Part I
Migration theory and historiography
1 Mobility in the Roman world
New concepts, new perspectives
Introduction
Moses Finley’s model of face-to-face societies prevailed until the end of the 1980s, during which time mobility was considered as quantitatively marginal, due to the costs and deficiency of transportation. Scholars had, of course, studied migrations such as Greek colonisation during the eighth–seventh centuries or the so-called barbarian invasions, but these were considered exceptional, and migration was defined only as a long-distance, collective, and uprooting movement. These two conceptions have vanished: the history of migration is written based on a long-term perspective, and the concept of migration has been vastly expanded. To follow up on our examples, Greek colonisation is now seen as the continuity of migration from Late Bronze Age mobility, while late barbarian migrations are considered particular episodes in the long history of voluntary and forced settlement into Roman territory attested since the second century bce.
A seminal article by Robin Osborne was crucial in renewing the debate at the beginning of the 1990s.1 Using archaeological data and a comparison with mediaeval and modern England, he called upon historians to better estimate the degree of human mobility in ancient societies. While he acknowledged the difficulty in interpreting some sources, he concluded in a very new way, ‘Even in a community whose overall population level is stable and which is not suffering from any major internal change or reorganisation, it is quite likely that half the population will have arrived within the last ten years and that half the population will depart within ten years’.2
The article was based on two premises. First, the necessity to historicise causes and forms of migration: the same social structures do not produce the same effects. For example, he proposed that while, as we might expect, Russian serfs did not move very much in the nineteenth century – and this was probably the case of the Spartan helots too – our sources clearly show that the mediaeval serfs were far from being immobile, and this can be said too of Greek and Roman slaves involved in trade. Conversely, the landowners, deemed to be immobile, move in some contexts – especially according to the land market – while the workers who do not own land and whom we would expect to be mobile, may in the same context remain immobile. The second observation concerns the relative nature of distance. Distance must be defined in relation not only to the means of transportation but also to spatial representations: what we call local migration could, at certain moments, appear as really uprooting.3 In this context, mental maps are as important as real distances.4 This also explains why distance cannot be the only criterion of the distinction between local, interregional, or international mobility.
Recent historiography has rather confirmed Osborne’s propositions, and several studies suggest that pre-modern towns, cities, and rural zones could not compensate for their structural mortality without immigration.5 In the pre-modern Mediterranean, human mobility was therefore a structural social principle. This is a vision that the ancients themselves would have acknowledged, as can be shown by the memory of archaic migrations in the ancient collective consciousness6 and, above all, by the importance that travel narratives had in literary texts such as the epic of Odysseus, that of Aeneas, the trip of the Egyptian Ounamon in the eleventh century bce and by the numerous legends of foundation – until Stoic philosophers and Christians hypostatised the figure of the migrant.
Today, nobody would seriously deny the role of migration in the making of the Roman Empire, as well as the role of the Empire in developing mobility.7 The methods used in this field of research are highly diversified. Even if the results are still debated,8 archaeological and osteological evidence have provided new data on the movements of people or on mixed marriages;9 epigraphic sources offer many mobility stories,10 and legal sources show how Roman law progressively took into account the movement of people. A large number of new sources have emerged and have been explored from various points of view: demography,11 technology,12 cultural contacts,13 economy, and migration policies, to which I shall return. For these reasons, we may well say that classical scholarship has come to a real migratory turn.
The idea of the persistence of migration in the ancient Mediterranean does not, however, presuppose any linearity or continuity of the practices. First, important ruptures can be identified that undoubtedly had effects on human mobility. Some ruptures were ‘structural’, like the unification of Italy in the first century bce, or the Pax Romana; others were circumstantial, like wars, civil conflicts, diseases, natural catastrophes, and so on. Second, mobility did not have the same impact on all Mediterranean regions.14 Third, behind the unifying vocabulary, realities concerning the forms of mobility or the status of migrants varied greatly. Consider the word ‘emigration’ and the various experiences it refers to – colonisation of settlement, deportation, flight, exiles, and the like – or the word ‘migrants’, too often wrongly conflated with foreigners to ‘foreigners’: this word makes sense in our modern societies where the national finds its counterpart in the foreign, but it occludes the multiplicity of the conditions of people who moved in the Roman world (incolae, hospites, advenae, inquilini, qui commorantur, etc.).15
New concepts
I would like first to present three conceptual shifts that have contributed to this new development in mobility studies and then to reflect on some perspectives they opened. These shifts are due to the influence of social sciences and of migration studies and to world and global history: these approaches bring greater attention to the history of connections within the global community16 and reject any master narrative, any unilateral and Eurocentric system of interpretation, challenging Braudel’s and Wallerstein’s models of world economy.17
The first conceptual shift is the inclusion of migration in a broader concept, that of mobility. While migration previously referred to the idea of crossing borders and uprooted experience, now historians study migration as a social process: as the historian Dick Hoerder wrote, migration starts with leaving one’s family and ends in a transnational process.18 In this perspective, causes and forms of mobility must be linked to an analysis of social structures, not only to economic and ecological fluctuations: by way of example, one could point to the relationship between geographical and social mobility in the Roman Empire and the impact of family structures on the difference between male and female mobility,19 or to chain migrations.20 This shift has led historians to extend their research in various directions. Studies of mobility now encompass a greater number of categories – soldiers and elite but also craftsmen, merchants, slaves, students, peasants, etc. More push and pull factors – not only natural resources or labour shortage, but also social promotion or education – are taken into account. All forms of mobility – not only departure but also returns or pendulum movement – are also now looked at.
The second shift is the stress on the movement itself and on the flow of people, not only on the place of arrival. Mobility is no longer considered just as an event that splits the life of migrants in two clearly defined phases but as a durable process. Therefore, migrants are viewed as actors who, in a way, create their own territory throughout their peregrinations, within their own networks. This shift allows us to acknowledge the existence of individual logics and of a real culture of mobility, acquired through spatial experiences and composed of technical, legal, geographical knowledge, as shown in the Periplus of the Erythrean Sea. This concept of the culture of mobility can be applied to several cases: craftsmen, athletes, sophists, doctors, and merchants. Recent scholarship has shown how this latter group has been able to create and share a common legal and commercial culture.21
A third shift is the study of the structuration of movement, that is, the organisation of people who moved. Hence the attention to the so-called transnational communities and diasporas: this last term has lost its traumatic meaning and is now taken only to describe, according to the definition of Robin Cohen and Will Saffran,22 dispersed members of a community who, for some time, maintain links – real or through shared memories – with their place of origin before transforming themselves in their new place. The concept of diaspora has several productive effects. A number of people who move can be included in this category: colonial and imperial administrators, merchants, artists, traders – according to the model proposed by the anthropologist Abner Cohen and applied by Philip Curtis to slave traders in the Atlantic world.23 The migrants are not, then, studied only as minorities within the society into which they have migrated but as members of transnational networks: the link between identity and territory is therefore analysed in a very different way and reveals that civic identities were more open than previously thought. Finally, the category of diaspora emphasises the migrants’ capacity of organisation.
This focus on organisational patterns has had several consequences: it has led historians to emphasise the role of formal and informal associations in the process of mobility. Associations provided migrants with structures of hospitality, bases for social life, and even primary places of arbitration in cases of conflict within the migrant group or between the group and the authorities.24 These institutions formed compartmentalised networks and contribute to the understanding of connectivity as a segmental phenomenon.
The fact that a situation of mobility involves rigid structures also allows us to consider sedentary practices and mobility jointly. This is another area to which the most recent wo...