Historical accounts of mobility

Mediterranean textual accounts leave little doubt that mobility was frequent, especially for the elite men who wrote the overwhelming majority of sources. The detailed character of mobility is considerably more complex to understand. It appears that much movement was local and short-term, whether due to labour commitments (transhumance, war, trading, or elite travel between multiple residences) or social obligations (visiting family, friends, or patrons, religious journeys, or civil commitments).Footnote ²⁴ Elites, those connected to the military and enslaved people were likely to have experienced longer-term mobility, over longer distances, but patterns are challenging to discern, and some orthodoxies are being challenged by scientific studies.Footnote ²⁵ Maritime transport was the key mode of longer-distance travel, but was highly seasonal, being concentrated in the sailing season of May to September. In a recent overview, Woolf has argued for the significant majority of the Roman period Mediterranean population being ‘stayers’, with a relatively small percentage of ‘movers’.Footnote ²⁶ He emphasises that this is not a binary; stayers may well be locally mobile and would have been aware of wider connectivity of others in their communities. The picture that emerges from his summary is one of a highly connected Roman Mediterranean world where only around 5% of the population were mobile beyond their immediate locality. However, these individuals and groups had significant transformative effects on communities they encountered. Recent bioarchaeological work in the Mediterranean has begun to demonstrate great variability in the proportion of the population who could be considered ‘movers’, with much higher rates of mobility at Portus and in imperial period Rome than in rural areas, as might be expected, but also significant differences between other urban sites.Footnote ²⁷ Sample sizes remain small, and combined aDNA and isotope studies very rare, however, limiting the impact of these results in transforming narratives in the wider field. Nevertheless, they provide an important challenge to narratives based primarily on epigraphic and textual evidence, which reproduce strong gender and class biases.

For Britannia, textual sources allow us glimpses into mobility, some of which echo the consensus on the character of mobility in the Roman Mediterranean. The Claudian invasion saw significant numbers of legionaries and auxiliaries cross to Britannia, although estimates vary between 30,000 and 40,000 soldiers.Footnote ²⁸ The overall number of people moving into Britannia was probably somewhat higher, when one considers accompanying enslaved people, merchants, camp followers and families, albeit these may not have followed until a little later. This was the largest textually attested mass movement of people in the early Roman period relating to Britannia. The other main examples are the invasions by Julius Caesar,Footnote ²⁹ the possible movement to Britain of the Gallic leader Commios and his followers in the Late Iron Age as a consequence of the Gallic Wars,Footnote ³⁰ and, later, the movements of the Imperial field army to Britain in the third and fourth centuries, of which Septimius Severus’ army was the largest.Footnote ³¹ Also notable in the literature is Marcus Aurelius’ transfer of 5,500 Sarmatians to Britannia in the a.d. 170s, especially given the recent discovery of an individual of late second-century a.d. date with Sarmatian ancestry in Cambridgeshire.Footnote ³² Mobility from Britannia into mainland Europe is attested in the early Roman period through redeployment of major military units, and more often in the later period by the crossing of troops from Britain to the continent in support of various claimants of the imperial throne.Footnote ³³ The majority of these episodes are military, which reflects both the bias of written sources towards military history and the historical reality that a large proportion of the empire’s resources were dedicated to mobilising the army according to the priorities of the day; this military-associated large-scale movement regularly included large numbers of people from defeated communities.Footnote ³⁴

Beyond these larger-scale movements of people, a handful of individual journeys to and from Britannia are attested in the literary evidence. Other than those by Emperors, perhaps the most famous are those of Paul ‘the Chain’ in the mid-fourth century,Footnote ³⁵ and although somewhat after the end of Roman imperial control of the province, Saint Germanus of Auxerre in a.d. 429 and possibly a second occasion some years later.Footnote ³⁶ There are very few mentions of more ordinary journeys to or from Britannia in literary evidence, and thus we turn to epigraphy; the Vindolanda tablets, for example, show us that soldiers from the first cohort of Tungrians, based at Vindolanda, are recorded as being in Gaul.Footnote ³⁷

Epigraphy and textual sources

Epigraphy has been one of the primary means of engagement with mobility in the Roman world, not least because it is one of the only ways in which we can engage with biographical details of named individuals. In inscriptions individuals could represent themselves (or be represented by others), conveying what they wished to be known about their identities, their origins and their life courses. While published estimates of the total number of inscriptions from Britannia differ significantly,Footnote ³⁸ and the body of evidence is comparatively small for a Roman province, it is well documented.Footnote ³⁹ It is shaped by the limited spread of the epigraphic habit geographically, culturally and socially, and by practical factors such as the availability of good-quality stone for carving and the survival, reuse and recovery of inscriptions in the archaeological record.Footnote ⁴⁰ Epigraphic studies which have considered inscriptions solely as texts have also sometimes been disciplinarily isolated, though interdisciplinary approaches increasingly situate epigraphic materials within their wider archaeological, social and cultural context.Footnote ⁴¹

The distribution of surviving inscriptions from Roman Britain (both funerary and otherwise) is strongly correlated with military sites, veteran colonies and other towns with military or official associations.Footnote ⁴² The epigraphic habit does not seem to have been very widely adopted outside of these settings, perhaps because there had been no pre-existing tradition of writing or epigraphy in Britain before the conquest.Footnote ⁴³ The practice of epigraphy was therefore further restricted primarily to the Latin language — though the few exceptions to this rule are sometimes particularly illustrative (e.g. Barates, discussed further below). These restrictions mean that epigraphic studies of mobility have of necessity focused primarily on military and urban contexts. Tombstones dedicated to soldiers comprise the majority of such inscriptions in Britannia.Footnote ⁴⁴ A study of public votive epigraphy similarly found that where the occupation of the dedicant could be identified, 43% of the dedications were made by soldiers.Footnote ⁴⁵ Soldiers were an inherently mobile population, particularly in the early Roman period (somewhat less so as the frontiers settled, units became more stationary and local recruitment increased). Levels of mobility are difficult to assess, but where soldiers’ origins are recorded they are diverse: Rowland’s analysis of 94 inscriptions records 69 unique places of origin from across the empire.Footnote ⁴⁶ Origins differed depending on the form of service to which a soldier was recruited. The requirement of citizenship meant that for legions in the western provinces recruitment was initially largely drawn from Italians and those of Italian descent, while auxiliaries were primarily recruited from particular regions and drew from local recruitment.Footnote ⁴⁷

In funerary inscriptions, too, it was common among certain groups — particularly soldiers, officials and traders — to record the ethnic, provincial, or regional origin of the deceased.Footnote ⁴⁸ While these dedications were made by the living, it is an indication that the deceased’s origins were known and asserted during life, and it has been noted that where the identity of the living dedicant is known, this was often similar to that of the deceased.Footnote ⁴⁹ Interestingly, figurative tombstones dedicated by auxiliaries often asserted a visibly Roman military identity, emphasising their belonging to the community of the Roman army and their achievement of citizenship, such as the auxiliary veteran Dannicus, originally from Germania Superior but commemorated in Cirencester, whose tombstone shows him riding down his enemies,Footnote ⁵⁰ or Crotus, veteran of the Fourth Cohort of Gauls, shown possibly holding a will which only a Roman citizen had the right to contract.Footnote ⁵¹

Individuals with origins outside of Britannia also form a significant sub-group of the civilian corpus of inscriptions.Footnote ⁵² Recording a place of origin or ethnic affiliation clearly played an important role in the assertion of identity for those who found themselves living and working far from the place in which they were born. At Silchester, three inscriptions referring to a collegium peregrinorum (an association for non-citizens) perhaps suggest the importance of this self-definition.Footnote ⁵³ However, it should be noted that the proportionally higher representation of immigrants to the province in Romano-British epigraphy is influenced by a range of factors and is almost certainly not representative of the broader population. People from other parts of the empire where the epigraphic habit was more established were more likely to dedicate inscriptions, and aligning with a non-local origin in this frontier province may have indicated a higher status.Footnote ⁵⁴

The mobility of Britons within Britannia is noted epigraphically only in rare instances; the best known example is the funerary monument of Regina, a Catuvellaunian freedwoman commemorated at South Shields.Footnote ⁵⁵ The inscription and figurative depiction of Regina’s memorial articulate both her origins and status at the time of her death.Footnote ⁵⁶ It was dedicated by her husband and former owner, Barates, who was from Palmyra; the inclusion of a Palmyrene coda to the monument’s Latin inscription suggests that this identity remained important to him.Footnote ⁵⁷ The non-local origins of both Regina and Barates provide a tantalising glimpse into the extent of mobility in Britannia, particularly at its frontiers, and the complex relationships and power asymmetries that could arise from the interactions of mobile individuals.Footnote ⁵⁸

Inscriptions commemorating individuals with British ethnicities or places of origin are largely associated with the military, but this limited body of evidence is probably not representative of the real extent of Britons’ mobility either within or outside the province.Footnote ⁵⁹ Epigraphic evidence for mobility of Britons outside of Britannia is similarly limited, with c. 40 inscriptions and military diplomas of individuals with British origins known from the rest of the empire.Footnote ⁶⁰ Again, this does not represent the full extent of the mobility of British individuals throughout the Roman empire, only the surviving evidence.

The Roman state generated a vast quantity of written documentation, the remnants of which survive in rare taphonomic contexts, given that biodegradable materials were normally used.Footnote ⁶¹ Literacy was not restricted to the official and military class, and though estimations of the extent of literacy among the population vary widely, written evidence offers another key dataset for the exploration of mobility in the Roman period.Footnote ⁶² London’s Bloomberg Tablets provide early evidence; dating from the a.d. 50s to 80s, they provide evidence for around 92 named persons, all male, many of whom can be identified as mobile, though their identification as such is largely contextual: members of non-citizen auxiliary units of the Vangiones and Lingones, or citizen officers such as Julius Classicus, commander of the Sixth Cohort of Nervians.Footnote ⁶³

The Vindolanda tablets are the most famous corpus of written documentation of the Roman period in Britannia and provide contemporary evidence of the broad inter- and intra-provincial networks that characterised the Roman period, and of the army as a driver of mobility during this period. The documents date to the late first century and early second century, during which Vindolanda was garrisoned by Tungrian and Batavian auxiliaries; more than 60 known texts relate to the correspondence of Flavius Cerialis, the prefect of the Ninth Cohort of Batavians.Footnote ⁶⁴ Auxiliary units were composed of non-citizen soldiers, often recruited from the same community or region (at least initially — while such recruitment characterised the early formation of auxiliary units, this practice did not persist through the life of the unitFootnote ⁶⁵ ) and intentionally stationed away from their regions of origin.Footnote ⁶⁶ As well as strengthening the Roman army, targeted recruitment from particular groups also removed young men of fighting age from their home communities — an example of the wider Roman imperial practice of forced mobility as a tool of state control.Footnote ⁶⁷ The distribution of military diplomas across the Roman world shows that some auxiliary troops settled in the provinces to which they had been posted after their term of service was complete, though others returned to their homelands. This serves as a reminder that while mobility is often conceptualised as a binary between two points of a journey, movement in the Roman world need not entail a permanent exile and the individual lifecourse could encompass a range of mobilities.Footnote ⁶⁸

Though the Vindolanda tablets largely deal with soldiers and military affairs, there are references to other mobile individuals. Notably, one individual identifies himself as a hominem trasmarinum, a man from overseas.Footnote ⁶⁹ The letter is an appeal for redress for a beating received from a centurion, and the petitioner’s distinguishing of himself as from overseas may be an attempt to contrast his higher status and rights (whether legal or perceived) with that of the local British population, who are referred to derogatorily elsewhere in the tablets as Brittunculi, ‘wretched little Britons’.Footnote ⁷⁰

Beyond the Vindolanda tablets, some curse tablets hint at mobility into Britannia, chiefly through names. Almost all the personal names used therein are Latin, Celtic, or a combination of the two, and it is of course impossible to discern whether these individuals originated in Britannia or otherwise, although they should not necessarily be assumed to be locals.Footnote ⁷¹ ‘Roman’ names might be borne by those from Britannia or anywhere else within the empire, or even beyond, and some of the ‘Celtic’ names from tablets are paralleled in other provinces.Footnote ⁷² A curse tablet from the south Wiltshire temple bears the unusual Greek name of Hegemon, suggesting a non-native of Britannia, like the four names of Greek origin in the Bath tablets, and various Greek names from Roman London and elsewhere in the province.Footnote ⁷³ Greek names appear infrequently but regularly in other epigraphic media, although not all the people who bore them were necessarily Hellenic; Webster reminds us that enslaved people might be given a Greek name unrelated to their origin.Footnote ⁷⁴ The complexities of defining migration through onomastics are a salutary reminder of the limitations of reductive binaries in categorising people in Britannia.

Material culture

Until recent decades the identification of people moving into Britannia from the wider empire and beyond has largely depended on material culture. This has generally been from grave assemblages, though other forms of material culture have been used to explore the mobility of certain groups or individuals, such as craft workers with specialties which had little or no existing tradition in pre-conquest Britain, including stonemasonry, mosaic-making, or large-scale pottery production.Footnote ⁷⁵ It is also important to consider the ways in which the larger-scale urban and rural material infrastructure of Britannia, public or private, has also been used to argue for mobility within, or into, the province.

The transport of people, goods, animals and information throughout the Roman world was crucial to the functioning of the Roman state and the endurance of its imperial power. This was facilitated by a transport infrastructure which encompassed maritime, riverine and road transport, connecting areas under Roman control with the central authority.Footnote ⁷⁶ In Britannia, as in other provinces, the road network was systematically expanded as a means of bringing the province fully under Roman authority.Footnote ⁷⁷ The road network also supported the cursus publicus, the official communications infrastructure comprising rest and relay stations along major routes, which enabled the movement of state communications, government agents and goods — all at provincial expense.Footnote ⁷⁸

While the Roman road network was extensive, road transport was still onerous and expensive and it is often assumed that maritime or riverine transport would have been preferred where possible. As an island province, Britannia would have been particularly dependent on maritime connections with the wider Roman world, yet despite this there are few Roman shipwrecks known from British waters and none are of Mediterranean origin.Footnote ⁷⁹ Glimpses of the scale of cross-channel trade are provided by samian from the Pudding Pan wreck, though the ship’s exact location remains unknown. The Blackfriars 1 ship sank carrying ragstone from Kent to London, but may also have been seagoing, given the great volume of activity at the Roman port of London.Footnote ⁸⁰ The mobility of individuals engaged in ongoing cross-channel trade is further illustrated by the negotiator Britannicianus L. Viducius Placidus’ dedication of altars both at the shrine of Nehalennia near Colijnsplaat and at York,Footnote ⁸¹ and by the dedication in London by a moritix Londiniensium (a term of Celtic origin which seems to have referred to cross-channel traders).Footnote ⁸²

When considering the larger-scale archaeological evidence for mobility, a relatively rich seam of evidence is provided by the rural residences of the elite: villas. Influential scholars in the twentieth century argued that some architectural expressions of material wealth at villas in late Roman Britannia were such a departure from the level of previous rural wealth that they must have been stimulated by capital flight from Gaul. Rivet, Branigan, Applebaum and other mid-century villa scholars argued that the third-century crisis provoked migration by wealthy members of Gaul’s decurion class, who brought new forms of architecture, sculpture and, perhaps, social organisation, to the British countryside.Footnote ⁸³ In this framing, although not the chronology, their thesis echoes that of Ling and others in the field of art in Britannia, who essentially argued that the higher-quality sculpture and wall plaster art of the province was too sophisticated to be the product of British artists until a period of training and diffusion by immigrants from the continent, if it could be at all.Footnote ⁸⁴ There are now serious issues with the argument for migration triggering the floruit of villas in Britannia; the chronology of villas is now better established and simply does not support the relatively precise timeframe posited for the greater villas of Britannia being established. Rather, continuity is very much better evidenced almost wherever sites have been systematically investigated, and larger villas emerge at different times across the province.Footnote ⁸⁵ This is not to argue that villa owners could not be individually mobile; indeed, the only named landowner of Britannia, Melania the Younger, owned estates widely across the empire, and it is highly likely that other members of the highest social strata in Britannia moved between provinces.Footnote ⁸⁶

We can also consider those who served the elite of Britannia. Skilled workers were a mobile class in the Roman world, and their movements are well attested in the historical and epigraphic record. Medical practitioners in particular are known to have moved extensively throughout the empire; the close association of medicine with Greekness in the Roman world suggested mobility (whether real or implied to align with this expectation).Footnote ⁸⁷ There are several inscriptions in Britannia dedicated by doctors, inscribed in Greek by dedicants with Greek namesFootnote ⁸⁸. The distribution of collyrium stamps across the northern provinces may also suggest the movement of oculists and other medical practitioners on a more localised scale.Footnote ⁸⁹

Craftspeople also travelled significant distances to work on elite or state structures — Priscus, a stonemason from Gaul died in Bath (RIB 149), and artisans were sent from Britannia for the rebuilding of Autun in a.d. 296.Footnote ⁹⁰ However, itinerant workers, if not commemorated in epigraphy, may leave little evidence for mobility except in any influences or connections which may be identified in their products. The classic example from Britannia is the identification of so-called ‘schools’ or workshops of mosaicists from stylistic similarities and geographic distributions of their work.Footnote ⁹¹ Smith’s initial schema of urban officinae has been modified in the light of discoveries and meticulous scholarship to become a model of mobile groups of mosaicists moving around broad regions, overlapping and occasionally collaborating.Footnote ⁹² Some, such as those making black-and-white mosaics around Ilchester, ranged less broadly, whereas others, such as the Midland group, worked more widely. ErdkampFootnote ⁹³ discusses the importance of seasonal and temporary migration of labour, which is much more difficult to identify — ancient authors did not concern themselves with the movements of unskilled workers, and they leave fewer traces archaeologically.Footnote ⁹⁴ These studies are directly concerned with the seasonal migration of rural Italian smallholders to the city of Rome, and little comparable work has been undertaken on seasonal labour migration in Britannia, although local transhumance is well attested in later periods.Footnote ⁹⁵

Certain classes of material culture have been repeatedly used to investigate mobility in Britannia. Pottery has been a particular focus, as a product in its own right and as a proxy for various aspects of economic and cultural mobility: the movement of skilled potters following particular markets; the movement of goods and traders through the transport infrastructure of the Roman world; and latterly as evidence of cultural and social practices around food and drink which may indicate the presence of non-local groups.Footnote ⁹⁶ Legionary pottery at Usk, for example, was made in local fabrics but in forms more closely paralleled in the Rhineland. One explanation offered is that the pottery was manufactured by mobile potters (whether free or enslaved) catering to the legionary market, or by members of the legion itself, which had been stationed in Germany prior to the conquest of Britannia.Footnote ⁹⁷ Swan’s identification of distinctive vessels resembling North African types at sites in northern Britannia explores the presence of North African soldiers in the frontier zone, potentially linked to Legio VI Victrix and the Severan campaigns of the third century.Footnote ⁹⁸ The vessel forms — deep, ‘casserole’ dishes — suggest distinctive traditions of food preparation,Footnote ⁹⁹ though the extent to which the presence of these vessels indicates the presence of North Africans, or the transmission of a style which became increasingly popular across the Mediterranean, is debated.Footnote ¹⁰⁰

The presence of exotic materials has also been used to distinguish burials of potential non-local individuals, most notably the ‘Ivory Bangle Lady’, a high-status burial from York, whose burial included bangles in ivory and jet, an openwork mount bearing a Christian salutation, glass objects and glass and amber jewellery. Though her material assemblage marks her as distinctive, craniometric ancestry assessment (using FORDISC 3.0) suggested a mixed North African ancestry based on statistical comparisons with modern populations. However, using this approach, for archaeological material especially, is potentially problematic and isotope analysis was less conclusive and suggested regions across the Mediterranean or western Europe to be compatible.Footnote ¹⁰¹ An ivory leopard clasp knife also formed part of the high-status grave assemblage of an adolescent girl in Southwark, London, whose strontium and oxygen isotope results suggest a southern Mediterranean origin.Footnote ¹⁰² ‘Exotic’ materials such as ivory are not necessarily directly linked with North African heritage or origins, but perhaps indicate alignment with an elite identity in which cross-provincial connectivity and mobility was common.Footnote ¹⁰³ Such burials also highlight the potential mobility of women in the Roman world, an aspect which is often difficult to identify otherwise, given the limited means for women to convey agency publicly in Roman society.Footnote ¹⁰⁴ Variation in the distribution of beads and glass bracelets also suggests that female dress was more regionally distinctive.Footnote ¹⁰⁵ However, a note of caution is required in over-relying on oxygen isotope data to establish origins, particularly from warmer areas of southern Europe, Africa and Asia. There is increasing evidence that high oxygen isotope values that have in the past sometimes been interpreted as evidence for Mediterranean originsFootnote ¹⁰⁶ actually commonly occur in western Britain.Footnote ¹⁰⁷ In short, understanding of local and regional ranges is constantly changing as more data comes to light. In addition, issues of method standardisationFootnote ¹⁰⁸ and the effects of foodwaysFootnote ¹⁰⁹ on elevating ratios makes oxygen isotope analysis a complex proxy.

Brooches have also been widely employed in attempts to distinguish regional or ethnic origins.Footnote ¹¹⁰ Their primary use as fasteners for clothing renders them highly visible objects, conveying information about the personal identitie(s) of the wearer(s).Footnote ¹¹¹ While brooches were in use in Late Iron Age Britain, their use became more widespread following the conquest, and particular classes of brooch, such as the Aucissa, were certainly brought into Britain with the army.Footnote ¹¹² The use of certain forms of brooch to signal military identities persists throughout the Roman period, most notably with the later Roman crossbow brooch — a form which was widespread throughout the western empire and closely linked in iconography with military and administrative status.Footnote ¹¹³ Crossbow brooches form part of a material suite that has been linked to late Roman military dress, also including belt sets,Footnote ¹¹⁴ which are sometimes found in furnished inhumations in the north-west provinces. These were commonly identified as individuals recruited into the army from beyond the imperial frontiers (foederati), though this interpretation has been critiqued.Footnote ¹¹⁵ Graves furnished in this fashion are limited in Britannia, and the direct nature of the link between this material and the ethnic origin of the person with whom it was associated has been challenged.Footnote ¹¹⁶ Crossbow brooches and belt sets likely displayed status, prestige and power through claimed or actual links to Roman authority; the mixed origins of groups wearing them reflects the nature of late Roman military and state personnel.Footnote ¹¹⁷

Where small finds have been used to explore the extent of mobility in Britannia this has generally been in the context of burials and their associated material culture as a site of final deposition. In particular, this kind of study has focused on ‘intrusive’ burials and their associated assemblages, which are atypical in their immediate context but may have parallels in a distant context.Footnote ¹¹⁸ It should be noted, however, that identification of the ‘local’ mortuary practice often rests on an absence of distinctive material culture in contrast to the intrusive groups,Footnote ¹¹⁹ a somewhat recursive argument.

The cemetery at Lankhills, Winchester, is perhaps the best known example of the identification of immigrants to the province through artefact and mortuary practice evidence and forms a case study for the ways in which the identification and study of mobility in Roman Britain has developed over past decades. Lankhills forms part of the cemetery of the civitas capital of Winchester. Excavations from 1967–72 produced 451 burials;Footnote ¹²⁰ and a series of excavations by Oxford Archaeology 2000–05 produced a further 307 inhumations.Footnote ¹²¹ Clarke identified intrusive groups of fourth-century graves, defined by both burial rite and grave goods: in particular, one group of 16 burials whose grave goods and funerary practice were deemed to have parallels in Pannonia. While there was variation even within this group, consistent characteristics included bronze bracelets worn around the left wrist (in the case of females), cross-bow brooches at the right shoulder and belt-fittings at or near the waist (in the case of males), and coins provided but never placed in the mouth. Additionally, beads without parallel in Britain were found in several of the female graves around the neck.Footnote ¹²² While the criteria for distinguishing these graves have been debated,Footnote ¹²³ the conclusion that these represented a distinctively different community was widely cited in literature on Romano-British archaeology, burial practice and migration.Footnote ¹²⁴ The identification of distinct burial rites and the search for parallels across long distances as proof of non-local origin is problematic, however, given factors such as the taphonomic processes at work on burial assemblages. At Lankhills in particular, subsequent isotope analysis (explored in full below) has strikingly demonstrated the insecurity of identifying migrants by these means.

All this should caution against a direct correlation between material culture and ethnicity. Material culture and burial rites are the product of complex interactions of numerous factors.Footnote ¹²⁵ Similarly, other characteristics — profession (e.g. military), age, gender, religion — also play an important role in the formation of identities, and may be expressed through material culture and burial rite.Footnote ¹²⁶ Furthermore, the durability of small finds in metal meant that their use-lives could continue far beyond their period of manufacture: Roman objects are found in Anglo-Saxon burials,Footnote ¹²⁷ and other objects were reworked and recycled into new forms.Footnote ¹²⁸ The meaning and cultural associations of these objects likely shifted over the period of their use.

Isotope analysis

As discussed above, isotope analyses make it possible to test hypotheses about mobility within, and into, Britannia, although this kind of direct evidence has only been possible in recent decades. This discussion will introduce the principal techniques and summarise the state of the field of isotope studies on mobility in Britannia.

Oxygen (δ ¹⁸O) and strontium (⁸⁷Sr/⁸⁶Sr) isotope analyses on bioapatites (e.g. dental enamel) are among the most commonly used proxies to study mobility in past human populations. Oxygen isotope ratios (δ ¹⁸O) mainly reflect the oxygen in ingested water, secondly that from respiration and thirdly from the metabolism of food.Footnote ¹²⁹ The δ ¹⁸O value of the water ingested is controlled by evaporation and condensation processes that are influenced by several environmental and anthropogenic drivers. The location of the source water (e.g. distance from coast, altitude, etc.), and its environment (humidity, seasonality, temperature, etc.) play a vital part in this.Footnote ¹³⁰ For example, waters from warmer areas are typically enriched in ¹⁸O as the lighter ¹⁶O will evaporate first, resulting in higher δ ¹⁸O values.Footnote ¹³¹ The δ ¹⁸O value in an individual thus reflects the geographical location of their water source. Comparing the individual’s δ ¹⁸O value with what is expected around their burial environment will therefore provide information on whether the individual is compatible with the region or not.Footnote ¹³² The assumption that past populations obtained their water locally is vital in this. Additionally, human interferences, such as food and beverage preparation,Footnote ¹³³ nursing and weaning,Footnote ¹³⁴ or physical activity,Footnote ¹³⁵ have also been shown to affect δ ¹⁸O values.

Strontium isotope (⁸⁷Sr/⁸⁶Sr) ratios in an individual mainly reflect the ⁸⁷Sr/⁸⁶Sr of the plant part of the diet.Footnote ¹³⁶ These plants primarily reflect the bioavailable fraction of the strontium in the soil in which they were cultivated. The soil, in turn, obtains its strontium through the weathering of the underlying lithology. The ⁸⁷Sr/⁸⁶Sr ratio of the lithology depends on the type of bedrock, the original concentrations of rubidium 87 (which decays into ⁸⁷Sr), strontium 87 and strontium 86, and age of the bedrock. For example, in general, plants cultivated on younger bedrocks such as basalt will have lower ⁸⁷Sr/⁸⁶Sr values than those cultivated on older bedrocks such as granite.Footnote ¹³⁷ Plant ⁸⁷Sr/⁸⁶Sr may also be affected, however, by environmental aspects such as the ‘forest effect’, which can elevate ratios,Footnote ¹³⁸ dust,Footnote ¹³⁹ or sea spray.Footnote ¹⁴⁰ The latter is due to the continuous mixing of oceanic and continental bedrock strontium throughout the ocean which gives it a global ⁸⁷Sr/⁸⁶Sr value of about 0.7092.Footnote ¹⁴¹ Consequently, the ⁸⁷Sr/⁸⁶Sr ratios and concentrations of the plants growing within c. 5 km (depending on climatic factors) from the ocean,Footnote ¹⁴² or foods preserved in sea salt,Footnote ¹⁴³ are substantially affected by the marine environment.

One of the earliest isotope studies on Roman Britannia by Budd et al. combined oxygen, strontium, and lead isotope (Pb) analyses on human dental enamel from Roman Winchester (n = 4) to explore the use of such isotope systems in mobility studies.Footnote ¹⁴⁴ This pilot study was a success, demonstrating that diagenetic alterations had a minimal effect on the isotope results. The combination of three isotope systems thus provided key knowledge on the individual’s environment during life, and using a combination of these isotope systems (with or without lead) is still routine in mobility studies today. As our understanding of the effects of anthropogenic activities and environmental processes on isotope data improves and as our knowledge of biosphere variation across isotope proxies advances, so do the interpretations. For example, after more research into climate and oxygen isotopes, Budd et al. revised their original interpretations made in 2001,Footnote ¹⁴⁵ in which climate change was considered to have caused the human δ ¹⁸O values being higher than the environment. Instead, mobility from the western Mediterranean or North African coastal area became the more likely explanation for all four individuals. These studies around the beginning of the twenty-first century were important in the progress of the use of isotopes in mobility studies in general, but the findings also gave key insights into the lives of individual inhabitants of Britannia.

Isotope analysis is especially helpful in testing assumed correlations between ‘exotic’ material culture and mobility. A good example is the intrusive group at the cemetery at Lankhills, Winchester, introduced above.Footnote ¹⁴⁶ Evans and colleaguesFootnote ¹⁴⁷ tested the hypothesis of non-local individuals at the site by analysing oxygen (δ ¹⁸O) and strontium (⁸⁷Sr/⁸⁶Sr) isotopes in 18 fourth-century individuals, of whom 9 were in the ‘Pannonian’ group. Only 5 out of the 9 buried with such objects, and 2 of the 9 buried with traditions more local to Lankhills, had isotope results inconsistent with being raised in Britannia. Interestingly, the authors identified the origins of the individuals to be dispersed throughout central Europe and as far south as Italy. The absence of a link between origin and burial rites was later confirmed by Eckardt et al. Footnote ¹⁴⁸ This demonstrates that the presence of non-local objects within a burial does not mean the individual themselves originated from elsewhere. Cultural habits may have been passed on to the following generations or fused with the habits already present in Roman Britain.

Eckardt’s Diaspora Communities in Roman Britain projectFootnote ¹⁴⁹ found that a substantial number of individuals buried in York,Footnote ¹⁵⁰ Catterick,Footnote ¹⁵¹ GloucesterFootnote ¹⁵² and WinchesterFootnote ¹⁵³ were likely not local, based on strontium and/or oxygen isotope results. Some of them probably came from outside Britannia and some (from York) were hypothesised to have originated from as far as Northern Africa, based primarily on cranial morphometric techniques (FORDISC 3.0), in combination with artefacts found at the burial site and elevated oxygen isotope values. Importantly, the individuals identified to be non-locals were a mixed group of men, women and children. The results of the Diaspora project confirmed that mobility and migration were important in shaping the cultural landscape of urban sites in Britannia, and that migration encompassed a range of demographics.

One isotope study of Roman migration in Britannia, however, identified a different pattern of migration to others previously published. Most of the individuals buried at Catterick were compatible with the same region, based on their oxygen and strontium isotopes.Footnote ¹⁵⁴ While acknowledging the presence of similar isotope ranges across central Europe, Chenery et al. concluded local origins to be the most likely explanation despite the presence of Germanic or Danubian artefacts. A subsequent study by Crowder et al. Footnote ¹⁵⁵ aimed to determine a more precise origin for individuals buried with Danubian artefacts in Britannia, whether they were considered non-local or not by Chenery et al. (Reference Chenery, Eckardt and Müldner2011). Using oxygen and strontium isotope data, Crowder and colleagues suggest the Transylvania region to be the most appropriate origin for many of the individuals from the Diaspora project with similar isotope values. In general, it is largely accepted that multiple regions within Europe and beyond may overlap in oxygen and strontium isotope ratios, and more comprehensive studies on bioavailability in the Pannonian region are required to support these interpretations. Nonetheless, the presence of eastern European artefacts on sites in Britannia emphasises the importance of considering Empire-wide mobility, and addressing the challenges of equifinality in isotope biosphere mapping.

Towns which were higher in Britannia’s administrative hierarchy (e.g. the citizen settlements (coloniae), and to a somewhat lesser extent the civitas capitals), can be expected from historical and archaeological evidence to have had more mobile populations due to their greater connectivity with the wider empire. Isotope studies by Millard et al.,Footnote ¹⁵⁶ Redfern et al. Footnote ¹⁵⁷ and Shaw et al. Footnote ¹⁵⁸ undertaken on cemeteries in Roman London have corroborated this, identifying that about 25%, 29% and 7–15% of the respective study samples had oxygen, or strontium and lead isotope ratios compatible with origins in continental Europe or the Mediterranean. Other large urban sites, such as York,Footnote ¹⁵⁹ GloucesterFootnote ¹⁶⁰ and Winchester,Footnote ¹⁶¹ show similar proportions of non-local individuals. In contrast, fewer migrants have been identified at rural sites, albeit in a smaller number of studies. In Cambridgeshire, as will be discussed further below, of 64 individuals analysed for δ ¹⁸O by Scheib et al. (Cambridge and Duxford)Footnote ¹⁶² and Wiseman et al. (Knobb’s Farm),Footnote ¹⁶³ only 2 were considered to have isotope values incompatible with an origin in Britannia.

Beyond Cambridgeshire, few mobility isotope studies have been undertaken on small towns and rural areas, and of those undertaken, many have small sample sizes. For example, the studies by Budd et al. Footnote ¹⁶⁴ on Mangotsfield, Gloucestershire, and O’Regan et al. Footnote ¹⁶⁵ on Dog Hole Cave, Cumbria, had sample sizes of 2 and 7, respectively. Other publications have focused on exceptional burials, such as the military burial from Dorchester, Oxfordshire,Footnote ¹⁶⁶ or the single apparent Sarmatian burial from Offord Cluny, Cambridgeshire, which showed no indication of being unusual until aDNA analysis was undertaken.Footnote ¹⁶⁷ Such small or exceptional samples cannot be considered representative of the local populace. Furthermore, many regions are still underrepresented in isotope mobility studies. For example, while Wales has been pointed out as a possible origin for several outlying individuals,Footnote ¹⁶⁸ studies on Roman burials from Wales are extremely scarce, partially as a result of acidic soils and a limited burial record. Faillace and Madgwick published the only study from Roman Wales with oxygen and strontium isotope values for five individuals from Five Mile Lane, Vale of Glamorgan,Footnote ¹⁶⁹ of whom two appear to be non-local. Drawing firm conclusions on the extent of mobility in Britannia from currently published isotope datasets is therefore challenging. However, although more research is needed, the presence of non-local individuals — whether from within Britannia, or from beyond the province — at almost all sites studied show that movement was common, although the proportion of the population engaged in longer-distance mobility may have varied significantly through both time and space.

Smaller numbers of studies have also utilised isotopes other than oxygen (δ ¹⁸O) and strontium (⁸⁷Sr/⁸⁶Sr) to identify mobility or non-local individuals. For example, lead (Pb) has also been used to explore origins.Footnote ¹⁷⁰ Similarly, a small number of studies have attempted to determine non-local individuals through the use of carbon (δ ¹³C) isotopes — identifying individuals consuming C₄ crops (e.g. millet) that were not cultivated in Britannia.Footnote ¹⁷¹ However, it should be noted that a number of authors (e.g. Müldner et al. Footnote ¹⁷² and de HinghFootnote ¹⁷³ ) have also suggested that C₄ crops may have been imported from the continent — therefore suggesting the movement of foodstuffs rather than the individuals themselves. However, finding an individual with a C₄ signal in Roman Britain is rare. Müldner and colleagues, as well as other studies that have found such individuals,Footnote ¹⁷⁴ therefore usually suggest that the individuals themselves came from outside Britannia as a more likely explanation. This is likely the case, as C₄ foods would have to form a major part of the diet for an extended period to show in collagen isotope values.Footnote ¹⁷⁵ Another isotope approach, sulfur (δ ³⁴S) has become markedly more common in archaeological research in the last decade,Footnote ¹⁷⁶ but it remains rare in studies of Roman mobility and represents a profitable approach for disentangling issues of equifinality in origins. For example, the study by Nehlich et al. Footnote ¹⁷⁷ applied sulfur isotopes to examine weaning patterns in Roman Oxfordshire rather than mobility. That being said, compared to oxygen and strontium, sulfur isotope analysis is still in its infancy and debates on the comparability of past populations to modern sulfur maps are still ongoing.Footnote ¹⁷⁸ For this reason, even if sulfur isotopes are analysed, the interpretations are still often inconclusive.Footnote ¹⁷⁹

As isotope provenancing is an exclusive approach rather than a designative one, using multiple proxies can refine resolution. For example, analysis of an individual from a late Roman rural cemetery in Gravesend, Kent, was found to have ⁸⁷Sr/⁸⁶Sr and δ ¹⁸O results consistent with the local range; however, the same individual’s δ ¹⁵N and δ ¹³C values indicated a diet including a significant proportion of C₄ proteins, in contrast to the local diet.Footnote ¹⁸⁰ This example also illustrates the problem of equifinality — the possibility that two regions can produce isotope values within the same range — and the importance of biosphere mapping for understanding local isotope signatures. Multi-isotope analysis is increasingly used to overcome this issue and to enrich more widely our understanding of mobility, whether human or animal.Footnote ¹⁸¹

The impact of lead (Pb) on individuals has been studied through the analysis of lead isotopes and concentrations.Footnote ¹⁸² During the Romano-British period, lead was increasingly found in the environment,Footnote ¹⁸³ and used in everyday household objects such as cosmetics,Footnote ¹⁸⁴ vessels,Footnote ¹⁸⁵ and even in food products to make them sweeter.Footnote ¹⁸⁶ This cultural change is observable in the lead concentrations of an individual’s skeletal remains, and this has been used as an opportunity to study mobility on several occasions. For example, Scott et al. Footnote ¹⁸⁷ found that the population of Roman London had high lead concentrations which would have affected their health.Footnote ¹⁸⁸ However, as these data were obtained from bone apatite which is strongly affected by diagenesis and prone to exchange contaminants with the burial environmentFootnote ¹⁸⁹ the reliability of the data is uncertain. Montgomery et al. Footnote ¹⁹⁰ published >200 dental enamel samples dating to the Neolithic to the late Medieval period, from 33 sites in Britain, Ireland and Rome.Footnote ¹⁹¹ The dense crystallinity of enamel compared to the apatite structure of bone makes it less susceptible to burial environment contaminants, thus providing more reliable data.Footnote ¹⁹² Montgomery and colleaguesFootnote ¹⁹³ then compared the data against the concentrations and lead isotopes from several lead ore locations in the west and north of the UK. By doing this, they found four additional individuals in York to have originated from the Mediterranean basin, and one woman from Roman London likely to have originated from Rome.

It is not only studies on the mobility of humans in Britannia that have employed isotope analysis. The movement of animals, either as imported exoticaFootnote ¹⁹⁴ or domesticates for food,Footnote ¹⁹⁵ has been explored through strontium and sulfur isotope analysis. In some instances, animals provided a useful proxy for patterns of human movement as their faster-developing teeth and relatively short lives mean greater temporal precision and less bioavailable averaging can be achieved. These studies have generally demonstrated strong evidence for the movement of animals, sometimes over vast distances, occasionally from beyond, but especially within, Britannia.

While isotope analyses of Roman mobility are increasing, there remain biases in site selection and individual sampling that reduce the representativeness of these studies. Research has largely focused on military and urban cemeteries, and selected groups already identified as exceptional through distinctive burial practice or grave assemblages (e.g. the Lankhills burials with ‘exotic’ material culture,Footnote ¹⁹⁶ or the ‘Headless Romans’Footnote ¹⁹⁷ ). Geographical biases also exist, which are partially the result of modern development and archaeological practice (e.g. towards areas of central and south-east England where large-scale construction projects are concentrated), and partially the result of taphonomic factors (e.g. acidic soils largely precluding bone preservation in Scotland, Wales and the south-west peninsulaFootnote ¹⁹⁸ ). There are also chronological biases which are difficult to overcome — with the study of mobility largely dependent on grave assemblages there is a bias towards inhumation, which had largely replaced cremation by the later third century,Footnote ¹⁹⁹ meaning that the earlier Roman period has been investigated less intensively than the third and fourth centuries. However, the relatively recently development and validation of methods for ⁸⁷Sr/⁸⁶Sr on cremated remainsFootnote ²⁰⁰ provides new potential, albeit without the benefit of multiple isotope proxies. For example, as part of the bigger Decoding the Dead project by the Colchester Museum, a cremation cemetery was studied through strontium isotope analyses. Among the individuals are the cremated remains of an older male who was buried in the ‘Colchester Vase’. While the intensely decorated vase itself was likely produced locally, Davis et al. found that the individual was not from the Colchester area.Footnote ²⁰¹ However, without the addition of other proxies such as oxygen, whether the individual arrived from within or outside Britain remains uncertain.

Isotope analysis remains a rapidly evolving sub-discipline of archaeological science. While certainly no longer in its infancy, having been deployed in bioarchaeology for almost half a century,Footnote ²⁰² it is arguably still in its adolescence. The integration of multiple isotope proxies has transformed both the resolution and the confidence of interpretations relating to mobility and issues of equifinality, where different regions have the same bioavailable range in an isotope system, are being overcome more frequently.Footnote ²⁰³ However, challenges with method standardisation and data reliability and comparability pose interpretative challenges, particularly in relation to oxygen isotope analysis.Footnote ²⁰⁴ In addition, understanding of bioavailabilityFootnote ²⁰⁵ and the range of variables that impact on isotope variationFootnote ²⁰⁶ is advancing all the time and interpretative revision of legacy data may often be required. The integration of multiple isotope proxies with wider evidence for mobility undoubtedly provides the best pathway to confident interpretation.

aDNA

In recent years, multi-isotope analysis has been supplemented by ancient DNA (aDNA) analysis. Where isotopes can provide information on childhood origins and lifetime mobility, aDNA can give insights into kinship, descent and genetic ancestry. Aided by recent advances in sequencing technologies and laboratory methods, aDNA can provide a holistic view of past mobility, relatedness and genetic ancestry — and has to date provided significant contributions to our understanding of many archaeological periods.Footnote ²⁰⁷ Richards et al., in one of the very first papers to use aDNA analyses to answer archaeological research questions, highlighted the potential genetic complexity of Britannia’s population — ‘the mixed ethnic composition of the subjects of the Roman Empire’.Footnote ²⁰⁸ Yet few studies of Britannia’s populations have been undertaken, although the volume of sampling undertaken in recent years is likely to change this significantly in the near future. The majority of aDNA work undertaken on Britannia thus far has focused instead on unusual individuals, including possible victims of infanticide and Roman Londoners buried with exceptional objects.Footnote ²⁰⁹

The largest study published to date, and also one of the few studies focused on rural sites, is that by Scheib et al., which generated genome-wide data from 52 individuals from eight cemeteries in Cambridgeshire.Footnote ²¹⁰ The authors suggest that communities in Britannia saw a low genetic impact from the Roman occupation, with little genetic ancestry differentiation between Roman, Bronze Age and Neolithic individuals. This is in contrast to smaller aDNA studies of urban or unusual burials from Britannia which have often identified non-local individuals. For example, Martiniano et al. analysed seven individuals from York, one of which appeared to have Middle Eastern ancestry,Footnote ²¹¹ and Silva et al. identified Sarmatian ancestry in an isolated burial from Cambridgeshire.Footnote ²¹² Interestingly, however, oxygen (δ ¹⁸O) isotopic data generated from the same Cambridgeshire individuals did suggest a number of non-local individuals.Footnote ²¹³ This therefore highlights the need for future studies to sample the rural population of Britannia more extensively than has hitherto been the case, but also the need to combine multi-isotope systems alongside aDNA data in order to gain a more detailed understanding of Roman mobility.

Another notable study is that undertaken by Redfern et al., who investigated four individuals from Roman London using aDNA analysis in tandem with oxygen, strontium and lead isotope analyses, alongside contextualising the burials through analysis of their material culture and site context.Footnote ²¹⁴ These analyses indicated that one female individual grew up in Britain but had genetic ancestry from somewhere within eastern Europe, the Caucasus or central Asia.Footnote ²¹⁵ One male had ambiguous results indicating he may have had genetic ancestry beyond Britannia, or may have been local. Another male had Black African ancestry but spent his childhood in the London region, and the final individual, a 14-year-old girl, grew up around the Mediterranean and moved to Britannia around the age of 9. The study indicates a high degree of migration into Britannia and a diverse population, but the individuals analysed were drawn from London, possibly the most cosmopolitan urban settlement in the province, perhaps alongside York.Footnote ²¹⁶ Rural populations may not have had the same levels of mobility.

Most recently, Walton et al. Footnote ²¹⁷ have reported aDNA data alongside macroscopic craniofacial analysis and isotope data for an individual known as ‘Beachy Head Woman’ (BHW), believed to be recovered from the coastline near Eastbourne. BHW serves as an important reminder of the development of aDNA technologies within the last decade, and how significantly our interpretations of even a single individual can shift. Initial low-level sequencing of BHW in 2017 suggested an east Mediterranean origin, posited to be Cyprus. However, more recent aDNA data generated using capture array technology has instead suggested a strong genetic affinity to individuals from Roman Britain — thereby indicating local British ancestry, and supported by ⁸⁷Sr/⁸⁶Sr values.

The future of aDNA studies in Roman archaeology is rich with possibility, given the complexity and geographical spread of the empire, its connectivity and the suite of material evidence available. The potential of aDNA for investigating aspects of human populations beyond mobility, such as genetic relatedness, genetic diversity, population structure, disease and phenotypic traits has also barely been explored in the Roman period. In particular, recent developments in kinship analysis now mean it is possible to determine genetic relatedness between individuals at both an intra- and inter-site level. While these methods determining levels of genetic relatedness have been employed extensively to prehistoric human skeletal material, they are yet to have been applied to individuals from Roman Britain. In future, considering genetic relatedness alongside broader narratives surrounding mobility and gender has the potential to enable a deeper understanding of Britannia. Lastly, there is additionally vast scope for investigating not only human populations but also domestic animal and plant genetics in the period, given the outstanding existing aDNA work on selective breeding, species introductions, domestication, commensal animals and human–animal relations.Footnote ²¹⁸ However, it is important to note that, particularly in framing results to the public and press, lessons can be learnt from other periods where aDNA has been more extensively employed.Footnote ²¹⁹ As the field of ancient DNA has developed and becomes more interdisciplinary, significant progress has been made in improving collaborative culture and presenting more nuanced narratives to the public. It is vital we learn from these experiences in Roman archaeology, particularly given the centrality of the period in public narratives.

The future of studying mobility in Britannia

The examples discussed here from across the suite of evidence of mobility in Britannia demonstrate both the historiographical development of Roman mobility studies and the difficulty of truly holistic work in the field, given the breadth and complexity of the data and their history of collection and study. Interdisciplinary work across archaeology, history, biosciences and beyond is therefore necessary to provide more comprehensive insights into mobility;Footnote ²²⁰ it is insufficient to rely primarily on one line of evidence, whether epigraphic, genetic, or whatever else, to advance a general narrative, given the richness of the data and historiography. Evidence sources operate at different scales and have complementary strengths: therefore ambiguities in one proxy can be clarified using others.

This is the aim of our Arts and Humanities Research Council-funded Roman Britannia: Mobility and Society project (RoBMobS). Drawing together cutting-edge aDNA, multi-isotope analysis (Sr, O, S, C and N) and scientific dating with expertise in artefactual and stratigraphic research, we will work in partnership with archaeological units and museums to develop multi-scalar and temporally rich narratives of the many mobilities within, and into, Britannia through the Roman period. Through collation of all previous isotope and aDNA data from the province, and by undertaking isotope analysis of up to 400 individuals and aDNA screening of 200 individuals, from non-military inhumation cemeteries across the province, we will test the interpretative potential of an integrated approach. This research will provide the first large-scale multi-proxy study of this kind for any Roman province, and in targeting non-military and principally non-urban cemeteries will provide a more balanced view of mobility both within and beyond Britannia.

We have argued that Roman archaeology has a problematic inheritance of its long-term entanglement with modern imperialism, with the embedded reification of ‘Roman’ culture and widespread conceptualisation of Romanisation explaining cultural change promulgated by that ideology and perpetuated by the biases towards the survival and study of material culture in elite or more durable media. This exploration has, we hope, shown that parsing the ancient and modern biases in our datasets, such as the privileging of male, high-status and often military individuals in epigraphy, or the preferential study of exceptional or unusual burials or urban centres in much bioarchaeological research, is just as key to successful future studies of mobility in Britannia as the creation of truly interdisciplinary research groups to exploit the richness of Roman archaeology’s data. In future, we must, if our collective consensus that Britannia was a colonial encounter is to be meaningful, focus the cutting-edge tools of our discipline on the people at the sharp end of that encounter, the rural poor.Footnote ²²¹ There is a deep and ironic echo of older historiographies in the methods of bioarchaeology being used to enrich narratives of the upper classes of Britannia.Footnote ²²² Work on urban, elite and military assemblages must be rebalanced through investigation of the mobilities of the rest of Britannia’s population. New narratives will emerge from holistic research into these communities, which will surely disrupt current understandings and provoke new questions.

Focusing on mobility in this review leaves us open to critique for privileging the ‘movers’ over the ‘stayers’.Footnote ²²³ It has also been argued that theoretical frameworks such as globalisation and glocalisation, with their emphasis on connectivity and the variably shared ‘Roman’ cultural milieu of much of higher-status Roman society, simply reproduce Romanisation under another name.Footnote ²²⁴ We argue that this review has again demonstrated that mobility was fundamental to society in Britannia, whether or not an individual moved. Given the improving precision of multi-isotope analytical techniques and UK biosphere mapping, modern bioarchaeology may be able to contribute significantly to dissolving that dichotomy through multi-scalar nuance. Indeed, if conducted in combination with detailed artefactual study, stratigraphic analysis, scientific dating and chronological modelling, the widely neglected temporal dimension of changes in mobilities through time can also be greatly enhanced.

Much valuable work has been undertaken to draw out the complexities and nuances of group, individual and other identities articulated by material culture, epigraphy and burials in Britannia. Thomas Booth argues that ‘when addressing simple questions of demographic history, genome-wide palaeogenetic analyses reasonably take precedence over the archaeological record, as genetic evidence is a direct measure of population change, rather than a potentially poor proxy’.Footnote ²²⁵ For questions of demography focused on ancestry, this is surely the case, just as isotope analyses provide direct evidence of individual mobility within the lifecourse. However, these issues are far from the limit of our questions. Whether or not one was a migrant or descendant of migrants in the Roman world is not the same question as whether one identified, or was identified by others, as such. Nor have questions of social and economic transformation during the Roman period often been framed so simply as cultural change being caused by population replacement.Footnote ²²⁶ Rather, thanks to the rich archaeological and historical dataset of the period, the critical questions revolve around the character of the encounters caused by the historical conditions of Rome’s expansion, control and transformation. To emphasise the importance of exploring these in a spatially and temporally multiscalar way, engaging with the nuance of individual identities and biographies, whether they include mobility or not, as well as investigating the colonial experience of Britannia in the Roman period at a provincial level, is to argue that archaeologists asking questions of the Roman period should be empowered by the bioarchaeological revolution in collaborative ways, rather than disenfranchised.

Future work on the topic should meet Woolf’s three requirements of archaeological studies of mobility in antiquity.Footnote ²²⁷ Firstly, to differentiate between different kinds of migrants and different kinds of mobility. Secondly, to investigate changes over time in mobility, rather than simply acknowledge it as a structural fact; this latter is one consequence of less-than-nuanced readings of Horden and Purcell’s grand narrative of Mediterranean ecological diversity and socio-economic connectivity, and is essentially reductive.Footnote ²²⁸ Thirdly, to make quantitative claims about mobility; how many, how often, how far, and how many did not move. While formulated for the Mediterranean, these sensible criteria are highly relevant to Britannia, and achievable given the evidence and techniques available to archaeology today. To these we would add that, if undertaking primary sampling of human remains, a partnership with the custodians of those individuals is vital to maximise the value to today’s society of the destruction of human remains. Completing the research loop through prioritising timely dialogue and resourcing the renewal or creation of public-facing interpretive materials relating to the individuals is ethically vital. Without this, researchers abdicate responsibility and leave an opening for the misuse of results, without assisting museums with the specialist knowledge to respond to external criticism.

As this interdisciplinary field develops, we anticipate greater integration of scientific techniques and increasing interdisciplinary work across the old boundaries of arts and science. Some work will shift from publishing smaller datasets or individuals to focus on population dynamics and better chronological resolution, as well as site- or region-based studies. As datasets grow larger, scientific data will become more reasonably able to be compared to large provincial epigraphic datasets to provide overviews of mobility. Woolf argues that migrants would mostly have been male, and highlights Haley’s epigraphic study of mobility in Spain and Wierschowski’s in Gaul.Footnote ²²⁹ Both studies suggest a migration rate of around 5%, and that most migration was local or regional, and by men. These studies deploy large datasets, identifying 649 and 715 migrants respectively. The well-known biases of epigraphic commemoration are not considered at length in these studies, and it will be fascinating to see if these patterns are borne out in large-scale scientific studies.Footnote ²³⁰ So far, for Britannia, estimates of longer-distance mobility vary significantly. The gender narrative in Woolf’s argument is reminiscent of those around women in studies of military communities, before the pioneering work of Lindsay Allason-Jones and others demonstrated conclusively the presence of women in military spaces from archaeology.Footnote ²³¹ Given the evidence of mobility of the less wealthy from historical sources, it is debatable whether the practice of founding society-wide narratives for mobility on epigraphic evidence can be sustained in light of the scientific datasets currently being generated.

As bioarchaeological techniques, alongside artefactual analysis and scientific dating and modelling, refine our understanding of how people moved and when they did so, we will be better empowered to answer the questions of why people moved and how they understood themselves in a social context. After all, mobility is a social behaviour conducted in specific contexts; as Gerrard emphasises, if mobility is bluntly defined as those who die in a different place from where they were born, this can dehumanise these individuals’ lives, so we must more subtly explore this behaviour’s complexities at different temporal and spatial resolutions.Footnote ²³² This future and continued work on mobility within Britannia will therefore enable other significant questions to be addressed: for example, the extent to which the rural poor became increasingly tied to the land as the Roman period went on;Footnote ²³³ whether pre-conquest elite groups maintained wealth and power in Britannia, and if this varied by region; and the changing demographic profile of supposed ethnic auxiliary units in the Roman army. All of these have much wider implications for understanding the imperial encounter in Britannia than simply characterising mobility, and can be greatly enhanced through using the latest bioarchaeological techniques alongside historical and archaeological analysis in current theoretical frameworks. There are undoubtedly significant challenges accompanying the adoption of bioarchaeological approaches into our studies of mobility in the Roman period in a holistic, nuanced manner, but we are convinced that the opportunities to address critical questions anew to Roman archaeology, with relevance to much wider academic and contemporary debates, must be taken.