Routes to slavery in the Roman World: a comparative perspective on the archaeology of forced...

21
Routes to slavery in the Roman world. Jane Webster Servi autem aut nascuntur aut fiunt Slaves are either born or made (Dig. 1.3.4). Not only was the slave denied all claims on, and obligations to, his parents and living blood relations but, by extension, all claims on and obligations to his more remote ances- tors and his own descendants. Truly he was a genealogical isolate. Formally isolated in his social relations with those who were living, he was also culturally isolated from the social heritage of his ancestors. To be sure, he had a past, but a past is not a heritage. for even sticks and stones have a past. Slaves differed from other human beings in that they were not allowed freely to integrate the experiences of their ancestors into their lives. 1 A Roman buyer surveying captives in the slave market at Rome, Delos or Ephe- sos would be in lile doubt as to the point of origin (natio) of any prospective purchase. Because ethnic origin was felt to offer a yardstick to the character, skills and potential of the enslaved individual, Roman law obliged dealers to disclose the natio of a slave at the point of sale. 2 Enslaved ‘others’ were subject to the same forms of ethnic stereotyping as their free counterparts, and judged accordingly. Thus some peoples were regarded as ‘born for slavery’, 3 whilst others were felt to be preferable for specific tasks. According to Varro, for example, slave families from Epirus made good overseers, and Gauls made very good animal-herders. 4 Ethnic origin was also a material consideration for slave owners in the sense that owning too many slaves with a shared ethnicity was felt to lead to ‘household difficulties’ 5 and — in the worst case scenario — to invite rebellion. 6 Thus, a slave owner purchased captives with an eye to their point of origin precisely in order to limit ethnic bonding among his slave familia. Slave owners may have aspired to suppress cultural memory, ethnic affiliation and ethnic self-identification amongst their slaves, but whether they were successful in doing so is another issue entirely. The tension between the recognition of natio at point of sale and its aempted suppression thereafter offers a starting point for this paper, which will consider coerced migration and enslavement within the Roman world. I will focus both on those carried from their homelands into slavery and also on their creole descendants. The laer were born (not borne) into slavery and were not therefore migrants, but, in the impor- tant sense delineated by O. Paerson, 7 were deracinated from an ancestral culture. My aim is to consider some of the ways in which documentary sources and material culture studies may be brought together in exploring Roman understandings of slave ethnicity; in identi- fying the ethnic origin of enslaved individuals; and in recognising possible identification strategies amongst Roman slaves. 1 Paerson 1982, 5 (emphasis my own). 2 Dig. 21.1.31. 3 Cic., Prov. cons. 2.5.10 on the Syrians. 4 Varro, Rust. 1.17.5 and 2.10.4. 5 Varro, Rust. 1.17.5. 6 See Wiedemann 1981, 198-206, on common ethnic origin as a factor in the First Slave War of 136-132 B.C. 7 Paerson 1982, 5: see quotation at the head of this article. Routes to slavery in the Roman world: a comparative perspective on the archaeology of forced migration

Transcript of Routes to slavery in the Roman World: a comparative perspective on the archaeology of forced...

Routes to slavery in the Roman world.A comparative perspective on the archaeology of forced migration

Jane WebsterServi autem aut nascuntur aut fiunt

Slaves are either born or made (Dig. 1.3.4).

Not only was the slave denied all claims on, and obligations to, his parents and living blood relations but, by extension, all claims on and obligations to his more remote ances-tors and his own descendants. Truly he was a genealogical isolate. Formally isolated in his social relations with those who were living, he was also culturally isolated from the social heritage of his ancestors. To be sure, he had a past, but a past is not a heritage. for even sticks and stones have a past. Slaves differed from other human beings in that they were not allowed freely to integrate the experiences of their ancestors into their lives.1

A Roman buyer surveying captives in the slave market at Rome, Delos or Ephe-sos would be in little doubt as to the point of origin (natio) of any prospective purchase. Because ethnic origin was felt to offer a yardstick to the character, skills and potential of the enslaved individual, Roman law obliged dealers to disclose the natio of a slave at the point of sale.2 Enslaved ‘others’ were subject to the same forms of ethnic stereotyping as their free counterparts, and judged accordingly. Thus some peoples were regarded as ‘born for slavery’,3 whilst others were felt to be preferable for specific tasks. According to Varro, for example, slave families from Epirus made good overseers, and Gauls made very good animal-herders.4 Ethnic origin was also a material consideration for slave owners in the sense that owning too many slaves with a shared ethnicity was felt to lead to ‘household difficulties’5 and — in the worst case scenario — to invite rebellion.6 Thus, a slave owner purchased captives with an eye to their point of origin precisely in order to limit ethnic bonding among his slave familia.

Slave owners may have aspired to suppress cultural memory, ethnic affiliation and ethnic self-identification amongst their slaves, but whether they were successful in doing so is another issue entirely. The tension between the recognition of natio at point of sale and its attempted suppression thereafter offers a starting point for this paper, which will consider coerced migration and enslavement within the Roman world. I will focus both on those carried from their homelands into slavery and also on their creole descendants. The latter were born (not borne) into slavery and were not therefore migrants, but, in the impor-tant sense delineated by O. Patterson,7 were deracinated from an ancestral culture. My aim is to consider some of the ways in which documentary sources and material culture studies may be brought together in exploring Roman understandings of slave ethnicity; in identi-fying the ethnic origin of enslaved individuals; and in recognising possible identification strategies amongst Roman slaves.

1 Patterson 1982, 5 (emphasis my own).2 Dig. 21.1.31.3 Cic., Prov. cons. 2.5.10 on the Syrians.4 Varro, Rust. 1.17.5 and 2.10.4.5 Varro, Rust. 1.17.5.6 See Wiedemann 1981, 198-206, on common ethnic origin as a factor in the First Slave War of

136-132 B.C. 7 Patterson 1982, 5: see quotation at the head of this article.

Routes to slavery in the Roman world: a comparative perspective on the archaeology of forced migration

J. Webster46

I will draw on historical and archaeological studies of the African diaspora, the dispersal of people of African birth as a result of the early modern Transatlantic slave trade (c.1500-1900). That trade is, of course, far more comprehensively documented than its Roman counterpart, yet whilst some historical archaeologists in the Americas are beginning to make sustained use of the mass of demographic data collected by economic historians, the archaeological study of early modern slavery remains as firmly focused on issues of slave experience as on questions of slave origins. This is a point we need to bear in mind in the Roman world, where an interest in the archaeology of slavery is finally beginning to emerge.8

Statistical starting points

The Americas: the transatlantic slave-trade database

My title is borrowed from a groundbreaking volume of papers, Routes to slavery: direc-tion, ethnicity and mortality in the transatlantic slave trade (1997),9 which marked the first flowering of a hugely ambitious project to synthesise archival data on known transatlantic slave-trading voyages from c.1500 to 1900. Between these years, more than 12.5 million Africans were carried into slavery on European slave ships — the largest coerced migration in modern history (fig. 4.1). From the European perspective, this was a well-documented process, and a wide range of archival sources, from shipping registers to colonial customs records, facilitate the study of demographic patterns over the long history of the transat-lantic trade in human cargoes. The slave-trade ‘numbers game’10 first began in the 1960s, when P. Curtin published an influential ‘census’ of the transatlantic trade.11 In the early 1990s, leading statisticians proposed creating a single, multi-source dataset, combining all the archival data uncovered since the 1960s. This project, hosted by the Du Bois Insti-tute for Afro-American Research at Harvard University, came to fruition in 1999 with The trans-Atlantic slave trade: a database on CD-ROM,12 synthesising information relating to more than 27,000 slaving voyages. Data-collection continued post-publication. In 2007, in a move coinciding with the 200th anniversary of the abolition of the British slave trade, the transatlantic slave trade database (TSTD) was updated and made freely available on an open-access website.13 It now contains information on over 35,000 transatlantic slaving voyages (an estimated four-fifths of all those made).

Among modern historians, the TSTD is facilitating fine-grained studies of demographic trends in the trade in W African slaves. In particular, the database is being used to confirm, or reveal, links between specific African groups and discrete New World regions. To give just a few examples of the kinds of patterns emerging, interrogation of the database reveals that two-thirds of all Africans embarked in the Bight of Biafra were landed in the British West Indies, whereas nine-tenths of those arriving in Bahia (Brazil) originated from the Congo/Angola region. Large numbers of Yoruba from Nigeria were shipped to Cuba, and there were similar strong links between Sierra Leone and the Gullah region of South

8 See Fentress et al. 2005; Webster 2005 and 2008.9 Eltis and Richardson 1997a.10 Eltis and Richardson 1997b, 1.11 Curtin 1969.12 Eltis et al. 1999.13 Voyages: the transatlantic slave trade database – http://www.slavevoyages.org/tast/index.faces

Routes to slavery in the Roman world: a comparative perspective 47

Carolina.14 In short, it is becoming clear that different African peoples “tended to flow in one dominant channel” across the Atlantic.15

Whilst the potential archaeological importance of the TSTD cannot be overstated, there is nevertheless considerable disagreement among historians as to the extent to which the dataset supports the study of slave ethnicity — and ethnogenesis — in the Americas. For some historians, as D. Eltis and D. Richardson put it:

The middle passage was more than a route to slavery that separated Africans from their past: it was a series of cultural highways that linked the history of transatlantic slave socie-ties to the history of specific peoples of W and W Central Africa.16

Others feel that, whilst the TSTD dataset does reveal clear links between some African and American regions, the underlying complexities of the long-term diasporic movement of different African peoples are such that caution is in order.17 For example, captives often made lengthy overland journeys to the coast, and for that reason the exit port for these individuals is no clear guide to their ethnic origin.

Perhaps because of these divergences of opinion, few archaeologists have engaged directly with the TSTD in the decade since it first appeared. Some signs of closer use of the dataset are emerging, however, and C. Fennell’s groundbreaking work on diaspora,

14 Eltis and Richardson 1997b.15 Morgan 1997, 125.16 Eltis and Richardson 1997b, 10. For studies advancing this argument, see Thornton 1992 and

Hall 1992. For a study drawing directly on the TSTD, see Chambers 1997. 17 In Morgan’s view (1997, 125), the “homogenizing tendency of stressing cultural unity in Africa,

and of seeing the dominance of particular African coastal regions or ethnicities in most American settings, is at variance with the central forces shaping the early modern Atlantic world”.

Fig. 4.1. Overview of the slave trade out of Africa (1500-1900) (Eltis and Richardson forthcoming; by permis-sion of Yale University Press).

J. Webster48

cosmology and ethnogenesis18 will be drawn on throughout this article. Else-where, N. Norman is studying the movement of people through the slaving port of Ouidah in Benin, and has drawn on data from the TSTD to refine infor-mation on flows of imported goods into hinterland sites currently under excava-tion.19 What the TSTD offers, in essence, is a new string to an existing bow: a poten-tial way for archaeologists to build on ongoing work, combining interrogation of the dataset with excavated data to reveal the macro-ethnic affiliations of individual slaves and potentially (subject to the cave-ats above) of entire slave communities.

The Roman world

Images of bound war-captives, destined for the slave markets, are far from uncommon in the Roman world (fig. 4.2).20 Ancient authors provide a good deal of information on the sources of, and trade in, Roman slaves, but these data differ qualitatively from those available for scholars of the transatlantic trade. We know that there were well-established centres of exchange, including Rome, Delos and Ephesos,21 but, in place of customs records, shipping registers and the like, we have large numbers of brief and almost certainly exag-gerated ‘after the fact’ notes on the influx of slaves at given points in time. These figures generally relate to Roman military campaigns. Thus, Gracchus reportedly enslaved 80,000 Sardinians in his campaign of 177 B.C.,22 Julius Caesar is reported as enslaving one million prisoners in the course of the Gallic Wars,23 and Trajan is said to have enslaved 500,000 Dacians in A.D. 105/106.24 Useful though they are,25 these references hardly amount to a comprehensive statistical database.

The situation faced by classicists is admirably summed up by W. Scheidel in his recent discussion of the problems faced by ancient historians attempting to explore the Roman slave supply. He notes that, in contrast to the Americas:

18 Fennell 2003 and 2007.19 See Norman 2009. I am grateful to N. Norman (College of William and Mary) for discussing this

work with me in advance of its publication.20 Representations of captive ‘barbarians’ are discussed in Ferris 2000. For cast bronze images of

shackled captives from Roman Britain and the Rhine-Danube limes (as in fig. 2), see Jackson 2005; as he suggests, these artefacts may well be emblematic symbols of a ‘frontier zone’ slave trade in war-captives. Most are casual finds, or discoveries made by metal detectorists, but Jackson cautiously advances a 3rd-c. date for this distinctive group of artefacts.

21 For recent archaeological studies of possible venues for slave sales at Delos and in Italy, see Coarelli 2005, Fentress 2005 and Trümper 2009.

22 Livy 41.28.8.23 Plut., Caes. 15; App., Kelt. 1.2.24 Lydus, Mag. 2.28.25 Textual sources on the Roman slave supply are collated in Thompson 2003, 1-46, and discussed

in detail by Scheidel 2007.

Fig. 4.2. Bronze ‘bound captive’ figurine found near Andover (Hants.) in 2007. Possibly 3rd c. A.D. (© PAS / Winchester Museums Service).

Routes to slavery in the Roman world: a comparative perspective 49

… hardly any genuine statistics are available, and historians face two similarly unpalatable options. Thus, we may decide to eschew speculative quantification altogether and focus on what our sources readily provide — that is, qualitative impressions of the prevalence of slave-ownership and the provenance of slaves. … Conversely, we may choose to advance broad probabilistic estimates of the demand for slaves and the likely weight of different sources of supply. … Any meaningful discussion of the Roman slave supply must seek to combine both approaches for the fullest possible picture.26

How many people lived in slavery in the Roman world? Scheidel’s own most recent estimate is that some 7-10% of the imperial population (between 5 and 8 million people) were slaves. He has calculated a hypothetical population of 1.3-1.9 million urban and 3.5-6.5 million rural slaves, against a population of 6.5-7.5 million urban-dwellers and 49-52 million rural free.27 It is generally agreed that Late Republican expansionism marked the high point in the import of ‘foreign’ slaves into Italy, and that imports slowed under the Empire. Even so, it has been estimated that the empire-wide demand for slaves per annum stood at half a million.28 (Vernae, the children of slaves, born into slavery, will be consid-ered at a later point in this article.)

Despite all these difficulties, it is possible, if only on a very broad scale, to pinpoint dominant channels in the Roman slave-supply over time. A series of major campaigns brought vast numbers of slaves onto the market at particular moments in time. Thus, the Second Carthaginian War (219-202 B.C.) brought a huge influx of N African slaves into Italy over a very brief period, whilst 55,000-60,000 captives are reported for the fall of Carthage in 146 B.C.29 The destruction of the Cimbri and Teutones in 102/101 and the Mithradatic Wars in Asia Minor from the 80s to the 60s B.C. will also have led to a massive intake of slaves; and so on. The difficulty, as noted above, is that, from the moment of sale, the ethnicity of these individuals was masked, either unintentionally or deliberately, by their owners. The fact that captives entering slavery were stripped even of their names is an important factor here.

Obscuring ethnicity: slave-naming practices

The Americas

Africans carried to the Americas on European ships were renamed by their eventual purchasers. These ‘given’ names derived from a wide variety of sources: they include place-names, day-names, Biblical names, and indeed classical names such as Cato, Flavia, Pompey and Caesar.30 In the Caribbean, it was not uncommon for plantation slaves to have an additional or ‘country’ name, utilised exclusively within the slave community,31 but the extent to which slaves played a part in naming offspring born into slavery, and the extent to which primary or ‘country’ names of African derivation may be seen as a meaningful guide to ethnic origin or affiliation, are topics of some debate.32

26 Scheidel 2007, 2.27 Ibid. 6 and Table 1. Scheidel (2005) estimates that annual imports into Italy averaged between

10,000 and 20,000 during the last two centuries B.C.28 Harris 1980, 118. Scheidel (2007, 6) estimates that some 250,000-400,000 slaves were required

annually to maintain numbers.29 Scheidel 2007, 6 and Table 2.30 The classic study is Cody 1987.31 Handler and Jacoby 1996.32 Ibid.; Cody 1987. Contra, Burnard 2001.

J. Webster50

Few certain instances of the birth names of African-born slaves survive in the Americas. The recently-augmented TSTD dataset contains a unique record of 67,000 slave-names, all recorded during the ‘illegal era’ (that is, after the abolition of the British slave trade in 1807), when the British Navy signed treaties with various countries to establish courts of mixed (or joint) commission to adjudicate suspected slave ships stopped by its naval cruisers. The names of the captives found on board these vessels, spelled with the help of African interpreters, were entered into the Registers of liberated Africans, compiled in Sierra Leone and Havana between 1819 and 1845.33 In addition to recording personal names, these registers record individual characteristics such as age, height, gender and, in over 12,000 cases, places of origin. As G. Ugo Nwokeji and D. Eltis put it:

…the new data provide a basis for identifying the region of origin of each recaptive, with-out traversing the minefield of European identifications of ethnicity that have plagued attempts to pin down the homelands of Africans in the Americas. In contrast to many plantation records in the Americas, the ethnic or regional basis of many of the names is recognizable, and makes it possible to identify broad groupings of peoples and in some cases sub-groupings on which the slave trade probably drew.34

Whilst these registers do not facilitate understanding of the movement and self-identification of slaves prior to 1807, they do provide data on ethnic origin of a kind entirely lacking for the Roman world.

The Roman world

As a res (thing), the Roman slave had no right to a nomen, but slaves did, of course, have personal names. Like their counterparts in the Americas, slaves in the Roman world were ‘given’ their personal name, either by their sellers or by their new owners follow-ing acquisition. Whim, fashion and a general preference for Greek names all informed the naming process; as a result, slave-names are a poor indicator of actual provenance. According to Varro, slaves sold at the market at Ephesos could be arbitrarily renamed after the trader, or the region in which they were purchased, or the city where they were bought.35 But slave-names were derived from a host of other sources too: historical and mythological characters, animals, plants, geographical terms, and so on. Greek names dominate the record not just because many slaves came from the Hellenistic East, but also because Greek names and other foreign-sounding ‘barbaric’ names were overwhelmingly reserved for slaves, and parents of free-born children tended to avoid them.36 In his survey of more than 28,000 attestations of the names of slaves and ex-slaves in the city of Rome, H. Solin37 documents some 500 ‘barbarian’ names, many of which are possibly, but not cer-tainly, indicative of the natio of the bearer.38 For example, 12 names, including Anina, Ibu and Monica, are listed as ‘African’ by Solin.39 But at the same time, his study also reveals that around two-thirds of all attested metropolitan slaves bore Greek names, and most oth-ers bore Latin ones. It is noteworthy that many of the most commonly attested slave-names

33 Ugo Nwokeji and Eltis 2002a and 2002b.34 Ugo Nwokeji and Eltis 2002a, 192.35 Varro, Ling. 8.9.21.36 Solin 1977 and 2003; Mouritsen 2005.37 Solin 1996.38 See Westermann 1955, 96, for examples of papyri revealing a lack of relation between the

provenance of a slave and his/her name.39 Solin 1996, 611.

Routes to slavery in the Roman world: a comparative perspective 51

at Rome were of Latin origin (Table 4.1), but the key point, as Scheidel suggests,40 is that the auspicious or otherwise cheerful names, such as Felix, Primus/Prima and Eros which topped the popularity rankings, were all customarily assigned without regard for eth-nicity. Explicit ethnic attestations are uncommon, and those we do have tend to confirm Scheidel’s point. Thus, the slave girl ‘Fortunata’ named on the first Roman deed of sale of a slave to have been found in Britain (a writing tablet discovered in London at No. 1 Poultry in 1994) has a typical Latin slave-name, but is stated in that document to have come from Jublains (NW Gaul) (natione Diablintem).41 42

In both the Americas and the Roman world, names are thus of very limited use in addressing questions of slave origin. Can the bodies of slaves tell us more?

The bioarchaeology of slavery

The modern world

The modern discipline of African-American bioarchaeology dates to the 1980s;43 it emerged in the context of an ongoing debate among economic historians (stimulated by such works as R. Fogel and S. Engerman’s Time on the cross44) concerning nutrition stress, disease, and mortality rates among slaves. Many archaeological studies from the 1980-1990s focused on cemetery populations, exploring precisely these topics.45 The focus of this body of work was on the quality of African-American life, as seen via diet, mortal-ity and health. This strand of research continues to be very important,46 and it is only in recent years that diasporan skeletal remains have been interrogated for the informa-tion they might provide on migration, as opposed to, or in addition to, life experience. Thus far, ‘migration’ isotopes have principally been used in the Americas to identify

40 Scheidel 2007, 15.41 Tomlin 2003.42 Based on Solin 1996, 680 and the revision of Solin’s figures in J. Bodel’s (2003) review.43 Blakey 2001.44 Fogel and Engerman 1974.45 See the 8 papers arising from the ground-breaking 1985 symposium Afro-American biohistory:

the physical evidence, subsequently published in the Am. J. Phys. Anth. 74.2 (1987). 46 See, e.g., Rankin-Hill 1997; for an excellent study of skeletal remains from a South African

cemetery, Cox et al. 2001.

4.1

J. Webster52

individual Africans displaced from their homelands in childhood and so likely to have made a diasporic voyage into slavery. Several important isotopic studies of this type have also been undertaken in South Africa. For example, analysis of 25 skeletons buried at Fort Knokke following the 1818 wreck of the Portuguese slaving brig Paquet Real suggested that the dead originated from the Makua, Yao and/or Marivi peoples of Mozambique.47 Simi-lar work has been undertaken on skeletons from Newton Plantation on Barbados, the most extensively excavated slave-cemetery in the Caribbean. This site was first excavated in the 1970s48 and numerous pioneering biocultural analyses were subsequently published, yielding new information on the health of the Newton Plantation slaves.49 New excava-tions were undertaken at Newton in 1997-98, and isotopic analysis performed on the 32 skeletons recovered at that time has been used both to reconstruct dietary histories and to trace life trajectories, in some cases back to their African origins.

If we turn to the USA, a battery of scientific techniques, ranging from craniometric analysis to dental morphology and both isotope and mtDNA analysis, has been brought to bear on the human remains from the New York African Burial Ground, a cemetery for enslaved and free Africans rediscovered in 1987.50 A key goal of this sophisticated inter-disciplinary study was to establish the origins of individuals interred at the site. To give just one example of the findings, W African molecular genetic affinities have been estab-lished for 48 individuals, and macroethnic affiliations (with the Fulba, Yoruba, Hausa and Mandinka peoples) for 14 of those.51 Elsewhere, however, scholars have suggested that, whilst mtDNAs of recent African ancestry in the Americas and Eurasia can be traced to broad geographical regions within Africa, considerable caution is needed when assess-ing claims to be able to trace the ancestry of particular lineages to discrete localities within modern-day Africa.52

The New York African Burial Ground is also notable as a site of contested heritage, and for the lengthy and ultimately successful struggle by an African-American descendant community to have academic work on the skeletal remains relocated to a university with a majority African-American clientele.53 That dialogue, and others like it,54 remind us that, throughout the Atlantic world, descendant communities have played a major part in pro-moting the archaeology of early modern slavery. No similar imperative exists to promote the study of slavery in the classical world, of course, fuelling the myth that archaeology has little to contribute to the study of an ‘invisible’ social class.55

Slavery, ethnicity and the corporeal exterior

Several of the skeletons from the wreck of the Paquet Real and from the New York Afri-can Burial Ground exhibited dental modification, their teeth having been chipped or filed

47 Cox and Sealy 1997.48 Handler and Lange 1978.49 Corruccini, Jacoby et al. 1987; Corruccini, Aufderheide et al. 1987; Corruccini, Brandon and

Handler 1989; Jacobi et al. 1992.50 Blakey 2001 and the papers collected in Blakey and Rankin-Hill 2004.51 Jackson et al. 2004.52 Salas et al. 2004.53 This process is summarised in La Roche and Blakey 1997.54 For a case study from South Africa, see Shepherd 2007.55 As Morris 1998, 193, has noted with reference to the archaeology of the excluded (including

women and slaves), these groups are ‘invisible’ only in the sense that we have failed to look for them.

Routes to slavery in the Roman world: a comparative perspective 53

into decorative shapes.56 These findings emphasise that indelible markers of African cul-ture, such as tooth modification and scarification, could literally be borne on the bodies of captives crossing the Atlantic on slave ships. Indeed, both tooth modification and cic-atrisation (scarification) have been recognised as signature markers of ‘salt water’ slaves — individuals born in Africa rather than in the Americas.57 These so-called ‘country marks’ were systematically recorded in the Registers of liberated Africans discussed above, and also in the slave registration returns of some of the British West Indies sugar islands after 1807,58 precisely because they were regarded at the time as identifiers of African birth. It may be noted in this context that efforts are currently under way to develop a new online resource named The African origins portal, based on the Registers of liberated Africans, which will make available to scholars all the data on ethnic origin (including cicatrisation marks) recorded in these registers but not yet available via the TSTD Voyages website.59

From the bioarchaeogical perspective, cicatrisation is, of course, problematic, in that, whilst dental modification is visible on the skeleton, scarification is not — a point that fore-grounds the importance of combining textual, pictorial and skeletal data in studying the ‘slave body’ at any time in the past. In this context, contemporary newspaper advertise-ments have played an important part in the study of ‘country marks’ on 18th- and 19th-c. slaves in the Americas.60

The Roman world

The body of work discussed above is beginning to demonstrate that, even for a region as vast and ethnically diverse as pre-modern Africa, it is becoming possible to ‘drill down’ from the homogenising notion of the enslaved ‘African’ to isolate ethnically-specific ‘routes to slavery’ in the Americas. This brings us to the equally vast, and equally diverse, Roman world: will we ever be able to do the same there? The notion of a ‘bioarchaeology of Roman slavery’ can be quickly passed over, for the simple reason that archaeologists have yet to identify or excavate a Roman slave cemetery. We cannot even be sure which (if any) categories of slave were buried in discrete cemeteries. As large numbers of inscriptions testify, the slaves (and freedmen) of wealthy urbanites shared in the family tomb. Signifi-cant numbers of slaves turned to each other for commemoration,61 sometimes establishing burial clubs (collegia) or guilds for that purpose,62 but little indeed is known regarding the burial of rural slaves, one of the least epigraphically visible strata in Roman society.63

56 Paquet Real: Cox and Sealy 1997, 208-9 and 218; New York African Burial Ground: Goodman et al. 2004, 216-65.

57 See, e.g., Handler 1994 and Gomez 1998.58 Higman 1995, 21-22.59 Supra n.13. Information on the African Origins Portal project kindly supplied by L. Milewicz

([email protected]).60 Handler’s (1994) work on 18th- and 19th-c. advertisements for runaway slaves in Barbados has

demonstrated that, in all cases where runaways were stated to bear ‘country marks’, they were also stated to be of African birth.

61 Saller and Shaw 1984.62 For a British example, see RIB 1436, a tombstone from Haltonchesters: the slave Hardalio’s

tombstone was set up by a collegium of fellow-slaves.63 Samson 1989. As Thompson (2003, 124) notes, the 55 Flavian 2nd-c. burials discovered in a

cremation cemetery outside the enclosure wall of a villa at Köln-Müngersdorf may represent the remains of slaves or farmworkers dispatched with few grave goods, and without cremation urns. As ever, it is impossible to be certain that these individuals were slaves.

J. Webster54

Slavery, ethnicity and the corporeal exterior

In the Americas, somatic distance rendered slaves (and all persons of African descent) readily identifiable. Simply, they were not white. In the Roman world, by contrast, there was no close correlation between somatic type and servile status.64 Rare but significant comments by classical authors suggest that Graeco-Roman writers and artists had encoun-tered various forms of permanent bodily alteration (including cicatrisation and tattooing) amongst Rome’s ‘others’, both white and black.65 For example, cicatrisation marks are present on the terracotta head of a Sudanese woman from the Fayum,66 but not all blacks in the Roman world were slaves and there can be no certainty as to whether the individual depicted was a slave or a free woman.

Dress was no guide to servile status either: although some wealthy individuals required their slaves to wear the equivalent of livery, there were no forms of clothing uniquely asso-ciated with slaves.67 Yet in some intriguing ways the corporeal exterior of the slave was a focal point for (involuntary) status marking. At the point of sale, for example, the feet of foreign slaves were marked with chalk,68 and ancient authors refer to the practice of shaving or close-cropping the hair of slaves.69

In the Atlantic world, marks of ownership, made with a branding iron, were inflicted on the bodies of captives shortly before they made the journey into slavery. In the Roman world, hot-iron branding was used to mark animals, not humans, but stigmata in the form of tattoo marks were etched on the bodies of both delinquent slaves and criminals.70 This practice, and the use of judicial torture,71 remind us that the slave was, as R. A. Bauman puts it, “answerable with his body for any infraction”.72 By the Hellenistic period, delin-quent slaves were tattooed on the face or forehead with the name of their offence, their faces etched with a narrative of their misdemeanours.73 Unfortunately, tattoo marks survive upon ancient bodies only in the most exceptional circumstances, leaving little prospect for a Roman bioarchaeology of slave stigmata.

In colonial contexts of all kinds, the human body, including the body of the slave, was both a natural and a social symbol,74 and the relationship between punishment and bodily marking briefly sketched here may suggest that in the Roman world, as in the Americas, permanent bodily alteration was regarded as a marker of deviant status, literally stigma-tising criminals and delinquent slaves.

An ‘internal diaspora’?

Many (perhaps the majority) of those living in slavery in the Americas and the Roman world simply had the misfortune to be born the children of slave mothers. They crossed

64 As Thompson (2003, 104) puts it, “the overwhelming majority of slaves in Roman society was always white”.

65 For textual and archaeological evidence for bodily marking in Britain, see, e.g., Carr 2005.66 Snowden 1970, 22-23, fig. 3.67 Bradley 1994, 97.68 Plin., NH 35.199.69 Including Petron., Sat. 102-6 and Apul., Met. 9.12.70 Jones 1987.71 On the torture of slaves to obtain confessions, see Bauman 2000, 116-20.72 Bauman 2000, 118.73 Jones 1987, 148.74 On the colonised body, see Rothschild 2008.

Routes to slavery in the Roman world: a comparative perspective 55

neither ocean nor continent on a diasporic journey into slavery, and they had no first-hand knowledge of their parents’ ancestral culture. As I will suggest below, questions of ethnic identity, and its material expression, become particularly complex when we consider such slaves, usually known as creoles in the Americas and as vernae in the Roman world.

The ‘internal slave supply’ in the Roman world

What proportion of Roman slaves were the children of slaves, born into slavery (vernae)? This remains a fraught issue, one much-studied by ancient historians yet largely ignored by archaeologists. It is generally agreed that Late Republican expansionism marked the high point in the import of ‘foreign’ slaves into Italy, but after this point the extent to which metropolitan and provincial slave-labour requirements were met through ‘natural reproduction’ (or breeding) is much-debated.75 Scheidel has estimated the biological replacement rate as being as high as 80% in Late Republican Italy.76

One of the best-known funerary monuments in Britain comes from South Shields, the easternmost fort on Hadrian’s Wall (Noy, above,p. 00).77 This monument was commis-sioned by Barates of Palmyra in memory of his wife Regina. The Syrian-born Barates is a fine example of the voluntary Roman migrant; but what of Regina? Despite being a former slave, she was of Catuvellaunian origin (natio). She reminds us that there were many ways in which Britons might have become slaves in Britannia. First, of course, some may have been enslaved at the time of the conquest, during the steady expansion of the province, or following an insurrection. Second, as suggested above, it was common practice in many provinces to top up the slave supply from internal (rather than external) sources: these sources included vernae, orphans, exposed infants and other foundlings, children sold as a result of poverty, self-sale for debt, and penal condemnation to slavery.78

As Scheidel has noted in discussing slavery in the provinces,79 it is likely that some of those referred to as ‘slaves’ in the literature (as, e.g., in the report that in A.D. 280, a wealthy pretender from the Maritime Alps was able to arm 2,000 of his own slaves in a bid for the throne)80 should probably be seen, not as freely alienable chattels, but rather as dependants bound by local traditions of subservience. In this context, students of the Late Pre-Roman Iron Age have long argued that a slave trade existed in Britain before the conquest, and that this trade supplied the Roman market.81 If we take this point further, it seems reason-able to assume (as was also the case in the Hellenistic world and, much later, in W Africa) that those pre-conquest communities developing a slave trade would have done so in part because forms of slavery already existed at home. Textual evidence points to a variety of pre-Roman ‘Celtic’ social relationships placing individuals in ambiguous but certainly ser-vile positions that might, either as a result of increased slave trading for Rome or following conquest by Rome, have been redefined simply as slavery. Referring to Gaul, Caesar notes that Gallic peoples were oppressed by debt and servitude, and committed themselves to the service of the élite.82 He also remarks that the élite had the same rights over these peo-

75 See, e.g., Scheidel 1997; contra, Harris 1999.76 Scheidel 1997, 166. More recently he has reduced this figure to 50%; Scheidel 2007, 17.77 RIB 1065.78 See Birley 1979, 145-50.79 Scheidel 2007, 5.80 HA, Firmus 12.2.81 See, e.g., Fitzpatrick 1989.82 Caes., BG 6.13.

J. Webster56

ple as masters over slaves. Caesar commented on a number of ‘social contracts’ between individuals in Gaul, noting, for example,83 that beloved servi et clientes had once been burnt at the conclusion of funeral formalities. In his account of the defeat of the Sotiates,84 he speaks of one Adiatunnus, who took action with 600 devotees known as soldurii. In life, Caesar remarks, these men enjoyed all benefits offered by the comrades to whose friend-ship they had committed themselves, but if violent death befell their fellows they either endured the same misfortune or took their own lives.

In some provinces foundling infants may also have contributed in an important way to the slave body. The raising of foundlings as slaves is documented in Roman Egypt, and Scheidel has suggested that enslavement of exposed babies may conceivably have been the leading domestic source of free-born slaves in the mature Empire.85 The practice of fosterage was known in pre-Roman Britain, and the fosterage of foundlings after the conquest might well have produced subordinate sub-classes whose servitude would have continued.

We cannot know precisely how the British-born slave Regina became a slave, though it seems likely that she was slave-born. How should we conceptualise her experience, and that of other slaves in the Roman world whose route to slavery did not involve migration? How, if at all, can archaeologists isolate and study these individuals? The archaeologist I. Lilley has recently advanced the concept of ‘internal diaspora’, positioning indigenous peoples, as well as settlers and their descendants, as diasporic.86 He has drawn on the work of the anthropologist J. Clifford, who argued that

[contemporary tribal peoples]. who have been dispossessed of their lands or who must leave reduced reserves to find work, may claim diasporic identities. Inasmuch as their distinctive sense of themselves is oriented toward a lost or alienated home defined as aboriginal (and thus ‘outside’ the surrounding nation-state), we can speak of a diasporic dimension of con-temporary tribal life.87

Lilley has himself explored the notion of internal diaspora with reference to both con-temporary Australasia88 and the Lapita dispersal in the western Pacific, some 3,000 years ago.89 There is much of value in this concept for archaeologists working in colonial con-texts of all kinds. At the very least, it foregrounds the critical attributes of a diaspora,90 and presents the possibility that colonized native minorities, uprooted literally or figura-tively by the colonial system, may be considered as diasporic. It would be interesting to bring this idea to bear on subaltern peoples in the Roman world, such as the Batavians, who appear to have re-invented their ethnic identity in response to Roman rule.91 Never-theless, Lilley works in settings that preclude explicit discussion of indigenous slavery, and it is highly debatable whether his formulation of diaspora should be brought to bear

83 BG 6.19.4.84 BG 3.22.85 Scheidel 2007, 10.86 Lilley 2004 and 2006.87 Clifford 1994, 309.88 Lilley 2006.89 Lilley 2004.90 Defined by W. Safran (1991) as dispersal, memories or myths of a homeland, distinction from

the host society, an ethic of eventual return to the homeland, and an individual and collective continuing relationship (direct or vicarious) with the homeland.

91 Roymans 2004.

Routes to slavery in the Roman world: a comparative perspective 57

upon the slave-born. O. Patterson’s well-known concept of natal alienation, or genealogical isolation92 — the universal condition of the slave across time and place — is surely more appropriate.93 Natal alienation was not simply a condition of ‘foreign’ slaves. To para-phrase M. I. Finley, all slaves, whether foreign-born or recruited from the local community, were reconceptualised as outsiders.94 Despite having being born in Britain, then, the ex-slave Regina was natally alienated. No formally-recognised ties bound her to her living or ancestral Catuvellaunian kin. In that sense she died in genealogical isolation, a long way from home. And yet her natio was noted by her husband on her memorial — a reminder that, whatever the efforts slave traders and owners made to suppress and deny cultural memory and ethnic self-identification among slaves, these things remained vitally impor-tant to slaves and ex-slaves themselves.

Very little work has been done by ancient historians or archaeologists on the impli-cations either of indigenous slavery or of an increasing percentage, over time, of locally recruited (or creole) slaves. Put another way, would the ‘material signature’ of these cate-gories of non-migrant slave be the same as that of ‘foreign’ slaves? What changes might we expect to see over time, with rising numbers of individuals being born into slavery, and having no direct experience of the homelands of their forebears?

If we turn to the archaeology of slavery in the Americas, we find a body of archaeological work engaging with precisely with these questions.

The material world of the slave

Africanisms, creolization and ‘ethnogenic bricolage’ in the Americas

Historical archaeologists in the USA and Caribbean have long been interested in isolating the practices of ‘salt water’ slaves (those who had made the journey into slavery via the Atlantic crossing known as the ‘middle passage’). In part, this is simply because these first-generation slaves carried with them a cultural memory of their native cultures, and were more likely to retain clear markers of African practice (known as ‘Africanisms’ in the United States) than were their creole descendants, who lacked direct knowledge of African traditions and practices. At the same time, however, it has also been appreciated that slave culture in the Americas was in no sense static, and that an understanding of long-term creolization processes cannot develop simply by focusing on ‘salt water’ slavery.

For some 30 years now, archaeologists have been recognising strategies by which captive Africans and their New World-born descendants maintained and adapted their traditions and customs in the Americas. Examples range from studies of African motifs on clay pipes, through work on colonowares (slave-made ceramics), to the recognition of African architectural elements in plantation architecture.95 The ‘Africa’ revealed by these studies has justly been accused of lacking specificity, subsuming the awesome ethnic com-plexity of West and West Central Africa within a single label. The best of these studies, however, have made careful use of historical ethnography in identifying parallels between New World practices and ethnically-specific practices within West Africa. For example,

92 Patterson 1982, 5-6; see also the quotation at the head of this paper.93 See Patterson ibid. 111.94 On the slave as outsider, see Finley 1968.95 For an excellent discussion of artefacts and ethnicity in African-American archaeology, see

Perry and Paynter 1999.

J. Webster58

J. Handler has drawn on ethno-historical accounts of West African customs and traditions in identifying macro-ethnic affiliations for practices found in several ‘unusual’ graves at Newton Plantation, Barbados.96

C. Fennell’s work on cosmology and ethnogenesis has gone a step further, combining anthropological theory with archaeological data to reveal the emergence of new, ‘blended’ cosmological symbols, indicative of emergent group identities, among dispersed peoples of different origins in Haiti and Brazil.97 At the heart of Fennell’s work is an analysis of the use of ‘core’ symbols within diasporic cultures. Fennell describes a core symbol as

[one used] to express fundamental elements of a group’s cosmology and sense of identity within the world. Core symbols can be communicated through spoken words and ritual performances and are often depicted in tangible, graphic form through renderings in mate-rial culture.98

He distinguishes further between emblematic and instrumental versions of core symbols: the former summarizing the identity of a cultural group as a cohesive unit (the cruxi-fix, Star of David), the latter comprising a simplified or abbreviated form of the avatar, used for individual rather than public or group purposes.99 Much of his book is concerned

with a particularly important core symbol among diasporic Africans: the BaKongo cosmogram (fig. 4.3). Among the BaKongo people of West Cen-tral Africa, this symbol served as an emblematic expression of identity, summarising a broad array of ideas and metaphoric messages that comprised their sense of identity within the cosmos. Versions of this ‘core’ symbol have been found across the plantation belt of the United States and the Carib-bean. Many are etched on the bases of curved, bowl-shaped objects ranging from slave-made pottery through to European manufactures.100 In Haiti, Cuba and Brazil, this and other instrumen-tal symbols from a number of cultures (BaKongo, Yoruba, Fon) were combined in new ways, in a

process Fennell defines as “ethnogenic bricolage”: creative interaction amongst individu-als raised in different cultures, but coming together in new settings, often at the geographic crossroads of multiple diasporas.101 African-Caribbean slaves obviously fall into this category, but so do colonial migrants of other kinds, as revealed through Fennell’s excel-lent analysis of the use of Hexerei (witchcraft) symbolism by German-Americans in colonial Virginia.102

Can these ideas be applied to the Roman world? Rome itself, and cities and emporia throughout the empire, were certainly, to repeat Fennell’s term, the “geographic crossroads

96 E.g., Handler 1996 has linked a unique prone burial at Newton with practices documented amongst the LoDaaga of Ghana.

97 Fennell 2007.98 Ibid. 6.99 Ibid. 8.100 Examples of the former can be found in Ferguson 1992.101 Fennell 2007, 9.102 Ibid. 106-12.

Fig. 4.3. The BaKongo cosmogram (Fennell 2003, 6, fig. 2; courtesy C. Fennell).

Routes to slavery in the Roman world: a comparative perspective 59

of multiple diasporas”. It may, of course, be objected that slaves were only one of many deracinated or diasporic groups in the Roman world, and that as a result it would be diffi-cult to identify emblematic and instrumental symbolism unique to slaves. Yet slaves were not the only migrants in the colonial Americas (hence Fennell’s study of German-American magico-religious symbols), and it is clearly possible to isolate emergent instrumental sym-bols amongst slave populations there. Perhaps the point here is that we should worry less about identifying practices ‘unique’ to slaves, and think more in terms of identifying the continuum of material culture discourse shared by migrant groups of all kinds within the Roman ‘melting pot’.

Having made this point, however, it remains the case that a significant number of arte-facts, from tombstones to clay roofing-tiles, bear common ‘servile’ names and can in fact be associated closely with slaves in the Roman world. Uncertainties remain as to the exact status of these individuals: as J. A. Bodel has emphasised, the 6 most common slave-names in metropolitan Rome (Table 4.1) also rank among the 18 most common Latin cognomina in general use, suggesting that in everyday address many Roman slaves and ex-slaves may have been onomastically indistinguishable from their free-born peers.103 Similarly, the three most common Greek slave-names at Rome (Hermes, Eros, and Onesimus) are also the three most frequently-attested Greek cognomina there.104

This uncertainty has resulted in the convention that inscriptions on monuments and artefacts are attributed to slaves only where status is explicitly attested (for example, through use of the word servus). Where less certainty is demanded, the number of attesta-tions rises very considerably. H. Solin’s onomasticon of metropolitan Rome slave-names105 registers more than 28,000 examples of some 5,500 names of slaves and ex-slaves from the earliest period down to the 5th c. A.D. Solin’s study remains little known outside the circle of ancient historians working on slavery (in much the same way that Scheidel’s important body of work is rarely referenced by archaeologists). It is true that the criteria adopted for the inclusion of individual names in this onomasticon remain unspecified, and that this limits the level of ‘certainty’ that the person referred to in any given attestation is indeed a slave. But certainty is an archaeological luxury at the best of times, and far more could and should be made of artefacts bearing common slave-names than is usually the case.

There are so few certain examples of slave-made or slave-used material culture in the Roman world that it seems perverse not to make more use of artefacts inscribed with common ‘servile’ names than we currently do. As Table 1 reveals, in the city of Rome alone there are over 1000 attestations of the names Hermes/-a, Eros-otis and Onesimus/-e — Greek names overwhelmingly reserved for those of servile status. Solin’s onomasti-con, principally based on CIL VI, does not include similar attestations on instrumentum domesticum and so on, and, as I have suggested elsewhere with reference to graffiti,106 the occurrence of ‘servile’ names on a variety of artefacts — from clay tiles to parietal graffiti (fig. 4.4), from defixiones to pot sherds — at least presents opportunities for archaeologists to begin sustained study of the symbols and discursive strategies associated with a mate-rial culture that, in the majority of cases at least, is likely to have been created or modified

103 Bodel 2003.104 Solin 2001, 312.105 Solin 1996.106 Webster 2008.

J. Webster60

by slaves. Certainly, far more could be done with the many examples of graffiti and dipinti that we can quite confidently attribute to slaves either on contextual grounds (as in Mou-ritsen’s work on the House of the Menander)107 or because they include common ‘servile’ names. For example, I know of no work that has focused on the symbols that (albeit rarely) accompany graffiti bearing servile names. Can any of these symbols be traced to specific ethnic groups and belief systems within the Roman world?

This brings me back to Fennell’s work on core symbolism and ethnogenic bricolage among diasporic communities. There are strong grounds for suggesting that, in the Roman world as in the Americas, we may posit a strong link between ritual, ‘magic’, and mani-festations of core symbols.108 Roman defixiones offer some interesting possibilities here. Amongst the Late Roman curse tablets deposited in tombs or graves on the Via Appia, for example, is a series of tablets bearing drawings of human figures portrayed in bonds and surrounded by serpents (fig. 4.5). These tablets and their purchasers clearly fall into the category of freedmen and slaves, and their context concerns competition amongst chariot-eers at Rome.109 But the imagery reveals a complex cosmological bricolage, amalgamating horse-head spirits (daimones), Christian Gnosticism, and the worship of Osiris. It is tempt-ing to dismiss artefacts like these as ephemera, reflecting behaviours at the very fringe of mainstream Roman belief and practice, yet it is precisely these oddities which may prove invaluable in exploring the potential emergence of new, shared, belief systems at the cross-roads inhabited by Rome’s unfree.

Fennell correlated the BaKongo cosmogram with a specifically Congo-Angolan (rather than generally ‘African’) diaspora by constructing an ethnographic analogy based on 16th- to 19th-c. accounts of West Central African belief and practice. Identification of potential ‘core’ identity markers among Roman slaves cannot proceed by looking forwards from our period in this way, but we can compare sideways, to the symbolic systems of peoples dominating the external slave supply at specific points in Roman history. This is where

107 Mouritsen forthcoming.108 It is not without interest that Columella’s advice on the ideal management of slaves hints at

efforts to curb slave agency in relation to ‘superstition’ and magic: for example, Columella (Rust. 1.8.5-6), counselled that the master must be responsible for the performance of all rites in the household, and that on rural estates the vilicus should not carry out sacrifices without permission of the master. Similarly, witches and seers should not be admitted to estates, lest slaves’ minds be filled with superstitious ideas (Rust. 1.8.7). See also Giacobello 2008 on secondary lararia as potential foci for religious agency amongst slaves in Pompeian households.

109 Gager 1992, 67-72.

Fig. 4.4. Graffito with the personal name Fortunata and associated portrait from a thermopolium at Pompeii (Langner 2001, no. 189; permission?).

Routes to slavery in the Roman world: a comparative perspective 61

entwinement with texts of a different kind, and with the work of historians of the Roman slave-supply discussed above, should properly come to the fore. Unfortunately, the study of ancient slavery is severely impeded by a lack of collaboration between ancient histori-ans and archaeologists, and by the resultant failure to “marry words and things”110 that historical archaeologists of the modern world regard as essential to the development of a nuanced archaeology of slave-owning societies. One way forward here might be to focus upon objects that marry words and things in a literal way: artefacts inscribed with ‘servile’ names.

Finally, I should note that, whilst space has precluded discussion of the material cul-ture and symbolism of freedmen, the material world of the ex-slave is particularly relevant

110 Hall 2000, 16-17.

Fig. 4.5. Late Roman defixio from the Via Appia, Rome (Gager 1992, fig. 9; by permission of Oxford University Press).

J. Webster62

to the programme of enquiry into naming and symbolism that I have proposed. Freed-men are far more epigraphically visible than slaves, and their epigraphic commemoration has attracted considerable interest amongst historians. As H. Mouritsen has argued in a series of influential studies,111 the vast majority of those commemorated epigraphically at Rome were former slaves, individuals united by their experience of both slavery and manumission. At the core of that experience, Mouritsen argues, lay a new sense of personal and family security, clearly reflected in the funerary monuments of freedmen and their descendants. The act of monumentalisation was, in this sense, genealogical: the celebration of the secure family unit. The fact that freedmen dominate the epigraphic record for Rome, Italy and some of the provinces is a reminder, in other words, of the keen interest amongst the deracinated in genealogy, security and roots. Ultimately, it may prove to be amongst freedmen that any search for ‘ethnogenic bricolage’ will prove most fruitful. Despite their disparate origins, ex-slaves were united by the fact of slavery, and at Rome at least there is some evidence to suggest that freedmen worked actively to form communal bonds within the social circle in which they moved. For example, few freedmen married freeborn wives, yet it seems inconceivable, given the material success of many freedmen, that they would have been unable to find freeborn spouses had they wished to do so. Similarly, a surpris-ing number of freedmen gave their freeborn children Greek cognomina despite their servile connotations. What emerges here, Mouritsen argues, is the suggestion of a “distinct freed-man community”,112 and it may be here, amongst those with a clear psychological need for genealogical security, that new, shared ethnic identities might also have emerged.

Conclusion

It is likely that more than 100 million people were enslaved in the millennium during which the Roman empire rose and was eclipsed.113 Not one of these men, women and chil-dren were voluntary agents of their own displacement. This point needs to be borne in mind as dispora becomes a more inclusive term, embracing voluntary migrants and even, as discussed above, non-migrant indigenous peoples. As diaspora studies move ‘beyond the black Atlantic’ and the so-called ‘victim’ diasporas of the Jewish and West African peoples,114 it is important to recall that diasporas did (and do) have victims: to imagine otherwise is to do a fundamental disservice to the experiences of the millions who have struggled to transcend the structures of slavery, making a space in the world in a place not of their own choosing. Roman slavery, so easily overlooked in mainstream archaeological thought, will only recede further into the background if we lose sight of this point.

ReferencesBauman, R. A. 2000. Human rights in ancient Rome (London).Birley, A. R. 1979. The people of Roman Britain (London).Blakey, M. 2001. “Bioarchaeology of the African diaspora in the Americas: its origins and scope,”

Ann. Review of Anthropology 30, 387-422.Blakey, M. L. and L. M. Rankin-Hill 2004. The New York African Burial Ground, skeletal biology, final

report (New York). http://www.africanburialground.gov/FinalReports/HUsbrABG_Ch5_Rvsd.pdf

111 See especially Mouritsen 2005, on which the discussion here is based.112 Mouritsen 2005, 57. See also George 2005, 40, on natal alienation and the importance attached

to family by former slaves.113 Scheidel 2007, 18.114 Lilley 2004 and 2006.

Routes to slavery in the Roman world: a comparative perspective 63

Bodel, J. 2003. Review of Solin 1996 at BMCR 2003.01.03.Bradley, K. 1994. Slavery and society at Rome (Cambridge).Burnard, T. 2001. “Slave naming patterns: onomastics and the taxonomy of race in eighteenth-century

Jamaica,” J. Interdisc. Hist. 31, 325-46.Carr, G. 2005. “Woad, tattooing and identity in later Iron Age and early Roman Britain,” OJA 24,

273-92.Chambers, D. 1997. “‘My own nation’: Igbo exiles in the diaspora,” in Eltis and Richardson 1997a,

72-97.Clifford, J. 1994. “Diasporas,” Cultural Anthropology 9, 302-38.Coarelli, F. 2005. “L‘Agora des Italiens: lo statarion di Delo?’’ JRA 18, 197-212.Cody, C. A. 1987. “There was no ‘Absalom’ on the Ball plantations: slave-naming practices in the

South Carolina Low Country, 1720-1865,” AHR 92, 563-96.Corruccini, R. S., K. Jacobi, J. S. Handler and A. C. Aufderheide 1987. “Implications of tooth root

hypercementosis in a Barbados slave skeletal collection,” Am. J. Phys. Anth. 74, 179-84. Corruccini, R. S., A. C. Aufderheide, J. S. Handler and L. Wittmers 1987. “Patterning of skeletal lead

content in Barbados slaves,” Archaeometry 29, 233-39. Corruccini, R. S., E. Brandon and J. S. Handler 1989. “Inferring fertility from relative mortality in

historically controlled cemetery remains from Barbados,” AmAnt 54, 609-14. Cox, G. and J. Sealy 1997. “Investigating identity and life histories: isotopic analysis and historical

documentation of slave skeletons found on the Cape Town foreshore, South Africa,” Int. J. Hist. Arch. 1, 207-24.

Cox, G., J. Sealy, C. Schrire and A. Morris 2001. “Carbon and nitrogen isotopic analyses of the under-class at the colonial Cape of Good Hope in the eighteenth and nineteenth centuries,” World Arch. 33, 73-97.

Curtin, P. 1969. The Atlantic slave trade: a census (Madison, WI).Eltis, D. and D. Richardson (edd.) 1997a. Routes to slavery: direction, ethnicity and mortality in the trans-

atlantic slave trade (London).Eltis, D. and D. Richardson 1997b. “The ‘numbers game’ and routes to slavery,” in Eltis and Rich-

ardson 1997, 1-15.Eltis, D. and D. Richardson forthcoming. An atlas of the transatlantic slave trade (New Haven, CT)Eltis, D., S. D. Behrendt, D. Richardson and H. S. Klein (edd.) 1999. The transatlantic slave trade: a data-

base on CD-ROM (Cambridge).Fennell, C. 2003. “Group identity, individual creativity, and symbolic generation in the BaKongo

diaspora,” Int. J. Hist. Arch. 7, 1-31. Fennell, C. 2007. Crossroads and cosmologies: diasporas and ethnogensisis in the New World (Florida).Fentress, E. 2005. “On the block: catastae, chalcidica and cryptae in Early Imperial Italy,” JRA 18, 220-34.Fentress, E., J. Bodel, F. Coarelli, P. Braconi and G. Pucci 2005. “Selling people: five papers on Roman

slave traders and the buildings they used,” JRA 18, 181-240. Ferguson, L. 1992. Uncommon ground: archaeology and early African America 1650-1800 (Washington,

DC).Ferris, I. 2000. Enemies of Rome: barbarians through Roman eyes (Stroud). Finley, M. I. 1968. “Slavery,” in D. Sills (ed.), International encyclopaedia of the social sciences vol. 14

(New York) 303-13.Fitzpatrick, A. 1989. “The uses of Roman imperialism by the Celtic barbarians in the later Repub-

lic,” in J. Barrett, A. Fitzpatrick and L. Macinnes (edd.), Barbarians and Romans in north-west Europe (BAR S471, Oxford) 27-54.

Fogel, R. W. and S. L. Engerman 1974. Time on the cross: the economics of American negro slavery (Bos-ton, MA).

Gager, J. G. 1992. Curse tablets and binding spells from the ancient world (New York).George, M. 2005. “Family imagery and family values in Roman Italy,” in ead. (ed.), The Roman family

in the Empire (Oxford) 37-66.Giacobello, F. 2008. Larari pompeiani: iconografia e culto dei lari in ambito domestico (Milan).Gomez, M. 1998. Exchanging our country marks: the transformation of African identities in the colonial and

antebellum South (Chapel Hill, NC).Goodman, A. et al. 2004. “Isotopic and elemental chemistry of teeth: implications for places of birth,

forced migration patterns, nutritional status, and pollution,” in Blakey and Rankin-Hill 2004, 216-65.

Hall, G. 1992. Africans in colonial Louisiana: the development of Afro-Creole culture in the eighteenth cen-tury (Baton Rouge, LA).

J. Webster64

Hall, M. 2000. Archaeology and the modern world: colonial transcripts in South Africa and the Chesapeake (London).

Handler, J. 1994. “Determining African birth from skeletal remains: a note on tooth mutilation,” His-torical Archaeology 28, 113-19.

Handler, J. S. 1996. “A prone burial from a plantation slave cemetery in Barbados, West Indies: possi-ble evidence for an African-type witch or other negatively viewed person,” Historical Archaeology 30, 76-86.

Handler, J. S. and F. W. Lange 1978. Plantation slavery in Barbados: an archaeological and historical inves-tigation (Cambridge, MA).

Handler, J. S. and J. Jacoby 1996. “Slave names and naming in Barbados 1650-1830,” William and Mary Quarterly 53, 685-728.

Harris, W. 1980. “Towards a study of the Roman slave trade,” in J. H. D’Arms and E. C. Kopff (edd.), The seaborne commerce of ancient Rome (MAAR 36) 117-40.

Harris, W. 1999. “Demography, geography and the sources of Roman slaves,” JRS 89, 62-75.Higman, B. 1995. Slave populations of the British Caribbean 1807-1834 (Kingston).Jackson, F. L. C. et al. 2004. “Origins of the New York African Burial Ground population: biological

evidence of geographical and macroethnic affiliations using craniometrics, dental morphology, and preliminary genetic analyses,” in Blakey and Rankin-Hill 2004, 150-215.

Jackson R. 2005. “Roman bound captives: symbols of slavery?” in N. Crummy (ed.), Image, craft and the classical world (Instrumentum Monog. 29, Montagnac) 143-56.

Jacobi, K., D. Cook, R. Corruccini and J. Handler 1992. “Congenital syphilis in the past: slaves at Newton Plantation, Barbados, West Indies,” Am. J. Phys. Anth. 89, 145-58.

Jones, C. P. 1987. “Tattooing and branding in Graeco-Roman antiquity,” JRS 77, 139-55.Langner, M. 2001. Antike Graffitizeichnungen (Wiesbaden).La Roche, C. J. and M. L. Blakey 1997. “Seizing intellectual power: the dialogue at the New York Afri-

can burial ground,” Historical Archaeology 31, 84-106.Lilley, I. 2004. “Diaspora and identity in archaeology: moving beyond the black Atlantic,” in

L. Meskell and R. Preucel (edd.), A companion to social archaeology (Oxford) 287-312.Lilley, I. 2006. “Archaeology, diaspora and decolonization,” J. Social Arch. 6, 28-47.Morgan, P. 1997. “The cultural implications of the Atlantic slave trade: African regional origins,

American destinations and New World developments,” in Eltis and Richardson 1997a, 122-45.Morris, I. 1998. “Remaining invisible: the archaeology of the excluded in Classical Athens,” in

S. J. Joshel and S. Murnaghan (edd.), Women and slaves in Greco-Roman culture (London) 193-220.Mouritsen, H. 2005. “Freedmen and decurions. Epitaphs and social history in Imperial Italy,” JRS

95, 38-63.Mouritsen, H. forthcoming. “The inscriptions of the insula of the Menander,” in H. Mouritsen, J. M.

Reynolds and A. Varone, The insula of the Menander at Pompeii, vol 5. The wall inscriptions (Oxford).Norman, N. 2009. “Powerful pots, humbling holes and regional ritual processes: towards an archae-

ology of Huedan Vodun, ca. 1650-1727,” African Arch. Review 26, 187-218.Patterson, O. 1982. Slavery and social death: a comparative study (Cambridge, MA).Perry, W. and R. Paynter 1999. “Artifacts, ethnicity, and the archaeology of African Americans,” in

T. Singleton (ed.), I, too, am America: archaeological studies of African-American life (Charlottesville, VA) 299-310.

Rankin-Hill, L. 1997. A biohistory of 19th century Afro-Americans. The burial remains of a Philadelphia cem-etery (Westport, CT).

Rothschild, N. 2008. “Colonised bodies, personal and social,” in J. Robb and D. Boric (edd.), Past bod-ies (Oxford) 135-44.

Roymans, N. 2004. Ethnic identity and imperial power: the Batavians in the Early Roman Empire (Amsterdam).

Safran, W. 1991. “Diasporas in modern societies: myths of homeland and return,” Diaspora 1, 83-99.Salas, A. et al. 2004. “The African diaspora: mitochondrial DNA and the Atlantic slave trade,” Am. J.

Human Genetics 74, 454-65.Saller, R. P. and B. D. Shaw 1984. “Tombstones and family relations in the Principate: civilians, sol-

diers and slaves,” JRS 74, 124-56.Samson, R. 1989. “Rural slavery, inscriptions, archaeology and Marx: a response to Ramsay MacMul-

len’s ‘Late Roman Slavery’,” Historia 38, 99-110.Scheidel, W. 1997. “Quantifying the sources of slaves in the Roman Empire,” JRS 87, 159-69. Scheidel, W. 2005. “Human mobility in Roman Italy, II: the slave population,” JRS 94, 1-26.Scheidel, W. 2007. “The Roman slave supply,” Version 1.0 May 2007 Princeton/Stanford Working

Routes to slavery in the Roman world: a comparative perspective 65

Papers in Classics. http://www.princeton.edu/~pswpc/pdfs/scheidel/050704.pdfShepherd, N. 2007. “Archaeology dreaming. Post-apartheid urban imaginaries and the bones of the

Prestwich Street dead,” J. Social Arch. 7, 3-28.Solin, H. 1977. “Die Namen der orientalischen Sklaven in Rom,” in N. Duval and H.-G. Pflaum

(edd.), L’onomastique latine (Paris) 205-20.Solin, H. 1996. Die stadtrömischen Sklavennamen. Ein Namenbuch I-III (Stuttgart).Solin, H. 2001. “Griechische und römische Sklavennamen. Eine vergleichende Untersuchung,” in

H. Bellen and H. Heinen (edd.), Fünfzig Jahre Forschungen zur antiken Sklaverei an der Mainzer Akademie, 1950-2000. Miscellanea zum Jubilaeum (Stuttgart) 307-30.

Solin, H. 2003. Die griechischen Personennamen in Rom (2nd edn., Berlin).Snowden, F. M. 1970. Blacks in antiquity: Ethiopians in the Greco-Roman experience (Cambridge, MA).Thompson, H. 2003. The archaeology of Greek and Roman slavery (London 2003).Thornton, J. 1992. Africa and Africans in the making of the Atlantic World 1400-1680 (Cambridge). Tomlin, R. 2003. “‘The girl in question’: a new text from Roman London,” Britannia 34, 41-51.Trümper, M. 2009. Graeco-Roman slave markets: fact or fiction? (Oxford).Ugo Nwokeji, G. and D. Eltis 2002a. “Characteristics of captives leaving the Cameroons, 1822-1837,”

J. African Hist. 43, 191-210.Ugo Nwokeji, G. and D. Eltis 2002b. “The roots of the African diaspora: methodological considera-

tions in the analysis of names in the Liberated African Registers of Sierra Leone and Havana,” History in Africa 29, 365-79.

Webster, J. 2005. “Archaeologies of slavery and servitude: bringing ‘New World’ perspectives to Roman Britain,” JRA 18, 161-79.

Webster, J. 2008. “Less beloved. Roman archaeology, slavery and the failure to compare,” Arch. Dia-logues 15, 103-23.

Westermann, W. L. 1955. The slave systems of Greek and Roman antiquity (Philadelphia, PA).Wiedemann, T. 1981. Greek and Roman slavery (London).