In 30 AD or thereabouts, a Roman merchant called Gargilius Secundus purchased a cow from a man called Stelus, a non-Roman who lived near the modern Dutch town of Franeker across the River Rhine. A record of this transaction, which cost 115 silvernummiand was witnessed by two Roman centurions, just happens to survive. One modern commentator has called it ‘banal’, and so it was: small-scale and entirely unremarkable. If it happened once on Rome’s European frontier, it happened a thousand times. The reason for thinking so is straightforward. Especially in the early period, but also later on, large numbers of Roman soldiers were stationed right on the imperial frontier. They represented a huge source of economic demand. In the first century AD some 22,000 Roman soldiers, a mixture of legionaries and auxiliaries, were established on the territory of only 14,000 or so indigenous Cananifates in the northern Rhine region alone. The latter could not possibly supply the soldiers’ demands for foodstuffs, forage, and natural materials such as wood for construction and cooking, or leather. A legion of 5,000 men required approximately 7,500 kilos of grain and 450 kilos of fodder per day, or 225 and 13.5 tonnes, respectively, per month. Some of the soldiers’ needs were supplied directly from the imperial centre, but this was cumbersome and logistically problematic. Where they could, the imperial authorities preferred to pay cash and let local suppliers meet the troops’ demands.47
Trade and Control
Throughout the Roman period, therefore, the frontier zone of the Empire had a huge requirement for primary agricultural products of all kinds and there is every reason to suppose that non-Roman suppliers played a major role in meeting it. This was still the case in the fourth century, where the pages Ammianus devotes to the Alamanni again make interesting reading. After his victory at Strasbourg, the Emperor Julian was in a position to impose virtually whatever terms he wanted on the defeated Alamannic kings. All the treaties differed in detail, but they had in common demands for foodstuffs, for raw materials such as wood for construction purposes, for wagons and for physical labour to carry out rebuilding projects. On the back of his victory, Julian could simply requisition these items, but even in less favourable circumstances they were still required by the Roman army, and presumably had to be paid for. Whether paying or not, the Roman army was a constant source of economic demand for any neighbouring Germani.
None of the items mentioned in Julian’s treaties is archaeologically visible. You can’t identify – because they couldn’t survive – traces of Germanic-grown wheat, Germanic-felled timber, leather cured by the Germani, or items constructed by Germanic labour. They were all, however, real enough, and show up in more indirect fashion in the huge expansion of agricultural production that we have observed in Germanic Europe in the Roman period. Some of this extra food was consumed by the new kings and their retinues, and some by Germania’s own expanding population, but a further – perhaps even the original – stimulus to production was provided by the Roman army. For one thing, there is close chronological coincidence between the arrival of Roman demand on the fringes of Germania and the rural intensification. The earliest of the new villages, such as Feddersen Wierde and Wijster, also grew up in regions from which it was relatively easy to ship agricultural products by water to the mouth of the Rhine and then upstream to the river’s military installations. As much recent literature has rightly emphasized, and as has been shown to be the case along all of Rome’s borders, the frontier acted in some ways more as a zone of contact than, as you might initially expect, a line of demarcation dividing the Empire from its immediate neighbours.48
In the case of the Germani, Rome may have acted as a source not only of extra economic demand, but also possibly for some of the ideas and technology that made agricultural intensification possible. At Wijster and Feddersen Wierde, higher yields seem to have resulted from a more systematic integration of arable and pastoral agriculture, using animal manure to sustain the fertility of the wheat fields. More generally, it involved the adoption of more sophisticated ploughing techniques and equipment. Where and how, exactly, these ideas spread remains to be studied, but both the more efficient ploughs and the better-integrated farming regimes were well known in Roman and La Tène Europe, much of which the Empire swallowed up in the first century BC (Chapter 1), long before they spread into Germania, and these areas may have inspired the Germanic agricultural revolution.
Other goods produced in Germania were also in demand in the Roman world. The occasional loan word and literary reference identify some specific products. Goose feathers for stuffing pillows and particular kinds of red hair dye were two such items. Much more important than any of these, though, was the demand certainly for two, and probably three, other raw materials. The one that is not so certain is iron. There is no specific evidence that pig iron was shipped in large quantities south and west across the frontier from Germanic Europe. But the vast quantities of iron produced at the two main Polish sites far outstripped any amount that can have been required for local use. Possibly, this iron was being circulated within just the Germanic world, but it is entirely conceivable that it was also being processed to satisfy Roman demand. Of the other two materials, there is no doubt. The first is amber: solidified sap from submerged trees washed up on the Baltic Sea coast. Amber is one of few loan words taken over from Germanic languages into Latin, and we know that the Romans took a huge interest in this product for making jewellery. In the time of Nero, a senatorial mission even went north to investigate its origins, and the amber route from the Baltic – in both its main branches, one striking south to the middle Danube at the legionary fortress of Carnuntum, the other going east of the Carpathians to arrive at the Black Sea ports (Map 2) – was well known to Roman authors.49
At least as important, though less discussed in our sources, was the demand for Germanic manpower. This took two main forms. First, recruits were always needed for the Roman army. The so-called barbarization of the Roman army used to be one of the main reasons given for the decline of the Empire. The point is at best partly mistaken. From the time of Augustus, at least half of the total army – all its auxiliary formations – was always composed of non-Romans, a substantial number of whom were recruited from the Germanic world. All that happened in the late Empire was a recategorization of military units, which saw the distinction between citizen legionaries and non-citizen auxiliaries partly collapse. Nothing suggests, either, that there were more Germani in percentage terms serving in the Roman army in the fourth century than before, or that the army was any less reliable for their presence – it is normally considered likely, anyway, that legionary recruiters had, in practice, been ignoring for some time the requirement that only citizens should be drafted. Throughout the Roman period, therefore, there was a huge demand for Germanic recruits, and many turn up in the epigraphic record. From narrative sources we know that these men were recruited in two ways. Some were individual volunteers, deciding to follow a potentially lucrative career path in the Roman army. Many others, however, had no choice. Again Ammianus is explicit. A forced draft of recruits was part of most of the peace treaties he records between the Empire and different barbarian groups. Not only did you have to supply labour and foodstuffs to buy your way back into the Empire’s good graces after a defeat, but you also had to give over a portion of your young men for service in the Roman army.50
Manpower from Germania also entered the Empire in another form: slaves. We have no detailed account of the operation of the slave trade in the Roman era, such as we get from Arab authors for its counterpart of the ninth and tenth centuries (Chapter 10). So there is no information on the identity of the main traders, on the areas from which they tended to take victims, and on whether, as later, there were any major slave markets inside Germania, where slaves could be traded on to middle-men or directly to Roman merchants. But the slave trade was a constant phenomenon of the Roman era, and there is one powerful testimony to its importance. Germanic languages have as one of their basic word-stems for trade and merchants a series of terms deriving from the Latinmango. But in Latin mango meant not a merchant in general, but very precisely a slave trader. The Roman merchants first and perhaps most often encountered by Europe’s Germanic-speaking populations, therefore, were probably traders in human flesh.51
Overall, an excellent case can be made that the new opportunities for trading with the much wealthier Roman Empire, which suddenly opened up around the birth of Christ with the expansion of Rome’s European frontiers northwards, played a major role in stimulating the evident economic development of Germania in the early centuries AD. According to Caesar in the mid-first century BC, the Germani of his day had little interest in trading with Roman merchants, and only allowed them into their territories at all in the hope that they could sell them captured war booty. If that had really been the case in the middle of the first century BC, the situation evolved rapidly. By the end of the first century AD, trade was so common across the Rhine frontier that Roman silver denariiwere being used as a medium of exchange by the Germanic tribes on the east side of the river. It is likely enough, indeed, that much of the silver found in Germania in Roman times – in the form, for instance, of intricate fibulae – represents the reworking of metal from such coins, many of which remained in circulation right down to the fourth century. And while (for reasons we will return to in a moment) it is not the case that every frontier grouping was trading so heavily with the Empire as to be using Roman coins, this certainly happened periodically, throughout the Empire’s existence. As a phenomenon, it shows up in the presence of relatively dense concentrations of low-value Roman coins from particular periods in areas fairly close to the frontier, such as those of the fourth century found along some of the old Roman roads east of the Rhine which still existed in the Agri Decumates – a triangle of territory between the Upper Rhine and Upper Danube – then under Alamannic control; or, further east along the Danube, within regions bordering the Roman province of Moesia Superior.52
Equally striking is the fact that throughout the Roman period the Empire’s immediate neighbours were interested in obtaining trading privileges with the imperial merchants, privileges which Rome usually kept under tight control. Even when the fourth-century Gothic Tervingi wanted to sever most of their ties with the Empire, it was part of the resulting agreement that two designated trade centres continued to operate. A huge amount of archaeological evidence confirms the impression given by the literary sources. Roman goods of all kinds have been found in large quantities in most of the major excavations conducted on Germanic sites from the first four centuries AD.
There are distinct chronological and geographical patterns to the finds. The first two centuries AD, for instance, saw a huge explosion in the quantity of Roman goods present within Germania in many areas of the immediate frontier zone, up to about a hundred kilometres from the defended line, both on settlement sites and deposited in graves. Fine pottery (terra sigillata), bronze ornaments and glass have all been unearthed in substantial amounts, alongside the Roman coins we have already mentioned. In the first- and second-century levels of the site of Westrich, for instance, which is far from untypical, Roman manufactures account for about a third each of the excavated pottery and metalwork. But while common in some places, this pattern does not apply to the regions of the northern Rhine frontier, between the Rhine and the Weser, where Roman materials of this date are much less plentiful. Moving beyond the immediate frontier zone to the area up to the River Elbe, the pattern is again slightly different. Here Roman goods are present in large quantities, but they tend to concentrate in particular areas. The region of the River Saale in modern Thuringia, for instance, has produced one striking concentration. Others have been identified around the tributaries of the Upper Elbe in Bohemia (heartland of the Czech Republic), and south of the Lower Elbe and the Middle and Lower Weser (both in Lower Saxony). The other identified concentration is along the North Sea coast. Moving still further away from the frontier, Roman goods are present only in smaller quantities, but there are still a few identifiable concentrations such as Jakuszowice in southern Poland, the Gudme/Lundeburg complex in Scandinavia, and in eastern Denmark.53 In general terms, there is more than enough material to show that the Germanic economy was mobilized in the early centuries AD in part to pay for large quantities of attractive Roman imports. But how are we to explain these concentrations?
Part of the answer lies in logistics. The fact that a wagon of wheat doubled in price for every fifty Roman miles travelled emphasizes how difficult and expensive land transport was in pre-modern times. Hence relatively low-value items – such as pottery, bronze and glass – were only ever likely to move comparatively short distances unless water transport or some other mitigating factor intervened. The fact that even spreads of Roman goods have been unearthed only within the immediate frontier zone, then, is not surprising. Transport may also explain some more particular phenomena. The possibility of water shipment probably allowed relatively distant places like Feddersen Wierde to be involved in supplying the Roman army of the frontier, and as the coin distributions suggest, the old Roman road networks of the Agri Decumates perhaps still facilitated trade in the fourth century, even after the area had fallen under Alamannic control. Logistics, however, will not explain everything.
A second line of explanation requires us to look more closely at the mechanics of trade in the Germanic world, and the role played in Germanic society by the Roman goods received in return. If Caesar is to be believed, there was originally some Germanic resistance to trade with the Empire. But this was quickly and entirely overcome to a point where possession of Roman goods came to be associated with high social status. Analyses of the types of goods found together in richer burials have demonstrated a powerful correlation from the late first century AD between the presence of large numbers of everyday items of local manufacture, clearly expensive items of local manufacture (such as weapons and jewellery), and Roman imports. Thus Roman imports quickly came to be part and parcel of demonstrating social pre-eminence. Again, this is not surprising. Roman imports were exotic and had to be paid for by giving something in return to a Roman merchant. They were bound, therefore, to possess a certain cachet. It is also a further dimension of the phenomenon we have already observed. As in modern globalization, the benefits of ancient Germanic economic development were not enjoyed evenly, but concentrated in the hands of kings and their retainers; so, as one might expect, more Roman imports ended up in their hands.
This point is worth dwelling on, because while it might again seem entirely natural from a modern perspective, it is also telling us something important about how the new exchange networks operated. As soon as you stop to think about it, such an outcome can only be reflecting the fact that kings and their entourage were organizing the profits accruing from economic development for their own benefit. On one level, possession of military muscle enabled kings to exact a percentage of the new agricultural surplus now being generated. They could then use this not only to feed their retinues but also to trade on to the Roman world, getting precious metal coins, or wine and olive oil, or whatever else they desired, in return.
But military muscle was also crucial to securing the lion’s share of the profits from some of the other new trade flows. Think about the slave trade. Slaves do not volunteer. Someone was rounding them up in Germanic society to sell them on to the Roman traders, and this will not have been a peaceful process. This line of thought also suggests, incidentally, a further possible context for the massacred retinue unearthed at Ejsbøl Mose. If they were a slave-trading outfit, you can quite see why such methodical fury was vented upon them. And even the amber trade was no gentle process of wandering along the Baltic shore picking up whatever had washed up overnight. One of the most startling finds to emerge from northern Poland in recent years has been a series of wooden causeways, many kilometres long, establishing a network of routes across boggy territory near the Baltic Sea. Carbon-14 and dendrochronology have established that these were laid down around the birth of Christ and then maintained for the best part of two hundred years. They have been interpreted, surely correctly, as servicing the northern end of the Amber Route. But all this took a huge effort. In other words, it must have been enormously worth someone’s while to go to this much trouble. In return for their effort, they were clearly receiving a substantial cut of the profits from the trade, presumably in the forms of tolls of one kind or another. Interestingly, ‘toll’ in Germanic languages is another loan word from Latin, suggesting that the concept did not exist among the Germani before the Empire became their immediate neighbour. And, of course, where taking this kind of percentage from a trade flow was so obviously profitable, others would have been interested in a share of the action. Here again, military strength counted. You could employ it to force those of lesser status to do the physical work of building and maintaining the causeways, and also to prevent any other armed group from taking over what was clearly such a nice little earner.54
Contrary to the bland neo-classical platitudes of 1980s-style trickle-down theories, economic development is not always or not straightforwardly, at least, a good thing. Increasing wealth in Germanic society during the Roman period set off major and in some cases seriously violent struggles for its disproportionate control. In some developing areas of the economy, the adverse effects were perhaps not so bad. It is notoriously hard to tax agricultural production, and higher outputs were anyway dependent upon having plenty of labour available, at least for arable agriculture, so that the demands of kings and their warbands, some of whom may anyway have been recruited from the wealthier farmers, perhaps did not impinge too heavily. Other aspects of economic development, however, were much nastier for those caught up on the wrong side: slaves obviously, but I wonder too about iron-mining since, in the Roman world at least, being condemned to the mines was a form of capital punishment. And even at the top end of society, the struggle to control the new wealth could have serious consequences. Ejsbøl Mose is one of over thirty weapons deposits known from the bogs of northern Europe, most of which were laid down between 200 and 400 AD – explicit testimony to the level of violence set loose in the Germanic world for control of all this burgeoning wealth. Moreover, there is no reason to suppose that these struggles were limited to just those areas that happened to have convenient bogs and lakes available for disposing of the defeated. Tacitus refers to a first-century votive ritual which involved hanging the dead and their weaponry from trees. Weapons deposits of this kind would not survive to be excavated by archaeologists, and I am inclined to think that accident of survival is the reason for the direct evidence of violent competition being confined to areas around the North Sea, rather than that the proximity of water made the Germani of this area particularly quarrelsome.55
It is not a new idea to discuss trade with the Roman Empire when trying to understand the transformation of Germanic society in the early centuries AD. But, as has reasonably been pointed out, trade on its own never looked like a powerful enough explanatory mechanism, since large quantities of Roman goods have not turned up everywhere. The case for the importance of such trade becomes much more convincing, however, when you factor in not just the new wealth flows themselves, but the consequent struggles for their control. It was this knock-on effect, rather than the mere existence of the new wealth, that had the really transformative effect. Various groups within the Germanic world responded dynamically to the fact that the new wealth existed by seizing control of its profits, and, in doing so, helped remake the sociopolitical structures of the world around them.
This extra dimension of argument belongs alongside those which in the area of post-colonial studies have attracted the general label ‘agency’. The point here is that earlier analyses (‘colonial’ rather than ‘post-colonial’ ones, as it were) tended to explore the effects that more developed societies have upon less developed ones in too passive a fashion. The fundamental point of ‘agency’ (although much ink has been spilled over more precise definitions) is to stress that indigenous groups respond to outside stimuli by taking hold of certain possibilities (and not others) for their own reasons and according to their own priorities. In this instance, we see exposure to the economic opportunities presented by contact with Rome taking a number of forms, and being seized on in different ways by different groups. Some learned to expand agricultural production, some exported iron or amber, and still others set up slave-trading operations. Not only did the consequent increase in inequality provide the economic basis for larger political confederations by the fourth century, but this is also reflected in the patchy distribution of Roman goods observable in the archaeological record. The particular concentrations of goods in the intermediate zone up to the Elbe were presumably created by Germanic groups able to dominate some specific new flow of wealth out of the Roman Empire, which they used to pay for the items found by archaeologists. The beneficiaries of the slave trade of the ninth and tenth centuries, for instance, are certainly visible archaeologically through the fruits of their trade, as well as being identified in the historical texts (not true of the Roman era), so it is not unreasonable to apply the same principle to the Germani around the Roman Empire.56 But even adding in a dynamic indigenous response to the existence of the new wealth flows doesn’t come close, in my view, to establishing the full extent of Rome’s role in the transformation of the Germanic world. For that, we also need to explore how the Empire set about maintaining long-term stability along its frontiers.
The Art of Client Management
In 1967 some gravel-digging in the River Rhine itself, close to the old Roman city of Civitas Nemetum (modern Speyer), led to the discovery of loot from a Roman villa. Careful excavation over the next sixteen years reconstructed the full story. The finds were there because late in the third century some Alamannic raiders had been trying to get their booty back home across the Rhine when their boats were ambushed and sunk by Roman river patrol ships. Called lusoriae, the latter were light, oar-driven warships equipped with rams and a well armed crew. An everyday kind of frontier story, except for what the raiders were trying to get home. They had with them an extraordinary seven hundred kilograms’ worth of booty packed into three or four carts which they were rafting across to the east bank of the Rhine. On close inspection, the loot proved to be the entire contents of probably a single Roman villa, and the raiders were interested in every piece of metalwork they could find. The only items missing from the hoard were rich solid silverware and high-value personal jewellery. Either the lord and lady of the house got away before the attack, or else the very high-value loot was transported separately. In the carts, however, was a vast mound of silver-plate from the dining room, the entire equipment from the kitchen (including 51 cauldrons, 25 bowls and basins and 20 iron ladles), enough agricultural implements – everything from pruning hooks to anvils – to run a substantial farm, some votive objects from the villa’s shrine, and 39 good-quality silver coins.57
The nature of this extraordinary hoard makes clear the depth of the problem facing the Empire in one dimension of frontier relations. We naturally think of barbarian raiders being interested in gold and silver, and plenty of rich plundered objects have turned up over the years from various hoards of the Roman era. But the total range of desirable goods was massively wider. Because the economy of the Germanic world was so much less developed than its Roman counterpart, all of these goods were directly useful to the raiders, or could be sold on to someone else, whether Alamannic farmer or housewife, or even to an Alamannic smith for reworking. This is just about the most vivid illustration of the kind of booty your average raider targeted ever to be unearthed, but historical sources make it clear that banditry, perhaps often on a smaller scale than this amazingly comprehensive house-clearing exercise, was endemic all along Rome’s frontiers.
The fact that the legions’ advance had halted at different moments in the first century, broadly along the line of the Rivers Rhine and Danube, did not mean, therefore, that lands beyond the frontier could be left to their own devices. On the contrary, there was a huge propensity for cross-border raiding, the natural result of two very different levels of economic development sitting side by side. Nor, as has sometimes been argued, did the Empire go suddenly from attack to defence. Frontier security demanded a much more proactive response, and throughout most of its history Rome maintained a general military superiority all along its European frontiers, backed up by aggressive diplomacy. These policies turned its closest neighbours effectively into client states.58
The methods used remained pretty constant throughout the life of the Empire, and had profound effects upon patterns of sociopolitical development within the Germanic world. For an excellent case study from the fourth century, we can turn to Ammianus’ account of the response of the Emperor Constantius II to trouble on the Middle Danube in the years 358/9. Constantius’ first step, like every emperor before him, was to establish military superiority. Starting just after the spring equinox, when the opposition thought they were still safe, he threw a pontoon bridge over the Danube and came upon the Sarmatians unexpectedly. The results were nasty:
The greater number, since fear clogged their steps, were cut down; if speed saved any from death, they hid in the obscure mountain gorges and saw their country perishing by the sword.
In the following weeks, the campaign was quickly extended to the neighbouring Quadi and all the other frontier groups of the region. The Emperor then used this military superiority to dictate what he hoped would be a lasting diplomatic settlement. One by one, the groups and their leaders came, or were forced to come, to hear the Emperor’s judgement.
Not all groups were treated in the same way. To some Constantius =showed favour. One prince of the Sarmatians, Zizais, had mastered the script:
On seeing the emperor, he threw aside his weapons and fell flat on his breast as if lying lifeless. And since the use of his voice failed him from fear at the very time when he should have made his plea, he excited all the greater compassion; but, after several attempts, interrupted by sobbing, he was able to set forth only a little of what he tried to ask.
Barbarians were expected to show subservience to the divinely ordained might of Rome, as Zizais was perfectly well aware, and as the iconography of barbarians on Roman coins and monuments emphasized. Barbarians were always presented lying down in submission at the bottom of any pictorial scene, often literally under an emperor’s feet (Plate 7). The Sarmatian’s approach may well have been calculated, therefore, and it produced the desired result. Constantius decided to restore the political independence of Zizais’ followers, who had been held as junior partners in an unequal coalition, and raised the prince himself to the status of independent king. Rearranging the political alliance systems currently in operation on this part of the frontier after the fashion that best suited Rome’s interests was, in fact, was one of Constantius’ chief preoccupations. This meant breaking up over-large and therefore – from a Roman perspective – potentially dangerous alliances. Where Zizais gained, others lost. Araharius, a king of the Quadi, was denuded, despite his protests, of the services of his Sarmatian underking Usafer, who, like Zizais, was restored to independence. Sometimes the interference could be much more violent. Another tactic, which occurs three times in the twenty-four years covered by Ammianus’ narrative, was to invite potentially problematic frontier dynasts to dinner and then either murder or kidnap them.59
Aside from political restructuring, various other measures were enacted: securing economic returns for the Empire on the military effort it had just expended, combined with strictures to enforce the new settlement once the legions had withdrawn. Some measures were standard, such as extracting drafts of young men from the groups submitting to him to serve as military recruits. This, as we have seen, was one of several ways in which young Germani had entered Roman armies throughout the Empire’s existence. Hostages were also extracted from each of group, usually young men of high status. They were not treated as prisoners, exactly, once on Roman soil, but were sometimes executed when agreements broke down. Any Roman captives were also returned to imperial soil. In other respects, the details of agreements differed. According to the amount of blame the emperor decided to allocate to any particular group for the original trouble, it might have to supply labour, raw materials and food; or it might, on the other hand, be granted privileged trading status. Diplomatic subsidies were, in addition, a standard feature of Rome’s diplomatic armoury. In the past, some historians have doubted this, supposing payments to barbarian leaders to be a sign of Rome’s military weakness in the late period. This is mistaken. We would call such subsidies ‘foreign aid’, and they were utilized throughout Rome’s history, even after major Roman victories. After he crushed the Alamanni at Strasbourg, for instance, Julian granted the defeated kings annual subsidies. The reason is simple. Subsidies helped keep in power the kings with whom Rome had just made its agreements. As such, they were an excellent investment.60
Apart from all this diplomatic detail, one further preoccupation emerges from Constantius’ intervention. The Empire did not want the immediate hinterland of its frontier to become too crowded, for two reasons. First, this would mean that there were too many groups with an opportunity to raid Roman territory. Second, as the establishment and reorganization of all the over- and underkingships shows, frontier groups were always in political competition with one another, and their jockeying for position stood more chance of spilling over into violence on Roman soil when there were more groups playing the game. In this instance, Constantius and his advisers eventually decided that a key part of the new settlement was to make one group of Sarmatians, the Limigantes (again, a coalition), move away from the immediate frontier zone. This was not something the Limigantes wished to do, so further military intimidation was required and duly delivered. After two of their subgroups, the Amicenses and Picenses, had been brutalized, the rest surrendered and agreed to depart. The region seemed set for peace – but not quite yet. A year later, in 359, some of the Limigantes returned, saying that they would prefer to move into the Empire itself, as tax-paying tributaries, rather than continue to occupy their assigned lands so far from the frontier.
What happened next is rather mysterious. Ammianus blames it all on the Limigantes’ bad faith, but then he would. An agreement in principle seems to have been reached. The Sarmatians were to be allowed across the river and to enter the imperial presence, Constantius having returned to the region with his army. Then, at the crucial moment, something went wrong. Instead of surrendering, the Sarmatians attacked the Emperor, or so Ammianus says, and the Romans responded:
So eagerly did our forces rush forth in their desire to . . . vent their wrath on the treacherous foe, that they butchered everything in their way, trampling under foot without mercy the living, as well as those dying or dead; and before their hands were sated with slaughter of the savages, the dead lay piled in heaps.
Perhaps the Limigantes did act in bad faith, or Constantius maybe wanted to put down a clear marker that his orders had to be obeyed – or, just as likely, the tragedy resulted from mistrust and confusion. But throughout its history, the Empire did on occasions use the acquisition of outside population groups as one technique for managing the frontier. While the consequent gain to the Empire in terms of taxpayers and potential soldiers was part of the calculation, so too was a concern to prevent potentially dangerous overcrowding.61
This portfolio of methods was applied very generally. Occasional major military interventions made it possible to construct region-wide diplomatic settlements, which broke up dangerous coalitions, identified and rewarded friends and punished enemies, while a mixture of stick and carrot – the fear engendered by punishing campaigns and hostage-taking combined with targeted foreign aid and trading privileges – was used to make sure that the new settlement held beyond the short term. The methods were effective, but not, of course, perfect. From a Roman perspective, their success can be measured in terms of the life expectancy of the settlements. By my reckoning, the average fourth-century diplomatic settlement on the Rhine and Danube frontiers lasted about twenty to twenty-five years – one generation, in other words – per major military intervention. This was probably a fair return on the amount of force expended, and about as much as could reasonably be expected. It is important to understand, however, that the whole system was sustained by occasional but decisive Roman campaigning. The frontier groups were part of a Roman world system, but terms and conditions were not arrived at by free, mutual agreement. Rome consistently used military force to maintain its preponderance.
The methods of Roman diplomacy are fascinating in themselves and have their own scholarly literature. They also advanced the transformation of Germanic society. To understand why this was so, we must again reckon with populations on the far side of the Roman frontier as active agents in the story. Roman diplomacy certainly had some important direct effects, but that is not the whole story. Groups and individuals within Germania responded in a variety of ways to the stimuli applied by the totality of Roman foreign policy over four centuries, and this response is just as important as the original imperial interference.
The transformative potential of one aspect of Roman diplomacy has received due attention over the years: annual subsidies. These could take the form not just of cash or bullion, but also of highly valued Roman commodities, such as intricate jewellery or richly woven cloths. In the Byzantine era, foodstuffs unavailable in the target economy were sometimes used, and this may have been the case in earlier eras. The point of the subsidies, as we have seen, was to reinforce the power of a reasonably compliant frontier king, so that he would have a real stake in maintaining peace on the frontier. Subsidies tended to strengthen existing monarchies. But it is important to realize that, like the amber or slave trades, diplomatic subsidies represented a major flow of new wealth into the Germanic world, and, as was also the case with the profits of trade, the appearance of new wealth sparked off competition among potential recipients. Losing their subsidy may have been one element in the Limigantes’ unwillingness to be resettled further away from the frontier, an extra downside in being demoted (in Rome’s eyes) from overkingdom to underkingdom status. Certainly, any diminution in the size or quality of the annual gifts could cause crisis, as it did when Valentinian unilaterally reduced those of the Alamanni in 364, and we have specific examples of groups moving into the frontier region precisely to overwhelm the current recipients of any subsidies and receive them in their place. Competition for the control of the flow of subsidies thus multiplied its transformative effect, and meant that Rome was sometimes left awarding gifts to the victors in struggles beyond its capacity to control.62
But subsidies were only part of an overall Roman diplomatic strategy whose other aspects also had powerful effects. Take, for instance, the periodic military interventions, which seem to have averaged out in the fourth century at about one substantial campaign per generation in each sector of the frontier. These interventions classically took the form of burning down everything you could find until the local kings came into the imperial presence to make their submission, when all the diplomatic manoeuvring and subsidy reallocation would begin. The economic effects of these burnt-earth interventions are worth careful consideration. We have no precise information from the fourth century, of course, but an interesting analogy is provided by medieval estate records from areas subject to similar levels of terrorism. Those of the Archbishop of York’s lands, subject to cross-border raiding from Scotland in the fourteenth century, for instance, show that it took revenues – a decent proxy for ‘output’ – a full generation to recover. This was because raiders, alongside grabbing moveable goods that might be easily replaced, also targeted the capital items of agriculture such as ploughing animals (approximating, in the medieval context, to tractors), which were very expensive, not to mention housing and other major items. The costs of replacing all this meant that revenues were reduced for twenty or more years.
If you factor this kind of economic effect back into the pattern of Roman frontier strategy, then, particularly in periods and areas where conflict was fairly constant, living next to the Roman Empire would be a substantial hindrance to economic development, and this is again suggested by the archaeological record. Alongside the other frontier areas where Roman imports became plentiful in the early Roman centuries, for instance, the Rhine/Weser region stands out as an exception. Few Roman imports have been found there and settlement remained much less dense until the later second century. This reflects the particular hostility between many of the groups of this region and the Empire, the Rhine/Weser being the heartland of the Cherusci and of Arminius’ rebellion which destroyed Varus’ legions in the Teutoburger Wald in 9 AD. The one area in the fifth-century west that seems to have enjoyed economic expansion at a time when the wheels were otherwise generally coming off the west Roman economy was the territory of the Alamanni, where there is good evidence of deforestation and of the expansion of agriculture and settlement, and hence by implication of population expansion as well. To my mind this is not surprising, since the contemporary reduction in the power of the west Roman state meant that it had stopped burning down Alamannic villages once per generation and regularly stealing agricultural surpluses. It was also in the fifth century that the observable tendency towards political unification among the Alamanni reached its climax, with the emergence finally of a single, unchallenged king. Again, this is not so surprising given that Rome’s countervailing interference, bent, as we have seen, on regularly removing emergent dominant figures, had ceased to be effective.63
It is also worth thinking about this and all the other aspects of Roman diplomatic strategy from an Alamannic – or general frontier-client – perspective. The regular destruction of villages could only have caused huge resentment, and Ammianus often refers to ill-feeling towards Rome on the other side of the frontier. In fact, even the less violent aspects of Roman intrusion, creating as it did winners and losers, must have been highly resented by the losers. The kind of grovelling expected in public ceremonies and so well mastered by Zizais can’t exactly have been welcome to those from whom it was required. And while Zizais may have been happy to have his political independence established, his former overking, who lost command of established rights over Zizais’ followers, can only have been hugely irritated. Ammianus records, likewise, that another former overking, Araharius, was angered when he was denuded of his subjects. Additionally, the Empire would occasionally decide – as in the case of the Limigantes – that particular barbarian groups could no longer carry on living where they had long been established, and, as we have seen, was happy to use terror to enforce that decision. This is only one of a series of high-handed actions on the part of the Romans that appear in Ammianus’ narrative. Valentinian I, for instance, altered agreements unilaterally when it suited him, both lowering, without consultation, the value of annual gifts made to Alamannic leaders, as we have seen, and constructing fortifications where it had previously been agreed that none would be placed. There are also hints in the sources that emperors would arbitrarily swap around ‘favoured ally’ status in a region so as to ensure the requisite level of subservience. Most ferociously, emperors were happy to authorize the elimination of frontier kings who posed too great a threat. The picture of Roman frontier management which emerges from all this is clear enough. The regular burning of neighbouring villages was backed up by a repertoire of aggressive diplomatic manoeuvres, which did not stop short of assassination.
If you consider all this from a non-Roman viewpoint, it becomes apparent that we need to factor into the equation a weight of oppressive Roman domination. The resentment among the many on the receiving end shows up in several different ways in the historical narratives. At the lowest level, it is evident in the willingness with which frontier groups engaged in petty and grander larceny. Raiding across the frontier was very general, and of course represented yet another Rome-emanating flow of new wealth to be squabbled over, and whose control might have transformative political effects in the Germanic world. More strikingly, resentment lay at the heart of the willingness of would-be dynasts to mount larger-scale rebellions, whether that of Arminius in the first century (whose explicit cause was taxation demands) or that of Chnodomarius in the fourth, where feelings ran high enough, as we have seen, for a sitting king, Gundomadus, to be ousted for refusing to participate.
A major factor to take into account when trying to understand the transformation of Germanic societies in this period, therefore, is four centuries’ worth of ill-feeling caused by Rome’s heavy-handed military and diplomatic aggression.
Two lines of explanation have recently been offered for the militarization of the Germani in the Roman period, evident in the increasing deposition of weapons: one, that the Germani were serving in increasing numbers as Roman auxiliary soldiers; second, that Roman campaigning east of the Rhine increased the status of warriors. As has rightly been observed, though representative of opposite reactions to Roman power – the first to its opportunities, the second to its threat – the two explanations are not remotely incompatible. Different elements among the Germanic population surely did respond along each of these lines, perhaps even the same persons at different points in their lives.64 I would only stress that the negative reaction to Roman power must be taken seriously, and its role in political consolidation acknowledged.
For militarization, as we have seen, went far beyond burying the dead with weapons. A whole new language for political leadership evolved in the Roman period, which stressed the importance of war. Rulers became war leaders literally by definition and this transformation wasn’t just achieved by force. The Germanic political community in the late Roman period still involved many others beyond kings and their immediate retinues, and the consent of this (freeman?) community to the process of political consolidation represented by the rise of military kingship was required. Here again, positive and negative worked happily side by side. A militarily effective king, as many have argued, was one more likely to win Roman recognition as a good partner to do frontier business with, and hence attract worthwhile subsidies and gifts. But he was also someone – like Athanaric and Macrianus – who was inherently more capable of resisting the more outrageous demands and intrusions of Roman imperial power. These two figures, it seems to me, show both the importance of anti-Roman sentiment and the limits to its expression by the fourth century. Both gained esteem and power in their own societies by resisting Roman intrusion, but both were willing enough to do deals when the Empire – for whatever reason – backed off and offered more acceptable terms.65 They vividly illustrate the tightrope that even the prime beneficiaries of the unfolding processes of political centralization among the Germani had to tread.