I A RISING TIDE

The last quarter-century has witnessed a transformation in approaches to the study of mobility in the Roman Empire. Social scientific frameworks, especially network analysis and the concept of connectivity, have granted explanatory force to population movements.Footnote ¹ Technological advances in bioarchaeology, archaeobotany, archaeozoology and related fields provide ever finer-grained views of the role of migration in interactions between people and their environments.Footnote ² The digital turn has produced new tools for research and visual representations, such as Walter Scheidel's web-based ORBIS: The Stanford Geospatial Network Model of the Roman World, which allows users to simulate communication costs and routes while controlling for complex variables like season and manner of travel.Footnote ³ Finally, increased use of non-Roman and non-ancient comparanda has allowed for productive theorisation about poorly attested mobilities.Footnote ⁴

Equally important, though less discussed, is the impact of postcolonial studies on the politics of disciplines concerned with the ancient Mediterranean. Postcolonialism as an analytic began to take hold in the 1980s and 1990s, when scholars such as Edward Said, Gayatri Spivak and Malek Aloulla focused on the British and French empires and made a case for rethinking how knowledge is made, organised and evaluated.Footnote ⁵ In the discipline of Roman history, researchers from historically excluded groups have made two demands that bear on the present and the study of the past: that the discipline adopt approaches that respect marginalised subjects, and that it become a more inclusive community with regard to the scholars therein. These demands are indivisible, since the racialisation of groups in the present-day is partly rooted in interpretations of Greek and Roman taxonomies of power and domination.Footnote ⁶ The present generation of Roman historians includes those who fruitfully draw on intersectionality, cultural appropriation and related frameworks, and lead conversations about the geographic and cultural myopias that accompany traditional definitions of ‘classics’ and ‘the classical’.Footnote ⁷ Consequently, research on mobility in the Roman world is shifting its focus on Roman citizens and elites to include non-Roman, non-elite, coerced and women actors.Footnote ⁸ Borderlands and zones of contact increasingly command attention and replace ideas of stark, spatial binaries. In addition, sustained interest in the Black Sea, Indian Ocean and Atlantic Ocean suggests the possibility of unsettling the traditional dominance of the Mediterranean.Footnote ⁹ The margins have moved a little closer to the centre.

The implications of the developments outlined above are multi-dimensional. A region once viewed as static is now seen to be a dynamic world in motion. At the same time, by paying more attention to systems of power in the Roman world, historians have developed a more nuanced appreciation of relationships between mobility, identity and agency and produced better knowledge about the empire's connections with distant regions. The role of mobility in a wide range of social, cultural, political and economic phenomena has become a large, exciting and important area of study.Footnote ¹⁰

The volumes reviewed in Section II reflect many of these methodological and political trajectories. With The Boundless Sea: Writing Mediterranean History (2019), Peregrine Horden and Nicholas Purcell respond to two decades of discussion about their The Corrupting Sea (Reference Horden and Purcell2000), which transformed the study of mobility in the premodern Mediterranean. Kasper Grønlund Evers, in Worlds Apart Trading Together: The Organisation of Long-Distance Trade between Rome and India in Antiquity (2017), and Federico de Romanis, in The Indo-Roman Pepper Trade and the Muziris Papyrus (Reference de Romanis2020), examine trade between the Mediterranean and Indian Ocean worlds. And throughout Migration, Mobility and Language Contact in and around the Ancient Mediterranean (Reference de Romanis2020), edited by James Clackson, Patrick James, Katherine McDonald, Livia Tagliapietra and Nicholas Zair, contributors interrogate the relationship between population movements and language use in the Roman world and, notably, the peoples of pre-Roman Italy.

Each of these publications shows that the tide of interest in Roman mobility continues to rise. This poses the eternal question of what comes next. Of particular interest to me is the prominence of the concept of connectivity, which Horden and Purcell introduced in The Corrupting Sea. As Section III observes, studies on mobility in the Roman world tend to implicitly presume that connectivity was accessible, desirable and beneficial to everyone in the Mediterranean. This position, as Section IV explores, obscures the effects of Roman imperialism. Although the rise of Rome was mediated by Mediterranean connectivity and in turn generated novel connections and networks, it also transpired through the enactment of countless separations and detachments. These disconnections, so to speak, remain comparatively unexamined. As Section V articulates, our neglect is partly due to our own ease of movement in the Mediterranean. It hides the absence of such ease for others in the past and today, and leads Roman historians chronically to underplay the region's connections to places remote from it. In light of the Mediterranean's imperial histories, we must remain vigilant in how we define and characterise it and equally aware of the role that our relationships to the region play in debates about it. To quote Michael Herzfeld: ‘To say that the Mediterranean “does not exist” is as silly as to argue that facts themselves “do not exist”. They exist in the sense that they are representations of something experienced in the phenomenal world.’Footnote ¹¹ We have arrived at yet another moment in which we should ask: which Mediterranean do we choose to bring into being?

II REVIEWS

In The Corrupting Sea, Horden and Purcell presented a new vision of the premodern Mediterranean. They argued that its regional coherence was due to a distinctive regime of risk, logic of production, topographical fragmentation and connectivity. The Corrupting Sea (henceforth CS) was the self-described successor of Fernand Braudel, who in 1949 argued that the sixteenth-century Mediterranean possessed a ‘unity and coherence’ that allowed historians to write longue durée histories of the region in addition to histories of individuals, peoples and events within it.Footnote ¹² Of the elements that make up Horden and Purcell's fourfold model, connectivity has been the most influential for how researchers think about the Mediterranean. The authors imported the concept from locational analysis to explain the various ways in which its microregions cohered internally and with each other.Footnote ¹³ They further argued that the degree to which connectivity characterised the region made it distinct from any other comparable area. Use of the concept now ranges from the strict sense that Horden and Purcell originally intended to a generalising descriptor for connectedness.Footnote ¹⁴ As Brent Shaw observed, CS is ‘one of those manifest watersheds in the study of antiquity’.Footnote ¹⁵

The Boundless Sea (henceforth BS) is not the sequel that CS promised, but a response to the discussion and debate it provoked. It consists of a preface and twelve essays; eleven are reprints with light revisions, and all are listed as co-authored, including those that did not originate so. Broadly speaking, the essays fall into one of three kinds of response. Some, like chs 2 and 3, are part of long-term conversations between the authors and their critics, including those who argue that they veer into Mediterranean exceptionalism with their claims of the sea's distinctiveness. Their reply essentially amounts to an agreement to disagree about whether they present ‘the Mediterranean as the product of, or even as too closely associated with, the hegemonial systems of its history’ (46). The second set of essays considers ways to integrate the Mediterranean into global histories. In ch. 1, the authors describe the relatively new subfield of the history of seas and oceans, which they term ‘the new thalassology’. In this context, they consider ways to define the Mediterranean and contrast it with other bodies of water. They conclude by insisting that the premodern Mediterranean was indeed singular. Subsequently, chs 8, 11 and 12 compare the Mediterranean with northern Europe, the Sahara and Asia. The third kind of response treats the problem of conjoining cultural, social, economic and political histories with approaches to Mediterranean ecologies and logics of communication. Thus, ch. 10 examines periods in which political conditions put people to sea, as when Rome's conquest of the Mediterranean prompted the emigration of tens of thousands of Roman and Italian negotiatores, or businessmen, from Italy. Rightly, the authors argue that this population (which Sections III and IV discuss further) ‘articulate[s] a world of connectivity which maps onto what we call the Mediterranean’ (174).

Because the essays that constitute BS appear to be randomly ordered, the volume lacks a strong narrative arc. In addition, the absence of maps, charts and diagrams hinders the explanatory potential of its prose. Even so, it is a rich resource. BS contains a plethora of bibliographies and notes, and its case studies are chronologically, geographically and culturally diverse, encompassing the Roman and the Greek, the pagan and the Christian, and the ancient, medieval and early modern. It also offers succinct outlines of its predecessor's primary arguments and good-faith overviews of the debates CS prompted. Consequently, BS provides a concise introduction to their project and its influence. As importantly, the authors take pains to speak directly to critics like Brent Shaw and Michael Herzfeld.Footnote ¹⁶ The field needs more public dialogue about the politics of individual scholars’ work, and Horden and Purcell's articulation of these disagreements in print helps to normalise a reflexive disciplinary culture.

Where BS is primarily concerned with the Mediterranean, the monographs of Federico de Romanis and Kasper Evers draw attention to its connections with the Indian Ocean world. Both argue that the Mediterranean's internal connectivity did not prevent the intensive circulation of people and goods between the Roman world and the Indian subcontinent.Footnote ¹⁷ Harnessing the heuristic potential of connectivity, they elucidate links between the Mediterranean Sea and Indian Ocean. In doing so, they demonstrate how much researchers can reveal of the Roman world's wider connections when they decentre Rome. Simultaneously, they add to a growing volume of work on Indo-Roman links.Footnote ¹⁸

De Romanis's monograph centres on the fragmentary Muziris Papyrus (P.Vindob. G 40822). An informative and challenging document, the papyrus's recto text details a loan contract, signed in Alexandria, between a merchant and a financier for a commercial operation to Muziris, which has been tentatively located in or near Pattanam in South India. Its verso text, which is written in a different hand, records the customs duties levied on Indian cargo imported by the Hermapollon. The papyrus is a valuable source of information for the structure and organisation of trade between Rome and the Indian subcontinent. Using ancient, medieval and even early modern evidence, de Romanis reconstructs its logistical, economic and anthropological contexts and locates his interpretation of its texts therein.

The volume's thirteen chapters are divided into four parts. Part I details the challenges of navigating the Red Sea corridor; the geography and nature of pepper production in the Indian subcontinent; and the ancient literary sources most useful for interpreting the surviving parts of the papyrus's texts and reconstructing its missing sections. Part II focuses on the papyrus's recto text by exploring when the loan was made and for how long; what the loan was to cover; and the stipulations overseeing its repayment. Part III unravels the logics of the verso text's assessment of the Hermapollon's cargo by determining routine procedures for maritime loan agreements in the Greek and Roman worlds. The same section of the book contrasts the cargo and cargo-carrying capacity of the Hermapollon with ancient and sixteenth-century comparanda. Finally, Part IV articulates the stipulations that governed customs duties on Indian commodities as they entered the Mediterranean. It also contextualises the professions and identities of the loan's co-signatories.

De Romanis's contributions are wide-ranging (so much so that he himself provides a numbered list of conclusions in the introduction). I emphasise four. The first is the elucidation of the timeline for the repayment of the loan and the assessment of the Hermapollon's cargo. By determining routine procedures for maritime loan agreements in other Mediterranean contexts, he shows how lenders, borrowers and traders adapted their practices to trade with the Indian subcontinent. Second, he provides a thorough and updated account of the geography of pepper farming, and the gatherers, traders and kings involved in its production and export. In doing so, he dismantles two related arguments: that there was little systematic control of Indian emporia by local governments and that the Indian contribution to exchanges and transfers of commodities with the Roman world did not reach a level of complexity that merits categorisation as trade.Footnote ¹⁹ Third, de Romanis offers a rare, detailed examination of the contingency of premodern mobility. In ‘Bridging Disconnected Seas’, he stresses that Mediterranean connectivity did not preclude relationships with distant places. Then, with intricate detail and maps, he shows the interplay of environmental, meteorological and human factors that impacted mobility through the Red Sea corridor, which was crucial to movement between the Roman world and Indian subcontinent. Finally, the monograph as a whole brings the picture of Indo-Mediterranean trade in the Roman period to a level of detail that enables in-depth comparison with trade between these regions in later periods. De Romanis has opened a door to even more ambitious longue durée studies.

Evers's Worlds Apart complements de Romanis's monograph with its focus on the role of trade and social networks in the systems of production, distribution and consumption that connected the Roman world and the Indian subcontinent. The volume draws on a range of evidence from 30 b.c.e. to the end of the sixth century c.e., a period that spans Augustus’ annexation of Egypt and the abandonment of the harbour of Berenike and eventually Egypt itself. The broad geographic and chronological scope of the study is punctuated by case studies.

The first two chapters of Worlds Apart summarise the history of the study of Indo-Roman trade and the location of Evers's monograph therein. Ch. 1 outlines the politics of the book and their implications for his conclusions. Evers rejects studies that present Rome and actors from its sphere of influence as dominant in exchanges with the Indian Ocean world. He points out that Romanocentric approaches express colonialist ideals that ascribe an arbitrary hierarchy of value to peoples and cultures. In ch. 2, he sets aside New Institutional Economics to favour a bottom-up approach that leads him to focus on demand in Rome and the provinces for products from the Indian subcontinent, and the private associations (collegia, synodoi and so on) that participated in the import of Indian commodities. Chs 3–5 focus on networks formed by private associations of craftspeople and traders in Italy and the empire's provinces. Evers argues that such associations participated directly in long-distance, reciprocal exchanges with the Indian Ocean world by acting as a bridge between craftsmen and traders, and by processing goods from the East at Egypt's Red Sea ports. Chs 6–8 turn an eye eastward. Evers details the diverse associations and individuals in Egypt that handled the receipt, taxation and westward shipment of goods from the Indian subcontinent, as well as the operation of Roman, Egyptian, Axumite and Indian networks from Southern Italy to India. The volume concludes by showing the involvement of Indian associations of merchants and associations of craftsmen (nigama and shreni, respectively) in the production and shipment of goods to the Roman world.

As a whole, Evers seeks to demonstrate the mechanisms that Roman-world associations used to augment group cohesion, enforce contracts with strangers and maintain long-distance networks. In doing so, he categorises some associations and networks as diasporas. His use of this term merits some attention. Evers does not explicate whether or how shared ethnicity or origins differentiated the behaviour of diaspora groups from that of other communities. Yet the designation of a population as a diaspora has limited heuristic value if the population's migration from a shared place of origin is the sole ground for doing so: persecution and other forms of oppression have long been important to distinguishing the behaviours of diasporas from other migrant groups.Footnote ²⁰ A clear elaboration of Evers's use of the term — even one that rejects the significance of oppression for identifying a group as a diaspora — would have strengthened the theoretical underpinnings of the volume.Footnote ²¹ Ultimately, however, the absence of a definition does not detract from Evers's contributions. He uncovers the integral role of non-state institutions in long-distance trade networks; the impact of those networks on stylistic preferences in Italy; and the availability of items from India, like pepper and perfume, to non-elite consumers in remote parts of the Roman Empire. In addition, his discussion of Indian associations will be welcome to those who study associations of the Roman world that were involved in trade. Moreover, maps at the end of the book offer helpful visual guidance for a study of such scope.

In contrast with the broad chronological and geographic scope of Evers and de Romanis, the essays in Migration, Mobility and Language Contact in and around the Ancient Mediterranean, edited by James Clackson, Patrick James, Katherine McDonald, Livia Tagliapietra and Nicholas Zair, are microhistorical and mainly concerned with Mediterranean contexts. The volume aims to demonstrate how linguistic data can fill out the picture of migration in pre-Roman Italy and the Roman Empire. This evidence, the editors point out, has received far less attention than archaeological, osteological and epigraphic material. Yet it can reveal much about the effects of individual and population movement on language use and, in turn, what language use reveals about mobility (4).

The book's introduction (ch. 1) and ten essays assess evidence for pre-Roman Italy and the Roman world. In ch. 2, for example, Daniele Maras marshals epigraphic and literary evidence to explore scenarios in which interethnic mobility and interaction in Italy led to the appearance of non-Etruscan components, like roots and full names, in Etruscan nomenclature. In ch. 3, Elena Isayev reviews the widespread migration and settlement of Romans and Italians in the last two centuries b.c.e. to underscore the routine nature of long-distance mobility in the early Roman Empire and the involvement of Romans and Italians in it. She juxtaposes this evidence alongside the Plautine corpus to demonstrate the playwright's creative responses to the new mobilities that emerged in his lifetime. In ch. 8, Francesco Rovai contextualises a collection of 26,000–27,000 seals from Delos's House of Seals in the island's cultural and linguistic environment and its connections to communities further afield. On the basis of these seals, he shows the wide-ranging linguistic repertoire of the clientele — including Romans’ and Italians’ preference for Latin for self-identification — and ties it and the ‘cosmopolitanism’ (177) of this clientele to the extraordinary, unprecedented boom of immigration to and economic activity on Late Hellenistic Delos. In ch. 10, Patrick James draws on linguistic evidence to assess why Romans and Italians journeyed to Egypt before its annexation by Rome. He argues that these travellers were likely to be engaged in overlapping sacred, military and commercial activity, and emphasises their commercial ties beyond Egypt. As on Delos, the use of Latin by Romans overseas materialises as a strategy for self-differentiation. The author takes as his point of departure an intriguing graffito from Philae which dates to 116 b.c.e. and represents the earliest Latin inscription from Egypt. Its author, one Gaius Acutius, scrawled on a Ptolemaic structure at the sanctuary of Isis that he ‘arrived here first’: hoc venit primus.Footnote ²²

Like BS, this volume falls short when it comes to maps. It contains just one, that of Isayev in ch. 3, which shows long-distance journeys by characters in Curculio, Persa and Poenulus (65). Maps would have vastly improved the legibility of arguments already highly technical in their analysis. But apart from this oversight, this is an exceptional collection of work. By examining the effects of population movement on language, it furnishes new data to historical linguistics and mobility studies. Its focus on fragmentary languages in Italy also gives new detail to our picture of interactions among the peoples of Italy. Recent work has emphasised that a complicated latticework of conflict and collaboration informed their relationships with each other before Rome's conquest and influenced interactions among Romans of different Italian origins after the Social War.Footnote ²³ This insightful volume raises the possibility that mobility shaped how perceptions of self in relation to place transformed from the peninsula's pre-Roman era to Augustus's tota Italia and beyond.

III MOBILITY, CONNECTIVITY AND THE CIRCULATION OF NEGOTIATORES

The volumes reviewed in Section II demonstrate the continued prominence of concepts which pertain to connection and connectedness in the study of mobility in the Roman world.Footnote ²⁴ While they have played a role in this strand of scholarship since the days of Braudel, they have enjoyed exceptional attention in the last twenty-five years through the development of frameworks like social network analysis, centre-periphery theory, globalisation theory and connectivity.Footnote ²⁵ As the publications above show, their application to regions apart from the Mediterranean has enormous and relatively untapped potential.

Here, however, I return to the Mediterranean to explore the stakes of the prominence of connectivity in the study of mobility in the Roman Empire. Research on mobility in the Roman Empire has tended to portray connectivity as a positive attribute of the Mediterranean, though not always explicitly or deliberately. Yet there is abundant evidence for the disconnectivities, so to speak, of mobility in the empire. I use the term disconnectivity to describe the severance of ties and destruction of networks that resulted from phenomena that have been categorised as instances of connectivity and connection.Footnote ²⁶ Although ‘isolation’ might seem a more established rubric to describe such disconnections, ‘disconnectivity’ puts them in considered tension with connectivity.Footnote ²⁷ Here and in Section IV, I focus on the unprecedented expansion of Roman power in the last three centuries b.c.e. It is in this period of imperial activity that the contrast between connection and disconnect is particularly vivid.Footnote ²⁸

As ever, the debt our discipline owes to researchers from historically excluded backgrounds is particularly great. The place to begin is the work of Dan-el Padilla Peralta. In his recent edited volume, he describes Horden and Purcell's presentation of connectivity and the discipline's overwhelming approval of the concept as a ‘celebration’. He juxtaposes that description with the reminder that the connectivities in the Roman Empire were just that: connectivities in the Roman Empire.Footnote ²⁹ In an essay titled ‘Epistemicide’, he looks to Boaventura de Sousa Santos to draw out the relationship between connectivity, mobility and erasure:

In the passage above, Padilla Peralta quotes a very recent publication by Frank Daubner. The fuller context from Daubner is as follows:

Macedonian abstention from the massacre of 88 b.c.e. is not, as Daubner claims, necessarily a sign of non-violent relations among Romans and Macedonians. Is it not possible that Macedonians struggled to re-establish pre-Roman connections of unity and solidarity after Rome divided the region into new administrative units?Footnote ³² Daubner does not consider the difficulties that their new circumstances may have presented, even though the appearance of Romans and Italians across the eastern Mediterranean disrupted pre-Roman patterns of settlement, political organisation, economic activity and mobility.Footnote ³³ Lisa Eberle and Enora Le Quéré have shown that the extent of Roman and Italian land ownership and exploitation in the Greek East in the second and first centuries far exceeded what researchers have appreciated.Footnote ³⁴ Even without their conclusions, the number that Valerius Maximus gives for Romans and Italians killed at Mithridates's behest — 80,000 — demands (even if he exaggerates) that we recognise this reality.Footnote ³⁵

Daubner's reference to Caesar's account of the Roman community at Lissus also merits a closer look. At Bciv. 3.29, Caesar writes:

Daubner misapprehends the power dynamics at Lissus. The lines above do not show the town's Roman community, which Caesar calls a conventus civium Romanorum, helpfully mediating between locals and Roman authorities. They show the control that these Romans maintained over the town's political loyalties, regardless of local preference.

To re-quote Horden and Purcell: the collective movements of negotiatores in this period ‘articulate[d] a world of connectivity which maps onto what we call the Mediterranean’ (174). Many such businessmen formed groups like that in Lissus which engaged in concerted action under epithets like conventus civium Romanorum, cives Romani qui negotiantur and Ῥωμαῖοι οἱ πραγματευόμενοι. Wealthy, numerous and widespread, they garnered prestige and power as corporate bodies and for the individuals that formed them.Footnote ³⁷ Inscriptions reveal their frequent interaction with non-Roman individuals and communities.Footnote ³⁸ Yet the networks and connections that we see therein cannot have been founded on equitable relations. Literary accounts suggest that the influence of Roman communities like the conventus of Lissus was not rare. During a visit to Corduba, Caesar thanks the Roman citizens there for ‘for their enthusiasm in keeping the town under his control’.Footnote ³⁹ Ps.-Caesar reports that in a speech at Utica, Caesar censures a similar group of Romans and confiscates their money and property. At the same time, he thanks the people of Utica for their support, revealing that the local Roman population steered the loyalties of the town against the will of its people.Footnote ⁴⁰ Similar circumstances emerge at Zama, Thapsus and Hadrumetum.Footnote ⁴¹ The most notorious episode occurs in the previous century. In Sallust's account of the leadup to the Jugurthine War, he describes events in Cirta while the city was under siege from Jugurtha in 112 b.c.e. A group of individuals whom he calls Italici coerce Adherbal to surrender to Jugurtha. Adherbal yields, but not because he thinks any good will come of it: he knows the Italici will make him do it anyway. After seizing the town, Jugurtha executes him and the Italici.Footnote ⁴²

Daubner's view has a lineage. He cites Ramsey MacMullen's claim in Romanization in the Time of Augustus that Roman and Italian civilians ‘moved or lodged where they pleased, while fitting in not too badly’.Footnote ⁴³ This opinion is similar to that of Robert Errington, who in 1988 argued that the ‘peaceful penetration of Greek social and state institutions by Rhomaioi’ had favourably transformed Greek economic, social and cultural institutions, on the one hand, and Roman and Italian identity on the other: ‘[they] often remained in their chosen Greek city so long and lived there with such enthusiasm that they obtained local citizenship’.Footnote ⁴⁴ An earlier generation of historians includes Edward Gibbon, Francis Haverfield, René Cagnat and other colonial-era writers who saw Rome's presences overseas as essentially beneficial. Not coincidentally, their narratives resemble that of some Romans. ‘In fact’, Cicero declares in the Verrines, ‘our Roman businessmen (negotiatores) are linked with the Sicilians in the closest way by daily interaction, material interests, common sense and friendly rapport’.Footnote ⁴⁵

Cicero wanted to convey that for the most part, Romans and non-Romans got along. But this was a dubious claim to be making within two decades of the Mithridatic massacre. Elsewhere in the Verrines, Cicero himself gestures at widespread anger at Roman greed and injustice.Footnote ⁴⁶ In Pro Lege Manilia, he states, ‘It is difficult to say, Quirites, the extent to which we are held in hatred among foreign peoples due to the degeneracy and abuse of those whom we have sent to them with military command these last years’.Footnote ⁴⁷ In Pro Fonteio, Cicero conveys the ubiquitous financial power of negotiatores in the first century b.c.e.: ‘No Gaul conducts any business without the aid of a Roman citizen; not one coin in Gaul changes hands without being entered in the accounts of Roman citizens.’Footnote ⁴⁸ As Padilla Peralta would have us consider: to what extent were non-Romans free actors? We must ask if locals could choose not to work with Romans and remember that peace is not the same as the absence of physical violence.

IV REPUBLICAN DISCONNECTIVITIES

Compared to other peoples of the Mediterranean — including their neighbours the Etruscans — the Romans were late to long-distance mass-migration. Until the first war with Carthage, their movements were focused on terrestrial and riverine routes in the Italian peninsula.Footnote ⁴⁹ But by the end of the Third Macedonian War in 168, Mediterranean geopolitics and Roman mobility were permanently altered. A wave of civilian emigration from Italy swept the sea and remade patterns of circulation in and beyond it. Often forgotten but important early victims of their mobility were the people of Delos. Following the conclusion of the Third Macedonian War, Rome granted administration of the island to Athens and made it a free port to punish Rhodes for lacklustre support in the war. In addition, Rome expelled the island's native population via embassy. Polybius reports that those expelled migrated to Achaea, where they received citizenship.Footnote ⁵⁰ Their land was likely leased out or turned over to the Sanctuary of Apollo.Footnote ⁵¹ In their place arrived Romans and Italians from Italy.Footnote ⁵²

Here we see the complexity of mobility and connectivity under Roman imperialism. Politically motivated displacement was not new to Delian history. In an earlier era of Athenian control, Athens expelled the local population in 422, but expellees returned a year later.Footnote ⁵³ The expulsion under Rome, however, had permanent consequences. Long before the island's Roman phase, and despite the fifth-century expulsion under Athens, Delos had been part of an international network. As a major cultic centre, its place in this network generated a regional islander identity and a specifically Delian identity.Footnote ⁵⁴ But its ties in this network were permanently cut — disconnected, let us say — in 167/166. Thereafter, the international network of which Delos was a part took on a radically different shape. Interlocking social, cultic and financial relationships among Romans, Italians, Greeks, Phoenicians and others on and beyond the island grew on a scale that was novel to the history of mobility and networks on Delos and the Mediterranean itself.Footnote ⁵⁵

The Romans and Italians on Delos were deeply involved in trade and finance. Their language use in inscriptions shows strategic navigation: Greek to show their place in Hellenic society and relationships with local and regional elites, Latin to emphasise their ties to Rome.Footnote ⁵⁶ Attuned to their recently acquired ease of movement, they deliberately drew attention to where they were and where they had been. For example, in the last half of the second century b.c.e., two groups of Romans and Italians — similar to those which Caesar and Ps.-Caesar would later describe — installed a statue to honour Lochos, son of Kallimedes, kinsman of Ptolemy VIII Euergetes II and Queen Cleopatra, on the dromos of the Sanctuary of Apollo. The dedication on its base described them as ‘the Roman shippers (ναύκληροι) and merchants (ἔμποροι) who, in the capture of Alexandria, were treated benevolently by King Ptolemy VIII Euergetes II, god’.Footnote ⁵⁷ Around 99 b.c.e., a statue base went up in the so-called Agora of the Italians.Footnote ⁵⁸ Its Latin and Greek dedication described its authors as ‘the Italians who were at Alexandria’.Footnote ⁵⁹ As inscriptions from across the Greek East and even Italy attest, the practice extended well beyond Delos.Footnote ⁶⁰ In light of these migrants’ awareness of place, the graffito by Gaius Acutius that P. James discusses takes on a new look. The most fanciful explanation for primus — namely, that Acutius believed he was the first Roman citizen or Latin speaker to make it as far south as Upper Egypt, and that he knew the import of that position — begins to seem credible.

Italy, Philae, Alexandria and Delos were worlds apart. Italy is at the sea's midpoint. To the east lies Delos, at the centre of the Cyclades. Alexandria perched on the Egyptian coast and the island of Philae sits in the Nile in Upper Egypt, just north of where the Aswan High Dam now stands. Mediterranean connectivity made travel across the varied, difficult topography that separated these places possible for those with the means. Yet connectivity alone was not responsible for the arrival of migrants from Italy to Delos and Egypt and the rise of their wide-ranging networks. The connections adumbrated above depended on disconnections like that of Delians from their homeland and the social, political and cultural worlds of which they were a part. Such disconnections were essential to the mechanics of empire.

If we shift our gaze a little further back in time, we see that Rome was not new to the expulsion of native populations.Footnote ⁶¹ In Italy, after a battle with the Senones in 284 b.c.e., Rome drove survivors from their territory.Footnote ⁶² In 268 b.c.e., it expelled the Picentes and resettled them at the gulf of Paestum.Footnote ⁶³ In 187, Marcus Aemilius Lepidus defeated Ligurian Friniates and other Ligurian groups and ‘removed a multitude from the mountains into the plains’.Footnote ⁶⁴ From 180 to 179, Ligurian Apuani were transferred to and resettled in Samnium.Footnote ⁶⁵ Also in 180, Quintus Fulvius oversaw the resettlement of Ligurians to lowlands in an unspecified area.Footnote ⁶⁶ These few examples suffice to show that forced population movement in northern Italy under Rome took a variety of forms, ranging from expulsions that left victims to wander to state-funded resettlement schemes that foreshadow the Republic's later land policies.

Of a transfer of Apuan Ligurians from Liguria to Samnium in 180, Livy writes, ‘There was among the Samnites public land (ager publicus) belonging to the Roman people, which had belonged to the Taurasini.’Footnote ⁶⁷ Before their transfer, the Apuani begged the consuls Marcus Baebius and Publius Cornelius that they ‘not be forced to give up their household gods, the settlements in which they had been born, the tombs of their ancestors’.Footnote ⁶⁸ Liguria and Samnium may have shared some topographic features, but the deportees’ new home was unfamiliar to them. This robbed the Apuani of the specific places in which to conduct their rites and the means to pass them on.Footnote ⁶⁹ New forms of land management would have further contributed to their epistemic losses. Lowland areas also gave Roman authorities an ease of surveillance and movement that Apennine topography did not.Footnote ⁷⁰ Together, these factors imposed a fixity upon the Apuani that was previously unknown to them.Footnote ⁷¹

Presumably, the region that these Apuani were made to occupy was partially emptied of Samnites during a previous phase of Roman conquest. Territory in the north acquired new residents and forms of land use as well. Newcomers from central and southern Italy arrived to inhabit settlements that were very different from those of northern peoples. New colonies studded the region.Footnote ⁷² Many more Romans migrated north to occupy land outside these state-sponsored foundations.Footnote ⁷³ Centuriation, drainage and the construction of canals and roads accompanied these waves of migration.

Roman roads are a particularly well studied index of continuity and change. Despite their fame, even in antiquity, for their transformation of landscapes, they often followed existing routes and did not impose perceptible changes in archaeological evidence for pre-Roman local settlement, trade and social patterns.Footnote ⁷⁴ Such is at least one conclusion about the Via Postumia, which was constructed from Genoa to Aquileia in 187.Footnote ⁷⁵ Yet Roman roads were powerfully different from earlier routes. Durable construction material shortened journey times and made road travel more reliable, thereby transforming the experience of mobility within northern Italy and between it and the rest of the peninsula.Footnote ⁷⁶ Calgacus's speech in Tacitus’ Agricola, though imaginary, suggests the suffering that roads caused native populations that were forced to build them.Footnote ⁷⁷ Milestones and other monuments declared Rome's primacy.Footnote ⁷⁸ In northern Italy, the creation of a new political space was clear to locals.Footnote ⁷⁹ Between 187 and 131, Rome constructed the Via Aemilia, Via Postumia, Via Annia and Via Popilia. Together, these roads enclosed the Po plain.Footnote ⁸⁰ The Sententia Minuciorum, which records the settlement of a dispute between the Genuates and the Langenses Viturii, thought to be a Ligurian tribe, suggests that northern peoples understood the role of the Via Postumia in demarcating their movements and use of land. The dispute seems to have concerned the occupation of ager publicus and grazing rights in territory through which the Via Postumia ran.Footnote ⁸¹ The Via Sebaste, which was also constructed in a recently and fiercely rebellious area in Pisidia, provides a useful comparison. Built in 6/5 b.c.e., it reshaped local mobility by connecting a group of new Augustan colonies and ignoring preexisting local centres. The decision to route the Via Sebaste thus could be explained by the fact that these centres were already connected by a regional network. But as Stephen Mitchell and his co-authors point out, it was a tool for facilitating imperial aims.Footnote ⁸²

The long-term outcome is indisputable. The flourishing northern Italy that late republican and early Augustan authors describe came into being through repeated and heterogeneous disconnections. The extent to which mass death, enslavement and coercive resettlement diminished local populations may be why the region seemed vacant to the wealthy Roman elites who acquired property there.Footnote ⁸³ Our reliance on Romanocentric literary sources and uncertainty around the question of whether Ligurians were actually distinct from Gauls is also telling: we hardly know who they were.Footnote ⁸⁴ We should contextualise the strenuous efforts of the linguists who contributed to Clackson et al.'s volume within this ignorance. The fragmentary nature of the languages they examine is the outcome of Roman imperial expansion through the matrices of Mediterranean connectivity. Refugee and settler, Aeneas was indeed the quintessential ur-Roman.

V ‘THE ROMAN MIRACLE’

The intellectual trajectory of which Daubner, MacMullen and Errington are a part arises from the unwavering commitment to an ideal that Marcel Bénabou called ‘the Roman miracle’ in the opening lines of his 1976 treatise La résistance africaine á la romanisation:

The discipline of Roman history, Bénabou argued, legitimised Europe's past and present colonial projects, and rendered Rome into Europe's ideal image of itself. The process called Romanisation was central to the creation of this image. In studies of Roman Africa, it portrayed native North Africans as passive beneficiaries of Roman society and culture.Footnote ⁸⁶ Bénabou has not had the last word on cultural change in Roman Africa, of course, but La résistance africaine remains relevant.Footnote ⁸⁷ It problematised triumphalist accounts of the effects of the crucible of mobility and Roman imperial activity. A glance at recent literature on Rome's road system, for example, regularly yields positive rhetoric to describe its effects in antiquity and even today.Footnote ⁸⁸ To quote again from MacMullen on the early Roman Empire: ‘A thing of wonder, indeed!’Footnote ⁸⁹ The miracle is alive and well.

In ‘Epistemicide’, Padilla Peralta urges us to do more than consider how we think about Roman history: we must contemplate how we feel about it.Footnote ⁹⁰ The same is true of the Mediterranean. As Horden and Purcell point out, seas and oceans provoke strong emotions (177). In his introduction to the first edition of La Méditerranée, Braudel wrote: ‘I have loved the Mediterranean with passion, no doubt because I am a northerner like so many others in whose footsteps I have followed.’Footnote ⁹¹ Braudel penned these lines in 1949, but it is easy to imagine encountering the same sentiment in recent scholarly prefaces, the result of the journeys of western-origin researchers to and in the region.Footnote ⁹² Now we can stretch the idea of disconnect to the present day. Inextricably tied to the miracle that Bénabou condemned is the relationship that today's historians of Rome have with the Mediterranean. Their — our — mobility is its own Grand Tour, anchored by foreign schools like the French Academy in Rome, the American Academy at Rome and British School at Athens. Several were, in fact, founded in the era of the Grand Tour, and they functioned to realise the imperial ambitions of the nations they represented.Footnote ⁹³ This is brutally apparent in cases like the involvement of the American School of Classical Studies at Athens in displacing the populations of the neighborhoods of Vrysaki and Vlassarou (whose numbers included refugees who had relocated to Greece from Asia Minor in the 1920s) and dismantling their homes to excavate the site of the agora of classical Athens.Footnote ⁹⁴

These discrepant mobilities are part of a broader spectrum of imperial phenomena. Today, the central Mediterranean Sea is the deadliest part of the water for undocumented migration, a mass grave for those who drown before they can reach inhumane conditions on Lesvos and Lampedusa. From 2013–2014, Italy monitored the movements of these migrants and prevented their entry with Operation Mare Nostrum, which was named for how the Romans expressed the political unity wrought by domination: mare nostrum, ‘our sea’.Footnote ⁹⁵ The Italian operation was succeeded by Mos Maiorum, a collaboration with the EU's Frontex Border and Coast Guard Agency.Footnote ⁹⁶ Many of the migrants policed by these regimes of control originate in African nations in regions previously subjected to Italian imperialism, which was also inspired by the Roman Empire's control of the sea. ‘Roma antica sul mare’, Mussolini declared in 1926, to conclude a lecture in which he asserted the importance of maritime power.Footnote ⁹⁷ The Roman Mediterranean formed an important ideological strand of Fascist romanità and the rise of a new imperial Italy.Footnote ⁹⁸

In the discrepant mobilities of the modern Mediterranean, we can discern the cosmopolitanism that we impose on the Roman sea. Rovai in Clackson et al. describes Late Hellenistic Delos as a centre of ‘cosmopolitanism’, for example, and Claudia Moatti has used the concept to frame her studies of migration to the city of Rome and in the empire as a whole.Footnote ⁹⁹ In their and many other Mediterraneans, individuals form pluralistic communities as a result of living and moving freely.Footnote ¹⁰⁰ This traditional formulation of liberal cosmopolitanism is not just misleadingly depoliticised: it is dangerous. It feeds into Mediterranean exceptionalism, which is in turn bolstered by the connectivity that made such allegedly free movement possible.

In CS, Horden and Purcell alleged ‘the end of the Mediterranean’ as a subject of interest to historians and geographers.Footnote ¹⁰¹ Famously, the opposite occurred: over the last two decades, Mediterranean studies have surged. Much of this interest can be traced back to CS and its claim that the degree of connectivity that characterised the premodern Mediterranean was unsurpassed. But there is no proving such a claim — nor is there a way to make it without upholding what the authors call in BS ‘the hegemonial systems of its history’ (46). When they write that ‘[t]he Mediterranean has some claim to be the great original of seas as the subject of history’ (2), they do not see that the resources allocated to the study of other seas have been comparatively limited. Were that not so, we might better perceive the connectivities of the Indian Ocean, for example, or the Pacific Rim. Such an outcome demands that we continue the process of shattering our attachment to the status of our discipline, the peoples and places we study, and, crucially, the movements that link them together and break them apart.