. Excavation is responsible for recovering our textual knowledge of the two most thoroughly documented units in the Roman imperial military: cohors ix Batavorum stationed at Vindolanda near Hadrian’s Wall circa A.D. 100 (Bowman 1994; Bowman and Thomas 2003); and, at the opposite end of the Empire, cohors xx Palmyrenorum, based at Dura-Europos,2 Syria, circa A.D. 200–250 (Welles et al. 1959; Kennedy 1994).
Like the surviving textual evidence, the wider archaeological record exhibits characteristic, sometimes dramatic, strengths but, as we will see, also suffers from weaknesses, limitations, and frustratingly large lacunae. Of its nature, archaeological evidence is anonymous. It is also, like the textual record, highly fragmentary, depending on the vagaries of deposition, survival, recovery, and research. Interpretation of such incomplete evidence presents its own difficulties, and in trying to do so archaeologists, like ancient historians, are conditioned by their personal prejudices and cultural baggage. However, among its greatest strengths is its ability to tell us about aspects of classical warfare on which texts are silent, or at best partial, in both senses of the term: incomplete and biased. Perhaps most important here is its capacity to give us a view, undistorted by Greco-Roman prejudices or misunderstandings,3 of the many “barbarian” antagonists of classical polities, silent or silenced in that they have left us few if any texts of their own. In the field of war, archaeology can especially help us better understand the world of soldiers and warriors. It is particularly suited to creating social, economic, and technological histories.
However, archaeology also contributes to the writing of narrative history. Indeed, there are cases where astonishingly informative direct physical evidence of war has been recovered, generating archaeological accounts which stand direct comparison with the most vivid surviving ancient eyewitness descriptions of men in battle.
ARCHAEOLOGY OF BATTLE AND COMBAT
An outstanding example of the potential of archaeology at its best is provided by the case of the siege of Dura-Europos, a Roman-garrisoned city of Hellenistic origins on the west bank of the Middle Euphrates overlooking Mesopotamia (figure 4.2). In or aroundA.D. 256, Dura was invested and destroyed by the Sasanian Shah Shapur I. This siege is completely unattested in classical sources (James 1985). Yet, purely from archaeological excavations, the story of the struggle for the city can be told in dramatic detail which stands direct comparison with that of another major Sasanian siege known, by contrast, only from a literary source: Ammianus Marcellinus’s eyewitness testimony of the siege and fall of the fortress city of Amida on the Tigris (modern Diyarbakir, Turkey), 350 km (220 miles) north of Dura, at the hands of the second Shapur a century later, in 359. This is among the most remarkable of all textual accounts of ancient warfare (Amm. Marc. 18.8–19.8).
Figure 4.2. The city of Dura-Europos, Syria, showing the location of the intramural Roman military base (top), and the position of the Sasanian assault ramp from the final siege (bottom left). Plan of the Franco-Syrian Mission to Dura-Europos, adapted by Simon James.
Ammianus, a Syrian-Greek officer trapped with many other fugitives at Amida, identifies the Roman contingents defending the city, seven legions and other units. He also details the composite nature of the Sasanian army, incorporating many subject and allied contingents, including Chionite Huns. His thrilling description records the main operations of the siege, undertaken beneath storms of artillery projectiles, sling-bullets, and arrows. Ammianus graphically conveys the ordeal of the defenders: the sufferings of the wounded, the impossibility of burying all the dead, and an epidemic blamed on the putrefying corpses. The struggle for the city involved psychological warfare, the two sides competing to shout the praises of their rulers, while the Sasanians attempted to intimidate the defenders into surrender by shows of numbers, including massed cavalry and elephants.
Both sides used torsion artillery. Ammianus cites examples of the accuracy and power of light arrow shooters in picking off individuals (like the son of Shapur’s ally, the Chionite King Grumbates, shot through the breastplate), and describes how five machines rapidly cleared a tower of seventy Sasanian archers who had infiltrated the defenses by night, some bolts killing two men. The Sasanians constructed siege towers to allow their artillery to command the walls, while raising siege ramps to get over them. Ammianus details the ripostes of the defenders, making sorties to disrupt the siege, and countering the ramps by heightening their own defenses. These eventually collapsed under their own weight—unless, unbeknownst to Ammianus, they had been undermined—leading to the fall of the city. Under cover of night, Ammianus escaped the ensuing slaughter through an unguarded postern. Captured survivors joined the columns of prisoners deported to Persia, and Amida was left an abandoned ruin over which the Emperor Constantius subsequently wept (Amm. Marc. 20.11.5). The siege lasted seventy-three days.
In the archaeologically recovered case of Dura, occurring a century earlier, we do not know for sure which Roman units were trapped there (the military papyri relate to events from two years to decades before the siege), or the names of any individuals caught up in the fighting. However, in some ways we know more details of what happened than we do at Amida, and without the selectivity, possible exaggeration, or dramatic editing of an Ammianus—although conversely we know less of the sequence of events. However, the nature and extent of siege works argue for a period of weeks or months (Du Mesnil du Buisson 1936 and Leriche 1993; James 2011b).
At Dura the Sasanians built a huge mud-brick-walled siege camp, far larger than the city, on the plain facing Dura’s relatively vulnerable western wall (the others being protected by cliffs). They raised a sophisticated siege ramp to try to get over the southern end of the western wall, undermining a troublesome adjacent tower to eliminate its use as an artillery platform. As a result of new excavations we can trace exactly how the ramp was built, no mere mound of earth but a solid box construction, walled and paved with fired brick, its regular incline intended for a wheeled siege machine. The Romans sought to counter this, as later at Amida, by heightening the city wall and, behind it, deepening the earth rampart which had recently been added to strengthen the defenses against siege. Meanwhile, both sides also dug further mines here, the Persians seeking (with unclear outcome) to enter the town, the Romans (certainly successfully) to destroy the attackers’ ramp.
This engineering duel, probably ending in stalemate, was part of a multipronged Sasanian assault which doubtless included attacks on the relatively vulnerable river gate (now wholly lost to the Euphrates), and certainly saw direct assaults on the heavily fortified western (“Palmyrene”) gate. Here the Sasanians also used siege machines, to which the Romans responded with a tornado of artillery stones, iron bolts, and incendiary weapons plus, perhaps, sorties (a Roman shield boss was found in the debris from the fighting, well outside the gate).
The Sasanian strategy also involved another mine, intended to bring down Tower 19 and an adjacent stretch of city wall, to make a breach viable for assault by troops charging across the plain (figure 4.3). This operation was detected in progress by the Romans, who dug a countermine to intercept and foil it. The resulting underground struggle led to deposition of perhaps the most astonishing material evidence yet found for ancient warfare in action.
On excavation, the Roman countermine revealed intense burning and a scatter of equipment ending in a dense tangle of some twenty skeletons, still with their armor. From the nature and disposition of these gruesome remains, it is possible to determine in fine detail what happened (figure 4.4). The Persians had tunneled under the hard limestone surface layer of the gypsum plateau on which Dura stands, and had dug upward into the foundations of the city wall, encased in the recently added external mud-brick glacis and internal inner-earth rampart. They were removing the lower courses of the wall and tower, replacing them with wooden props which would later be fired to bring down the superstructure. The Romans dug a timber-propped tunnel through their own earth rampart to seize and hold this expanding Persian chamber. But the Sasanians heard them coming and were waiting for them. When the Romans broke into the Persian mine, about twenty were killed, almost certainly by asphyxiation; like contemporary Roman writers, the Sasanians knew the Hellenistic trick of filling an enemy mine with choking fumes (cf. Livy 38.7; Polyb. 21.28; Polyaenus, Strat. 56.7), doubtless employing the sulfur and bitumen which, minutes later, they used to destroy the Roman countermine.
The surviving Romans fled, and the Persians, shutting off their smoke-pot, entered the Roman gallery, intent on collapsing it to prevent further interference with their mine. However, to stop the Romans reentering their countermine before its timberwork could be set alight, the Sasanians dragged the Roman dead and dying, along with their shields, toward the entrance, and stacked them into a wall across the tunnel. Then they piled straw, wood, and Roman cloaks against the roof props, and used sulfur and bitumen to start an inferno leading to collapse of the Roman mine, one Sasanian warrior perishing in the process (James 2011). The Sasanians then completed their own mine, and fired it.
Figure 4.3 The siege operations at Tower 19, Dura-Europos (see Figure 4.2). Top left, the Sasanian approach tunnel and sap beneath the tower and adjacent city wall was intercepted by a Roman countermine from within the city. Inside the Roman tunnel were found a pile of dead Roman soldiers, and nearby a Sasanian. His iron helmet (left) shows signs of Roman influence (mail, reinforcing plates), and in turn was a prototype for Roman helmets of the following century. This illustrates convergence in military technology and practices across this frontier, seen also in the shared methods of siege warfare exhibited at the site. Simon James, after Du Mesnil.
Figure 4.4 Schematic diagram, to approximate scale, showing the stages of the fight for the mines at Tower 19, Dura-Europos. 1. The defenses of the city, showing the wall and towers reinforced with a mud-brick glacis, and earth rampart behind. 2. The Sasanian sappers reach, and begin to undermine, the wall inside the rampart. The Romans hear them, and begin a countermine. 3. The Romans break into the Sasanian gallery, intending to prevent its completion. 4. The Sasanians are ready for them with a smoke generator, which rapidly fills the Roman mine with choking fumes, killing about twenty. 5. The Sasanians enter the Roman gallery, and prevent Roman re-entry with the bodies of the dead while they set it on fire. 6. The Roman gallery partially collapses, allowing the Persians time to finish and fire their mine. Drawn by Simon James.
Although the Roman countermine failed, their earth-strengthened defenses held anyway: the undermined wall and tower sank about a meter, but stayed upright. The tower floors collapsed, entombing several horse armors and a painted “scutum” (James 2004), but there was no practicable breach here. Stalemate.
We do not know how the Sasanians finally entered the town, but the defenders made a last stand in the military cantonment in the northern quarter of the city. Recently a row of iron bolt heads was found, still lying by the base perimeter wall where they had been placed to serve a catapult positioned to shoot down one of the streets (James 2007). But this last line of defense was, in turn, overwhelmed. Bodies were found lying where they had fallen inside some of the barrack rooms. As would befall Amida a century later, Dura was sacked, the defenders slaughtered or deported, the site abandoned.4
The best archaeological testimony for ancient warfare, then, can rival that of the best texts in detail and drama. However, Dura is exceptional in a number of respects, the product of a rare concatenation of circumstances (but then, the vivid detail of Ammianus is equally exceptional). Unlike Amida, Dura—in what became a contested no-man’s-land between Rome and Persia—was never substantially reoccupied, permitting survival and archaeological accessibility of so many traces of the siege operations, neither disturbed nor overbuilt beyond reach of archaeologists. Further, local environmental conditions were conducive to remarkable organic survivals like complete painted wooden shields and papyri.
There are other remarkably well-preserved and highly informative sieges sites, famously the Roman works round the Zealot stronghold at Masada, Israel, which site, like Dura, was not later disturbed (Richmond 1962; Yadin 1966; Goldfus and Arubas 2002). Other striking examples of siege archaeology include Gamla, Israel (Syon 2002), and Olynthos (Lee 2001).
Archaeology also produces dramatic insights into some “open-field” battles. A major discovery was the scene of the destruction of Varus’s three-legion army by Germans under Arminius in A.D. 9. It occurred around modern Kalkriese near Osnabrück (Schlüter 1999; Schlüter and Wiegels 1999; Wells 2003). The battlefield, initially located by recovery of coins and metalwork using metal detectors, has been more fully explored through archaeological survey and excavation. This work has now revealed a great deal about the actual course of this notorious Roman disaster. Scatters of artifacts across the landscape, pits containing the subsequently collected remains of the dead, and remains laid down during the fighting itself have provided a fairly clear picture of the shape of events. The Romans were in a long column which was led at this point into a narrow space between boggy ground on their right, and a forested hill on their left. Overlooking the Romans, in the cover of the trees, Arminius’s men had constructed an earth rampart, behind which they hid and from which they launched their attack on the unsuspecting legionaries, who had neither time to arm nor space to deploy. They were slaughtered, along with their civilian followers and even their pack animals: the body of a mule, still wearing its bell, lay by the rampart.
However, archaeology generally proves less informative for the majority of ancient battles, especially open-field conflicts (as, indeed, most textual accounts are inferior to those of Ammianus on Amida, or the battles of Strasbourg and Adrianople: Amm. Marc. 16.12; 31.12–13). The reasons are clear. Unlike modern wars with high-explosive munitions, few ancient battles other than sieges left lasting traces. The Roman habit of entrenching camps and digging earthworks makes their military actions relatively detectable, but even these offer limited information, being often almost devoid of artifacts the result of battlefield clearance. Soldiers and other scavengers normally picked up anything of value, and so nothing much larger than arrowheads was likely to be left. Anything missed, in vegetation or shallow-buried, was vulnerable to weather, rust, and the plow. However, Roman projectiles, boot nails, and other militaria preserved by special soil conditions, recently discovered at Harzhorn, Germany, attest a forgotten third-century battle deep inside Barbaricum (Wiegels et al. 2011).
The same considerations also applied to the bodies of the dead: men, horses, even war elephants. Occasionally we find some remains of soldiers, as at Kalkriese and Krefeld, Germany, which has also produced graves of horses killed in fighting in A.D. 69 (Reichmann 1999). The latter may offer clues to the shape of the fighting, since heavy and noisome equine cadavers are likely to be rolled into pits dug where they had fallen. However, perhaps most remarkable is just how rarely we recover potentially the most important archaeological evidence for the nature of ancient front-line combat: the bodies of slain combatants, bearing their death wounds (James 2010). Given the millions who perished in Rome’s wars alone, it seems astonishing that, as yet, we have not identified and scientifically studied a single mass battlefield grave. As with other remains, this is due to battlefield clearance (the winners usually recovered their own dead, for disposal elsewhere or to cremate), to the likely shallowness of any hasty on-field burials (thereafter vulnerable to the plow) and, in the case of enemy dead left to rot where they fell, complete obliteration even of teeth due to weathering. Consequently, it becomes less surprising that no convincing archaeological traces have yet been identified even of so vast a slaughter as Cannae.
ARCHAEOLOGY AND NARRATIVE HISTORIES OF WARS AND CONQUESTS
It is often unappreciated just how fragmentary the Greco-Roman historical record is, by comparison with that for more recent times. Even for the relatively well-documented early Roman Empire, it is peppered with lacunae large enough in which to “lose” battles, campaigns, even entire wars.5 Material evidence sometimes proves invaluable in reconstructing the courses of campaigns and conquests, and even political strategy. Excavations at Waldgirmes, Germany, have confirmed that Augustan commanders actually did commence foundation of the first cities in his new province east of the Rhine before the disaster of A.D. 9 (Becker et al. 2003), proving beyond reasonable doubt that permanent occupation and civil development were intended, not simply military operations to protect Gaul.
At a finer scale, in the case of the conquest of Britain, there has long been uncertainty over whether, in the early 70s, the provincial governor Cerealis Governor Cerealis was simply harrying the powerful Brigantes of northern England to neutralize them while the conquest of Wales was completed, or whether he was trying to conquer them outright. Tacitus, our source, has been suspected of deliberate ambiguity, downplaying the achievements of Cerealis to inflate the subsequent glories of his father-in-law, Agricola (Tac.Agr. 17; Salway 1989: 136–7). Precise dendrochronological (tree-ring) dates of A.D. 72–3 for the cutting of timbers used in the first Roman fort at Carlisle prove Cerealis was establishing garrisons deep inside Brigantian territory (Birley 1973; McCarthy 2002: 69–71).
Roman Europe has produced the most extensive and intensively explored archaeological evidence for ancient campaigns, conquests, and military occupation, at the scale of “theaters of operation” beyond battlefields—“conflict landscapes” (Coulston 2001). Here, in the first two centuries A.D., particular early imperial campaigns left entrenched “marching camps” across the landscape. Principally through aerial photography, individual camps and even sequences have been found, sometimes allowing us to track the passage of particular armies on the ground, notably in northeast Scotland; however, these ephemeral traces prove frustratingly difficult to date closely (Jones 2012; figure 4.5).
Campaigning early imperial armies also began establishing their familiar playing-card-shaped “forts” and “fortresses” (Wilson 1980). These modern designations are actually misleading since, while defended stoutly enough to resist surprise attacks, they were not designed as strongholds intended to survive sieges, but were winter quarters and depots, and springboards for offensive operations: they are better termed “bases.” Consequently, their distribution can be hard to interpret in terms of campaign strategy, because their individual locations were ad hoc compromises determined according to multiple considerations, logistic and political as well as strategic. Some were sited to control communications junctions, or to watch settlements or other potentially troublesome concentrations of newly conquered people. Others were located for convenience of supply by friendly groups, to exploit areas of good pasture for cavalry horses, or to be on navigable waterways for bulk supply of foodstuffs brought from more distant regions (legions daily consumed grain by the ton). And some were indeed sited as “jumping-off points” for future campaigns.
Figure 4.5 Selected archaeological evidence for Roman military operations in Northern Scotland. Hollow symbols: Roman marching camps of various periods (those reaching towards the Moray Firth probably represent Agricola’s operations culminating in the battle of Mons Graupius). Solid symbols: Flavian military installations, including Inchtuthil legionary base, the ‘glen blocking forts’ and the location of Elginhaugh. Drawn by Simon James.
For example, excavation of Flavian installations in Scotland attest the nature and disposition of garrisons in the territory conquered by Tacitus’s father-in-law Agricola, notably at Inchtuthil, most northerly of all legionary bases (Pitts and St. Joseph 1985), and the “glen-blocking auxiliary forts” at the edge of the Scottish highlands (figure 4.6). These last, for example Fendoch at the mouth of the Sma Glen (Richmond and McIntyre 1939), were doubtless partly to contain the unconquered peoples beyond (Fendoch was overlooked by a watchtower sited to see up the glen), but also to form starting points for anticipated future conquest of the Highland massif which never happened. These sites also reveal what happened after Tacitus’s panegyric on Agricola—our main historical source—turns its attention away from the island with the governor’s recall. Excavation shows that Inchtuthil was never finished, but demolished and abandoned during the later 80s, according to coin finds.
These are all valuable cases, and others could be cited, not least building inscriptions which allow us to date many Roman bases and frontier works to precise years. However, most archaeological sites produce few or no epigraphic finds, while only extremes of waterlogging or desiccation preserve dated military “paperwork” or timbers for dendrochronology. More often we are obliged to estimate foundation, rebuilding, and destruction dates from stratified coins and pottery. We have to make judgments on the basis of the latest minted coins present, or the range and forms of pottery on the site, for example red glossy terra sigillata (Samian) vessels, with rapidly changing forms, decoration, and makers’ stamps that make them datable to a decade or two. Such data give us fuzzy dates for things like the building of bases, for example, “Claudio-Neronian” and “Flavian.” In broader archaeological terms such dates are unusually precise (and the envy of prehistorians often facing chronological uncertainties measuring centuries). However, twenty to thirty years is a very long time in warfare, and greater than the lifetime of many early imperial military bases. In places like Britain it is, therefore, virtually impossible using archaeological evidence to draw provincial maps of which bases were occupied in a given decade, let alone a particular year; redeployments and troop movements occurred at rates too rapid to resolve through this kind of data. Archaeological evidence, then, may sometimes help us to build year-by-year campaign maps, but most often it helps us understand the shape of conflicts, conquests and military occupations, that is, the nature of warfare in more general terms, rather than its detailed course.
Figure 4.6 Inchtuthil, northernmost legionary base in the Roman empire, abandoned incomplete in the later 80s A.D.. Reconstruction by Simon James.
ARCHAEOLOGY AND THE ORGANIZATION OF ARMIES
Archaeological testimony can also provide important direct information about the organization of armies, partly through numbers of bases, but especially through their conformation. It can sometimes tell us about the anatomy of units and, more generally, how they worked. Many early Imperial Roman bases were constructed to accommodate single regiments, often identifiable from inscriptions recovered on site, and in their layout “wrote” the structure of the unit onto the landscape. This is seen most clearly in the obsessively regular Flavian and earlier Antonine examples in the northern provinces. For example, trenching and aerial photography of the buried foundations of Inchtuthil, apparently intended to replace Chester as the base of legio XX Valeria Victrix, reveal the disposition of barrack blocks for all ten cohorts, including expanded accommodation for the first cohort, increased in size in Flavian legions (figure 4.6: Breeze 1969; Pitts and St. Joseph 1985).
Many fort plans, like that of Inchtuthil, are actually based on limited trenching and, using plans of other sites plus any air photos or geophysical data, “joining the dots,” usually with a ruler. The result can be significantly to exaggerate the regularity and orderliness of fort plans, tending to idealize them. With regard to forts and the Roman military generally, ruler lines, actual and metaphorical, may often be more in the minds of modern scholars than real. The Flavian base at Elginhaugh, Scotland, is a rare example of a timber-built installation almost completely excavated to modern standards (figure 4.7; Hanson 2007). Its actual plan is noticeably “shaky,” irregular in line, and also in conformation. For even with a complete plan, it has still proved difficult to decide the nature and composition of its garrison (thought most likely to have constituted part of a cavalry regiment).
Such early, timber-built installations of odd sizes or conformation, lacking epigraphic evidence are, then, harder to understand than Inchtuthil. Another example is the first-century base at Longthorpe, England, too small for a whole legion and far too big for an auxiliary regiment: it may have served as the winter quarters of a “mixed brigade” comprising a legionary vexillation and attached auxilia (Frere and St Joseph 1974). However, such vexillation fortresses tell us something of ad-hoc tactical formations during active conquest phases.
For other contexts and periods, things are much more obscure even than this. “Playing-card forts” exist in the East, but marching camps are hardly known, while legions in provinces like Syria and Egypt were stationed within or adjacent to major cities. We know vanishingly little of most of these bases due to lack of fieldwork, or because most such sites have been continuously occupied ever since, and so are deeply buried beneath medieval buildings or modern suburban sprawl. Only two urban bases have been explored in any detail, both relatively late and in abandoned Syrian cities: a Tetrarchic example at Palmyra (Baranski 1994), and that at Dura-Europos, consolidated in the 210s (Rostovtzeff 1934; Pollard 2000; James 2007; see figure 4.2). The latter occupied the northern third of the city, with a headquarters building (principia), military baths, and amphitheater inserted. However, most of the base interior comprised civilian houses converted to military accommodation startlingly irregular by comparison with earlier European playing-card forts, and therefore much more difficult to read in terms of the formations they were intended to accommodate—which the Dura papyri and inscriptions (above) suggest comprised one or more legionary vexillations, cohors xx Palmyrenorum, and perhaps other auxiliary contingents.
Figure 4.7 Elginhaugh, a Flavian turf-and-timber base discovered by aerial photography in the 1970s, and extensively excavated in advance of construction work. Note the underlying regularity, but absence of straight lines. Note also the annex; what went on immediately outside the ramparts remains generally understudied. Simon James, after Hanson 2007.
For the organization of most other ancient armies, the archaeological record is much more obscure than this. No others were as obsessed with building, and with expressing regularity through layout, as the early Imperial Roman military. Late Roman bases have more varied, less regular, and frequently ephemeral interior buildings which are hard to interpret in terms of unit organization. Greek or Macedonian campaign infrastructure is hardly known. The Sasanians, and perhaps before them the Parthians, built systems of military bases on their frontiers, especially facing Central Asia, but these are barely explored yet (Nokandeh et al. 2006). Indeed, even the all-conquering Republican Roman army remains virtually invisible in these regards. Among the earliest remains of this kind are the siege camps around Numantia, Spain, dating to the second century B.C. (Dobson 2006). And even early Imperial Roman base plans still pose problems. Notably, it has often proved difficult to identify where cavalry units kept their horses (probably because most were normally kept outside the walls), and only recently have special cavalry barracks finally been securely identified, for example at Wallsend (Hodgson and Bidwell 2004).
Archaeology of military bases, then, sometimes provides assistance with understanding the anatomy of military formations. However, it can provide valuable qualitative and quantitative information regarding other aspects of military organization, such as provision and maintenance of infrastructure, services, and supply systems. For example, deposits of animal bones and seeds can indicate sources of foodstuffs and the organization of their provision. Weed seeds contaminating grain may prove to come from distant environments, revealing the operation of long-distance water-borne supply systems rather than reliance on local production. Exceptional preservation conditions can provide astonishing detail on the plant food, fodder, and fuel consumed by military contingents, as at Mons Porphyrites, Egypt (Veen and Tabinor 2007). The kinds of bones preserved and details of butchery can show whether meat was supplied to military bases on the hoof, or already processed through curing. Such details can also suggest where the animals were coming from: the cattle consumed at the Claudian base at Alchester were evidently of native breeds, procured fairly locally (Stallibrass and Thomas 2008). And in the Roman case, it tells us something of the disposition of hospital facilities, and the medical system which kept Imperial armies operational (Davies 1970; Baker 2004).
ARCHAEOLOGY AND “THE OTHER SIDE OF THE HILL”
Paradoxical as it may seem, the most important contribution archaeology makes to the study of Greco-Roman warfare may come from the light it sheds on non-classical peoples, not only on “barbarians at war,” but societies as a whole, in peace as well as conflict.
It may seem obvious that, to understand any war, it is essential to look at both sides; to research the world wars, for example, historians now routinely access the political and military archives of all principal combatants. From the outset, at all levels, from “grand strategy” to fighting methods, armies have to respond to the capacities and customs, strengths and weaknesses of their foes. However, in the study of ancient armies and warfare, seeking to comprehend both sides is still too rarely attempted. Much scholarship on the Roman military and frontiers, for example, often still seems mesmerized by the internal workings of the “Roman war machine” to the virtual exclusion of the world in which it operated.6
To understand a phenomenon such as Alexander’s Macedonian army or the Imperial Roman military at war, then, we need to consider it in the context of its actual and potential foes. However, from the textual record this is often effectively impossible. The Hellenistic states fought unlettered “barbarians” from the Balkans to Afghanistan, as did the Romans in Western and Central Europe, from the Republic’s wars in Hispania to Imperial invasions of Caledonia, to the conflicts with Germans, Sarmatians, and Huns. Even for clashes with empires like Carthage or Parthia, surviving written testimony is universally biased toward the classical viewpoint because, being both literate and commonly the victors, Greeks and Romans wrote the histories. For example, our understanding of the classical Greeks’ duel with Achaemenid Persia is hampered by relative lack of texts surviving from the Persian side. Later, Parthian records were suppressed by the Sasanians, whose own historical traditions have left only limited traces via conquering Islam. Copious surviving writings of the Jewish and Christian traditions provide a tantalizing independent window on Hellenistic and Roman military domination, but are of limited help, being more concerned with religious matters than recording historical events.
In general terms, archaeology provides a fundamental contribution in giving a “voice” to these antagonists of Greeks and Romans to societies silent—or silenced—in ancient textual records. Take, for example, our knowledge of Rome’s rise to power over the multiple indigenous peoples and Greek colonists of Italy during the fourth and third centuries B.C. We lack writings from Samnite, Etruscan, or Cisalpine Gaulish perspectives to place alongside the Greek and Latin accounts which present the history of the period overwhelmingly from the Roman viewpoint. However, conversely, the archaeological record for Romans at war before the second century B.C., in terms of weapons, other military remains, or visual representations, has proved to be very meager, while by contrast we have striking archaeological evidence for the martial culture of other peoples in Italy.
These are known to us through weapons and armor they buried with their dead, or offered to their gods, and through some remarkable visual representations of martial gear, warriors, and armed violence in action. The Samnites and Lucanians of the south buried some males with full panoplies (Italia 2000), while the Gauls of the Po plain, active participants in peninsular wars from at least the 390s B.C. when they notoriously sacked Rome (Williams 2001), also often interred arms with their dead (Vitali 1991).
Etruria and Campania have also produced spectacular decorated tombs with martial themes, while Lucanian examples from Poseidonia (Paestum) depict funeral games, especially single combats thought to be the origin of Roman gladiatorial contests. Such paintings and finds of armor confirm that the splendor of Samnite arms recorded in Roman histories was not invented or exaggerated by Latin authors simply to make ideological contrasts with Roman austerity (see below).
These material remains help us to keep non–Greco-Roman peoples in view as actors in the processes which led to Roman dominion over Italy. They were not simply gladius fodder for uniquely aggressive Romans, as some accounts have been inclined to treat them. Such evidence helps remind us that all the peoples of Italy and the Po plain were bellicose; Roman military power in the later Republic did not emerge from a vacuum, through any special, precocious, gods-given martial talent wielded against hapless victims: rather, it was the outcome of long apprenticeship in a school of very hard knocks.
Wider archaeological research also calibrates, corrects, and often falsifies outright received conceptions of the classical world’s “barbarian foes” derived from ancient texts. For example, to Greeks and Romans, Gauls were barbarians par excellence, stereotypically portrayed as semi-nomadic savages who lived by the sword and by plundering: timeless primitives, animal-like in their lack of self-control, their warriors headhunters, their priests addicted to human sacrifice. Polybius, for example, decried the lack of arts (tekne) of the Po valley Gauls (Polyb. 2.17.10). These, together with their more numerous transalpine cousins, were regarded as teeming hordes posing an ever-present menace which had burst into Italy before (Gauls sacking Rome 387/6 B.C., and rampaging through Etruria in 225 B.C. until defeated at Telamon), and might again. They induced among Romans an abiding terror Gallicus (Williams 2001). But how far was all this “racist” denigration and atrocity propaganda intended to belittle prehistoric-becoming-protohistoric peoples who, from the annexation of the Po plain during the Punic Wars to Caesar’s advance to the Atlantic, became victims of Roman imperialism?
Archaeological discoveries, old and new, certainly confirm the importance of the martial aspect of Gaulish societies. Some buried arms with their dead, or sacrificed them in rivers and lakes. Indeed, the extensive offerings recovered from Lake Neuchâtel at La Tène, mostly arms, gave the culture of the Gauls its archaeological name (for the swords from the site see De Navarro 1972).
The Gauls’ reputation for head-taking in war has also been confirmed by a gruesome discovery at Ribemont-sur-Ancre in northern France (Brunaux 2001). Here, at what seems to have been a shrine, a large wooden rack was erected in the early third century B.C., and from it were suspended the bodies of scores of warriors, bearing their arms—but minus their heads. Whether these men all fell in battle, were massacred as prisoners, or were formally sacrificed, is unclear. Such evidence, deposited far from the Mediterranean and before Rome began to interfere significantly inside Gaul, confirms, in case there was any doubt, that Gaulish armed violence was real, and cannot be solely blamed on stresses induced by proximity to an aggressive and interventionist Roman Empire.
Archaeology tends, then, to vindicate the Gallic reputation for bellicosity and fearsome behavior.7 However, with regard to other aspects of Gaulish culture, archaeological research has revealed that views based solely on ancient texts are often wildly wrong, leading to mistaken understandings of the military interactions, and intertwined political and socioeconomic dynamics, between Romans and Gauls.
These misunderstandings range from battlefield realities to the highest levels of strategy and politics. So, for example, Polybius presented a ludicrous picture of Gauls charging Romans and slashing downward with their long iron swords, only to have these bend on impacting a shield, requiring the warrior to withdraw and stamp them straight (Polyb. 2.33)! He also claimed these swords lacked a point, and were good for nothing but slashing (2.33, 3.114, 6.39). However, finds of actual Gaulish swords show that, while some indeed lack points, many had them. Similarly, metallurgical analysis shows that about a third were of a quality poor enough for bending to be a real danger—but most were much stronger (Pleiner 1993: 156–9). Polybius exaggerated on both counts, to make contrasts with very well-made Roman weapons used, especially against Gauls, for thrusting.
More important than such technical matters is the wider picture which archaeology is now giving us of “barbarian” societies, without having to see them through the jaundiced or uncomprehending eyes of Greco-Roman observers. For example, studies of settlements, agriculture, and economies have been fundamental to proper appraisals of the late Iron Age Gaulish societies which Caesar conquered. Many prove to have been populous, productive agrarian societies with sophisticated arts (if not of a kind valued by Greeks like Polybius). Far from timeless, they were changing very rapidly, some evolving into states with constitutions, laws, coinages, and tax systems (Haselgrove 2006; Collis 2007). Transalpine Gallic societies were rapidly conquered and became well integrated as provinces, not because they were hapless primitive sword fodder, but quite the opposite: because they were large and complex enough—that is, already sufficiently like the Greco-Roman Mediterranean, in matters of war, politics, and socioeconomic organization—to be amenable to integration. Many possessed fixed assets and political interests vulnerable to military threats, and powerful elites with whom Caesar could negotiate.
In contrast, Germania east of the Rhine proved intractable to Roman conquest under Augustus, resulting in the Varian disaster. Thereafter, as the northern frontier systems crystalized, Roman rhetoric represented the Germani, like the Gauls, as timeless benighted primitives, beyond the pale of civilization. However, archaeological research shows that, during the early centuries A.D., major socioeconomic developments were also under way across Barbaricum from the Baltic to the Black Sea (e.g., among the Goths: Heather 1996). These saw more productive agricultural regimes, growing populations and material wealth, and increasing social stratification. These changes probably arose from internal dynamics, but were shaped by proximity to the Roman Empire, primarily through the effects of war and unrecorded diplomatic exchanges. The long reach of Rome can be seen in excavations of settlements, graves, and, most spectacularly great bog offerings of military equipment, especially from Denmark (Ilkjaer 2000; Jørgensen et al. 2003). These suggest the rise of statelets as far as Scandinavia, with war bands evolving into organized armies, equipped with the latest military technology: Roman pattern-welded swords (Biborski and Ilkjaer 2007). Many such weapons, silver coins, and other Roman artifacts are thought to represent long-range diplomatic efforts to manipulate barbarian polities, to foster wars among them by selective subsidies and “military aid,” thereby reducing pressure on the Roman frontiers (Erdrich 2001). Whatever the truth of this, such archaeological remains indicate Rome was having significant effects even on regions virtually beyond her ken. They suggest that, across Barbaricum, Rome was inadvertently shaping the societies that would ultimately overwhelm the western empire.
THE ACCOUTREMENTS OF WAR
Arms and armor are dealt with elsewhere, but one other example of Roman-era martial material culture may be cited here to show archaeology’s particular contribution to understanding the practical realities of ancient warfare. This is the breakthrough in our understanding of Roman-era saddles.
It is well known that the stirrup was not introduced to Europe until after the fall of the Roman West, and it has therefore been widely assumed that, compared with the shock cavalry of medieval and later times, stirrupless ancient cavalry could only have been of relatively limited effectiveness. Because they could not have had a very secure seat, they would have been restricted in the force with which they could wield a sword or thrust with a spear, for fear of falling. However, brilliant studies of archaeological remains of Roman saddles, combined with meticulously made experimental reconstructions and practical experimentation, have effectively proved this wrong (Connolly and Van Driel-Murray 1991).
Fragments of Roman saddles (partial leather coverings, and sets of metal pommel plates), and some depictions, have long been known. It was apparent that these saddles had four upstanding pommels, two in front of the rider’s legs and two behind, but the details remained obscure. To work out how they were actually constructed, Carol Van Driel-Murray, an expert on ancient leather, studied the shaping, stitching, stretching, and wear patterns on the surviving fragments of saddle coverings. She worked closely with Peter Connolly, experimental archaeologist and illustrator, who produced a series of trial reconstructions until the details perfectly matched the known construction techniques, reproducing the stretching and wear patterns. The result, based around a reconstructed wooden saddle tree whose shape was dictated by the details of the leather, was revelatory, especially when actually tried on a horse (see below figure 23.4). The purpose of the four pommels becomes immediately apparent: the rear two are tight behind the rider’s buttocks, preventing him from sliding backward, while the front pair project outward over the rider’s thighs, stopping him from sliding forward or falling sideways. The rider’s seat is as secure as with stirrups; he can slash with a sword or thrust with a spear to maximum effect. The problem, it turns out, is not falling off, but getting off, or on. Without stirrups, the projecting pommels require the rider to vault on, but extensive experience with replicas have shown this to be a skill rapidly developed.
It appears that this Roman saddle was adopted from the La Tène world or perhaps the Danube lands. Subsequent research has shown that it was also standard in the Partho-Sasanian world (Herrmann 1989). Like many other facets of contemporary equestrian warfare, including horse archery, perhaps armored lancers, and, later, the stirrup, this type of saddle probably originated on the Eurasian steppe. Consequently, such practical archaeological research regarding the function of the Roman four-pommel saddle has much wider relevance to the study of ancient warfare.
EXPERIMENTAL ARCHAEOLOGY AND CROSS-CULTURAL COMPARISON
The case of the four-pommel saddle provides a good illustration of the limitations of “armchair” study of ancient warfare, and a striking vindication of the value of reconstruction (perhaps better, simulation), or experimental archaeology. However, ethical considerations, not to mention laws, place limits on experimentation with replica equipment, and we obviously cannot simulate the elation and terror, agony and horror of ancient war. Such work, then, is largely limited to technical practical matters—but not wholly so (see below).
While some have been soldiers, few currently active researchers have direct experience of lethal violence, and scholarship, wider society, and the law would all doubtless look askance at civilian academics seeking to gain it (they usually limit their aggression to reviews of one another’s books). However, there is real benefit to be had from witnessing, or even better, engaging in some relevant practices short of actual bloodshed.
Experimental archaeology or reconstruction (as in the case of the Roman saddle), so-called reenactment, more general simulation, and cross-cultural comparison with more recent armies and wars, can all help us understand how things may have been done and—equally important—how they were not done. Critical comparison—considering similarities and differences—is a valuable tool. For example, Sabin has used Hollywood battle scenes as a heuristic device for reflecting on what ancient battle cannot have been like, in his attempts to understand the mechanics of actual combat (Sabin 2000).
Similarly, few today have observed the evolutions of large cavalry formations. However, I witnessed a pro-foxhunting demonstration by several hundred mounted protesters in England, which gave a hint of the visual and visceral impression which would have been made by an ala of cavalry; five hundred horse in column, even peaceable and unarmed, make a powerful impact on eye, ear, nose, and through the feet—the ground reverberated under hundreds of tons of muscle, bone, leather, and metal on the move.
Analogies drawn from modern military experience present their own dangers arising from cultural and technological differences between antiquity and modernity. However, they can be invaluable, especially with regard to basic physiological and environmental matters affecting soldiers. Hence Gabriel and Metz used data ranging from Napoleonic examples to modern military experiments to address matters rarely mentioned in classical sources: the impact on effectiveness and health of soldiers, simply of prolonged marching in column through dusty, very hot or very cold environments. Their work makes clear that classical armies did not have to cross the Alps to suffer significant rates of attrition in sick and dead, even before reaching the battlefield (Gabriel and Metz 1991: 104–09).
Ethnographic parallels provide a similar kind of cross-cultural comparison. For example, the ancient traditions of horse archery with a composite bow continued long enough to be thoroughly documented, and many of its skills survive among enthusiasts (e.g., in Hungary) and specialists. It was from an experienced bow maker, archer, and horseman, for example, that I learned how shooting is more accurate at a gallop than a slower jerking trot, because the horse provides a more stable “firing platform” (James 2004: 198).
Academic purists are rendered uncomfortable, if not tempted to snigger, by reenactors. This is a mistake (Croom and Griffiths 2000). There is much to learn from observing and talking to them, positive lessons and indeed reflections on where they are probably getting things wrong. For example, many reenactors are impressively skilled craftsmen who take great pride in their creations, which can help the archaeologist understand the traditions and sheer effort involved in creating ancient martial material culture. On the other hand, by comparison with archaeological examples, reenactors’ equipment is often clearly too well made—an observation which prompts valuable reflections on the standards of manufacture and likely durability of real ancient arms.
Figure 4.8 Representations of Roman soldiers of the first half of the third century A.D.: A. Tombstone of Aurelius Surus, bucinator of legio I Adiutrix (Istanbul Museum); B. Unknown, Rome (after Bishop and Coulston 1993); C. Tombstone of M. Aurelius Lucianus, Rome (after Bishop and Coulston 1993); D. Unknown, holding strap ends, funerary relief, from Herakleia-Perinthos (Istanbul Museum); E. Sasanian relief depicting Roman emperor, probably Valerian, Bishapur II (after Herrmann 1983); F. Sasanian relief depicting Roman emperor, probably Philip, Bishapur II (after Herrmann 1983).
Watching reenactors performing at “heritage events” is equally informative. It can, for example, be enormously impressive to see the power of replica stone-throwing and arrow-shooting torsion artillery. I have also found invaluable watching a full century of “Roman legionaries” put through their paces—and afterward discussing it with the “centurion” (it showed how hard verbal communication was, even for an officer with a loud voice and a small formation in ideal conditions, with so much noisy equipment). And I have found it highly instructive to try on replica armor, to feel the weight of sword and shield, to ride a horse, to vault into a Roman saddle, and to shoot with a composite bow.
Figure 4.9 Reconstructions of a Roman centurion, soldier and military tribune of the earlier third century A.D., based on archaeological finds from Dura-Europos and elsewhere, and on paintings and tombstones. Drawn by Simon James.
Perhaps the most insightful experience of this kind I have had was something deceptively mundane: simply to feel what it is like to walk and move while wearing a spatha (Roman long sword) on a baldric, and a cloak arranged in the style seen on many third-century depictions of Roman soldiers (figure 4.8). I did this primarily to create photographic references for a reconstruction painting I was making of the appearance of Roman soldiers from Dura-Europos (figure 4.9; James 2004: 256–9, plates 11–13). Like changing from jeans to a tuxedo, such dress makes one stand and move in different, particular ways. This physical experience viscerally reinforced theoretical ideas I was encountering on how artifacts do not just hang on bodies, but are active components in creating the feeling and materiality of human identity—not least of the warrior or soldier.
ARCHAEOLOGY OF FIGHTING MEN: CREATING WARRIORS AND ARMIES
In material terms, we can think of a human as a particular kind of primate compounded with artifacts (Schiffer 1999: 3). Appearance is one of the fundamental bases on which our identities are constructed. It comprises artifacts worn or carried, and treatment of the body, including modifications like tattooing or simpler grooming, especially of head and facial hair. The last merges into the matter of style, that is, how hair and garments are worn, and artifacts carried. All this in turn relates to how one stands and moves, and the acts one performs, while in such guise. (We would be startled to see someone in a business suit wielding a shovel, and not just because the clothing is impractical for the task: it would also contradict the general implications about the wearer’s identity and status signaled by the clothing.) Creating a particular manifestation of the body is fundamental to building a sense of who we are—to ourselves (literally, a feeling of who we are)—and, primarily visually, to others. Appearance is equally vital in signaling our sense of collective identity, of belonging to particular groups. As any sergeant major knows, soldiers’ appearance is a foundation stone for constructing and maintaining regiments and armies, developing their self-image, and making an impression on their enemies.
Similarly, in ancient war, appearance was of supreme importance to soldiers and armies in creating Us, and intimidating Them. The splendor of Samnite or Gaulish arms was intended to make the wearer feel good about himself, to impress his comrades, and to awe his foes. The seriousness with which the last point was taken is clearly documented in Roman officers’ concern to deflate enemies whose carefully cultivated appearance—glittering splendor, stature exaggerated with lofty helmet crests, and arms emblazoned with terrifying animals and monsters—was working all too well on the minds of their soldiers (Livy 9.40.1–6; 10.39; Frontin. Str. 1.18–19). They sought to burst the balloon by taking such men prisoner and stripping them of their finery—and so of their self-identity and constructed mystique (App. B Civ. 4.8; Plut. Mar. 16). Here subliminal, visceral, instinctive matters—visual intimidation—cross into conscious discourse, expressed most clearly in Roman rhetorical contrasts of the austere plainness their own arms with glittering Samnite panoplies, golden Gallic torcs, or (later) Parthian embroidered trousers, gaudiness to be dismissed as effeminate vanity (e.g., Livy 9.40.4–5). For Romans also used material means, albeit in rather different ways, to help construct their own sense of personal and collective identity as soldiers (milites).
Roman soldiers of the Middle Empire (ca. A.D. 200–250), for example, could know every man in their cohort or ala, at least by sight; these groupings constituted military communities of immediate daily experience. Legionaries would often see much, if not all, of their legion assembled in one place for ceremonies, exercises, or campaigning. However, no soldier—and few emperors, with exceptions like the peripatetic Hadrian—would ever see more than a small fraction of their commilitones (“fellow soldiers,” as emperors were expected to treat them). Rome’s milites as a whole constituted an “imagined community” (Anderson 1991), membership of which was signified, experienced, and made instantly recognizable through common martial culture. This was of course largely discursive, comprising shared ideology and values, expressed via special language (the sermo militaris, a demotic Latin dialect filled with military slang and technical jargon, much derived from the languages of foreign auxiliaries). However, martial identity was substantially articulated visually and physically, through posture, movement, and actions, employing common material culture, not just of combat, but of equestrianism and military dress. These physical traits made soldiers instantly mutually recognizable, and distinct from mere men with weapons (Petron. Sat. 82 describes the one encountering the other). They comprised a visual discourse of military appearance, style, and behavior, about which we can say a great deal, drawing on data from archaeology, texts, and visual representations, allowing us to some extent to get inside the minds of ordinary soldiers, their self-image, motivations, and world view.
The contrast between manly Roman plainness and effeminate barbarian ostentation was a rhetorical trope rooted deep in Republican Roman self-definition, and conscious differentiation from their foes. It persisted at least down to the Middle Empire. Martial dress in the third century A.D., known primarily from archaeological finds, paintings, and tombstones, evinced a quite complex code. It expressed traditional rhetoric of Roman soldierly austerity while toiling in dust and heat for the state (grayish trousers, yellow-brown sagum), but also, through the whiteness of his tunic and its purple detailing, asserted the soldier’s status among other free privileged males who wore the same color scheme.8
Even off duty and out of camp, milites materially expressed their special privilege of bearing arms on behalf of the state by routinely wearing swords, while their waist belts and sword baldrics themselves were invested with symbolic meaning as marks of martial status. Rhetoric of plain dress was transgressed in the special fields of weapon hilts, scabbards, belts, and cloak brooches. In these restricted but prominent zones, archaeology and tombstones reveal elaborate decoration: openwork ornament, use of silver for fittings or its widespread simulation through tinned bronze, inlay with niello or contrasting metals or brightly coloured “enamel.” Gods, religious symbols, and ideologically loaded messages were also sometimes literally written on the clothed body, belt fittings incorporating texts sometimes expressing individual prowess (FELIX VTERE, “use with good fortune,” implicitly presentation pieces; figure 4.9), incorporating a collective prayer for the soldier’s unit: “([Jupiter] Best and Greatest protect us, a regiment of fighting men all”: [trans. L. Allason-Jones; Bishop and Coulston 2006: 162]), or simply spelling out ROMA.
In such ways archaeology can give some voice to the concerns and aspirations of the ordinary soldiers of historical societies, especially Roman milites who in ancient literature are usually treated either as lumpen spear-carriers or as dangerous, volatile, and ignorant men, sometimes escaping control to become little better than armed mobs (James 2011a).
ARCHAEOLOGY AND MARTIAL CULTURAL EXCHANGES
A final major contribution from archaeology is another facet of its ability to see the other side of hill in ways texts cannot. In revealing the martial material culture of classical societies and of their “barbarian” antagonists, and in tracking these as they change through time, archaeology illuminates important processes of exchange and interaction across cultural boundaries in the field of warfare, largely or wholly unmentioned in the historical record.
To pursue the example of Roman military dress and arms circa A.D. 200–250, these are strikingly different from those of circa A.D. 100, familiar from Trajan’s Column. Unfolding during the second century A.D., the changes were so radical that they have been described as the “Antonine revolution” (Bishop and Coulston 2006: 128–48). This shift is not attributable to central authority. Roman soldiers’ equipment was not governed by modern-style technical specifications or written regulations, and was not centrally produced (at least before ca. A.D. 300: James 1988); rather, it was locally produced, its form and embellishment regulated by custom, convention, and traditions among the soldiers and their regiments, which evolved over time. However, archaeology helps reveal how the “Antonine revolution” came about, and why. Not least, it shows where the new types and styles came from, matters on which the texts are virtually silent. It turns out that, for the most part, this “revolution” did not represent Roman innovation, but drew on types of clothing and weapons taken from the martial cultures of Rome’s frontier provinces and peoples far beyond, adapted to varying degrees and melded with existing martial cultural traditions to suit the needs and tastes of Roman soldiers as a whole.
It is clear that many features of middle Imperial Roman equipment came from the Sarmatian peoples, through conflicts and exchanges in the Danube region and around the Black Sea. These included new military belt designs and an entirely different way of suspending the sword, using a slide or runner on the scabbard rather than lateral rings. Sarmatian models also extended to horse harness, probably the new draco (dragon) wind sock standard, and provided a source of inspiration for Roman lancers with armored horses. The influence of these Iranian-speaking peoples was so great that Coulston has written of the “Sarmatization” of Roman equipment (Coulston 2003).
However, there is good reason to think that the Euphrates frontier was equally important in the development of Roman martial culture (James 2006). Here, against the Parthians, the Romans first encountered both armored lancers and horse archers, and indeed were already employing the latter on the Rhine in the first century A.D. There are reasons to suspect that some widespread new Antonine belt designs, such as the ring buckle and the fashion for long belt ends hung in a curve to the hip, were copied from Partho-Sasanian sources, while the scabbard slide may equally have entered Roman usage via the Euphrates. Iranian influence appears to have continued to be strong into the late Empire: fourth-century Roman animal-headed buckles possibly, and more certainly the radical new styles of Roman helmet which appear around A.D. 300, can be attributed to Sasanian inspiration: a Persian example from the Dura countermine forms a clear prototype for them (figure 4.3; James 1986).
By no means least, between A.D. 100 and 200 there was a fundamental change in the basic clothing worn by Roman soldiers. The bare limbs of traditional Roman dress gave way to a long-sleeved tunic and long trousers, apparently with sewn-in feet like medieval hose. These garments, more suited to the extreme winters of central and northern Europe where the armies were operating, and more in keeping with indigenous eastern traditions of bodily modesty, were worn with the rectangular cloak (sagum), a northern “barbarian” garment Romans had long assimilated. In the second century they simply adopted the rest of an ensemble which seems to have been general male dress from Gaul and Germany to the Crimea (figure 4.10). As we saw, it was adapted to Roman tastes largely through its coloration. However, to Roman civilians of the interior, it was one feature among many which made Roman soldiers seem, increasingly and literally, outlandish.
This divergence between the soldiers of the frontier provinces and metropolitan civilians was already starkly seen during the civil wars of A.D. 69 when, on their arrival in the capital, the men of Vitellius’s northern army seemed to the people of the city to be shaggy aliens, and mutual misunderstandings and antagonism led to bloodshed (Tac. Hist. 2.88–89). Yet, paradoxically, this fission was the long-term result of one of Rome’s most ancient defining characteristics, one already identified by Polybius in the second century B.C. The Greek historian, who fought the Romans and later accompanied them on campaign, identified their exceptional readiness to adopting anything done better by others, especially in war (Polyb. 6.25). This tradition of openness, which famously resulted in Republican adoption of the “Hispanic sword,” was maintained by the Roman military down to the late Empire (cf. Polyb. 3.114, with Walbank 1: 209; the Suda, s.v. machaira, may preserve another fragment of Polybius describing Roman adoption of the weapon). It also extended to people. Her own foundation myths of mongrel origins underpinned a belief that any worthy male could become Roman, a practice extended from friendly foreign aristocrats to useful soldiers. Hence the evolving military equipment of the Imperial period was partly a result of continuing to adopt good ideas, but the mechanisms whereby these new traits entered Roman service were not only copying and capture. To a considerable extent, new gear came in on the backs of their original owners, recruited into the armies from newly conquered peoples, and even beyond the frontiers: the Romans recruited Germans, Sarmatians, Parthians, and many others in great numbers.
Figure 4.10 The ‘Antonine revolution’ in Roman military dress, and its primary source in the dress of northern ‘settled barbarian’ peoples. A. Tombstone of Faltonius, a legionary of the first century A.D. buried at Mainz. He still wears the traditional short Italian tunic, which leaves the limbs exposed, and the poncho-like paenula; B. a Danubian German of the second century A.D. from the Column of Marcus Aurelius at Rome. He is clad in a long-sleeved tunic, close-fitting breeches, and a sagum fastened at the shoulder by a brooch; C. Valerinus, a Praetorian guardsman buried at Rome in the early decades of the third century A.D. His dress is in all essential regards ‘northern barbarian’ in form; other, colored representations of soldiers show that close-fitting breeches were the norm by this period, and were probably originally represented here by paint. Drawn by Simon James.
Such soldiers were integrated into the armies and made Romans; yet many retained traditional tactics and arms considered useful in war, and these became accepted as part of the ever-evolving definition of what was Roman martial culture, spreading to other units. This was not, then, “barbarization” of the armies, for historical evidence shows how these soldiers of foreign origin came to think of themselves as Romans first. Rather it was a process of “cultural bricolage”—the taking and adapting of elements from more than one culture, and melding them into, and accepting them as, parts of another evolving one (Terrenato 1998).
That Vitellius’s soldiers seemed so alien to Romans of Rome was also partly due to another historical process. Augustus’s creation of a socially distinct professional military concentrated near the frontiers also inadvertently resulted in an increasingly demilitarized citizenry in the interior. It led to diverging civil and military manifestations of Romanness. That of the “core provinces” continued to draw on its Greek, Italian, and other Mediterranean roots. However, that of the soldiers on the frontiers—who also considered themselves fiercely Roman—was, in a sense, an “expatriate Romanness,” although with equal claim to be “real Rome,” for the frontier zones were where politics was increasingly focused, and where emperors and history were made. Yet paradoxically, in terms of ethnic origins, notions of martial masculinity, life experience, and material culture, Roman soldiers had more in common with the peoples around and beyond the frontiers, than with civilians of the interior of the Empire.
Roman martial culture, and especially the archaeology of arms and dress, illustrates how intimately connected Roman soldiers were with the peoples against whom they fought—and from whom they recruited (and, sometimes, to whom they deserted). The connectedness of Roman martial culture with that of Germans, Sarmatians, Parthians, and Sasanians reveals the frontiers not as peripheries, but as centers of zones, not only of violent conflict, but of intense, multidirectional cultural interaction and exchange, of weapons, ways, and men. Greeks and especially Romans at war, then, can only be understood as part of a much wider picture, a network of interactions reaching to the Baltic and Central Asia, not readily apparent from surviving texts.
CONCLUSION AND FUTURE POTENTIAL
Archaeology is a relatively young discipline, still discovering what it can do—and what it cannot. Classical archaeology in particular has been slow to develop fully, having grown up in the colossal shadow of centuries-old traditions of text-focused study of the Greco-Roman world. It faced a prolonged struggle for autonomy in its aims and methods from those of traditional classical studies, which long effectively imposed ill-fitting “text-driven” agendas on it. This remained true of the study of war and the military even longer than it did of civil aspects of classical archaeology. However, the discipline now has the confidence of finding its own voice (Alcock and Osbourne 2007), even if many text-based scholars have yet to comprehend this fully.
One recent example of archaeological discovery concerns what might seem a minor technical point, yet it falsifies a hallowed central image of ancient warfare, thereby providing a symbolic example of the potential contribution archaeology is making to the field. The famous gladius Hispaniensis, the “Spanish sword” of the Republican legionary, almost literally carved out the Empire in the last two centuries B.C., and is a Roman martial icon. This weapon, until recently known almost solely from literary descriptions, is routinely described as a “short, thrusting sword,” contrasted especially with much longer Gallic slashing blades. Yet actual examples of the gladius Hispaniensis, recently identified in existing museum collections, show that it was not short at all; it was actually as long as the earliest known imperial spathae, commonly characterized as long slashing swords, recovered from Scotland (Republican gladii: Quesada Sanz 1997; Connolly 1997; Flavian spathae from Newstead: Curle 1911: 183–5, plate 34, nos. 6, 7 and 13). Further, the Republican weapon looks as suited to cutting as to thrusting—and, returning to the texts, Polybius makes clear it was actually used in both modes (e.g., Polyb. 6.39), even if thrusting was tactically preferred against foes like the Gauls to foil the long reach of their blades. The case of the gladius Hispaniensis exemplifies a general point: our most cherished received ideas about the classical past are open to challenge from new research, not least in archaeology.
It also exemplifies the value of reexamining museum collections and excavation archives as an important source of new discoveries, alongside fresh field research. Excavations conducted by investigators are ever better primed to know what to look for, and how. We are expanding the range of kinds of evidence to seek, for example, using more advanced survey and prospection methods, including geophysical techniques, to recover hitherto inaccessible data from battlefields, and even wider “conflict landscapes.”
In terms of theoretical and methodological approaches to interpreting and understanding the data, as in most other branches of archaeology, borrowing concepts and approaches from cognate disciplines (here, anthropology as well as recent military studies), and overt, critical use of cross-cultural comparison, are likely to be increasingly important.
The martial archaeological riches of the early and middle Roman Imperial period, which have provided most of the examples presented here, are clearly the exception, not the rule. In antiquity the norms were to build more modest military infrastructure, andnotto deposit military equipment in the ground. To get at the archaeology of war for most other ancient classical and related cultural contexts, then, including classical Greece and much of the Hellenistic world, as well as the Roman Republic and (except for military installations) the later Roman Empire, we have to try harder, and deal with sparser remains. However, the arms in the tomb of Philip II and the cache of Hellenistic equipment from Aï Khanoum, Afghanistan, underline that the more we dig, the more finds we will make. And, indeed, as Aï Khanoum highlights, many regions remain untapped, or at best sketchily explored especially in the eastern half of the Greco-Roman world.
As we saw, archaeology is especially valuable for exploring the martial culture of the antagonists of Greco-Roman societies. Contexts like Samnite or Iberian “warrior graves,” Gallic shrines, and Scandinavian bog deposits offer archaeological windows on ancient warfare different from those of the Roman Empire, and continue to provide remarkable discoveries. The end of the Cold War is now leading to better understanding of the steppe peoples around and beyond the Black Sea. We can also hope that Western Asia will be further opened up to research, not least the Iranian heartland of the Achemenid, Parthian, and Sasanian empires. And this world had its own frontier systems, including the newly explored Gorgan Wall (Nokandeh et al. 2006; Rekavandi et al. 2007).
Rather than merely illustrating Greco-Roman texts, archaeology has shown it can range far beyond them on its own terms, in time, space, and kinds of information preserved. It may also help us to reflect back on surviving writings, to reinterpret them afresh. We will make the most effective advances in the study of ancient warfare where text-based and material-based studies work together, reflexively, exploiting their peculiar strengths in conjunction.
BIBLIOGRAPHY
Alcock, S. E., and R. Osborne (eds.). 2007. Classical archaeology. Oxford.
Anderson, B. 1991. Imagined communities: Reflections on the origins and spread of nationalism. London.
Baker, P. A. 2004. Medical care for the Roman army on the Rhine, Danube and British frontiers from the first through third centuries. Oxford.
Baranski, M. 1994. “The Roman army in Palmyra: A case of adaptation of a pre-existing city,” in Dąbrowa 1994: 89–98.
Becker, A., G. Rasbach, and S. Biegert. 2003. “Die spätaugusteische Stadtgründung in Lahnau-Waldgirmes. Archäologische, architektonische und naturwissenschaftliche Untersuchungen.” Germania 81: 147–99.
Biborski, M., and J. Ilkjaer. 2007. Illerup Ådal, Die Schwerter Und Die Schwertscheiden: 11, Textband, 12 Katalog, Tafeln Und Fundlisten. Moesgard.
Birley, A. R. 1973. “Petillius Cerialis and the conquest of Brigantia.” Britannia 4: 179–90.
Bishop, M. C., and J. C. N. Coulston. 2006. Roman military equipment. Oxford.
Bowman, A. K. 1994. Life and letters on the Roman frontier: Vindolanda and its people. London.
———, and J. D. Thomas. 2003. The Vindolanda writing-tablets: Tabulae Vindolandenses III. London.
Breeze, D. 1969. “The Organization of the Legion: The First Cohort and the Equites Legionis.” JRS 59: 50–55.
Brunaux, J.-L. 2001. “Gallic Blood Rites.” Archaeology March/April 2001: 54–57.
Bull, S. 2007. Triumphant rider: The Lancaster Roman cavalry tombstone. Lancaster.
Collis, J. 2007. “The polities of Gaul, Britain and Ireland in the late iron age,” in C. C. Haselgrove and T. Moore (eds.), The later iron age in Britain and beyond. Oxford, 522–28.
Connolly, P., and C. Van Driel-Murray. 1991. “The Roman cavalry saddle.” Britannia 22: 33–50.
———. 1997. “Pilum, gladius and pugio in the late republic.” JRMES 8: 41–57.
Coulston, J. C. N. 2001. “The archaeology of Roman conflict,” in Freeman and Pollard 2001: 23–49.
———. 2003. “Tacitus, Historiae I.79 and the impact of Sarmatian warfare on the Roman empire,” in C. v. Carnap-Bornheim (ed.), Kontakt—Kooperation—Konflikt: Germanen und Sarmaten zwischen dem 1. and 4. Jahrhundert nach Christus. Marburg, 415–33.
Croom, A. T., and W. B. Griffiths (eds.). 2000. Re-enactment as research: Proceedings of the twelfth international Roman military equipment conference, South Shields, 1999. JRMES 11.
Curle, J. 1911. A Roman frontier post and its people: The fort at Newstead. Glasgow.
Dąbrowa, E. (ed.). 1994. The Roman and Byzantine army in the east. Cracow.
Davies, R. 1970. “The Roman military medical service.” Saalburg Jahrbuch 27: 84–104.
De Navarro, J. M. 1972. The finds from the site of La Tène: Vol. 1, Scabbards and the swords found in them. London.
Dobson, M. 2006. The army of the Roman republic: The second century B.C., Polybius and the camps at Numantia, Spain. Oxford.
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.
Eckstein, A. M. 2006. Mediterranean anarchy, interstate war, and the rise of Rome. Berkeley.
Elton, H. 1996. Warfare in Roman Europe, A.D. 350–425. Oxford.
Erdrich, M. 2001. Rom und die Barbaren: das Verhältnis zwischen dem Imperium Romanum und den germanischen Stämmen vor seiner Nordwestgrenze von der späten römischen Republik bis zum Gallischen Sonderreich. Mainz.
Freeman, P. W. M., and A. Pollard, J. Bennett, Z. T. Fiema, and B. Hoffmann (eds.). 2000. Limes XVIII: Proceedings of the XVIIIth International Congress of Roman Frontier Studies held in Amman, Jordan (September 2000). BAR International Series. Oxford.
———. (eds.). 2001. Fields of conflict: Progress and prospect in battlefield archaeology. BAR International Series. Oxford.
Frere, S. S., and J. K. St Joseph. 1974. “The Roman fortress at Longthorpe.” Britannia 5: 1–129.
Gabriel, R. A., and K. S. Metz. 1991. From Sumer to Rome: The military capability of ancient armies. Westport.
Goldfus, H., and B. Arubas. 2001. “Excavations at the Roman siege complex at Masada, 1995,” in Freeman and Pollard: 2001: 207–14.
Hanson, W. S. 2007. A Roman frontier fort in Scotland: Elginhaugh. Stroud.
Harris, W. V. 1985. War and imperialism in republican Rome, 327–70 B.C. Oxford.
Haselgrove, C. C. (ed.). 2006. Celtes et Gaulois, l’Archéologie face à l’Histoire, 4: les mutations de la fin de l’áge du Fer. Actes de la table ronde de Cambridge, 7–8 Juillet 2005. Glux-en-Glenne.
Heather, P. 1996. The Goths. Oxford.
Herrmann, G. 1989. “Parthian and Sasanian saddlery,” in L. de Mayer and E. Haerinck (eds.), Archaeologia Iranica et orientalis: Miscellanea in honorem Louis Vanden Berghe II. Ghent, 757–809.
Hodgson, N., and P. T. Bidwell. 2004. “Auxiliary barracks in a new light: recent discoveries on Hadrian’s Wall.” Britannia 35: 121–57.
Ilkjaer, J. 2000. Illerup Ådal: Archaeology as a magic mirror. Højberg.
Italia 2000. Italia dei Sanniti: Guida all Mostra. Milan.
Nista, L., and S. Capini. 2000. Italia dei Samniti: Guida alla mostra. Milan.
James, S. T. 1985. “Dura-Europos and the chronology of Syria in the 250s A.D.” Chiron 15: 111–24.
———. 1986. “Evidence from Dura-Europos for the origins of late Roman helmets.” Syria 63: 107–34.
———. 1988. “The fabricae: State arms factories of the later Roman empire,” in J. C. N. Coulston (ed.), Military equipment and the identity of Roman soldiers. BAR International Series. Oxford, 257–331.
———. 2002. “Writing the legions: The development and future of Roman military studies in Britain.” AJ 159: 1–58.
———. 2004. Excavations at Dura-Europos, final report. Vol. 7: The arms and armour, and other military equipment. London.
———. 2006. “The impact of steppe peoples and the Partho-Sasanian world on the development of Roman military equipment and dress, 1st to 3rd centuries A.D.,” in M. Mode and J. Tubach (eds.), Arms and armour as indicators of cultural transfer: The steppes and the ancient world from Hellenistic times to the early middle ages. Wiesbaden, 357–92.
———. 2007. “New light on the Roman military base at Dura-Europos: Interim report on a pilot season of fieldwork in 2005,” in A. S. Lewin and P. Pellegrini (eds.), The late Roman army in the near east from Diocletian to the Arab conquest. BAR International Series. Oxford, 29–47.
———. 2010. “The point of the sword: What Roman-era weapons could do to bodies—and why they often didn’t,” in A. W. Busch and H. J. Schalles (eds.), Waffen in Aktion. Akten der 16. Internationalen Roman Military Equipment Conference (ROMEC). Xantener Berichte 16: 41–54.
———2011a. Rome and the sword: How warriors and weapons shaped Roman history. London.
———. 2011b. “Stratagems, combat and ‘chemical warfare’ in the siege-mines of Dura-Europos.” AJA 115: 69–101.
Jones, R.H. 2012. Roman camps in Britain. Stroud.
Jørgensen, L., B. Storgaard, and L. G. Thomsen (eds.). 2003. The spoils of victory: The north in the shadow of the Roman empire. Copenhagen.
Kennedy, D. L. 1994. “The cohors XX Palmyrenorum at Dura Europos,” in Dąbrowa 1994: 89–98.
Lee, J. W. I. 2001. “Urban combat at Olynthos, 348 B.C.,” in Freeman and Pollard 2001: 11–22.
Leriche, P. 1993. “Techniques de guerre sassanides et romains à Doura-Europos,” in F. Vallet and M. Kazanski (eds.), L’armée romaine et les barbares du IIIe au VIIe siècle. Rouen, 83–100.
McCarthy, M. 2002. Roman Carlisle and the lands of the Solway. Stroud.
Nicasie, M. 1998. Twilight of empire: The Roman army from the reign of Diocletian until the battle of Adrianople. Amsterdam.
Nokandeh, J., E. Sauer, H. O. Rekavandi, T. Wilkinson, A. A. Ghorban, J.-L. Schwenninger, M. Mahmoudi, D. Parker, M. Fattahi, L. S. Usher-Wilson, M. Ershadi, J. Ratcliffe, and R. Gale. 2006. “Linear barriers of Northern Iran: The Great Wall of Gorgan and the Wall of Tammishe.” Iran 44: 121–73.
Pitts, L. F., and J. K. St Joseph. 1985. Inchtuthil: The Roman legionary fortress excavations 1952–65. London.
Pleiner, R. 1993. The Celtic sword. Oxford.
Pollard, N. 2000. Soldiers, cities, and civilians in Roman Syria. Ann Arbor.
Quesada Sanz, F. 1997. “Gladius hispaniensis: an archaeological view from Iberia.” JRMES 8: 251–70.
Reichmann, C. 1999. “Archäologische Spuren der sogenannten Bataverschlacht vom November 69 n. Chr. und von Kämpfen des 3. Jahrhunderts n. Chr. im Umfeld des Kastells Gelduba (Krefeld Gellep),” in W. Schlüter and R. Wiegels (eds.), Rom, Germanien und die Ausgrabungen von Kalkriese. Osnabrück, 97–115.
Rekavandi, H. O., E. Sauer, T. Wilkinson, E. S. Tamak, R. Ainslie, M. Mahmoudi, S. Griffiths, M. Ershadi, J. J. Van Rensberg, M. Fattahi, J. Ratcliffe, J. Nokandeh, A. Nazifi, R. Thomas, R. Gale, and B. Hoffmann. 2007. “An imperial frontier of the Sasanian empire: Further fieldwork on the Great Wall of Gorgan.” Iran 45: 95–132.
Richmond, I., and J. McIntyre. 1939. “The Agricolan fort at Fendoch.” Proceedings of the Society of Antiquaries of Scotland 73: 110–54.
———. 1962. “The Roman Siege Works of Masada.” JRS 52: 142–55.
Rodger, N. A. M. 2004. The command of the ocean: A naval history of Britain, 1649–1815. London.
Rostovtzeff, M. I. (ed.). 1934. The excavations at Dura-Europos: Preliminary report on the fifth season, 1931–1932. New Haven.
Sabin, P. 2000. “The face of Roman battle.” JRS 90: 1–17.
Salway, P. 1989. Roman Britain. Oxford.
Schiffer, M. B. 1999. The material life of human beings. London.
Schlüter, W. 1999. “The Battle of the Teutoburg Forest: Archaeological research in Kalkriese near Osnabrück,” in J. D. Creighton and R. J. A. Wilson (eds.), Roman Germany: Studies in cultural interaction. Portsmouth, 125–59.
Schlüter, W. and R. Wiegels (eds.). 1999. Rom, Germanien und die Ausgrabungen von Kalkriese. Osnabrück.
Stallibrass, S., and R. Thomas (eds.). 2008. Feeding the Roman army: The archaeology of production and supply in NW Europe. Oxford.
Syon, D. 2002. “Gamla: City of refuge,” in A. M. Berlin and J. A. Overman (eds.), The first Jewish revolt: Archaeology, history and ideology. London, 134–53.
Terrenato, N. 1998. “The Romanization of Italy: Global acculturation or cultural bricolage?” in C. Forcey, J. Hawthorne, and R. Witcher (eds.), TRAC 97: Proceedings of the Seventh Annual Theoretical Roman Archaeology Conference, Nottingham 1997. Oxford, 20–27.
Veen, M. van der, and H. Tabinor. 2007. “Food fodder and fuel at Mons Porphyrites: the botanical evidence,” in D. Peacock and V. Maxfield (eds.), The Roman imperial quarries: Survey and excavation at Mons Porphyrites, 1994–1998. London, 84–142.
Vitali, D. 1991. “The Celts in Italy,” in S. Moscati, H. F. Frey, V. Kruta, B. Raftery, and M. Szabo (eds.), The Celts. London, 220–35.
Welles, C. B., R. O. Fink and J. F. Gilliam (eds.). 1959. Excavations at Dura-Europos: Final Report. Vol. 5, Part 1: The parchments and papyri. New Haven.
Wells, P. 2003. The battle that stopped Rome. New York.
Wiegels, R., G. Moosbauer, P. Lönne, T. Schwarz, M. Geschwinde and M. Brangs. 2011. “Eine römische Dolabra mit Inschrift aus dem Umfeld des Schlachtfeldes am Harzhorn (Lkr. Northeim) in Niedersachsen.” Archäologisches Korrespondenzblatt 41, 4: 561–70.
Williams, J. H. C. 2001. Beyond the Rubicon: Romans and Gauls in republican Italy. Oxford.
Wilson, R. J. A. 1980. Roman forts. London.
Yadin, Y. 1966. Masada. London.