Relief scene of Roman legionaries marching, from the Column of Marcus Aurelius, Rome, Italy, 2nd century CE / Photo by Barosaurus Lentus, Wikimedia Commons
Looking at how war-making profoundly changed Greek and Roman society.
By Dr. Harry Sidebottom
It is well known that the way in which a society makes war is a projection of that society itself. To give an example. The early 19th-century rise of Zulu desire for decisive battle, fought hand to hand, and resulting in the slaughter of enemy combatants and the incorporation of everyone else into the Zulu state was caused in large part by the rise of autocratic Zulu kingship, which sought to focus all political loyalty on the person of the king via the army.
This looks at things the other way round. It takes three examples that, it has been argued, illustrate that war-making profoundly changed Greek and Roman society.
The ‘Hoplite Revolution’ in Ancient Greece
Away of approaching the military aspect of the supposed ‘hoplite revolution’ in the Greek world ofthe late 8th and/or 7th centuries bc is to look at two ancient visual images of fighting. The first is a late Geometric, c. 735-720 bc, oinochoe (wine jug) from Athens (Figure 5).
The style of fighting appears fluid and individualistic. Some warriors are mounted on chariots, others are on foot. One of a pair of warriors, often identified as the mythic Siamese twin sons of Zeus.
Figure 5. Oinochoe (wine jug) from Athens, c. 735-720 BCE
The ‘twins’ have a square shield, others an oval with semi-circular cutouts at the sides (often referred to as a ‘Dipylon’ shield after the cemetery in Athens where many pots depicting such shields were found), while others have no shield. Fighting seems to proceed via throwing spears and close-quarter work with swords.
The fighting shown on this pot can be interpreted as matching closely the fighting depicted in Homer’s Iliad, in which it can be thought individual heroes, riding in chariots, and dismounting to fight, first by thrown spears, then with swords, dominate the battlefield, while the mass of their followers usually are reduced to spectators or victims.
The second image is on the famous ‘Chigi vase’ (named after a former owner). The top figure scene on this Protocorinthian, c. 650 bc, olpe (jug) shows men in battle (Figure 6).
I lore the style of fighting appears cohesive and communal. The warriors are arranged in serried ranks, and move in step. All are armed alike, with a round shield, and are about to fight with a thrusting spear.
The fighting shown on this pot can be interpreted as being the same as, or very close to, what became the distinctive Greek style of fighting later in the 5th and 4th centuries bc; the hoplite phalanx, which we know from the historians Herodotus, Thucydides, and Xenophon. The hoplite phalanx consisted of heavy infantry, organized in a close-packed ‘shieldwall’ several ranks deep, all carrying round shields, and fighting primarily with a thrusting spear.
Figure 6. Chigi vase from Corinth, c. 650 BCE
It is interpretations like those above that underpin the modern theory that the Greeks significantly changed their style of war-making some time before c. 650 bc, when the ‘Chigi vase’ was painted. The original battlefield is seen as a ‘primitive’ one, akin to those recorded by anthropologists in highland New Guinea. It is a “place of fear’ with little in the way of formations or tactics. In it individuals motivated by their personal desires to gain honour and avoid shame have a wide latitude in deciding when, where, and who to fight. As such, it is unsurprising that a small number of aristocrats decide the day. They have more to gain and lose in the honour/shame game, and have the best equipment (though few scholars would argue that they used chariots in reality). This is seen as being replaced by a ‘civilized’ battlefield, the key elements of which are two formations, the opposed phalanxes. Once the virtually identically equipped warriors are assigned their places in the phalanx (and it must be confessed that no one has any real idea how that happened), they have next to no choice in when, where, or who to fight. A desire to defend their community, their polls, now is added to personal interests as motivation to prevent them exercising one of the few choices left to them – to stop fighting, try to force their way through the ranks behind them, and to run.
This supposed military revolution has been seen as the cause of deep political and social changes. An outline of an argument by a scholar who made an important contribution provides a route into this area. Antony Andrewes argued that an increase in trade led to the creation of a new group of relatively wealthy non-aristocrats in Greek society. It was these people (principally farmers who benefited from better economic conditions) who constructed the hoplite phalanx. Once in the communal organization of the phalanx, now the main military weapon of the polis, this ‘middle class’ came to demand political rights, and thus supported, or at least did not oppose, the rise of tyrants, who overthrew the previously existing aristocratic regimes. After the fall of the tyrannies, it was the hoplites who dominated most Greek states. Although differing on many issues, such as how the phalanx was introduced (at a stroke, or gradually over time), who introduced it (rich non-aristocrats, aristocrats, or tyrants), and, above all, what the effects of its introduction were, several scholars agreed that there had been (l) a significant military reform, and (2) that this had some political and social repercussions. The ‘hoplite revolution’ had made it to ‘orthodoxy’.
It could not last. Joachim Latacz was an early opponent of the ‘orthodoxy’. In his reading, the warfare described in the Iliad was hoplite warfare. Few scholars have followed such an extreme line. But several have taken a ‘revisionist’ position. They tend to see ‘proto-hoplites’ on the Homeric battlefield. In their readings of the Iliad the impact of individual heroes is minimal, and the decisive factor is massed fighting by ordinary warriors. There is thus no room for a military ‘revolution’. Instead, at most, there was a gradual adoption of new items of equipment, and a slow trend towards uniformity. With the forefathers of the hoplites already playing the vital role in battle, the final ‘formalization’ of the hoplite phalanx could not have inspired important political and social changes. Needless to say, not all scholars have accepted these arguments, and the ‘orthodoxy’ has been re-argued.
How can such opposed interpretations exist? In part it is down to our evidence. Our knowledge of Greek warfare in the 8th and 7th centuries bc is poor. Despite the enthusiastic endeavours of ‘experimental archaeologists’, who create and use replicas, finds of weaponry tell us less than we might expect. We can never be certain that an item of kit was used in the most ‘sensible’ or ‘rational’ way.
Let us turn to supposed ‘pre-hoplite’ warfare, and look again at the Athenian oinochoe illustrated (Figure 5). Was the artist trying to give a realistic picture of contemporary warfare, and would contemporary viewers have tried to see it in those terms? What about the chariots, for whose use in war at this time there is no archaeological evidence outside art, or the ‘Dipylon’ shields which are only found in art, or the possible presence of Siamese twins on a battlefield? Much the same questions can be asked of Homer. While clever readers of the Iliad can make the warfare in it appear coherent, does that mean we have anything more than a coherent poetic or fictional world? We know that items of equipment in the Mad come from widely separated periods of Greek history; Mycenaean ‘tower shields’ jostle with what sound like contemporary hoplite shields. Might not the tactics also be an amalgam of different periods?
In a similar way, our grip on early hoplite warfare is slight. Poets of the 7th century do not always seem to describe hoplite fighting. Callinus of Ephesus talks of fighting with javelins, not thrusting spears (fragment 1). Tyrtaeus of Sparta in one fragment (11) gives what we think of as a classic description of a close-packed hoplite phalanx. Yet the fragment ends (line 35) with gumnetes, the ‘naked’ (lightly armed), crouching under the shields of the hoplites.
Look again at the scene from the Chigi vase, and imagine what would happen to the action in a moment’s time. The front ranks are poised for the killing blow. All four on the left will fall, as will four on the right. This will leave one warrior on the right, and the flautist on the left isolated between the second ranks. Or will it? If you look closely at the painting, you can see four warriors in the front rank on the left, but ten legs. Most warriors carry a second spear, which later hoplites did not, and many of these spears have a loop to aid throwing, whereas later hoplites thrust their spears. The ranks are not packed close behind one another, as most believe later hoplites were. If we did not know about hoplite battle in the 5th and 4th centuries bc, would we automatically interpret this scene as representing a clash of phalanxes several ranks deep?
There is a serious danger of taking what we know of later hoplite fighting, altering it, and retrojecting it into the past. We take the densely ordered phalanx of the 5th and 4th centuries bc, strip it of its supporting light troops and cavalry, as well as its relatively sophisticated tactics, project it onto the 8th and 7th centuries bc, and thus create a simple and ‘ritualistic’ phase of early hoplite war.
Given the challenges of interpreting the meagre evidence that we have, it is completely unsurprising that widely differing and opposed theories can be held about the nature of war in the Archaic Greek world, and its impact on society. The Greek hoplite phalanx was a phenomenon of the Greek polis. Every polis we know about ended up using them, and Greeks who did not live in a polis did not. Similarly, tyrants seem to have been a phenomenon confined to the polis. We do not hear of tyrants among Greeks who did not live in a polis. The connections between polis, tyrants, and hoplites remain agreeably open to reinterpretation.
What can be stated with confidence is that between c. 735 and c. 650 bc the Greeks changed how they thought about war. In this period the practice of burying men with weapons ended, except in remote areas. Again in this period the Greeks start to dedicate in sanctuaries both arms and armour, and miniature images of equipment and warriors, the most striking example of the latter being the tens of thousands of miniature models of warriors that have been found in the temple of Artemis Orthia in Sparta. Finally, around 650 bc there is an explosion of images on pots of men in hoplite equipment.
The ‘Agrarian Crisis’ in Italy
It is commonly held that large-scale warfare in the last two centuries bc caused an agrarian crisis in Italy, which in turn largely undermined Rome’s Republican government, and led to its replacement by the monarchic system we know as the principate. A neat way into this ‘traditional view’ is offered by the flow-chart created by ancient historian Keith Hopkins (Figure 7).
Figure 7. Model of the ‘agrarian crisis’ in Roman Italy
Let us follow the main lines of the model. The key ‘motor’ of the process is the top left box, continuous wars of imperial conquest. This leads to the plundering of conquered territories, the proceeds of which are imported into Italy as booty, taxes, and slaves. These (moving to the box below) fund the creation of large landed estates, which are worked primarily by slaves. The setting up of these (moving to the box on the right) entails the displacement of free yeomen (perhaps ‘peasants’ has less anachronistic overtones). The dispossessed peasants drift off either (following the two downward arrows) to the growing towns of Italy, above all to Rome, where they help form a market for the produce of the very estates that have replaced them, or to the army where they contribute to the wars of expansion which started the whole process. The peasants of Italy in effect were fighting for their own displacement. The Vicious circle’ finally came to an end in the reign of the first emperor Augustus (31 bc-ad 14), with the end of continuous expansionist war-making, the establishment of a professional army, and the implementation of a massive programme of settling Italian veterans in overseas provinces (top right box).
Elegant and evocative as it is, the flow-chart is a severely restricted synopsis of Hopkins’ arguments. It was (as is implicit in the flow-chart) the Roman elite who got the lion’s share of what was extracted from conquered territories. The highest social order in Rome, the senators, who were politically active landowners, got their cut, in the main booty and graft, via high military commands and provincial governorships. Some members of the second highest order, the equestrians, who were also landowners but less overtly involved in politics, acquired their share by tax collection and business activities. The elite did not just purchase land in Italy to create their estates. Peasant families were ejected from their smallholdings by force. The elite also enclosed public land (ager publicus). This indirectly removed peasants from the countryside. Peasant holdings tended to be so small that access to public land was necessary for their sustainability.
Classical beliefs that citizenship was legitimated by military service, that those who had a stake in the community were more likely to be loyal to it, and that farmers made the best soldiers explain why, probably until 107 bc, there was a property qualification to serve in the Roman legions, and that the vast majority called up were peasant farmers. In the 2nd century bc Rome’s big overseas wars meant that large numbers of Italian peasants were away from their farms for long periods of time. When they returned they found that their families had been forcibly thrown off their land, or that they had entered into debts which could not be serviced, or that they were denied access to necessary public land by enclosure. Many thousands would have become casualties and either not returned at all, or been injured to such a degree that they could no longer work as farmers.
The mass eviction of peasants is thought to have profoundly undermined the Roman Republic in two ways. First, many moved to Rome, where they formed part of the ever-growing disaffected urban poor. The very real discontents of the urban plebs (plebs urbana) allowed some senators, often themselves from the most aristocratic of families, to set themselves up as popular politicians (populates), who by championing the interests of the poor gained positions of great influence, and in so doing fatally fractured consensus politics among the elite. The first open break came in 133 bc when Tiberius Gracchus, as one of the Tribunes of the Plebs, forced through a scheme to redistribute ager publicus to landless Roman citizens. Our sources tell us that a major motive for his action was concern at the dwindling number and reluctance to serve of those available for recruitment into the army. Second, the pressure of military service on the declining number of landed peasants led to the abandonment of the property qualification. This is usually dated to 107 bc, and linked to the actions of Marius, one of the consuls of that year. Subsequently unpropertied legionaries obviously had no farms to return to after their service, so they began to look to their generals to force the government to find land for them on discharge. At the same time, the generals began to look to their troops to support them in politics. This reciprocity of interests between legionaries and their generals would eventually bring down the Republic, as armies were prepared to follow their commanders against the state.
This ‘traditional’ understanding of the ‘agrarian crisis’ recently has been repeatedly challenged by some scholars. We will look at two of the main lines of these ‘revisionist’ attacks here. First, it has been suggested that warfare in the 2nd century bc was not all that different from the 3rd, and that if it had been going to cause an agrarian crisis it would have done so earlier. Second, it has been argued that our archaeological evidence does not support the ‘traditional’ view.
Any arguments based on the demography of Republican Rome must be recognized as tentative. Our evidence is patchy and hard to interpret. Totally opposed conclusions can be reached: the citizen population of Italy was either in sharp decline or was rising fast. In some senses warfare in the 2nd century imposed less strain than it had in the 3rd. In the 2nd century the Romans never had to field at one time the number of troops that they had for the war against Hannibal (the Second Punic War, 218-201 bc), and wars were now not fought on Italian soil. But other factors had changed. Rome now had to keep permanent garrisons in some provinces: in Spain (and possibly Cisalpine Gaul, as northern Italy around the Po valley was called) throughout the century, and in Macedonia from 146 bc. Legions in theory were disbanded every winter and new ones enrolled the following year. It was impractical to release all the soldiers serving in Spain and take a new draft out from Italy annually. Legionaries were therefore enrolled not for a set period of time, but for the duration of a campaign. If called up for service in a ‘garrison’ army, that ‘campaign’ could seem never-ending. Legionaries in such armies could be away from their farms for years at a time. The actions of Tiberius Gracchus in 133 bc were an attempt to solve a problem that had been long in the making. In the 140s bc there had been such a strong perception of a crisis that even a politician of a staunchly conservative frame of mind had proposed reform. Although, as a reactionary hero, Laelius had won the nickname Sapiens (the ‘wise’) for then dropping the whole idea.
It has been argued that the archaeological evidence that we have does not support the ‘traditional’ picture of an agrarian crisis in Italy. For example, one of our main literary sources (Plutarch, Ti. Gracch. 8.7) says that Tiberius Gracchus was inspired to attempt reform on a journey through Etruria in Italy, when he saw the scarcity of free men, and that agricultural work was conducted by slaves. Yet archaeological survey of this region seems to show a landscape full of small farms rather than large estates. As he was a politician, it is easily accepted that Tiberius Gracchus exaggerated and generalized. But we should not assume that archaeologists automatically are immune from such behaviour. The vast majority of Italy has not been surveyed. The limitations of archaeology should be kept in mind. Often it is good at telling us how land was used (for example, to grow olives or grapes), but seldom can it tell us of the status of those cultivating the land (owner-occupiers, tenants, or slaves). The very poor, such as agricultural slaves, leave few traces, and thus are under-represented in archaeological surveys.
We must be careful not to create too rigid a dichotomy between the ‘traditional’ view and the ‘revisionist’ ones. No ‘revisionist’ would claim that no peasants were thrown off the land. Equally, no ‘traditionalist’ holds that all peasants were evicted. It is a debate about degree, not of kind. Various factors ensured that large estates could not take over all Italian land. Climate, topography, and soil were suitable for large estates only in certain areas; vines and olives flourished in western coastal Italy, pasture in the south. Unlike subsistence peasant farms (‘sufficers’), estates run as profit-making businesses (‘maximizers’) need access to markets. Bulk transport by land was economically inefficient in the ancient world. Large estates thus were limited to areas of proximity to their markets (towns), or water transport (navigable rivers or harbours). The need for labour on estates varied during the year: large numbers of workers were needed in harvest time, far fewer in mid-winter. It would be uneconomic for an estate to keep enough slaves to cover the peak times. Instead, they kept the minimum number of slaves, and hired in extra labour for busy times. It was thus in the interest of the large estates to ensure that some peasants remained on the land in their vicinity to be employed as occasional wage labourers. Large estates never became dominant in agriculture in Italy (in the sense of farming the majority of the land), but they did become the distinctive type of farm. The replacement of peasant farms by largely slave-worked estates was enough of a perceived problem for politicians of such different views as Tiberius Gracchus and Lealius to consider that a remedy was necessary.
The ‘Barbarization’ of the Roman Army
There is a popular view that the ‘barbarization’ of the Roman army led to the fall of the western half of the empire. It runs something like this. During the 4th century ad the large-scale replacement of indigenous units with barbarian units, commanded by their own tribal leaders and fighting in their native styles, and the influx of barbarian officers and men into regular Roman units, which adopted barbarian equipment and fighting techniques, combined to make the empire’s army less efficient, and more prone to desertion and treachery. In the 5th century ad ‘barbarization’ was curbed in the eastern half of the empire, but increased in the west. The east thus survived, while the west fell.
The concept of ‘barbarization’ recently has been the subject of repeated revisionist attacks. It has been suggested that only one in four of those officers and men in regular units about whose origins we can make a guess were barbarians, and that this number did not increase during the 4th and early 5th centuries. Also it has been argued that when whole regular units were raised from one barbarian tribe, these did not keep any corporate barbarian identity for long as replacement troops were not drawn from the original ethnic group. Similarly, it has been suggested that service in the regular army would lead to an individual recruit replacing his identity as a barbarian (Frank, Goth, or whatever) with that of a Roman soldier. On our available literary evidence it appears that barbarians in the army were no more liable to treachery or desertion than indigenous troops. It is pointed out that the Roman army had always adopted equipment and practices from its opponents, and that it is hard to see how some of those adoptions that we know from the late empire (such as wearing trousers or using the Barritus, a Germanic war cry) could have impaired efficiency. A contemporary commentator could claim that some borrowings from the barbarians had actually improved the performance of Roman troops. Vegetius (1.20) claims that copying from barbarians has been beneficial to the Roman cavalry.
The revisionist arguments, of course, are not above criticism. We have indications of ethnic origins for only a small percentage of officers, and a miniscule percentage of the rank and file. We are very seldom told the origins of an individual, and thus usually have to draw an inference from his name. As the revisionists admit, this is a very uncertain method. We know that some barbarians adopted Roman names, while some recruits from within the empire carried local names (Celtic, Thracian, or whatever), which can easily be taken to be barbarian. The model whereby a recruit neatly replaces his former identity with that of a Roman soldier can be doubted. A veteran described himself as ‘Francus civis, Romanus miles’, a Frank (‘citizen’) and a Roman soldier (ILS 2814). There was a contemporary perception that the army had been taken over by barbarians. In the 4th century ‘Goth’ was a colloquialism for ‘soldier’ in the Syriac language spoken in some eastern parts of the empire. Some contemporaries in certain circumstances could condemn the ‘barbarization’ of the army. Famously, Bishop Synesius of Cyrene in the early 5th century savagely criticized the emperor Arcadius for employing large numbers of barbarians in his forces (On Kingship, esp. 1091). The revisionist position assumes that barbarian ‘allies’ serving under their own officers and in their own style had no ‘knock-on’ effect to the regular army. It could be imagined that the effect of serving alongside units with looser discipline and a less committed attitude to training might have been deleterious to regular units. Finally, it must be remembered that the revisionist arguments are of relevance to a limited time; only up to either the great Roman defeat by the Goths at Adrianople in ad 378 or to c. 425. In the west by the mid-5th century, Roman field armies were thoroughly “barbarized’. In ad 451 the ‘Roman’ army that defeated Attila at the battle of Chalons was, depending on how one interprets our sources, either composed totally of barbarians, or the effective part was barbarian.
The revisionist arguments have shown that the “barbarization’ of the army will not do as a monocausal explanation of why the western Roman empire fell in the 5th century ad and the east survived. Other explanations must be explored: these include the west suffering both greater barbarian pressure and more usurpers of the imperial throne; its longer frontiers; lack of a virtually impregnable strategic capital like Constantinople; poorer tax base; failure to curb large landholding aristocratic families and to create a bureaucratic ‘service aristocracy’ akin to that in the east; and its failure to integrate army commanders into the imperial court. Yet some will still conclude that the ‘barbarization’ of the army had a role to play in the fall of the west.
Why do historical interpretations change?
The way in which these three sections are structured, a ‘traditional’ interpretation, followed by a ‘revisionist’ one, then a critique of aspects of the latter, illustrates that all historical interpretations are provisional and part of an ongoing process. There are many reasons why that should be so, three of which will be picked out here.
The discovery of new material is an obvious, although much less common than might be imagined, stimulation to new interpretations. Archaeological discoveries in Italy were a factor in reassessing the ‘agrarian crisis’ of the late Republic.
General shifts in thinking, or intellectual fashion, influenced by changes in current political or social circumstances, often provoke new interpretations of well-known bodies of evidence. The “barbarization’ of the late Roman army was first seriously studied by German scholars in the last quarter of the 19th and first half of the 20th century. Given the search for a corporate identity for the new German empire of the time, it is no surprise that these scholars tended to maximize the Germanic nature of Roman antiquity, and were predisposed to find a very ‘Germanized’ Roman army. As another example, it can be speculated that the turning against the traditional view of the ‘hoplite revolution’ was part of a general revulsion in the 1970s and 1980s against social-determinist, and above all Marxist, theories of historical change.
The final explanation offered here is rather more cynical: every generation rewrites history because it wants to get published and wants a job. Historians are trained to criticize the interpretations of others, and they do not, or should not, make a career rehashing the views of others. This, as one ancient historian put it, ‘is normal, cyclical, endogenous change, as a new generation of historians inevitably seeks to make “progress” in understanding and explaining the past by rejecting the dominant paradigms of its mentors’.