Barbarian Migrations and Socio–Economic Challenges to the Roman Landholding Elite in the Fourth Century CE
This paper draws attention to an important but neglected motivation for barbarian immigration into the empire in the fourth and early fifth century: its economic function as a source of agrarian labor. It argues that barbarian immigration into the empire must be seen against the background of specific economic challenges that the Roman landholding elites faced in the fourth century, particularly in its latter half. These challenges created a strong demand for agrarian workers whom the elites needed in order to maintain and increase productivity in a highly competitive environment. As a result, it appears that elite interest in increasing the pool of agrarian labor through barbarian immigrants was a much stronger motive of barbarian settlements in the fourth and early fifth centuries than is assumed in current scholarship.
When in 408 the Roman army (backed by a lot of Roman money) successfully drove a Hunnic force out of Thrace, a larger group of Sciri who were allied with the Huns fell into the hands of the empire. A couple of months later, Theodosius II issued the following instructions concerning their treatment:
… Therefore, we grant to all persons the opportunity to supply their private possessions with men of the aforementioned group. And all persons shall know that they shall hold those whom they have received by no other title than that of the colonate (ius colonatus), and that no one shall be permitted either fraudulently to take anyone of this group of coloni away from the person to whom he had once been assigned by deceit or to receive such an individual as a fugitive, under the penalty which is inflicted upon those [End Page 372] who harbor persons that are registered in the tax rolls of others (censibus adscriptus) or coloni who are not their own. (§1). Moreover, the landowners may use the labor of these captives like that of free persons (opera autem eorum … libera utantur) [lacuna]; and no one shall be permitted to transfer such persons, as though they had been given to him as a present, from the obligations of the tax rolls (a iure census) to that of slavery, or to assign them to urban service (sc. as domestic slaves). Those who receive such persons shall be permitted, because of the constraints in agrarian production, to retain them for a two-year period in any provinces they please, provided that these provinces are across the sea, and thereafter to place them in permanent homes, for which purpose their residence in the regions of Thrace and Illyricum shall be absolutely prohibited to them. Only within a five-year period shall it be permitted to make a transfer openly and freely within the confines of the same province. The furnishing of recruits shall also be suspended during the aforementioned twenty-year period. The distribution of these people throughout the transmarine provinces shall be made to those who so wish through petitions to apply to your court (i.e. that of the praetorian prefect of Oriens Anthemius, who was the recipient of the law).1
This text is famous for giving one of the most detailed accounts of the practical arrangements of barbarian settlements in the Roman empire.2 But its significance goes far beyond this. As a paradigmatic example, the law points to an important but neglected motivation of such settlements and of the barbarian immigration into the empire in the fourth and early fifth century in general: their economic function as a source of agrarian labor. This paper argues that the settlement of the Sciri as well as similar measures must be seen against the background of specific economic challenges that the Roman landholding elites faced in the fourth century, particularly in its latter half. These challenges created a strong demand for agrarian workers whom the elites needed in order to maintain and increase productivity in a highly competitive environment. As a result, elite landholders in the fourth century had a keen interest in increasing the pool of agrarian labor through barbarian immigrants, and they therefore welcomed the intake of new settlers—be they captives of war, chattel slaves, or "voluntary" immigrants who sought relief from economic hardship, warfare, or political pressure at home by gaining admission to the empire.
To be sure, such a use of barbarian immigrants was nothing new. The Roman state had a long history of settling barbarians on its territory; in [End Page 373] the fourth century alone such settlements are attested by the dozen.3 Moreover, the economic interest of the Roman landholding elites in barbarian immigration had always existed and has long been recognized by scholars, even though in most modern accounts the governmental interest in new taxpayers and recruits receives far more attention than the interests of private landholders.4 There is, however, reason to re-evaluate the phenomenon with regard to the fourth and early fifth century. While earlier scholarship tended to understand barbarian settlements in this epoch as a measure against a general demographic and economic decline in the Roman Empire, recent decades have seen a complete reversal of our picture of the late Roman economy. Also, in the wake of this reappraisal a new picture begins to emerge of how certain social and economic developments at the time affected the landholding elites of the empire. Against this background, the economic context and significance of barbarian immigration into the Empire in the fourth and earlier fifth century appear in a new light. With the argument outlined above, this paper offers a reinterpretation of the phenomenon that takes the reassessment of the socio-economic framework into account. In so doing, this paper also makes a case for bringing together two fields of study that scholars have rarely connected: the history of migrations and of the late Roman economy. The paper will first sketch the economic challenges to the landholding elites of the time in light of the new picture of the late Roman economy. In the second part it presents evidence that proves the link between these challenges and barbarian immigration into the Empire in the fourth and early fifth centuries. [End Page 374]
The Socio–Economic Challenges of the Late Roman Elite in the Later Fourth Century
Until late in the second half of the twentieth century, the commonly held view of the late Roman economy emphasized depression and decay. According to this model, demographic decline, tremendous fiscal pressure, and repeated warfare on Roman soil all contributed to ongoing economic subsidence beginning in the third century. In recent decades, this picture has changed dramatically. The notion of fiscal pressure, for example, has been rightly questioned, and evidence like the numerous references in legal sources to "abandoned land" (agri deserti), long read as indicators of unproductivity and demographic decline, has been reinterpreted. More importantly, both an increase in and reinterpretation of the archaeological evidence for settlement patterns, material culture and trade has made clear that many regions of the empire enjoyed considerable and at times extraordinary prosperity in the fourth century. In the Eastern Mediterranean or in North Africa this equalled or surpassed the levels of economic performance witnessed in the High Empire, and in certain places in the East this continued well into the sixth century. Commercialized agrarian production formed the basis of this prosperity and also fuelled other sectors like specialized industry and trade.5 Against this macroeconomic background—which cannot be explored in any depth here—a number of social and economic phenomena that have long been regarded as indications of crisis begin to appear in a different light. These phenomena as such are well known, but it is necessary to briefly restate them with a view to the challenges they created to the landholding elite. One is the issue of labor supply in the agrarian sector.
Famously, fourth-century imperial legislation is filled with efforts to inhibit the social and geographical mobility of the agrarian workforce by tying the peasants to the soil they tilled through a hereditary bond, a legal instrument that scholars traditionally call "the bound colonate." There is a long-standing scholarly debate as to whether the hereditary binding of coloni to their soil emerged out of the fiscal registration of taxpayers in the wake of Diocletian's tax reforms, or whether it grew out of private dependencies of tenants working a landlord's possession that were translated into an institution of public law. Many would now deny that "the colonate" was anything like a unified legal institution and that coloni represented a clearly defined status group; the [End Page 375] terms colonatus and colonus rather seem to have been umbrella words for a broad variety of legal statuses and agrarian labor relations. Equally controversial are the questions of how widespread the colonate was as a legal institution binding tenants to the soil of a landlord; how its extent compared to that of rural slavery, contract labor, classical locatio conductio-tenancies, and free smallholders; as well as what loss of freedom it practically entailed.6
Whatever the answer to these questions, the sheer abundance of relevant provisions shows that peasant mobility came to be regarded as an enormous problem in the fourth century. This clearly emerges from the harsh rhetoric of the legal texts concerning coloni. One law insists: "They may serve as slaves of the land (inserviant terris)—not just on the basis of a fiscal bond but under the name and in the legal condition of a colonus"; others talk of legally tied peasants as servi terrae or coloni servilis condicionis.7 Even if the binding of peasants to their status and soil may initially have been solely a matter of fiscal administration, this rhetoric leaves no doubt that soon more was at stake. On the one hand, the mobility restrictions reflect the concern of the imperial government to maintain a stable base of taxpayers. On the other, there can be no doubt that the intensification of legal efforts to bind peasants to their status and soil was driven by the interest of the landowning elite who solicited or, as members of the governing class of the empire, actually made these laws. As we shall see, fourth-century landowners had more than one reason to be concerned about peasant mobility.
The pertinent imperial legislation likes to censure the migration of peasants from their native soil as "absconding" or "flight," thus suggesting that [End Page 376] agency lay only with the peasants. A famous edict of Constantine, for example, thunders that "coloni who meditate flight must be bound with chains and reduced to a servile condition, so that by virtue of their condemnation to slavery they shall be compelled to fulfill the duties of a free man" (CTh 5.17.1). It is, however, revealing to look more closely at the social dynamics behind the "flight" of the coloni. The same Constantinian edict, for example, continues that "any person in whose possession a colonus that belongs to another is found not only shall restore the aforesaid colonus to his origo (that is, place of registration) but shall also assume the capitation tax for this man for the time that he was with him." Landowners profited twice from employing a peasant registered elsewhere: they not only benefited from additional labor power but also enjoyed that labor free of its concomitant capitation tax, provided they could manage to hide the fugitive coloni from the tax authorities. Indeed, reclaiming the capitation tax from the landowner who had housed the peasant was an often-repeated concern of the pertinent legislation.8 Landowners thus had a strong interest in harboring peasants belonging to someone else. The abundance of pertinent provisions implies that this was indeed a widespread practice. Some texts even imply that landowners engaged in active recruiting among the dependent workforce of their peers. A law of 386, for example, blames landlords for housing peasants belonging to someone else's sollicitatione, that is "after they had incited" the colonus to abscond.9 The multiplication of legal prohibitions against the "flight" of the coloni and their subsequent employment on the estates of other landlords, although rooted in a variety of concerns, thus seems particularly to have been motivated by a competition for labor among the landholding elite, with the repetition of laws against peasant mobility representing the voice of the losers in this competition.
Occasional glimpses into the realities of rural life in nonlegal sources confirm the same picture. One example is the letter of an anonymous large landowner in North Africa to a peer named Salvius concerning a legal dispute between the two over coloni working on the letter-writer's estate in around 400. Both were lawyers; during their legal training, they had together learned, [End Page 377] "by which right and according to which procedure coloni are reclaimed, who is entitled to bring an action, and who is not entitled to the result of an action," as the letter informs us.10 Such conflicts thus seem to have been a frequent issue in Roman courts. In the dispute under question in our letter, Salvius appears to have claimed that the coloni were registered on a piece of land he owned and had illegally "absconded." He therefore had demanded them back and threatened to take the issue to court. In response, the author of the anonymous letter denies Salvius's claims by arguing, among other things, that the land in the latter's possession where the coloni were allegedly registered was far too small to be the origo of so many people. Tellingly, however, he tries to convince Salvius to avoid the courts and accept a private meeting of the two to find a solution. Perhaps Salvius's claims were not as unfounded as the letter tries to suggest. In any event, it illustrates the seriousness of the competition for agrarian labor among the elite.
This struggle for agrarian labor power among the landowning elites is borne out by other phenomena. For example, fourth-century legislation repeatedly insisted that descendants of coloni retained this status, even if this was against the principles of the Roman law of status. For instance, by decree the offspring of a marriage between a daughter of a registered colonus and free man of non-coloni status was to inherit the mother's birth status even though, as a colona, she was technically a free person capable of entering a matrimonium iustum, whose offspring would normally inherit the father's status.11 These regulations appear to be geared to the interest of landowners in keeping their workforce stable. So does the increasing tendency to substitute the payment of cash for the obligation to furnish recruits, a privilege that was often granted to high-ranking senators at the time. The fervent protests of the landholding aristocracy of Italy against the general levy for the war against Gildo in 397, which forced Honorius to allow the commutation of recruits into cash, illustrates the interest of the landowning elite in keeping its agricultural workforce in place.12 Moreover, from around 400 onwards, [End Page 378] registered coloni were banned from military service, a rule that was explicitly introduced to "look after the rights of the landlords."13 In addition to obtaining imperial laws such as these, landowners also resorted to illegal practices. Numerous imperial decrees, for example, prohibit the harboring of deserters as agrarian laborers.14 One constitution intervenes against the practice of alienating land while keeping the workforce attached to it. Another bans what it calls an "old trick with regard to registered coloni" (quod in originariis saepe actitatum est): buying only a small parcel of land out of a large estate but taking the entire workforce of the estate with it.15 Evidently, the supply of agrarian labor was a pressing concern.
Traditionally, the abundant evidence for the "flight" of coloni has been read as an indication of the pitiful state of the peasantry in a climate of economic decline and fiscal depression. Confirmation for this was thought to be found in texts like an edict of Constantius II reporting that "a multitude of peasants located throughout Egypt have betaken themselves to the protection of those persons who are supported by their high rank of various degrees and even to dukes."16 Together with a number of similar passages, this text was taken as evidence that peasants fled pressing fiscal demands and greedy landlords by seeking the protection of powerful magnates, only to lose their independence and, in the long run, also their land, to these patrons. Scholars have long cast doubt on this gloomy picture, and the new assessment of the late Roman economy has further undermined it.17 Most recently, Dossey's work on the peasantry of Late Roman Africa has made a crucial contribution to calling the traditional picture into question.18 In her view, the peasantry of the fourth and fifth centuries fully profited from the economic boom of fourth-century North Africa. They fared better overall and were more integrated in the Roman world than they had ever been before—an interpretation based, among other evidence, on the density of settlements, the widespread use of high-quality ceramics, and the dissemination of high-quality building materials, including luxury amenities like bathhouses, some of them probably [End Page 379] provided by absentee landlords as incentives for their workers. Crucially, Dossey argues that it was precisely this prosperity that lay at the root of the social tensions between the North African peasantry and the landholding elite in the fourth and early fifth centuries, which also fuelled the Donatist controversy. In Dossey's view, the prosperity not only allowed the peasants to better their condition, it also helped them develop claims to fully participate in the Roman system and have rights within it. At the same time, Christian discourses provided them with a way to articulate their rights or grievances against the landowning elite in relation to matters such as debt. In sum, prosperity led to an increased self-confidence among the North African peasantry as well as to tensions in their relationship with landowners.
There is good reason to believe that Dossey's picture holds true for other regions of the empire as well. As mentioned above, economic prosperity was not restricted to North Africa, nor was Christianity. Indeed, similarly self-confident and demanding peasants are also attested elsewhere. Evidence for this comes, for example, from a sermon of John Chrysostom in 400 ce urging landowners to build churches on their estates in order to convert the peasantry. The text provides illuminating glimpses of the life on a typical large estate of the time: the peasants enjoy baths, taverns, and markets provided by the landlord; and Chrysostom implies that the landlords did this dia doxan, to enhance their prestige among the peasantry, in other words to ensure their loyalty. At the same time, Chrysostom seems to envisage tensions between peasants and landlords. For he warns the latter that all these amenities only make the peasants audacious, while a church and a priest on the estate would "drive away intemperance," "drunkenness," and "luxury," and would secure "peace (eirēnē) among the farmworkers" and "stability (asphaleia) on the estate."19 Chrysostom, it seems, presupposed a certain degree of uneasiness between peasants and landlords that threatened the smooth running of their estates and tried to sell them his church-building program as a remedy for that problem.
In his On patronage (Or. 47), delivered in the 380s, Libanius famously betrays a similar uneasiness when he describes the problems he had with the peasants in a village he owned in Syria. These, he claims, gained the patronage of a high-ranking military officer. Relying on this patronage, they bullied neighbors, took their property, and resisted the tax-collecting curials. They also used patronage to avoid their duties as tenants to their landlords. This is [End Page 380] where Libanius himself comes in: his own tenants had also refused to pay the rent, so Libanius brought the matter to court, and everything went well until the patron of the peasants intervened and stopped the investigation. There is reason to believe that Libanius stops short of telling us the whole story here,20 but in any event the text nicely illustrates the kinds of problems that landowners had, or could make others believe they had, with self-confident peasants playing games to their own advantage.
Against this background, the evidence presented earlier for peasant flight and patronage appears in a new light: These passages by no means show an indiscriminately desolate peasantry; rather, the sources at least partially attest to self-confident peasants actively seeking their own advantage in a booming economic environment. One of the strategies the peasants applied was to exploit the fierce elite competition for agrarian labor by moving to another estate in order to negotiate better conditions with the new landlord. This is the situation we find, for example, in a law of Valentinian I showing a broad variety of possible relations between a landlord and a "fugitive" peasant harbored on his estate: some peasants were exploited without receiving a fair remuneration for their work; but others, who managed to hide the fact that they were coloni belonging to someone else, were envisaged by the legislator as being able to enter a leaseholding contract and work the land for their own profit "as if out of their own will and like free men" or as being employed as wage laborers on a contractual basis. Conditions of work in the agrarian sector were thus at least partly a matter of negotiation and, to a certain extent, choice.21 On a revisionist reading, therefore, texts like those of Libanius or the laws on rural patronage appear to show that the peasants were trying to play the landlords against each other to their own advantage—even though in the long run they sometimes risked being absorbed into a powerful magnates' estate, as is attested in a number of texts.22
All this is not to say that life was easy for the rural population in Late Antiquity, and it would be foolish to deny that fiscal and economic pressure as well as oppression through the state and seigneurial authorities was a constant feature of rural life. Nonetheless it seems clear that in the fourth century the landholding elite was more than once confronted with a new and challenging situation: they not only had to fight against a notorious labor shortage but [End Page 381] also struggle with a workforce that often enough was able to assert its rights with self-confidence and, because of the competitive economic environment, was constantly threatening to leave in search of better conditions elsewhere. The bargaining power agrarian workers could have in this situation is nicely illustrated by an episode of the early 420s from Numidia: when the notoriously unscrupulous bishop Antoninus, who had been driven from his see at Fussala, decided to establish himself at a nearby senatorial estate, the coloni of the estate, resenting him, petitioned the absentee owner, a clarissimia femina. They threatened that "if she allowed this to happen, they would immediately leave the estate" (se continuo migraturos). The estate owner and the diocesan bishop, Augustine of Hippo (as he himself relates), promptly withdrew their support of Antoninus.23
Manpower shortage and a self-confident peasantry, however, were not the only challenges that the landowning elite faced. In the fourth century, the Roman Empire saw a massive expansion of the state apparatus. Because the upper echelons of this apparatus were mostly recruited from the municipal elites, local communities came to be confronted with a large group of active or retired high-ranking office-holders who became powerful patronage brokers in the cities where they resided. We have already encountered these powerful patrons in the law on peasant patronage cited above, where patrons were described as "persons who are supported by their high rank of various degrees and even dukes." Other laws mention members of the chancellery of a comes, magistri militum and comites, honorary proconsuls, vicars, and the like as patrons; in Libanius's speech it was a high-ranking military man.24 The existence of these high-ranking, influential patrons threatened the landowning elites not only because their patronage played a key role in mediating shifting relations between other landowners and peasants. These men also had a massive impact on elite landholding patterns. They were often well off as a result of the salaries they earned and the additional semi-official and sometimes illegal income they could acquire while in office. What is more, as Banaji has demonstrated, they also profited from the economic effects of changes in the monetary system and fiscal administration. After many decades of inflation and monetary instability, in the course of the fourth century the gold coinage, because of its stable value, became the dominant means of exchange for all larger transactions, including tax payment. According to Banaji, the social [End Page 382] consequences were dramatic in the highly monetized economy of the Roman empire, as these changes gave a significant economic advantage to people with access to gold. This economic, or monetary, advantage in Banaji's view can help explain why, in the long run, many freeholding farmers became dependent on rich landholding magnates. However, this economic development also affected the landholding elites: it shifted the dynamic in the land-market and gave those men a distinct advantage who, as a result of their office-holding in the imperial state apparatus, had regular access to gold through their salaries and other income. In fifth-century Egypt, this development led to the formation of large estates—the so-called "Great Houses"—in the hands of the new imperial aristocracy. And the same dynamic may also have contributed to the resurgence of aristocratic villae (as representational buildings and centers of large estates) elsewhere in the fourth and fifth century, particularly in the West.25
Contemporary complaints about honorati, imperial office-holders, putting pressure on the land market by buying out the land of the curial elite—their peers who had not entered imperial service—speak to the same socio–economic dynamic. The speeches of Libanius, himself a landowner, are full of such complaints. In Oration 49, for example, Libanius complains about "those who comfortably reap the harvest from the fields of the emperor"—that is, the office-holders in the imperial administration—turning up "from everywhere" to buy the land of councillors ruined by the fiscal demands of the empire. Oration 48 paints a similar picture: "If any of the honorati admires a councillor's estate, and proceeds from admiration to desire, it is no sooner said than done: the councillor is forced to sell, and the honoratus is all ready to buy." And, what makes it worse, much of this was done with the connivance of leading decurions: "In some cases, you personally put down the money, in others you stand in for persons of influence and curry favor at their tables by means of the property of the councillors".26 One of these land-grabbing honorati is directly attacked in Oration 62, another speech that engages with the social dynamics created by the enlargement of the imperial service. Libanius upbraids him for being wealthy because he used office-holding to enrich himself through usury at the cost of many a family, "unbending before floods of tears, bawling and bellowing: 'Sell your estates, sell your slaves, sell your family tombs, and, if you can, sell yourself too.'"27 [End Page 383]
These passages are polemical, and Libanius's accusations exaggerated. Wealthy decurions are well-attested throughout the fourth and earlier fifth centuries, and scattered evidence shows that they continued to exist well into the sixth century, even though the numbers might have decreased and many rich curial families were absorbed into the bureaucratic aristocracy.28 Moreover, evidence from Asia Minor, Egypt, Northern Syria, and North Africa, among other places, shows that small- and medium-sized landholdings continued to exist into the sixth century. Although a concentration of land is discernible in the documentary sources, such as the fiscal registers from mid-fourth century Egypt, the accumulation of property in the hands of the office-holding elite apparently was never as complete as earlier scholarship has maintained.29 However, it is beyond doubt that the competition for land was intense in the fourth century, particularly in its latter half, and it is very likely that this competition, in part at least, resulted from the social dynamics created by expansion of the imperial service.
This is suggested also by evidence beyond the speeches of Libanius. A law of 386, for example, ordains that estates (praedia rustica vel urbana) of city councillors can be sold off only if the provincial governor has been informed and assented to the transaction. According to the law, the purpose of the provision was to make sure that the seller was not "circumvented by the stratagems or overwhelmed by the power of the purchaser"30—this very likely meant a [End Page 384] high-ranking office-holder. But competition for land existed not only between honorati and curiales; the problem affected the landholding elites more generally. Evidence for this comes, for example, from a number of laws that assure the holders of perpetual leases of imperial domains—normally men of the provincial elite—that no one can cast them out of their tenancy by using his influence or bidding a higher rent. A constitution of 393 is illustrative of these provisions: "It has come about through the arrogance of wicked men that all the best lands are serving their greed for gain and profits, and only such inferior fields are left in the province as none of the aforesaid men deign to hold. …"31 The repeated appearance of the topic in legislation shows that such attempts to expel leaseholders of profitable land through a higher bid or other methods were common. To judge from extant legal evidence, so were other unlawful practices deployed in the competition for land.32 It is think-able—if ultimately impossible to prove—to link these texts at least in part to the dynamics described among the landholding elite in the later fourth century.
In sum, Roman landholding elites in the fourth-century experienced an intensification of the competition for wealth and property that forced them to increase the productivity and profit from their land in order to survive in this fiercely competitive environment. This development exacerbated the problem that landowning elites had with the labor supply. As economic growth largely depended on the availability of agrarian labor, they felt an ever-stronger pressure to keep their labor pool stable or, if possible, to increase it. At the same time, they were confronted with a self-confident peasantry that was able to exploit the fierce competition among landholders, the increased demand for labor, and the shifting relations of power and patronage within the elite. To be sure, none of these phenomena was entirely new in the fourth century; there had always been competition for land and labor among Roman landowners. But the density of the evidence discussed suggests that, for the above-mentioned reasons, the competition acquired a new quality and intensity in the fourth century. Landholders met these challenges with a variety of strategies. As to the land market, legislation was issued to enforce a closer surveillance [End Page 385] of land transactions and to prohibit illegal practices of land acquisition. With regard to the labor supply, landowners provided incentives to ensure the loyalty of their peasants; they solicited legislation to tighten control over the existing workforce; and they made an effort to enhance their share of the empire's agrarian labor pool by taking advantage of peasant mobility and by a variety of other, often illegal means.
Late Roman Landowners and the Uses of Barbarian Immigrants
In this situation immigration provided a remedy. It thus seems a priori likely that the landholding elites welcomed the intake of barbarians, whether voluntary or forced, as potential laborers. Examining the actual deployment of barbarian immigrants further strengthens this case. The evidence for these practical aspects of barbarian settlements is limited. Some of the newcomers appear to have received land as freeholders (often of uncultivated land, either agri deserti or virgin land). Others were settled on imperial land, possibly with a status similar to imperial coloni.33 In these cases it would appear that the interests of the Roman state in additional taxpayers and potential recruits shaped these settlements. And yet it is clear that individual Roman landowners also profited from the intake of barbarian laborers. For one thing, there is ample evidence that the slave trade between the barbarian world and the empire was flourishing in the fourth and earlier fifth centuries; Synesius of Cyrene famously grumbled that even moderate Constantinopolitan households were replete with "Scythian" slaves.34 It has long been recognized and has been re-stated forcefully in recent years that slaves remained an important source of labor in the agrarian sector as well: contrary to a widespread view in earlier scholarship, slavery was never entirely supplanted by other forms of bondage like the colonate. In many regions including Italy, the Aegean and Western Asia Minor, agricultural slavery remained at least one among several sources of labor. There is—to my knowledge—almost no positive evidence to prove that slaves of barbarian extraction were used as farm workers; but given the high numbers of slave imports into the empire this is more than likely.35 [End Page 386]
As regards larger groups of incoming barbarians, whether they were prisoners of war or immigrants by choice, there is clear evidence that Roman landowners did make use of them as agrarian laborers. The first source to prove this a panegyric to Constantius Chlorus delivered in 297 or 298 on the occasion of the recapture of Britain from the usurpers Carausius and Allectus some months before. One of Constantius's deeds that receives praise is the subjugation and subsequent deportation of Chamavi and Frisii from the Lower Rhine area, who apparently had been allied with Carausius. "With their wives and children," the panegyrist relates, "… they moved to land long since deserted (loca olim deserta) in order to restore to cultivation through their servitude what they themselves, perhaps, had once devastated by their plundering." The text then gives interesting details about the deployment of the captives: "In all the porticoes sit captive bands of barbarians"—men, old women and wives, boys and girls, vividly described by the panegyrist—"parcelled out to the inhabitants of your provinces for service (hos omnes provincialibus vestris ad obsequium distributos), until they may be led out to desolate lands (solitudines) assigned to be cultivated by them." And Gallia herself is put on stage, rejoicing that, "it is for me now that the Chamavian and Frisian plows, and that vagabond, that pillager, toils at the cultivation of the neglected countryside and frequents my markets with beasts for sale, and the barbarian farmer lowers the tax (annona)!"36 Private landowners thus clearly benefited from the immigrant workforce. To the audience the issue was evidently important enough that it was taken up again at the very end of the panegyric. There Diocletian as well receives praise for having transferred people from Asia to "the desert lands of Thrace," while the Augustus [End Page 387] Maximian is lauded because "laeti, restored by right of postliminium, and the Franks, admitted to our laws, have cultivated the empty fields of the Arvii and the Treveri." And once again the audience is reminded of Constantius's own benefactions: "Whatever land remained abandoned in the territory of the Ambiani, Bellovaci, Tricasses, and Lingones turns green again under cultivation by the barbarian."37
Further evidence for landowners profiting from barbarian immigrants comes from imperial legislation. The provisions of the law of 409 cited at the beginning of this article concerning the settlement of captive Sciri include that everyone may have the opportunity "to supply their private possessions (proprii agri) with men of the aforesaid race" by submitting an application to the office of the Praetorian Prefect. Moreover, the landowners are given remarkable latitude in deploying their share of the Sciri wherever it pleased them (provided they were not settled on the European side of the Bosporus). The government restricted itself to ensuring that the new workforce benefited the state through taxes and by enhancing agrarian productivity: the law stipulated that, "no one shall be permitted to transfer such persons … from the obligations of the tax rolls (a iure census) to that of slavery, or to assign them to urban service" (as domestic slaves)—as slaves, the Sciri would have benefited only the landowners who were free to sell them or deploy them in other businesses, while their status as coloni guaranteed that they worked to the advantage of both the landowners and the imperial treasury as well as contributed to the food supply. Moreover, the government even abstained from drawing the Sciri as recruits into the imperial army for twenty years. Although there is no reason to presume that barbarian settlements always followed the pattern described in this law,38 it is clear that in 409 the measure was aimed at benefiting Roman landowners. A contemporary of the settlement bears witness to this analysis. According to the church historian Sozomen, the Sciri were, "allocated at a low price, while others were given away into servitude (douleia, probably meaning tied coloni) as a present"—to deserving members of the elite and imperial protégés, one may suppose. Sozomen claims to have himself witnessed some of them working as farmers in Bithynia near Mount Olympus, in the hinterland of Constantinople where many elite members had their estates; there they "lived apart from one another, and cultivated the hills and valleys of that region."39 [End Page 388]
Less well-known but equally important as evidence for barbarian laborers benefiting Roman landowners is a constitution of Honorius issued in April 399:
Since persons of many nations seek the felicity of the Romans (sequentes Romanam felicitatem) and have betaken themselves to Our Empire, and since they must be provided with letic lands (terrae laeticae), no person shall obtain any of these lands except in accordance with Our written allowance (adnotatio). Since some men have either seized more land than they obtained from Us, or by the collusion of the chief decurions (principales) or of the defensores, or by rescripts surreptitiously obtained have acquired a measure of land greater than the official calculation (ratio) demanded, a suitable inspector shall be dispatched who shall recall whatever lands have been either delivered illegally or seized wrongfully by any persons.40
The constitution is one of the few sources furnishing information about the status und settlement of laeti, a term that is normally—and not least on the basis of this text—taken to denote voluntary immigrants who, as the constitution says, "seek the felicity of the Romans and have betaken themselves to Our Empire," as opposed to conquered and subdued captives of war (called dediticii or the like).41 Laeti settlements, attested in Gaul, Spain, and Italy (but not in Africa, implying that this law, addressed to Val. Messala Avienus, prefect of Italy and Africa, might have referred to a situation in Italy)42 appear to have been integrated to the administrative structure of the civitates, as our law and some other sources suggest.43 The constitution itself refers to [End Page 389] a situation where "by the collusion" of municipal magistrates and leading citizens and by way of "surreptitiously obtained imperial rescripts" "some" claimed more of the terrae laeticae than what the imperial administration had provided for them. Both the question of who exactly benefitted from the illegal rescripts as well as the implications of the law for the legal and economic conditions of the laeti remain debated. The most convincing reading is that the laeti were settled as coloni on public land—that is, on imperial or municipal land that then came to be called terrae laeticae—that was leased out to Roman landlords. In return for the rent they received from their laeticoloni, these landlords managed the land as well as the (potentially difficult) barbarian coloni. Importantly, they were held responsible for the taxes of the terrae laeticae and perhaps collected them themselves—a solution that sounds very Roman indeed in its practicality and cost-efficiency for the state. If this is true, the situation that prompted the constitution in all probability was not that some laeti received more land than others. Rather, it was Roman landlords who had, in collusion with members of the municipal elite, secured for themselves a larger share of the land with its coloni, while their Roman peers, who had been outmaneuvered, had countered by soliciting an imperial rescript against their actions.44 On this reading the constitution not only further supports the view that the Roman landholding elites benefited from this immigrant labor force. It also attests to a heated conflict over land and the barbarian workforce on it that was occurring within the municipal elite. The [End Page 390] constitution thus provides another example for the fierce elite competition for an agricultural workforce and land—and, indeed, for barbarian immigrants.
These texts thus provide explicit proof for the proposition that landholding elites across the empire profited from barbarian immigrants as agrarian workers and that they competed for control over them. Moreover, a number of sources reporting barbarian settlements in situations when the Romans were strong enough to dictate terms may refer to barbarian coloni and the like working on private land without explicitly saying so. The Historia Augusta, for example, reports of a multitude of "barbarian slaves," "Scythian farmers," and "Gothic coloni," as well as "slaves" filling the provinces after the Claudius Gothicus had vanquished the Gothic invaders in 269.45 This very much sounds like a reflection of fourth-century dealings with barbarian captives or refugees of war. The same source tells us that Aurelian planned to settle captive barbarians on fertile but derelict tracts of land along the Aurelian Way in Etruria and Liguria who would produce wine for the provisioning of the city of Rome—provided, however, that the owners of the agri deserti agreed. There have been different efforts to explain what kind of measure the author of the text envisaged, not least due to a textual problem with the passage. Some have understood the text to mean that Aurelian planned to take the land and then give it to barbarian settlers; others believe the landowners kept (or received) agri deserti and were to be given the settlers as coloni.46 It is impossible to judge the historical value of the story; but if the latter reading is correct, it may again reflect a retrojection of practices of the fourth century into a third-century context. [End Page 391]
To this we can add further sources. Ammianus reports that after the crushing defeat of the Taifali by the Roman general Frigeridus in 377 the survivors were sent to Italy "to farm the land around Modena, Reggio Emilia, and Parma"47—an arrangement that very much recalls the situation envisaged in the 399 law and thus may have benefited the local municipal elites. The same impression emerges from a passage in Eunapius mentioning Gothic captives of war in custody in Roman cities and households in 366 ce.48 Ausonius, in the Mosella (v. 9), mentions Sarmatians as coloni in the vicinities of Tabernae (Rheinzabern); even though the term can also mean coloni on imperial land or simply "farmers," it could just as well attest to dependant barbarians working Roman elite estates. After the defeat of Radagaisus's army in 406 it is reported that "the number of captive Goths was so large that a multitude of men was sold off (into slavery) for a solidus each, like cattle of the cheapest kind."49 Such a large number of slaves could hardly have been absorbed for household labor only. Not least, in some cases where barbarian immigrants are reported to have been settled as tributarii, this does not necessarily imply that they were freeholders paying tax (the normal meaning of tributum); in legal documents, tributarius could also be employed for dependent farmers ("coloni" or the like) when it pertained to their duties as taxpayers.50
Moreover, immigration as a source of labor also seems to have become a particularly strong concern in elite discourse in this period, especially in the last decades of the fourth century. The texts to prove this are, as we shall see, highly unusual and thus make a strong case that the use of barbarian immigrants as agrarian laborers gained particular importance in this time, not only in discourse but also as a social reality. One of these texts is Themistius's Oration 16, delivered in the presence of the emperor Theodosius I on the occasion of New Year's celebrations on January 1, 383, before the assembled elite of the Eastern empire. The official remit of the speech was to praise the new consul, the general Fl. Saturninus, and thank the emperor in the name of the senate for awarding him the consulate. But its true agenda was political and highly delicate: to justify the peace that Theodosius had made with the Goths in the autumn of 382—four years after the devastating battle of [End Page 392] Adrianople and after five years of constant warfare in Thrace and adjacent regions. Saturninus had been one of the chief negotiators of the peace.51
Theodosius, so Themistius, "was the first who dared entertain the notion that the power of the Romans now lay neither in weapons, nor in breast-plates, spears, and unnumbered manpower, but that there was need of some other power and provision, which … subdues all nations, turns all savagery to mildness and to which alone arms, bows, cavalry, the intransigence of the Scythians … yield": prudence and forgiveness "for those who had done wrong," in a word, peace.52 For, Themistius asks, even if it had been possible to utterly destroy the enemy, "would it have been better to fill Thrace with corpses or with farmers? … To progress through a wilderness or through cultivated land? To count up the number of the slaughtered or those who till the soil? To colonize it with Phrygians or Bithynians perhaps, or to live in harmony with those we have subdued? I hear from those who have returned from there that they are now turning the metal of their swords and breastplates into hoes and pruning hooks, and that while paying distant respect to Ares, they offer prayers to Demeter and Dionysus." Themistius further elaborates on the theme in a later passage: "Now the whole continent is settled, land and sea garland their leaders; … the roads are open, mountains are free from fear, plains now bear fruit, and the land around the Danube does not leap in the theatre of wars but devotes itself to sowing and ploughing."53
All this, like Themistius's encomium on the blessings of the peace in general, can easily be dismissed as empty rhetoric. Too obvious is the agenda to sell a peace that many in the audience found humiliating and detestable. Furthermore, the barbarian warrior-turned-into-farmer is an all too well-attested and long-standing stereotype in triumphalist Roman panegyric: to mention only a few examples from the fourth century, we have already encountered the theme in the panegyric of 297 or 298 on Constantius Chlorus. Similar ideas could be heard in a panegyric of 310 on Constantine, which marvels at the Franks who, "having been settled in the deserted regions of Gaul, now promote the peace of the Roman Empire by cultivating the soil."54 So too in a panegyric Libanius held before a local audience in Nicomedia ca. 345, in which he praised Constantius II for having rendered uncultivated land in [End Page 393] Thrace fertile by bringing in thousands of Persian war captives.55 The idea appears to be a topos.
At the same time, in no other extant imperial panegyric does the theme of barbarians cultivating land inside the empire take up so much space as in Themistius's speech of 383. Moreover, spoken in the high ceremonial context of New Year's celebrations and in the presence of the emperor and the greater part of the eastern elite, the words of the doyen of the eastern Roman Senate carried much weight and were surely carefully and deliberately chosen. In 383 Theodosius found himself in an awkward situation. He had concluded a controversial peace without any decisive victory that accorded the enemy land on imperial soil under conditions unusually favorable to a barbarian group. The peace by no means compensated the losses that five years of devastating warfare had inflicted upon the provincials (including, presumably, some members of Themistius's audience), to say nothing of the blow that such a peace dealt to the ideology of the empire and its elite. It is very likely that there was much resentment in the audience and perhaps even doubt about Theodosius's abilities as ruler.56 In such a situation, neither an experienced spin-doctor like Themistius nor his audience would have been content with platitudes. It thus seems likely that Themistius tried to justify the peace with arguments that the elite found convincing. Given its prominence in the speech, immigration as a source of agrarian prosperity would appear to have been such an argument, and apparently an important one. In view of the economic pressures on the elites described in the previous section of this paper this is hardly surprising. Themistius thus tried to capitalize on current concerns of the elite, even though the Gothic immigrants of 376 to 382 were not available as labor power for the elite—or at least not yet. And for those in the audience who did not want to let themselves be convinced so easily, Themistius resorted to a subtle threat: would it be better, he asked in the passage quoted above, to re-settle Thrace with Phrygians or Bithynians than with Goths? Phrygia and [End Page 394] Bithynia belonged to the hinterland of Constantinople and the heart of the Eastern empire, where many members of the elite present in the audience had possessions. Loosing agrarian laborers in that region through a large-scale relocation of the population clearly was not a welcome prospect. Themistius played on elite anxieties when trying to convince a critical audience of the landholding elite that the settlement of Goths was in its own best interest.
All this makes the panegyric of New Year's Day 383 strong evidence that barbarian immigration as a source of labor supply was an important theme in contemporary elite discourse. In this context it is perhaps significant that the same topic occurs also in another panegyric for Theodosius. When Pacatus Drepanius in 389 addressed the Western elite in the city of Rome, he also briefly touched upon the Gothic peace: The Goths, he claimed "were taken in our service and made themselves available—for you, emperor, as soldiers, for the land"—i.e. our land—"as farmers."57 It is likely that this was also geared at an elite discourse that welcomed barbarian immigrants as a source of labor.
Further evidence for such a discourse comes from the two laws from 399 and 409 discussed in the previous section. These texts were not just legal dispositions. Imperial legislation of the late empire always had a propagandistic aspect; emperors used legislation to communicate with their subjects and promote their policies through a medium that had very wide scope and, by virtue of being normative, conveyed the notion of a strong imperial commitment to the relevant topic. In more than a few cases it can even be shown that this effect was the primary reason for enacting an imperial decision as a general law rather than as a case-specific rescript.58 For example, the proliferation of laws on Kolonenflucht and related matters was probably at least in part due to an imperial policy to demonstrate awareness of a major elite concern.
That the two laws on barbarian settlements from 399 and 409 had a similar propagandistic purpose is a priori likely as the topic is unusual for legislation. While each of the dozens of settlements known from the fourth century alone must have created a mass of internal administrative communications, these never surface in the general and broadly publicized legislation that was collected in the Theodosian and Justinianic Codes. Apart from the two laws [End Page 395] of 399 and 409, elsewhere in the codes barbarian settlements are mentioned only in passing or in the context of other concerns.59 The decision to issue a general law dealing specifically with this topic thus was a statement of its own. It communicated to the elite that the imperial government was aware of a major concern of theirs, the need for agrarian workers, and that it was taking steps to cater to their interests. This is borne out by the historical context of these texts as well as the more specific concerns they address. The 399 law falls in the first years of the boy emperor Honorius when legislation massively courted the elites in other respects as well.60 In a climate of intensive competition for workforce and agrarian surplus, an imperial announcement that the government would not tolerate some individuals receiving privileged access to the immigrant workforce through illegal practices was probably well received among the landowning elite in Italy.
Similarly, the 409 law was issued only a few months after the seven-year old Theodosius II had ascended the throne. Through the distribution of the Sciri as coloni to private landowners, the massive concessions it made to the interests of the landlords, and the unusual decision to broadcast the measure in a general law, the new government arguably made an effort to win the sympathies of the Eastern elite by demonstrating its care for their concerns. It seems more than a coincidence that Anthemius, the praetorian prefect, was put in charge of dealing with landholders and apportioning the captive Sciri to them. Anthemius, the scion of a leading Eastern Roman family, an ex-consul, patricius, and prefect between 405 and 414, was regarded by contemporaries as the man pulling the strings at Theodosius's court.61 It is quite possible that he was behind staging the distribution of the Sciri coloni to win the favor of the governing classes of the East for his regime and to instill loyalty among the beneficiaries of his generosity. What is more, in distributing the captive Sciri to Roman landowners, the 409 law explicitly addresses the problem of manpower shortage in the agrarian sector—it is "because of the constraints in agrarian production" (pro rei frumentariae angustiis), that landlords receive almost complete license to deploy the laborers as they wished.62 By repeatedly [End Page 396] asserting that it was strictly forbidden, "to take anyone of this class of coloni away from the person to whom he had once been assigned or to receive such a one as a fugitive," the law also addressed a major concern of the elite we have encountered in the beginning of this paper. Read in this way as propaganda geared to elite concerns, both laws thus make a strong case that the landholding elite in the later fourth century regarded barbarian immigration as a remedy to their economic problems. Moreover, the fact that the issue appears in general legislation at all is, like the unusually broad treatment of the barbarian farmers in Themistius's Oration 16, an indication that the use of immigrants as agrarian laborers was a particularly characteristic (if surely not entirely new) concern of the later fourth and early fifth century. It is probably no mere coincidence that these texts coincide with the decades in which the social and economic dynamics described in the first part of this paper gained particular momentum.
The immediate conclusion that can be drawn from the arguments presented in this paper is that there is reason to think that the economic challenges faced by landowning elites in the Roman empire and their interest in agrarian labor power from outside of the empire played a considerable role in motivating (forced or voluntary) immigration into the empire in the fourth and earlier fifth centuries. More generally, this argument points to an aspect of the history of migrations in Late Antiquity that has been neglected in all the debates of recent decades on identity and ethnicity, on transformation vs. decline, or on confrontation vs. accommodation: the economic context and consequences of these migrations. Taking these questions more seriously has important consequences for our understanding of late antique migrations. Take, for example, the question of the causes and motivations for the immigration into the Empire. Scholarship on this question tends to focus on coercion and military pressure at home. But economic incentives may have played a larger role than is often assumed—on both sides of the border. The mass migrations of 376 or 405 and following are often attributed to Hunnic military pressure. In both cases, however, it has long been suspected that the traditional narrative of the Huns pushing forward entire populations cannot alone explain why some—and not all—Goths, Vandals, Suebi, etc. decided to make a move as radical as [End Page 397] giving up their homes rather than try to find an arrangement with the Huns.63 Hopes for economic gain, together with the knowledge of strong demand for agrarian labor power on the Roman side of the river in the fourth century and an established practice of Roman intake of laborers in that century might have had a stronger influence than is often assumed. Furthermore, with regard to the Roman side, it was not necessarily out of weakness (as has been argued) that the Roman government agreed to let in such a large group as the Tervingi in 376, or motivated merely by the prospect of gaining recruits, as Ammianus has Valens's advisers argue.64 The motivation for such an unusual move might have also been the economic interests of the landowning Roman elite—even if that particular investment proved to be a massive mistake.
This paper originated in the context of the Kollegforschergruppe 'Migration and Mobility in Late Antiquity and the Early Middle Ages' based at Tübingen University and funded by the Deutsche Forschungsgemeinschaft. I thank Paolo Tedesco and Lisa Eberle, both at Tübingen, as well as the Noel Lenski as editor and the anonymous referees of JLA for helpful remarks on an earlier version of this paper.
4. Recent contributions tend to emphasize the ideological, military (i.e. recruits for the Roman army or border defense), and fiscal interest on the Roman side, while the economic dimension and the role of the landholding elites are noted only in passing, if at all: cf. Lieu 1986, 486 ("the Romans had no coherent plan of settlement for these prisoners and did not seem to have any economic aim in their deportation beond using them as cheap farm-laborers"); Heather 1991, 158–165, at 160; Mirkovic 1993; Wirth 1997, 35–39; Whittaker and Garnsey 1998, 279–281; Heather and Moncur 2001, 26263.; Halsall 2007, 176 and 183. But see Charanis 1961, 150–51 (for Byzantium); Stallknecht 1967, 22–27, 30; Ste. Croix 1981, 248–49; Asche 1983, 112–22; Barnish 1986, 175–76 (for the fifth and sixth centuries); and, most importantly, Whittaker 1982 and 2004 as well as the magisterial survey by Modéran 2004 passim (but cf. his conclusions 390: "Les préoccupations économiques, prédominantes au temps du fondateur du principat, … semblent ainsi avoir été au IVe siècle … dépassés par le souci de renforcer les armées"). This paper builds on these studies but carries the argument further by identifying the Roman landowners as a driving force for immigration (and not just as beneficiaries of imperial foreign policy) and by situating it in the socio–economic context of the fourth century. Recently Ziche 2011 has argued that there was strong ideological resentment on the Roman side against barbarian settlements on Roman soil. This is based on some of the texts dicussed in the second part of this paper. In my reading, they imply exactly the opposite.
5. For general overviews of the new picture of the late Roman economy, see Whittaker and Garnsey 1998 for the fourth century and Ward-Perkins 2000 for the fifth and later; Chavarría and Lewit 2004; Lewit 2004; Harper 2011, 10–16; Banaji 2012; Grey 2012; Zerbini 2015 and other papers in Lavan 2015. For a reassessment of the agri deserti see the seminal article by Whittaker 1976 as well as Jaillette 1995; Grey 2007a and below n. 31.
6. The literature is vast and opinions are divided on almost all questions of detail. Thus the debate can be summarized here only in the briefest outline. The current state of the debate is best approached through the contributions to Lo Cascio 1997 and the overviews of recent scholarship in Scheidel 2000; Grey 2007b, 156–160; and Schipp 2009. For earlier scholarship, see Marcone 1988. The basis of all contemporary debates is the seminal articles by Carrié 1982 and 1983, arguing that the colonate, in the sense of a legal institution binding tenants to the soil of a landlord, is a "myth" of modern historiography and legal thought, while the limitations for coloni (which, in his view, were grossly exaggerated in scholarship) were purely a matter of fiscal registration that tied coloni to their origo. Many have supported Carrié's views in recent years, as exemplified by Grey 2007b (who also argues that colonus, etc., was an umbrella term for a variety of statuses and modes of agrarian production). But see Sarris 2011 and Lenski 2017, who strongly emphasize the limitations of coloni who were tied to the soil. Sirks 2008 steers middle ground; while he acknowledges that the colonate by the reign of Justinian had become a coherent legal institution and created a strong bond tying peasants to their place of fiscal registration, in his view the legal insistence on the free status of coloni, among other things, cautions against a picture of the coloni deteriorating into serfdom even in the sixth century. For the broad variety of agrarian labor relations that prevailed in the late Roman world see the summary in Grey 2012, 656–57. For slavery, see belown. 35. The role of wage labor has recently been underlined by Banaji 2001, ch. 8, and Sarris 2006, ch. 2–3, but see the review of Banaji 2001 by Kehoe 2003, esp. 714–6, and Freu 2013, arguing that wage labor was used mainly for specialized agricultural tasks.
7. CJ 11.53.1 (a. 371), 11.52.1 (a. 393); CTh 5.17.1 (Constantine).
8. CTh 5.17.1 (cited above); 11.1.7 (a. 361); 11.24.1 (a. 360); 11.24.6 (a. 415); CJ 11.48.8. In CTh 5.17.1 and 11.1.7 it is unclear whether the harborers of fugitive peasants had to pay the capitation tax to the state or to the former dominus, who had still been assessed for the fugitive. In CJ 11.48.8, as in CTh 11.24.1 (where the fugitives seem to be freeholders), it is explicitly the imperial treasury that receives the compensation.
9. CTh 5.17.2: Quisquis colonum iuris alieni aut sollicitatione susceperit aut occultatione celaverit. … The absconding of registered peasants because of the sollicitatio of another landowner (as opposed to sponte, "out of his own impulse") is also mentioned at CTh 5.18.1 (a. 419). In CTh 5.6.3 (a. 409): fraude aliquem abducere (as opposed to fugientem suscipere) seems to mean the same.
10. For the text (transmitted in three manuscripts of Orosius' Historiae adversus paganos and long falsely attributed to Sulpicius Severus) and a commentary see Lepelley 1989; the translation (from §2) is that of Sirks 2001. Competition over the labor of coloni (or their offspring) could also have been behind the conflict Sidonius Apollinaris had with a landowner in his neighbourhood over an abduction marriage, see his Ep. 5.19 with the reconstruction in Grey 2008.
11. For the marital law of the colonate see Munzinger 1998, 44–93. The earliest evidence for the above-mentioned aspects of the marital law for coloni in general is CTh 12.19.1 (a. 400) and 5.18.1 §3–4 (a. 419), referring to vetera constituta; with regard to imperial coloni, the rule can be traced back to the 360s, see Schmidt-Hofner 2008, 280–84.
12. For the aurum tironicum see Jones 1964, 614–19; Delmaire 1989, 321–29; Zuckerman 1998; Lenski 2002, 313–19. See Munzinger 1998, 40–42, for the elite as the driving force behind these regulations. On the Gildo affair, see CTh 7.13.13 and 14 (a. 397) and Symm. Ep. 6.58, 62, 64. On commutation into cash as a privilege to high-ranking officials, see for example CTh 7.13.15 (a. 402), 18 (a. 407), 20 (a. 410).
13. CJ 12.33.3 (around 400); 11.48.18 (a. 426): "because in that respect we look after the rights of the landlords and the honor of the state" (quia in hac parte et dominorum iuri et publicae consulimus honestati).
14. Cf. CTh 7.18 passim; CTh 7.18.2 and 4, for example, clearly concern larger elite estates under the supervision of a procurator and consisting of various tracts of land. Cf. Zuckerman 1998, 110–13 for the background to these desertions in forced conscription.
15. CTh 13.10.3 (Constantius II); CJ 11.48.7 (a. 371/375).
16. CTh 11.24.1 (a. 360).
19. Joh. Chrys. In Act. Hom. 18.4–5 (PG 60.147–48.). (I owe this reference to Rudolf Haensch.) The erection of a bathhouse by a landowner for his peasants ("to all the tillers of the land for the benefit of all") is also recorded in IGLS 4.1685 (sixth-century, from Androna in Apamene). A similar case probably in IGLS 4.1490.
21. CJ 11.48.8 (368/375). The conclusion drawn from this law might be confirmed by P.Oxy. 27: 2479, a petition from a registered farmer (georgos enapographos) who, having left his farm three years earlier, asked his former landlord to receive him back. Despite its submissive language, the point of the petition seems to be that the farmer negotiates with the landlord that he will not be forced to pay the taxes for the time of his absence. (I owe this reference to Paolo Tedesco.)
22. As, for example, in CTh 11.24.2 (a. 368), or 11.24.6 (a. 415).
23. Aug. Ep. 20*.10 (CSEL 88: 100). In the course of the affair (on which see the lucid reconstruction by Frend 1993, as well as Dossey 2010, 137–41) the peasantry several times resisted attempts to exploit them and actively intervened in their own interest (see especially §14, 17, 19).
25. Banaji 2001, ch. 3. For the connection with the villa culture of the time Sarris 2006, 121–26; Banaji 2012, 597–607. But contrast Vera 2012, 117, citing further critics of the idea, and belown. 29.
28. Evidence for the wealth of decurions is collected in Lepelley 1979–1981 (for fourth- and early fifth-century Northern Africa) and Laniado 2002 (up to the sixth century). For the fate of the curial order and the transformation of the governing class in the cities of the empire in the fourth and fifth centuries, see Liebeschuetz 2001 and the alternative model proposed by Schmidt-Hofner 2014.
29. Such is the conclusion of the general survey of landholding patterns in Oriens by Decker 2009, ch. 2, esp. 65–79. For the fourth-century evidence from the Hermopolite land lists and from Oxyrhxnchus, both in Upper Egypt, see Sarris 2006, 179–81. But as late as the sixth century a place like Aphrodito in Egypt seems to have been the home of many small- and medium-sized landholdings, see Wickham 2005, 411–419; cf. also Bransbourg 2016, summing up the debate at 308 and 402. The same is suggested by the large and carefully constructed houses and churches in the rich villages of northern Syria (mainly fifth to sixth centuries) as described by Tchalenko 1953–1958 and Tate 1992. See also Foss 1995, and the summary by Tate 1997 of his findings and historical conclusions, as well as Wickham 2005, 443–57. By way of example, see Theod. HR 17.3, a tale of a rich village in the Lebanon where only free smallholders lived; or by the world of small-scale perpetual leaseholders on an estate owned by a curial as apparent from the North African legal documents from the 490s known as the Tablettes Albertini, Weßel 2003, 30–44. The fourth-century census inscriptions from the Aegean and Asia Minor show a highly fragmented landownership including small and medium properties but also evidence for elite accumulation and the dominant role of the honorati, Harper 2008, 90–7 (see also 84–90 for their date); cf. Thonemann 2007, 472–75. But see also the warnings against underestimating the rise of the large aristocratic estates in the East voiced by Sarris 2004 and 2015. A balanced position is taken by Wickham 2005, 245; Decker 2009; Banaji 2012, 607–10.
30. CTh 12.3.1 (a. 386); cf. 12.3.2 (a. 423).
31. CTh 5.14.33. Parallels at CTh 5.15.15, 16, 21, CJ 11.71.5. The high social rank of leaseholders is explicit only in CTh 5.15.15, but generally likely as leases were given to the highest bidders. Cf. Whittaker and Garnsey 1998, 283. A similar practice can be observed elsewhere: As recently argued by Chouquer 2014, 132–33, the fiscal mechanism of adiectio sterilium (by which the tax obligation of uncultivated land was allocated to other possessions) was not necessarily a coercive instrument by which the Roman state secured its tax revenue; it often served the interest of the landowner, as the allocation apparently went along with substantial tax immunities. See CTh 5.15.14 = CJ 11.59.3 (a. 364). The abundance of laws concerning this mechanism, on which in general see Choquer 2014, 126–34, can therefore be interpreted as evidence for another way land-holding elites tried to gain control over additional land.
32. Cf. e.g. the provisions in CTh 4.22, 4.18.1, 2.26.5.
34. De Regno 20.2; cf. 21.1: "It is those from whom we get our slaves everywhere, because of their lack of land." For trade with barbarian slaves in Late Antiquity, see Harper 2011, 83–99; Lenski 2011, 193–94.
35. The case that slavery continued to play an important role has been stated most forcefully in recent years by Harper 2011 who goes so far as to claim that the Late Roman Empire continued to be a "slave society"; cf. Grey 2011b with a succinct survey of the scholarly debate and the most important contributions. But cf. Lenski 2017, questioning Harper's view of a "slave society" on the basis of a systematic survey of the North African evidence that shows a preponderance of coloni as agrarian laborers. A similar picture for Italy and elsewhere is drawn by Vera 2012 who argues that the evidence for large late Roman landholdings worked by slaves is limited to a few cases and points out that the use of slaves in agriculture in general made little difference as they worked under the same economic regime like coloni, i.e. as tenants paying rent to a landowner. Evidence for agrarian slavery is collected by Harper 2011, ch. 4; for the Aegean evidence, coming from the later fourth-century census inscriptions, see Harper 2008. A famous example for slaves working elite possessions from Italy is V. Melaniae Lat. 10.1 and 18.4. Barbarian slaves in agriculture: one case in point is Radagaesus's people, if admittedly under special circumstances: cf. n. 49. For Sozomen9.5.5–7 see below n. 39.
36. Pan. Lat. 8(5).8.4–9.3, trans. Nixon and Rodgers 1994, 121–22. For the historical background see Nixon and Rodgers 1994, 104–8 and 212–13, n. 28. Annona is often understood as (lowered) "price for food" (as a consequence of increased productivity through the additional workforce), but the word normally means "tax" in late Roman official language, and it is much more likely that the landholding elites rejoiced at lowered tax pressures (because of the enlarged number of taxpayers and/or the increased surplus their own workers would produce) than at a diminution of the profit they could realize from their products.
39. Soz. 9.5.5–7; translation Hartranft 1983, 422. The translation of Sozomen's passage reads, "Some of them were, therefore, sold at a low price; while others were given away as slaves for presents (τοὺς μὲν ἐπ' ὀλίγοις τιμήμασιν ἀπέδοντο, τοὺς δὲ πολλοῖς προῖκα δουλεύειν παρέδοσαν), upon condition that they should never be permitted to return to Constantinople, or to Europe, but be separated by the sea from the places familiar to them. Of these, a number was left unsold (ἄπρατον); and they were ordered to settle in different places." It is impossible to tell what status is implied for those left unsold. "Sold at a low price" and "given away as slaves for presents" sounds as if some of the Sciri were treated as slaves but could perhaps also refer to coloni, as indicated by the terms described in Theodosius's law. The vague phrasing might be a consequence of the blurred perception (and the secondary importance) of the subtleties of such status distinctions in everyday life.
40. CTh 13.11.10.
42. For the recipient, see PLRE 2: 760–61 Messala 3. Gothofredus 1736–1743, vol. 5, 148–49, ad CTh 13.11.9 (sic!) connects the law with Claud. Eutr. I, 377–383, mentioning negotiations of Suebi, Chauci, and Sugambri—the latter two being Frankish—who had "sued for peace" with Stilicho; he gave them leges et iura or reges and foedera and drew some of them as recruits into the army. The text, however, mentions no agrarian settlement.
43. Not. Dig. Occ. 42.33–44. gives a list of praefecti laetorum (and therefore laeti settlements) in Gaul, Italy, and the Iberian peninsula ordered by civitates; for the laeti in the civitas Lingonum it says that they are per diversa (sc. loca) dispersi (42.37).
44. Szidat 1995, building on Ste. Croix 1981, 248 n. 33. The phrasing of the surviving excerpt of the constitution is ambiguous as to whether the possessors of the terrae laeticae were laeti or Roman landlords. Even though the whole affair very much looks like a conflict between members of the Roman elite, it is, therefore, in principle possible that the ultimate source of the struggle and the machinations mentioned in the law arose from competition for land among laeti (presumably the more eminent persons among them) rather than among Roman landlords. And one could also speculate that the laeti, instead of being the tenants of Roman landlords, received the land as free-holders or as tenants of imperial domains who were directly subordinate to an imperial procurator or a similar official. Whatever the circumstances, in the end the (Roman) municipal elite must have had something to gain from colluding with the laeti and supporting them in their dealings with the imperial administration. It seems implausible that without their interference newly immigrated people successfully solicited an imperial rescript in their favor or, on the other hand, launched a protest against such a maneuver that would result in the present constitution. And the gain the Roman elite hoped for can hardly have been anything other than getting access to land or labor, however that functioned in practice. In one way or another, therefore, one comes to the abovementioned conclusion that the text attests to a competition among the Roman elite. Most recently, Janßen 2004, 113, proposed that, through giving larger estates to some of the laeti (as freeholders), the municipal elite wished to shift fiscal burdens on them or limit the number of newcomers. This is, however, already made impossible by the wording of the law, which leaves no room for doubt that at issue were claims to larger shares rather than the compulsion to take them: cf. amplius quam meruerant occuparunt; inprobe ab aliquibus occupata.
45. SHA Claud. 9.4. A passage in the letter to the Senate of Rome, SHA Prob. 15.2 (omnes iam barbari vobis arant, vobis iam serviunt et contra interiores gentes militant), does not refer to barbarian captives on Roman soil, as some take it, but to client states beyond the frontier, as is made clear by the last part of the sentence.
46. SHA Aurel. 48.2: Statuerat igitur dominis locorum incultorum, qui tamen vellent, *gratia* / *gratis* dare atque illic familias captivas constituere. Gratia, as given in one group of manuscripts, is grammatically impossible. The Loeb edition, as others, replaces it with pretia, implying that Aurelian bought the lands before giving them to the barbarians. Hartke 1951, 279–81, emends gratia to gratiam, arguing that Aurelian offered the landowners an abatement for the tax obligation but confiscated the land. Using the same emendation, Paschoud 2002, 216 holds that Aurelian freed the landowners from their tax obligation provided they accepted the settlers on their land; Modéran 2004, 364 agrees. Hohl, in his first Teubner edition (slightly altered in the later one), stuck to the alternative manuscript tradition gratis dare and emended qui tamen to quid tamen (cf. Hohl 1913, 407). By this reading, the landowners would be given land for free and receive the bar-barians to cultivate it. In these latter interpretations the question remains what use the landowners had from the barbarian coloni, as the wine they produced had to be delivered to the city of Rome: ut nihil redituum fiscus acciperet sed totum p.R. concederet. The system envisaged could have been that the landowner retained a share of the wine for independent commercialization.
47. Vivos omnes circa Mutinam Regiumque et Parmam Italica oppida rura culturos exterminavit: Amm. 31.9.4.
49. Oros. 7.37.17.
50. As in CTh 10.12.2 §2–3 (a. 368? 370? 373?); CTh 11.7.2 (a. 319); CJ 11.48.12 (a. 400); Sidon. Ep. 5.19; cf. also Modéran 2004, 369–70. On barbarian tributarii see Amm. 19.11.6 (status proposed fraudulently by the Limigantes in 357 in order to win the favor of the Roman responsibles) and 28.5.15 (Alamanni in the Po valley).
53. Them. Or. 16.211a–b and 212 a–b.
56. For the much-debated terms of the Gothic peace in 382 see Heather 1991, 158–165; Schulz 1993, 65–78, 178–79; Halsall 2007, 180–85. For contemporary critique and the tensions between Theodosius and the eastern elite with regard to the peace, see Errington 1996, 14–15: "It is clear from the unusual vehemence with which Themistius expresses his arguments that opposition was anticipated among the listeners"; "Theodosius had trouble with hard-liners in the Constantinopolitan Senate." Compare the contemporary episodes of anti-Gothic sentiment mentioned in Zos. 4.40, 4.26; Amm. 31.16.8. Indirect evidence for elite resentment against Theodosius's policy comes from Them. Or. 15, held in January 381; the text is a desperate attempt to mask the absence of military victory over the Goths with praise of Theodosius's achievements in the internal administration of the empire and repeated invitations to cooperation. Cf. Leppin 2003, 51–69 for the wider context of Theodosius's uneasy relationship with the eastern elite in these years.
57. Pan. Lat. 2(12).22.3, translation Nixon and Rodgers 1994, 473. See also Pan. Lat. 2(12).32.3, mentioning that the newly settled Goths exempted the provincials from furnishing recruits (against Maximus). A potential parallel can be found in Them. Or. 34.22 (ca. 384/385), with Heather and Moncur 2001, 304–10 for date and context and 327 n. 155 for a textual problem in this passage. Penella 2000, 34, tentatively suggests that some passages in Them. Or. 30, a (progymnasmatic?) encomium of agriculture might allude to the Visigothic treaty of 382 and that the whole piece may therefore have a political dimension; but this remains very speculative. I thank David Pitz for pointing me to this text.
59. As one subset of the soldiery, barbarian settlers are briefly mentioned in CTh 7.13.16; 7.20.12; 7.15.1; Nov. Val. 9 and Nov Theod. 24. CTh 7.20.10 mentions a praepositus laetorum among other officers. CTh 11.30.62 appeals by gentiles or their praefecti. CTh 3.14.1 interdicts marriages between gentiles and Romans. Nov. Sev. 2 mentions laeti among other corpora publicis obsequiis deputata.
60. Schmidt-Hofner forthcoming.
61. See PLRE 2: 93–95 Anthemius 1. For a contemporary assessment of his position see Soc. 7.1.1.
62. The phrase is sometimes understood as referring to a crop shortfall in 409 in the eastern empire and linked with sources attesting an increase in the price of foodstuffs at Constantinople in that year and to problems with the army supply in Palestine, see Lippold 1973, 964–65; Stathakopoulos 2004, 223. But it is by no means certain that these irregularities attest to a large-scale crop failure and food shortage; in fact both Lippold and Stathakopoulos adduce evidence suggesting that these irregularities had no natural cause but came from problems with shipping from Egypt. Not least in case of a massive crop shortfall, it would be astonishing if the landowners were given complete freedom as to where to employ the Sciri farm workers. Pro rei frumentariae angustiis is, therefore, better taken as a reference to more general conditions.
63. For example, Kulikowski 2007, 127–28 points to the fact that the Huns appear on the Danube only more than a decade later and concludes that Hunnic pressure must have built up much more gradually (see already Heather 1991, 135–42). Batty 2007 underlines that immigration from regions beyond the Danube was a constant feature of the history of the area throughout Roman times and also emphasizes the economic motivations of the migrants. Sarris 2015 posits that the primary aim of the Gothic immigrants of the 370s was to continue their accustomed lifestyle in a vibrant sedentary agriculturalist economy within the Empire.