Download file in PDF format: TRAC 2007: Identity in the Frontier: Theory and Multiple Community Interfacing (pp. 45–52)
Identity in the Frontier: Theory and Multiple Community Interfacing
Introduction Recent studies of identity in the Roman period have focused on identity as a substitute for romanization at the upper levels of society or at the level of the individual. It is for this reason that identity studies have tended to focus on changing identity in the late Iron Age and early Roman periods (e.g. Woolf 1998; Pitts 2005). While this is not a criticism of such scholars, I argue that identity is a complex social construct that has significance at all scales of social organization and cannot simply be restricted to processes of romanization. Given the diverse cultural groups and complex social hierarchies of the Roman period, not to mention the long-term political superiority of the empire, identity is a concept that must be carefully considered and explicitly applied in Roman studies. This paper therefore seeks to consider the utility of identity and identity studies in the frontier zone of northern England. Traditional studies in frontier identity have relied largely on the information available from inscribed stones, though this largely restricts the evidence to the second and third centuries. Any study of identity in the fourth century must rely on different data and explore different aspects of the concept of identity construction.
Identity in the Frontier – The Traditional Analysis The frontier zone of northern England provides a rich dataset for those studying identity thanks to the Roman habit of inscription. The hundreds of inscribed stones found at forts throughout the North, not to mention the rest of the Roman Empire, offer an unrivaled source of information about individuals.
Let us consider two inscribed stones from Hadrian’s Wall (described and illustrated in full in Allason-Jones 2002). Significant aspects of identity are bold-faced in the translation.
1. Tombstone of Victor (RIB 1064): D(IS) M(ANIBUS) VICTORIS NATIONE MAURUM [A]NNORUM XX LIBERTUS NUMERIANI [E]Q(U)ITIS ALA(E) I ASTURUM QUI PIANTISSIME PR[OSE]QU(U)TUS EST
‘To the spirits of the departed (and) of Victor, a Moorish tribesman, aged 20, freedman of Numerianus, trooper of the First Cavalry Regiment of Asturians, who most devotedly conducted him to the tomb.’
This second century tombstone indicates numerous aspects of this man’s identity, or those aspects deemed important by his former master. Notably these are his name, ethnicity (also linked to his geographic origins in this case), and his age. Furthermore, his social status is indicated, as is that of his master, both of whom are linked to a specific military unit.
2. Altar from Carrawburgh Mithraeum (RIB 1545): D(EO) IN(VICTO) M(ITHRAE) S(ACRUM) AUL(US) CLUENTIUS HABITUS PRA(E)F(ECTUS) COH(ORTIS) I BATAVORUM DOMU ULTI N(I)A COLON(IA) SEPT(IMIA) AUR(ELIA) L(ARINO) V(OTUM) S(OLUIT) L(IBENS) M(ERITO)
‘Sacred to the Invicible God Mithras: Aulus Cluentius Habitus, prefect of the First Cohort of Batavians, of the Ultinian voting-tribe, from Colonia Septimia Aurelia Larinum, willingly and deservedly fulfilled his vow.’
This altar, dated to the early third century, gives a number of aspects important to this man’s identity. His personal name and family (which includes an ancestor defended by Cicero in 66 B.C.), his rank and unit, his tribe, his geographical origins (which also reveals a degree of education/snobbery as the abbreviation of Larinum was the same used by Cicero), and most obviously his religion or rather his desire to appear dutiful in his religious behaviour, which may also indicate his social network, that of the devotees of Mithras, which Allason-Jones (2004) suggests was a cult as much about networking as faith.
These stones are but two examples of the information available about individuals. Both of these stones indicate that identity in the second and third centuries is multi-faceted. Name, ethnicity and geographical origins are basic elements in establishing the identity of these individuals. But age can be an important factor, too, as age can preclude or grant an individual certain social limitations or liberties. Unsurprisingly in the military zone and frontier, the rank and unit that the individual is linked to is also important. These facts, shared through inscription, form a basis for reconstructing the individual’s identity.
As a corpus, the inscribed stones tell us that the frontier was a rather cosmopolitan, multicultural zone. Officers and soldiers were posted from across the empire to fulfill a number of roles. A number of other individuals such as doctors, servants, slaves, and traders directly or indirectly associated with the military also lived in the frontier.
While inscribed stones are an undeniably rich source of personal data, they present two major problems in the construction of identity in the frontier:
1. At a conceptual level, inscribed stones do not allow a complete reconstruction of personal identity. Often, the information available about the individual’s immediate family, let alone any extended family, is limited, which would be important in any tribal society from inside or outside the Roman empire.
2. The limitations of such a large corpus of inscriptions are often unaddressed. There is a bias toward military officers or wealthy private individuals linked to the military. This is to be expected in the frontier, but leaves us wondering about the lower salaried soldiers or local (native) farmers. Furthermore, there are very few inscriptions that date to the later third and fourth centuries. These two problems are addressed in turn below.
Defining Identity Identity is a concept that must be considered in light of research in the disciplines of anthropology, sociology, and psychology. The social sciences have demonstrated that identity is not only complex but fluid and highly contextualized (Howard 2000; Hitlin 2003; Hogg et al. 1995). Individual identity is not monolithic; it is composed of numerous lesser identities that are themselves the creation of personality characteristics and social roles, and these lesser identities come to the foreground depending on situational circumstances (Jones 1997). To put it another way, individual identity is multi-layered, with the configuration of layers changing for each role or social situation that a person engages with.
Thus, a legionary legate of the early empire would be a (1) Roman citizen, (2) a man of a (3) certain family of (4) senatorial rank and (5) income. This man is someone’s (6) son, perhaps (7) sibling and (8) nephew or cousin. Additionally, he is likely to be (9) married and have (10) children of his own, not to mention (11) servants and slaves. Each of these familial relationships informs the man’s sense of identity. Further input comes from his (12) education and (13) previous jobs/postings. As a (14) legionary legate he commands the legion, (14a) acts as a judge, (14b) serves as a patron for junior officers, (14c) commands the resources of the legion in terms of labour and supplies, yet he himself is a (14d) subordinate of the provincial governor and emperor.
This is a simple listing of 18 roles, which does not take into account pyschological or physical characteristics that would contribute to the man’s personality, or the personal experiences that further temper his character. At any given time, the legionary legate is a military commander, husband and father, patron, master, and/or judge. While each role cannot be seen as unique and psychologically separate, each role will have its own behavior pattern that may differ radically from that of another role. Each of these roles defines the man’s identity, though circumstances dictate how that identity will be perceived by himself and others.
Achieving an understanding of individual identity to this extent is an ideal, and it is almost impossible for an archaeologist or historian to achieve. But it is important to recognise the amount of information we are lacking or not incorporating into our studies and analyses of identity. For the individual, we are incredibly reliant on written information – inscriptions, contemporary biographies or references in textual sources. This information is highly biased. Inscriptional evidence, at least in Britain, is generally limited to the military zone; ancient writers usually only mention the ‘movers and shakers of the day – emperors, generals, rebels and usurpers, barbarian leaders, bishops, the occasional eunuch and the like – or gloss over an individual with a very basic identification. These biases increase the importance of sources like the Vindolanda tablets that provide insights on the mundane existence of daily life.
Archaeological evidence is limited in the amount of detail it can reveal about individual identity. Small finds research has been used to elucidate ethnic origins of certain units, most recently at Brougham (Cool 2004). Unfortunately, artefacts often limit studies of identity to gender and ethnicity, or occasionally distinguish between military and civilian, though this latter separation is increasingly being recognised as overly simplified. However, archaeology is well suited for considering evidence at a general level, for example social class. This is fortuitous, because as the social unit increases in scale (from individual to settlement, tribe, social class, etc.) identity becomes a simpler construct. A settlement is a scale of occupation that archaeologists are familiar with, and this scale of social formation can also be linked to the concept of community.
Community Studies A significant focus of anthropological and sociological research since the late nineteenth century has been the concept of the community and communitas – the sense of belonging to a community and the practices that membership incorporates. Despite the tradition in anthropology and sociology, community studies have not been incorporated into European archaeology (though see numerous studies on ‘identity’ to which this is related: e.g. Giles 2000; Jones 1997; Mattingly 2004; Roymans 2004; Shennan 1989; Wells 2001).
There are multiple sociological definitions of community. The traditional definition, encapsulated by Tönnies’ (1955; first published 1887) Gemeinschaft, understood a community to be a set of social relations most clearly achieved in traditional, rural village life. This definition has been strongly criticized and reworked since Durkheim’s review in 1889 (Delanty 2003: 36). The most recent and influential reconsideration of community was articulated by Cohen (1985), who argued that communities were a symbolically constructed reality, in which community membership and identity was recognized in contrast to outsiders. There are two notable problems, however, with Cohen’s symbolically constructed community. First, the emphasis on the exclusive nature of communities detracts from the fluidity and dynamic nature of community formation and identification. Second, he does not recognise that communitas can take violent forms and that violence is often a marker of community boundaries (Delanty 2003: 47–48).
The multitude of sociological definitions indicates that community is not an exact term or concept. Indeed, there is a multitude of types of communities, with numerous approaches by which to study these differing communities (Bell and Newby 1974). There is no fixed and static notion of community.
‘Community’ stands as a convenient shorthand term for the broad realm of local social arrangements beyond the private sphere of home and family but more familiar … than the impersonal institutions of the wider society … [Crow and Allan 1994: 1; reworking Elias’ 1974: xix definition].
Put another way, community is a mode of social organization and a manner of belonging that can often be expressed in symbolic forms, but community is not an exclusively institutional arrangement (Delanty 2003:49). Different types of community can be distinguished according to their contrasting features and characteristics, what Stacey (1974) would identify as a local social system and what archaeologists often classify as identity.
As communities are generally considered as smaller scaled social formations, larger than families but smaller than tribes, it is possible for numerous communities to co-exist within larger social and political formations. This can take the form of several geographically distinct communities or different types of communities within a close proximity, or both. Thus, the Roman Empire can be said to have contained innumerable communities of different types dispersed throughout it at any one time interacting, or rather interfacing on a regular basis. The term interfacing is preferred to interaction or interrelationships because interfacing requires acknowledged interaction between two or more parties, even if one party is a more passive recipient.
Frontiers, as complex transitional zones of contact between the Roman Empire and barbarian societies (Whittaker 1994), contained a number of different interfacing communities. Roman military communities are an obvious example and have been the most frequently studied group in the frontier, though rarely in the context of community studies. In addition, the development of towns in frontiers under the aegis of imperial domination and administration indicates that there were also urban community formations. Outside of the towns and distinct from the military were rural, “native” settlements that can generally be taken to indicate the presence of agrarian peasant communities. The final form of community present is the barbarian communities, which may or may not have varied considerably to native agrarian peasant communities in the frontier. Each community had different traditions, practises, institutions, and values, but these communities also would have come into contact with each other.
Community Identity and Interfacing As noted above, there is limited evidence for rural agrarian communities in the frontier in the late Roman period. In fact, most of the data from the fourth century is from excavations at military installations. Therefore, of necessity, the primary focus here will be on military communities. The military community was formed around a core of soldiers, whose sense of communitas was created and maintained in a number of ways: through drills and instruction in arms; use of Latin language and literacy; the provision of food, shelter, arms, and other facilities (e.g baths); religious and social rituals; dress code and other regulations; membership in a designated unit; and hierarchical organization, not to mention physical and psychological conditioning to perform violent actions. However, the military community was not exclusively made up of soldiers. Numerous individuals of differing social stations and occupation can be considered as part of the military community. This includes the dependants and families of soldiers, as well as those individuals that were in the frontier providing services and/or goods that supported the military (James 2001).
By the fourth century, the regiments along Hadrian’s Wall had been posted at each fort for at least a century, if not closer to two centuries. The fixation of frontier postings was reinforced by Constantine’s separation of mobile field armies from static frontier armies. This latter group, the limitanei were engaged in duties to maintain and police the frontier and its installations (Elton 1997: 204–205), and long-term occupation of the same area in some ways facilitated such duties. Knowledge of the local and regional landscape could be counted on, and ideally a supply infrastructure could be more easily guaranteed. Furthermore, the management or exploitation of the surrounding countryside could be thoroughly achieved. Yet, there are a number of social implications entailed by the long-term occupation of the region.
One factor related to this is the extent to which the military community was socially (if not economically) self sufficient. Was the non-soldierly population of the military community – the servants, dependants, merchants, etc. – large enough to provide the soldiers with everything that was needed? Primarily, the issue here is demographic. Were there enough prospective partners or spouses for the soldiers or their dependants? Could the military community provide the necessary number of recruits and other people to maintain regimental and support staff numbers? If not, to what extent was the local civilian population a source for necessary logistical support or prospective partners? Did any codified regulations or perceived mores restrict or encourage such behaviour? In either case, a soldier would have important familial relationships in the area of his posting, whether these were limited to the immediate vicinity of the fort/installation or perhaps somewhere else in the frontier. In this way, a soldier would identify the frontier region as his home and not only his official posting.
For the rural peasantry, the initial military occupation may have been brutal and exploitative (Mattingly 2006). After two centuries of such occupation, however, would the local regiment still be seen as a bunch of exploitative, bullying outsiders? Even if the level of exploitation from the military never decreased, surely after at least 8 generations such occupation would be seen as a matter of fact. In many cases, it is possible that the situation could have improved. At worst, an ‘us-and-them’ mentality may have existed that is typical of city and rural hinterland relations in which everyone acknowledges that they share the same home, but prefer not to mix socially. At best, the rural peasantry had come to see the local regiment as composed of their own sons, nephews, cousins, or grandchildren, with such kin-relations minimizing or reducing animosity. Elsewhere, I have argued that the theory of occupational communities is a rewarding manner in which to consider the changing military identity through the centuries, and that in the later fourth century the limitanei in the frontier became increasingly localised (Collins 2006). While they may have been soldiers of the Roman Empire, they were in fact soldiers of northern England. Archaeological evidence bears this out in a number of distinctive structural traits typical of late Roman northern forts (Collins 2007). These traits indicate not only a transformed physical existence in the forts, but also a changed cultural experience. For example, principia were the official headquarters building in a fort, containing offices, the regimental shrine, and a large hall and courtyard for official assemblages. By the later fourth century, these official functions were supplemented with other functions, such as domestic occupation, metal working, and butchery. At South Shields, the principia forecourt even incorporated a Christian church (Bidwell and Speak 1994). Cool (2000) has also demonstrated a shift in the artefactual assemblage of the late fourth century North, though at this point it has yet to be determined if this is regionally distinct. Unfortunately there is little evidence from rural farmsteads for comparison. The rural agrarian community is well known, archaeologically speaking, if not well understood. The settlements themselves typically retain similar characteristics such as numerous conjoined hut-circles and enclosures, an east-facing entrance, and often an enclosing ditch. These morphological characteristics can also be clustered regionally so that ‘rounder’ settlements are seen in the Eden Valley and Solway Plain while the rectilinear settlements east of the Pennines are distinguished by clusters of enclosed and unenclosed examples. In general, there are fewer artefacts from farmsteads in northern England. These are almost always a handful of glass objects (beads and bracelets) and the occasional sherd of Roman pottery. It should be noted that these artefacts tend to date to the second century, with little or no evidence dating to earlier or later occupation on site. There is a presumption that these settlements were occupied over the course of centuries (Clack 1982:381).
Hawkhirst in Cumbria, is a very notable exception and different to the typical native farmstead (Haverfield 1899; Simpson and Richmond 1936). Cropmarks observed through aerial photography indicated four ‘typical’ farmsteads in close proximity to each other (Cumbria SMR 305). Excavation focused on one of these farmsteads found evidence for ramparts formed by cobbles on a foundation of flags and structural evidence consisting of ashlar blocks, paving, and a series of small brick arches that the excavators interpreted as evidence for a hypocaust (Simpson and Richmond 1936). There was also a large number of finds, including a third century coin hoard, Huntcliff ware, a bronze lamp, a bronze statuette (possibly of Mercury), a bronze ornament inscribed with IOVIS, and a fourth century crossbow brooch. The high numbers of Roman objects Identity in the Frontier: Theory and Multiple Community Interfacing 51 and possible hypocaust suggest this site was not the homestead of a typical native farmer. Might this be the home of a local government official or someone with a good relationship with the local garrisons, like a trader or merchant? Or perhaps this is an example of a rural native elite, a type of which is hitherto unknown in the frontier this far north. Unfortunately, the uniqueness of Hawkhirst does not help in attempting to understand any relationships between the military and rural communities.
The extent to which the settlement at Hawkhirst is truly an exception is unknown. While many native farmsteads are known, not many have been excavated. It should be noted that from aerial photographs, Hawkhirst appears similar to most other native farmsteads. Those that have been excavated have not been subject to large open area excavations. Rather, certain features such as ditches, entrances, and walling have been the focus, leaving occupation floors and courtyards uninvestigated. Complete excavation of more rural sites could considerably change our idea of a typical farmstead.
Excavation and scientific study of burial evidence in the frontier would also be beneficial. This area of archaeology has been largely ignored in the frontier, with the exception of the inscribed memorial stones. Ideally, genetic profiling, stable isotope analysis, and other scientific techniques could provide physical evidence of discreet or mixed populations.
Conclusions The study of identity in the frontier is traditionally associated with information from inscribed stones, but this biases our study to the early Roman period. To assess identity in the late Roman period, we must consider evidence at a broader level – that of the community. Unfortunately, our datasets are largely limited to military communities. Further research focusing on rural settlement and burial activity (military and non-military) would help expand our knowledge beyond the military community and provide meaningful data by which to assess community identity. Finally, we must not forget that identity is a complex social construct that will have changed over the course of time, and that we should not focus exclusively on the Iron Age to Roman transition.
Museum of Antiquities, Newcastle University
Acknowledgements I wish to thank Lindsay Allason-Jones for reading and commenting on a draft of this paper, and Tony Wilmott, Nick Hodgson, and James Gerrard for offering additional comments at the conference session in which this paper was given. Any mistakes remain my own.